2023-01-01 16:21:56 <scorpion2185[m]1> is this  https://github.com/waycrate/swhkd being packaged by someone?
2023-01-01 16:22:50 <ikke> https://pkgs.alpinelinux.org/packages?name=swhkd&branch=edge&repo=&arch=&maintainer=
2023-01-01 16:24:51 <scorpion2185[m]1> since I didn't see it I asked in case someone was interested
2023-01-01 16:41:09 <scorpion2185[m]1> make: rustup: No such file or directory
2023-01-01 16:41:09 <scorpion2185[m]1> 
2023-01-01 16:46:05 <psykose> apk add cargo; cargo build --release
2023-01-01 16:58:10 <scorpion2185[m]1> are there gtk themes for lxappearance? I installed it from testing
2023-01-01 16:58:30 <scorpion2185[m]1> psykose: thanks i'll try that
2023-01-01 16:59:53 <Ermine> you need to look for gtk2 themes
2023-01-01 18:52:10 <scorpion2185[m]1> no adwaita-dark? I searched gtk2
2023-01-01 19:20:45 <scorpion2185[m]1> psykose: this time it complied some things but again that error. in what order do I run it after make setup, make clean and make ?
2023-01-01 19:21:05 <scorpion2185[m]1> s/in/Do/, s/what order do//
2023-01-01 19:21:30 <psykose> simply never call setup
2023-01-01 19:24:45 <scorpion2185[m]1> and no make clean and no make?
2023-01-01 19:59:59 <scorpion2185[m]1> where does it place the things that build?
2023-01-01 20:28:53 <ikke> If it's just make, most of the time in the same directory
2023-01-01 20:36:45 <scorpion2185[m]1> I am trying to package it, with only make I have that error : `make: rustup: No such file or directory`.  you should use make setup, make clean and make
2023-01-01 20:36:45 <scorpion2185[m]1> cargo build --release builds some things
2023-01-01 22:01:54 <paperowl> 10,99scorpion2185[m]199,99 what is your rust setup like on alpine? Did install via the rustup toolchain install script or the alpine packages. You may need the full toolchain to use rust make, which im not too familiar with. I did a cursory glance at the makefile for swhkd
2023-01-01 22:05:13 <psykose> it never tries to use rustup unless you do `make setup` so just don't
2023-01-01 22:06:23 <paperowl> Listen to him since he maintains the rust package
2023-01-01 22:13:44 <ikke> her
2023-01-01 22:14:44 <paperowl> My apologies 
2023-01-01 22:31:08 <psykose> i'll package it for you anyway
2023-01-01 22:31:21 <psykose> it needs pkexec and a functioning polkit to do anything
2023-01-01 22:31:29 <psykose> so does anyone remember offhand how to allow a user to pkexec :p
2023-01-01 22:34:41 <psykose> scorpion2185[m]1: you can use this in the meantime https://img.ayaya.dev/irKuuTcNRUpA
2023-01-01 22:58:11 <EtherMan> Hi. So I was wondering a bit about alpine networking. So there's a package ifupdown-ng-iproute2 which if I understand it, provides the vrf capabilities of ifupdown-ng? I installed that (and ifupdown-ng), and in my interfaces file, I have https://pastebin.com/Py2esVdd which as I understand the docs to mean that it now should make the address and gateway routes for that interface into table 10. I've now rebooted that VPS but that networking 
2023-01-01 22:58:11 <EtherMan> is now broken because while it does set up two ip rules for oif and iif for the vrf itself, eth0 never gets configured and no routes gets populated in table 10. What am I missing here?
2023-01-02 00:05:31 <minimal> EtherMan: did you look at this? https://ariadne.space/2021/09/13/efficient-service-isolation-on-alpine-with-vrfs/
2023-01-02 00:08:52 <EtherMan> Oh... "In future versions of ifupdown-ng, the routing rule setup will be handled automatically."... I thought that was working since the package was there ^_^;
2023-01-02 00:17:04 <minimal> EtherMan: there is 1 VRF-related commit in upstream git master that is not in any release yet
2023-01-02 00:17:36 <minimal> and so not in the Alpine package
2023-01-02 00:18:06 <EtherMan> And I have no idea if ifupdown-ng itself handles it. It was just the first I came across when looking at it :)
2023-01-02 00:19:19 <minimal> EthernMan: actually there are also 2 pending VRF-related PRs upstream
2023-01-02 05:07:43 <Guest483> I think I may have succeeded in doing a bootable data disk install but probably not. All I know is I unplugged my usb and rebooted and the computer boot into alpine, yet there are all the root directories I would expect from a full install, even though I ran setup-disk in mode data and ran setup-bootable afterwards. How can I tell if the OS is really booted into RAM? 
2023-01-02 05:08:14 <infinite[m]> Guest483: using miniroot to make one
2023-01-02 05:47:04 <Guest483> Hmmm, when I run mount, I see that the root directory (/) is mounted to tmpfs, so I think it is ephemeral (diskless). I would expect / to be mounted to /dev/sda2 or something if it was a system disk mode install.
2023-01-02 05:48:58 <ikke> Guest483: did you boot in the installation medium again?
2023-01-02 05:49:14 <ikke> oh you mentioned you unplugged your usb
2023-01-02 12:18:27 <EF^> Afternoon. Looking for a suggestion please.
2023-01-02 12:20:19 <EF^> On a x86_64 machine zith LUKS on RAID. Upgraded to 3.17 from 3.16. However, passphrase is rejected at boot.  Think this is the same as issue 13004
2023-01-02 12:21:35 <EF^> Booted on an Alpine USB key and can access LUKS using expected passphrase after assembling RAID etc.  Any suggestions on next steps please?  TIA
2023-01-02 12:28:03 <ikke> EF^: Did you check the luksDump output?
2023-01-02 12:34:00 <EF^> Hello ikke:  It is whirlpool using serpent-xts-plain64 as a cipher
2023-01-02 15:11:27 <psykose_> yeah that probably won't work
2023-01-02 15:25:57 <EF^> psykose_: re whirlpool issue? Any suggestions welcome
2023-01-02 15:29:16 <elly> is that cipher compiled into your kernel?
2023-01-02 15:36:00 <EF^> It is the stock kernel.  
2023-01-02 15:37:12 <elly> the stock kernel appears to have serpent and whirlpool (wp512) as modules, maybe they aren't in your initramfs?
2023-01-02 15:39:50 <morena> Hey, I have some issue with domain resolution on alpine. nslookup can't find answer for some subdomain, while tools like unbound or knot utils resolve subdomain, any idea?
2023-01-02 15:39:51 <minimal> elly: I'd expect EF^ has "cryptsetup" in their mkinitfs feature list and that will ensure that the Whirlpool module is included
2023-01-02 15:40:21 <elly> ah
2023-01-02 15:42:32 <minimal> morena: you'll have to give more information than that for anyone to be able to help you
2023-01-02 15:44:16 <elly> my very first thought would be /etc/nsswitch.conf or /etc/resolv.conf though
2023-01-02 15:45:21 <minimal> elly: musl does not use nsswitch.conf
2023-01-02 15:46:36 <elly> right, but other tools may
2023-01-02 15:46:38 <morena> the problem is I don't have more information. Spin up alpine server on digital ocean and when trie nslookup "irc.irc-nerds.net" I get no answer
2023-01-02 15:46:41 <elly> there's a warning at the top of it about that
2023-01-02 15:47:06 <morena> no matter what nameserver I set up in resolv.conf
2023-01-02 15:48:38 <EF^> minimal: was a workibg setup. Just upgraded. How does one force whirlpool to be included in initramfs or check it is?
2023-01-02 15:49:34 <minimal> does your /etc/mkinitfs/mkinitfs.conf file include "cryptsetup" in the features list? I expect so, otherwise LUKS in general would not work
2023-01-02 15:50:58 <minimal> morena: of course you can provide more info - you haven't even indicated which version of Alpine, you haven't shown the commands you have run (both working and not working) with their output, etc...
2023-01-02 15:52:37 <minimal> EF^: how did you upgrade exactly? (i.e. via which commands)
2023-01-02 15:52:44 <EF^> minimal: checked and yes, lvm, raid and cryptsetup all present:
2023-01-02 15:52:57 <EF^> minimal: apk upgrade -Ua
2023-01-02 15:53:58 <minimal> EF^: ok, just wanted to make sure the "apk upgrade" included the "-a" option, people often miss that
2023-01-02 15:54:40 <ikke> Not sure if it's related, but the issue that EF^ showed mentioned something about openssl1.1
2023-01-02 15:55:12 <ikke> https://github.com/openssl/openssl/blob/master/include/openssl/whrlpool.h seems to indicate it has been deprecated
2023-01-02 15:55:27 <elly> openssl isn't used in luks though surely?
2023-01-02 15:55:32 <EF^> The big change is the openssl version.
2023-01-02 15:56:26 <morena> minimal: 3.17 "nslookup irc.irc-nerds.net" - no answer. something like kdig (knot-utils) or unbound-host (unbound) resolve it without issue
2023-01-02 15:56:41 <elly> morena: does `nslookup -debug irc.irc-nerds.net` shed any light?
2023-01-02 15:57:42 <morena> no, it just shows me "Non-authoritative answer:" it's empty two times
2023-01-02 15:58:27 <minimal> morena: and what (type of servers) is your resolv.conf pointing to?
2023-01-02 15:58:50 <elly> hm, I have a digitalocean machine, let me check
2023-01-02 15:58:55 <morena> minimal: it is digital ocean what they add me, however nothing change if I change it to 8.8.8.8
2023-01-02 15:59:18 <elly> ya, it resolves without difficulty from my digitalocean host as well
2023-01-02 15:59:28 <elly> using their nameservers
2023-01-02 15:59:32 <morena> hmmm
2023-01-02 15:59:46 <morena> it's my second server the same shit ;/ no idea
2023-01-02 16:00:04 <psykose_> works for me too
2023-01-02 16:00:17 <minimal> I do notice when I did the nslookup on a Debian machine I saw ";; Truncated, retrying in TCP mode." at the top of the nslookup
2023-01-02 16:00:26 <minimal> musl only does UDP lookups...
2023-01-02 16:00:26 <elly> yeah, I see that also
2023-01-02 16:00:30 <elly> oh *really*
2023-01-02 16:00:39 <elly> because this domain has too many records for one UDP reply, it looks like
2023-01-02 16:00:49 <elly> I too am testing on a debian machine
2023-01-02 16:00:50 <minimal> there is work-in-progress to add TCP support, it'll be in the next release AFAIK
2023-01-02 16:01:03 <morena> so that can cause the issue?
2023-01-02 16:01:09 <morena> but  your machine works ;/
2023-01-02 16:01:17 <elly> my digitalocean machine is debian though
2023-01-02 16:01:26 <morena> oh
2023-01-02 16:01:29 <elly> and is doing a tcp lookup
2023-01-02 16:01:36 <morena> other os works well there
2023-01-02 16:01:39 <morena> just alpine 
2023-01-02 16:01:47 <minimal> yes, debian is switching from udp to tcp
2023-01-02 16:02:07 <minimal> morena: as I said, I believe that musl currently cannot do TCP dns lookups
2023-01-02 16:02:29 <morena> I thought something like that
2023-01-02 16:02:47 <morena> so bc. answer udp can't handle that or something like that?
2023-01-02 16:03:09 <elly> hmm, `dig +notcp irc.irc-nerds.net` still works just fine with DO's nameservers on debian though
2023-01-02 16:03:22 <elly> it gives back a (correctly truncated) UDP reply
2023-01-02 16:03:36 <morena> other tools works for me too
2023-01-02 16:03:45 <morena> but let's say irc client does not use these tools
2023-01-02 16:03:55 <morena> kdig unbound-host ... these works
2023-01-02 16:04:23 <morena> but packages don't touch these tools for resolving host
2023-01-02 16:04:45 <morena> even running unbound did not change a shit
2023-01-02 16:05:30 <elly> it sounds like you are about to add a few /etc/hosts entries and a big reminder to yourself to delete them later :P
2023-01-02 16:05:38 <minimal> morena: "apk add bind-tools"
2023-01-02 16:05:59 <minimal> that installs the "full" version of nslookup rather than using Busybox's version
2023-01-02 16:06:28 <elly> but if an irc client is just calling gethostbyname() or whatnot that won't help them, will it?
2023-01-02 16:06:29 <minimal> that works for me on a Alpine machine which showed the same problem beforehand
2023-01-02 16:06:44 <morena> thanks, yes, that fixed that crap ;/
2023-01-02 16:06:59 <minimal> so it is a Busybox nslookup limitation
2023-01-02 16:07:08 <morena> still weird
2023-01-02 16:07:19 <minimal> why? its a Busybox limitation
2023-01-02 16:07:44 <morena> that's pretty huge limitation to resolve domains
2023-01-02 16:07:49 <EF^> Anyone any ideas on how I can get a bootable box back?
2023-01-02 16:08:12 <minimal> morena: only for lookups that return a large number of replies
2023-01-02 16:08:34 <minimal> so its more of a "niche" limitation
2023-01-02 16:13:07 <minimal> EF^: you didn't have any disk space problems during the upgrade? (assuming you have a separate /boot partition)
2023-01-02 16:15:00 <EF^> boot has 9ù usage and is a seperate RAID1 array. 
2023-01-02 16:15:45 <minimal> EF^: BIOS-based system then?
2023-01-02 16:16:26 <EF^> yes with LUKS1 book etc
2023-01-02 16:16:37 <EF^> */boot
2023-01-02 16:16:51 <scorpion2185[m]1> can I make doas don't ask for root password?
2023-01-02 16:17:06 <scorpion2185[m]1> so when I am root i don't have to type it
2023-01-02 16:17:28 <morena> nopass 
2023-01-02 16:17:41 <minimal> scorpion2185[m]1: read the docs?
2023-01-02 16:17:44 <scorpion2185[m]1> psykose: nice
2023-01-02 16:18:15 <scorpion2185[m]1> permit nopass root
2023-01-02 16:19:44 <scorpion2185[m]1> that didn't work... aftet you edit /etc/doas.d/doas.conf no need to do anything?
2023-01-02 16:20:11 <ikke> no, should work immidiately
2023-01-02 16:20:29 <morena> no root, but user you use
2023-01-02 16:20:30 <minimal> morena: I think that the nslookup tool in bind-tools is compiled to use bind's own DNS libraries (which have TCP support), so bypassing musl's udp-only DNS support
2023-01-02 16:23:43 <scorpion2185[m]1> morena: I asked about root
2023-01-02 16:24:05 <morena> scorpion2185[m]1: as root you use doas?
2023-01-02 16:25:42 <morena> however it works for root too
2023-01-02 16:26:42 <scorpion2185[m]1> yes not  on purpose somebody put it in scripts how annoyin
2023-01-02 16:27:00 <morena> minimal: interesting, thanks for explanation
2023-01-02 16:27:05 <scorpion2185[m]1> it still asks it
2023-01-02 16:28:09 <morena> scorpion2185[m]1: "permit nopass root" then it does not ask for pass if root use doas
2023-01-02 16:30:01 <minimal> morena: more info: https://gitlab.alpinelinux.org/alpine/tsc/-/issues/43 and https://git.musl-libc.org/cgit/musl/commit/?id=51d4669fb97782f6a66606da852b5afd49a08001
2023-01-02 16:30:16 <scorpion2185[m]1> doas -C /etc/doas.d/doas.conf -u root cat
2023-01-02 16:30:16 <scorpion2185[m]1> permit nopass
2023-01-02 16:32:05 <morena> so musl issue, not busybox then?
2023-01-02 16:32:55 <minimal> busybox uses only musl for DNS
2023-01-02 16:33:08 <scorpion2185[m]1> I can use as a normal user doas without password but it still asks it for root
2023-01-02 16:33:13 <scorpion2185[m]1> *doas
2023-01-02 16:35:56 <morena> bc. that command
2023-01-02 16:35:57 <scorpion2185[m]1> does it work for you for root morena ?
2023-01-02 16:36:12 <morena> not that command, bc. with -C you are checing config there doas apply
2023-01-02 16:36:19 <morena> not for -u or something like that
2023-01-02 16:36:33 <morena> if you "doas -u root whatever" then it will work
2023-01-02 16:36:53 <morena> which you don't need bc. if you don't mention  user it is auto root
2023-01-02 16:38:05 <scorpion2185[m]1> but it asks the password for root, that command was to check root permission
2023-01-02 16:40:07 <scorpion2185[m]1> does it work for you? (As root)
2023-01-02 16:40:16 <morena> yes
2023-01-02 16:40:22 <morena> if I run it as root
2023-01-02 16:40:50 <morena> otherwise regular user does not have permission to check doas.conf
2023-01-02 16:41:51 <morena> so not sure if correct but logically like " doas -u root doas -C /etc/doas.d/doas.conf -u root shit
2023-01-02 16:58:02 <morena> hmmm elly you was right, it does not work in irc client ;/
2023-01-02 17:05:48 <morena> but, Openwrt also uses musl no? and from router with it I can resolve it
2023-01-02 17:06:35 <Ermine> It does
2023-01-02 17:07:10 <minimal> morena: resolve it how?
2023-01-02 17:07:18 <morena> with nslookup
2023-01-02 17:07:52 <minimal> right, and that is likely using bind-tools nslookup which uses bind's dns libraries
2023-01-02 17:09:06 <morena> no it uses busybox
2023-01-02 17:09:33 <minimal> yeah I just checked an openwrt box here and saw that
2023-01-02 17:12:58 <morena> oh but it uses musl 1.2.7
2023-01-02 17:13:07 <morena> so probably with that patch already?
2023-01-02 17:14:20 <minimal> morena: nope, looks like openwrt have their own patch for nslookup to NOT use musl's dns stuff but to use resolv library instead
2023-01-02 17:14:45 <minimal> https://github.com/openwrt/openwrt/blob/openwrt-21.02/package/utils/busybox/patches/230-add_nslookup_lede.patch
2023-01-02 17:14:46 <morena> oh
2023-01-02 17:15:20 <morena> so now everything makes sense
2023-01-02 17:15:30 <minimal> effectively they've replaced busybox's nslookup with their own modified version
2023-01-02 17:15:37 <morena> even it does not work ;/
2023-01-02 17:15:57 <morena> yes it would be weird if router can't resolve something
2023-01-02 17:16:25 <minimal> morena: what do you mean it does not work? you earlier said it DID work from openwrt
2023-01-02 17:18:36 <morena> minimal: not from alpine as irc client does not use bind tools for lookup
2023-01-02 17:18:56 <minimal> we were talking about openwrt though
2023-01-02 17:19:01 <morena> oh openwrt works
2023-01-02 17:19:17 <morena> intersting nnn file manager installed musl-fts ;/
2023-01-02 17:19:39 <minimal> yes anything on musl that is using the musl resolver will have the UDP-only limitation for now
2023-01-02 17:20:43 <vkrishn> musl 1.2.7 - is released ?
2023-01-02 17:20:54 <morena> no no
2023-01-02 17:21:06 <morena> it is that fts
2023-01-02 17:21:10 <ikke> Latest release: musl-1.2.3.tar.gz (sig)
2023-01-02 17:21:37 <morena> minimal: so that commit what you sent will be implement in musl in some future version?
2023-01-02 17:21:56 <Ermine> 1.2.4 is going to have dns over tcp afaik
2023-01-02 17:22:18 <minimal> well that was a commit in the musl git repo, so the next release then will have it
2023-01-02 17:22:28 <vkrishn> there is also tool 'drill'
2023-01-02 17:23:04 <morena> minimal: until then what? should I use windows? ;/
2023-01-02 17:23:11 <scorpion2185[m]1> morena:  I mean when you are root running `doas <cmd>` e.g. `doas apk update`
2023-01-02 17:23:23 <minimal> vkrishn: the issue is not to do with tools, it is to do with the system-wide resolver used by various programs
2023-01-02 17:23:37 <morena> scorpion2185[m]1: why would root run doas?
2023-01-02 17:23:56 <morena> but if  yo uput permit nopass root it should work without pass
2023-01-02 17:24:02 <scorpion2185[m]1> no
2023-01-02 17:24:22 <morena> yes
2023-01-02 17:24:24 <scorpion2185[m]1> not here at least 
2023-01-02 17:25:02 <morena> here it works
2023-01-02 17:25:27 <morena> "permit nopass root" it means root is permited to do anything without asking password with doas
2023-01-02 17:25:34 <morena> even root is permited anyway
2023-01-02 17:25:56 <morena> but whatever user you can insead of root there, it will not ask pass
2023-01-02 17:26:13 <morena> you maybe have other things in doas.conf?
2023-01-02 17:26:28 <morena> bc. it depends what is first line, what second ...
2023-01-02 17:26:36 <Ermine> scorpion2185[m]1: again, what are you trying to achieve?
2023-01-02 17:26:38 <morena> it goas that way
2023-01-02 17:27:04 <scorpion2185[m]1> no it was empty, but I sxmo.conf near doas.conf I'll check that
2023-01-02 17:27:24 <scorpion2185[m]1> * i have, bbl
2023-01-02 17:27:47 <morena> ha?
2023-01-02 17:29:53 <minimal> morena: well I guess you're stuck unless someone decides to backport the upstream musl changes (there's more than 1 related to TCP support AFAIK)
2023-01-02 17:30:22 <Ermine> morena: you also may want to try out dnsfunnel
2023-01-02 17:30:44 <minimal> even then a backport would only likely be in Alpine Edge, not 3.17
2023-01-02 17:31:37 <minimal> Ermine: yeah I had that at the back of my mind
2023-01-02 17:32:17 <minimal> not sure how it would handle that particular lookup
2023-01-02 17:32:52 <EF^> Okay, tried to rebuild the initramfs. No change
2023-01-02 17:33:07 <EF^> It is possible to downgrade?
2023-01-02 17:34:28 <Ermine> minimal: it uses libs6dns, so chances are that it will handle it differently. One can check this with tools from s6-dns package.
2023-01-02 18:15:01 <psykose_> you can also install bind-tools and nslookup from there also works
2023-01-02 18:19:25 <morena> psykose_: yes it works with bind-tools, but not for irc client for example
2023-01-02 18:28:20 <scorpion2185[m]1> sxmo.conf was interfering thanks
2023-01-02 19:23:54 <scorpion2185[m]1> how do you convert slurp into the option for wayshot -s ?
2023-01-02 19:26:05 <EF^> psykose_: fyi, your suggestion on gitlab to add legacy.so into the initramfs via a .files worked. I'll add some commentary tomorrow into issue 13004. Thank you.
2023-01-02 19:26:07 <scorpion2185[m]1> how do you specify geometry on wayland?
2023-01-02 19:33:51 <scorpion2185[m]1> wayshot -s "$(slurp)"
2023-01-02 20:26:26 <psykose_> https://img.ayaya.dev/rkNtXL2wtkEh
2023-01-02 21:33:33 <pj[m]> just saying, but reading doesn't hurt
2023-01-03 03:28:55 <uuidNuni1> is DNS on musl still a thing? i thought it was fixed already 
2023-01-03 03:29:52 <uuidNuni1> i never had DNS issues on alpine, i don't know in what cases it does not work
2023-01-03 03:31:46 <morena> uuidNuni1: in case I mentioned 
2023-01-03 03:32:25 <uuidNuni1> morena: sorry i am on phone, can you describe it again?
2023-01-03 03:32:43 <morena> uuidNuni1: try nslookup irc.irc-nerds.net
2023-01-03 03:33:22 <morena> uuidNuni1: they explained me, something like, too many IP in answers, udp can't handle it, need tcp, but musl does not do that
2023-01-03 03:33:34 <uuidNuni1> morena: i got a bunch of addrs 
2023-01-03 03:33:45 <morena> uuidNuni1: using busybox nslookup?
2023-01-03 03:34:03 <uuidNuni1> i think it from bind
2023-01-03 03:34:21 <morena> yes that works, but for example if you spin irc client, it will not
2023-01-03 03:34:27 <morena> https://git.musl-libc.org/cgit/musl/commit/?id=51d4669fb97782f6a66606da852b5afd49a08001
2023-01-03 03:34:30 <uuidNuni1> busybox nslookup returns nothing
2023-01-03 03:34:53 <morena> just copied what they sent me, fixed alredy but not yet in musl release
2023-01-03 03:35:04 <morena> uuidNuni1: yes, that's the issue
2023-01-03 03:35:17 <uuidNuni1> is that a busybox issue?
2023-01-03 03:35:28 <morena> musl I was told
2023-01-03 03:36:19 <morena> but tools like bind, ubnound-host kdig and similar will work
2023-01-03 03:36:27 <morena> but programs don't use them
2023-01-03 03:36:32 <uuidNuni1> nowerdays i use dig most of the time
2023-01-03 03:36:42 <morena> yes, but your irc client for example will not
2023-01-03 03:37:17 <uuidNuni1> so the solution is to install busybox alternatives
2023-01-03 03:37:50 <morena> I assuem the solution is to wait until musl is released with patch mentioned if one don't want to patch it
2023-01-03 03:38:42 <uuidNuni1> morena: you mean your irc client don't use the nslookup from bind?
2023-01-03 03:38:59 <morena> btw. I may not be correct, it's just how I understand it, but I don't give a shit much to be honest
2023-01-03 03:39:04 <morena> uuidNuni1: your  use?
2023-01-03 03:39:45 <morena> I think it don't use those tools at all
2023-01-03 03:40:03 <uuidNuni1> nslookup is a symlink to busybox by default, once you installed bind-tools, the symlink will be point to bind-tools's nslookup
2023-01-03 03:40:38 <uuidNuni1> unless your client hardcoded the command to "busybox nslookup"
2023-01-03 03:41:18 <psykose_> software doesn't call `nslookup`
2023-01-03 03:41:38 <uuidNuni1> ohh they call the api?
2023-01-03 03:42:12 <uuidNuni1> morena: what irc client are you talking about?
2023-01-03 03:42:23 <morena> I assume any will do the same
2023-01-03 03:42:53 <uuidNuni1> i use weechat on alpine. works fine  
2023-01-03 03:43:06 <uuidNuni1> and yes i have bind-tools installed too
2023-01-03 03:43:16 <morena> I think it calls some function in musl for dns
2023-01-03 03:43:41 <morena> and weechat did not bitch about dns?
2023-01-03 03:44:11 <uuidNuni1> weechat works fine. i have been using it all the time on alpine and other distroes
2023-01-03 03:44:23 <morena> with domain I mentioned?
2023-01-03 03:44:47 <uuidNuni1> well, i am on OFTC and librechat 
2023-01-03 03:45:01 <uuidNuni1> not using that server you memtioned 
2023-01-03 03:45:09 <morena> you can try connect if you run alpine now
2023-01-03 03:45:22 <uuidNuni1> ok
2023-01-03 03:46:07 <uuidNuni1>  /connect irc.irc-nerds.net, right?
2023-01-03 03:46:13 <uuidNuni1> let me try
2023-01-03 03:46:16 <morena> something like that I assume
2023-01-03 03:47:43 <psykose_> the specific issue is in cloudflare being a pile of garbage, but tcp fallback also fixes it, so meh
2023-01-03 03:48:14 <morena> cloudflare everywhere!
2023-01-03 03:49:14 <morena> psykose_: so they sent some "non standard" answer?
2023-01-03 03:50:06 <uuidNuni1> morena: weechat can't resolve irc.irc-ners.net 
2023-01-03 03:50:11 <uuidNuni1> you're right  
2023-01-03 03:50:43 <morena> ladies told me here everything, I had no clue ;/
2023-01-03 03:52:44 <uuidNuni1> holy crap
2023-01-03 03:53:22 <morena> yes
2023-01-03 03:57:38 <psykose_> they return noerror tc with an empty response
2023-01-03 03:58:10 <psykose_> for practically every client tc will retry with tcp, musl latest git included, but it doesn't make sense
2023-01-03 03:58:18 <psykose_> iirc they were supposed to have fixed that
2023-01-03 03:58:25 <morena> so question is, why people use it
2023-01-03 03:58:25 <psykose_> it makes no sense to return an empty record on truncation
2023-01-03 03:58:28 <uuidNuni1> what if you setup a local forward dns server with dnsmasq? 
2023-01-03 03:58:45 <psykose_> dnsmasq works since dnsmasq will just be using tcp and your queries will go to that
2023-01-03 03:59:04 <psykose_> just set 127.0.0.1 in resolv.conf and a usual dnsmasq setup and that domain will work
2023-01-03 03:59:08 <psykose_> just a workaround tho
2023-01-03 03:59:31 <uuidNuni1> yes, that's what i meant
2023-01-03 03:59:48 <uuidNuni1> also, you can setup a doh forward server with dnsproxy 
2023-01-03 04:00:13 <uuidNuni1> listen on localhost, forward to dns over https server
2023-01-03 04:01:07 <uuidNuni1> dnsproxy is easier than dnsmasq to be honest, just one line command, not configuration needed 
2023-01-03 04:02:51 <uuidNuni1> dnsproxy -l 127.0.0.1 -p 53 -u 1.1.1.1
2023-01-03 04:03:22 <uuidNuni1> or -u https://1.1.1.1/dns-query for doh
2023-01-03 04:03:40 <morena> obviously cloudflare again, why not yeah?
2023-01-03 04:04:02 <uuidNuni1> s/1.1.1.1/8.8.8.8/
2023-01-03 04:04:02 <alpine-meetbot> uuidNuni1 meant to say: or -u https://8.8.8.8/dns-query for doh
2023-01-03 04:04:12 <uuidNuni1> or your isp dns
2023-01-03 04:04:23 <morena> yes, understand
2023-01-03 04:04:56 <uuidNuni1> it's written in Go, so runs everywhere 
2023-01-03 04:06:12 <morena> well I did it different way, as I am not much into hacking system which can't resolve fucking domain ;/
2023-01-03 04:06:41 <uuidNuni1> lol  what did you do?
2023-01-03 04:07:57 <morena> destroyed droplet, spin up different OS ;/
2023-01-03 04:08:41 <uuidNuni1> doh is my favorate nowerdays , in my case, it is much faster than udp. because http2 has native multiplex and of cause https does not get noise like udp
2023-01-03 04:09:17 <uuidNuni1> i almost use DOH everywhere
2023-01-03 04:10:56 <morena> well localy I resolve dns recursively without middle man
2023-01-03 04:10:58 <uuidNuni1> ok, i think i may need to setup one on my alpine server. since the issue you mentioned today
2023-01-03 04:11:47 <uuidNuni1> morena: direct to the root server?
2023-01-03 04:11:54 <morena> yes
2023-01-03 04:12:09 <uuidNuni1> huge data to download i guess?
2023-01-03 04:16:40 <morena> other OS can handle it, not sure now about alpine/musl ;/
2023-01-03 04:18:09 <uuidNuni1> the issue is truncated reply on udp
2023-01-03 04:18:36 <uuidNuni1> as long as a domain does not reply too many addr, it is fine?
2023-01-03 04:19:08 <morena> or as long as your system is not alpine/musl ;/
2023-01-03 04:20:29 <uuidNuni1> yeah, that's indeed an issue. also i have encounter an isuse that if ipv6 is disabled on dns forwarder, musl can't resolve
2023-01-03 04:20:58 <uuidNuni1> lol, alpine can't even install :p
2023-01-03 04:21:36 <morena> easily fixable as I mentioned ;/
2023-01-03 04:22:09 <uuidNuni1> but i really like the simplicity and small footprint
2023-01-03 04:22:31 <morena> yes
2023-01-03 04:22:34 <uuidNuni1> let's wait for musl to fix this
2023-01-03 04:23:47 <uuidNuni1> i have some C program built against musl statically
2023-01-03 04:23:50 <morena> yes, 2023 the year of linux on desktop ;/
2023-01-03 04:24:11 <uuidNuni1> i don't know if that will broke
2023-01-03 04:24:36 <morena> let's wait for linux to fix this ;/
2023-01-03 04:25:34 <uuidNuni1> wait, there are so many musl statically linked programs over there 
2023-01-03 04:25:58 <uuidNuni1> why don't they have such dns issue?
2023-01-03 04:26:36 <uuidNuni1> maybe because they don't use musl's dns lib?
2023-01-03 04:28:25 <morena> if they don't have the issue I assume they do like bind unbound-host or whatever do job themself
2023-01-03 04:31:17 <uuidNuni1> https://news.ycombinator.com/item?id=28312935
2023-01-03 04:32:06 <morena> so they removing this feature now ;/
2023-01-03 04:32:28 <psykose_> removing what feature
2023-01-03 04:34:28 <uuidNuni1> https://git.musl-libc.org/cgit/musl/commit/?id=51d4669fb97782f6a66606da852b5afd49a08001
2023-01-03 04:34:34 <uuidNuni1> this commit? 
2023-01-03 04:36:41 <psykose_> yep
2023-01-03 04:38:02 <uuidNuni1> why don't alpine edge use master musl?
2023-01-03 04:38:31 <Danct12> too master
2023-01-03 04:38:32 <morena> I go sleep, I m tired of linux ;/
2023-01-03 04:38:51 <uuidNuni1> when a new musl released, alpine needs to rebuild every program in repo?
2023-01-03 04:38:56 <psykose_> no
2023-01-03 04:40:02 <Danct12> no, application built on musl or glibc even doesn't need to be rebuilt, unless there are some very big change that will break api
2023-01-03 04:43:01 <uuidNuni1> oh, i see, they are dynamiclly linked 
2023-01-03 04:44:19 <uuidNuni1> so, if i build a musl from master branch right now, that will fix the dns issue mentioned here? 
2023-01-03 04:44:37 <psykose_> sure
2023-01-03 04:45:31 <uuidNuni1> cool. let's go 
2023-01-03 04:47:40 <uuidNuni1> is malloc performance still a thing on musl?
2023-01-03 04:48:10 <uuidNuni1> people used to say musl is lower performant than glibc
2023-01-03 04:50:06 <psykose_> (didn't test if it did actually)
2023-01-03 04:50:24 <Danct12> yup, as long as it's dynamically linked, you can run the program with the latest version of glibc, that is if there aren't any api change that break application compatibility
2023-01-03 04:50:49 <Danct12> otherwise you may have to rebuild your program
2023-01-03 04:51:57 <uuidNuni1> or, if alpine cherry pick that commit into 1.2.3, is it doable?
2023-01-03 04:52:51 <uuidNuni1> i mean, do alpine devs want to do that? it's a work around before new musl release 
2023-01-03 04:53:56 <uuidNuni1> just one commit, right? it does not hurt 
2023-01-03 04:54:19 <uuidNuni1> have the devs thought about that?
2023-01-03 05:01:08 <paperowl> Uuidnuni1 from what I understand, musl will not be able to match glibc in terms of raw multithreaded allocs , because it does not and will not use a memory pool per thread. Rationale seems to be to keep musl overhead low for embedded systems . Correct me if im wrong
2023-01-03 05:01:54 <paperowl> But depends heavily on how and how much your program allocates
2023-01-03 05:02:15 <paperowl> Many people said no difference for them. You have to benchmark yourself
2023-01-03 05:03:25 <uuidNuni1> most of the time memory usage is less than 20% on my server :P
2023-01-03 05:03:55 <uuidNuni1>  i can't think of a way to make it 100% usage for my needs 
2023-01-03 05:05:26 <paperowl> Its talking to How it allocates, not the total allocation I believe 
2023-01-03 05:06:14 <paperowl> Glibc I think does the memory pool so when you need the alloc, its already available ahead of time. Musl has to do the alloc on demand 
2023-01-03 05:06:29 <paperowl> I think thats how it works, I just learned it
2023-01-03 05:08:52 <uuidNuni1> is that the case on mallocng? 
2023-01-03 05:08:57 <uuidNuni1> since 1.2.1
2023-01-03 05:09:36 <paperowl> But if you bench your program and see the alloc is the bottleneck .its trivial to use a different alloc for your program and not musls 
2023-01-03 05:09:50 <paperowl> I have no idea on that one 
2023-01-03 05:10:22 <uuidNuni1> LD_PRELOAD?
2023-01-03 05:10:35 <paperowl> Bottom line it seems musl brings some lightweight and basic functionality and if you want a better DNS resolver or malloc you bring your own 
2023-01-03 05:10:43 <uuidNuni1> i don't find a system wide conf to use another malloc
2023-01-03 05:10:56 <paperowl> Musl cant accommodate everybody since it is forced to be lightweight 
2023-01-03 05:12:07 <paperowl> You can choose your own if you are developing. If using other program im not sure yet how to change sys alloc
2023-01-03 05:12:57 <uuidNuni1> found it :There isn’t one. You can specify the library search path by creating or editing the file /etc/ld-musl-$ARCH.path, where $ARCH is the string identifying your architecture (look at what /lib/ld-musl-*.so.1 uses if uncertain). Paths can be separated by newlines or colons. The environment variables LD_PRELOAD and LD_LIBRARY_PATH are also supported for one-off cases.
2023-01-03 05:13:38 <uuidNuni1> so i can install mimalloc2 and set LD_PRELOAD to use it  
2023-01-03 05:15:34 <paperowl> Oh cool 
2023-01-03 05:17:21 <paperowl> Then you just set the var in your profile?
2023-01-03 05:17:31 <paperowl> Try it and let me know if your computer blows up
2023-01-03 05:17:38 <paperowl> For science 
2023-01-03 05:23:43 <uuidNuni1> yeah. l'll do some benchmark which i will never encounter
2023-01-03 05:25:32 <uuidNuni1> glibc has it's own issue too. if you want your program work everywhere, build it on a very old system like centos5 :p
2023-01-03 05:26:02 <uuidNuni1> if built on archlinux, then it won't work on debian or rhel
2023-01-03 05:26:04 <paperowl> the main concern is security right, its second to systemd in terms of linux bloat ?
2023-01-03 05:26:34 <paperowl> ID rather take simple and transparent than bloated and performant
2023-01-03 05:26:51 <uuidNuni1> me 2
2023-01-03 05:26:57 <paperowl> Are you using alpine as your daily driver BTW?
2023-01-03 05:27:09 <paperowl> I just came over to alpine a few days ago from debian
2023-01-03 05:27:33 <uuidNuni1> no. i don't use linux as daily driver, only server, headless
2023-01-03 05:27:47 <paperowl> Alpine teaches me a ton of stuff I didnt know using a bigger distro
2023-01-03 05:28:19 <uuidNuni1> well i learnt my knowledge from gentoo and arch. so alpine is familar to me
2023-01-03 05:29:02 <paperowl> Yea I think the glibc might be the biggest dealbreaker for most daily use 
2023-01-03 05:29:28 <paperowl> But I just use whatevers available . Not picky . Plenty of apks 
2023-01-03 05:30:18 <uuidNuni1> a fully working alpine is just around 200M
2023-01-03 05:30:35 <uuidNuni1> where others at least 1000M
2023-01-03 05:31:33 <paperowl> Im at 400 after installing and downloading some extra stuff
2023-01-03 05:31:52 <uuidNuni1> you're doing desktop? 
2023-01-03 05:31:59 <uuidNuni1> well, i do headless only 
2023-01-03 05:32:00 <paperowl> The memory usage is amazing tho. Debian would crash all the time on 8GB oom on firefox
2023-01-03 05:32:04 <Danct12> uuidNuni1, i can confirm this mostly
2023-01-03 05:32:06 <paperowl> Yes desktop
2023-01-03 05:32:17 <Danct12> if i build something on arch linux, then my binary just doesn't work on ubuntu :D
2023-01-03 05:32:22 <paperowl> Well laptop 
2023-01-03 05:32:38 <uuidNuni1> Danct12: glibc compatible issue
2023-01-03 05:32:44 <Danct12> yup
2023-01-03 05:32:49 <paperowl> Thats because glibc doesn't do static linking right
2023-01-03 05:33:47 <Danct12> my infrastructure is all built on way too modern and cutting edge stuff anyway
2023-01-03 05:33:59 <uuidNuni1> paperowl: musl does not do statically linking either
2023-01-03 05:34:09 <Danct12> so arch linux, debian sid, alpine edge or else
2023-01-03 05:34:39 <uuidNuni1> rhat's because glibc on debian is too old to handle the compatiblity
2023-01-03 05:36:47 <Danct12> yeah, it's 2.35 on ubuntu 22.04, while i'm using 2.36 on arch lol
2023-01-03 05:36:55 <paperowl> I believe musl did static linking, I think I just read online 
2023-01-03 05:37:05 <paperowl> Ill have to double check 
2023-01-03 05:37:23 <uuidNuni1> ldd vim on arch says libc.so.6 but that libc does not exist on debian that's the issue
2023-01-03 05:37:39 <uuidNuni1> glibc compatible issue
2023-01-03 05:38:11 <uuidNuni1> paperowl: you can, but it does dynamically linking by default on alpine
2023-01-03 05:38:59 <paperowl> Ah I see. Thanks
2023-01-03 05:39:17 <paperowl> Just started using linux 6 months ago.big newbie
2023-01-03 05:39:23 <uuidNuni1> although i did not know if a binary on alpine 3.17 works on alpine 3.11 or not
2023-01-03 05:40:18 <uuidNuni1> paperowl: ldd /usr/bin/some_command tells you 
2023-01-03 05:41:44 <uuidNuni1> if it is statically linked it says: statically linked
2023-01-03 05:41:59 <uuidNuni1> if not, it gives you the linked libs
2023-01-03 05:42:11 <paperowl> Appreciate the lesson
2023-01-03 05:42:18 <uuidNuni1> file /usr/bin/some_command works too
2023-01-03 05:42:31 <paperowl> very handy
2023-01-03 05:42:42 <Danct12> yup i just compiled a test program and seems like it works on ubuntu even when compiled on arch
2023-01-03 05:42:56 <Danct12> maybe that's because my program is waaaaaaaay too simple for any api breakage to come lol
2023-01-03 05:43:00 <paperowl> I had some program I ported over from debian that didnt work. Took me hours to figure out it was dynamic failure
2023-01-03 05:43:29 <uuidNuni1> paperowl: rebuild on alpine or install glibc on alpine
2023-01-03 05:43:35 <uuidNuni1> apk search glibc
2023-01-03 05:44:06 <paperowl> I plan to rebuild it someday. Its the bookmark function on text browser w3m lol
2023-01-03 05:44:10 <paperowl> Nothing too important 
2023-01-03 05:44:57 <uuidNuni1> apk search gcompat
2023-01-03 05:45:11 <paperowl> Btw I had some wierd issue with pseudoterminals yesterday. Someone told me I had way too many pts in /Dev/pts
2023-01-03 05:45:36 <paperowl> Like 240 or so, a bunch are spawned by xorg ,I3,tmux
2023-01-03 05:46:01 <paperowl> I wasnt quite sure how it was a problem..
2023-01-03 05:46:17 <paperowl> Like I have no control over those apps
2023-01-03 05:47:10 <uuidNuni1> paperowl: raise your nfile?
2023-01-03 05:47:11 <Danct12> gcompat is nice but it's not really a solution :)
2023-01-03 05:47:45 <Danct12> for example i had a few application that i compiled on glibc that just straight up segfault on alpine
2023-01-03 05:48:52 <paperowl> I could do that , just wasnt sure if 240 pseudoterminals meant anything 
2023-01-03 05:49:20 <uuidNuni1> paperowl: add this line to /etc/rc.conf : rc_ulimit="-n 1000000 -u unlimited"
2023-01-03 05:49:30 <uuidNuni1> and reboot?
2023-01-03 05:49:33 <uuidNuni1> i am not sure 
2023-01-03 05:49:50 <paperowl> OK will try thank you!
2023-01-03 05:50:00 <uuidNuni1> maybe a bad idea or not the right solution
2023-01-03 05:50:09 <uuidNuni1> if it has no effect remove 
2023-01-03 05:50:25 <uuidNuni1> that is used to raise nfile to 1million
2023-01-03 05:50:36 <uuidNuni1> for all programs
2023-01-03 05:51:45 <paperowl> 1 million is a nice number :p
2023-01-03 05:52:22 <paperowl> do you think big companies use alpine? Or just the usual Ubuntu rhel
2023-01-03 05:53:00 <uuidNuni1> 240 pts, and each pts has X file open, then, it is big number 
2023-01-03 05:53:09 <uuidNuni1> beyound default nfile which is 1024
2023-01-03 05:53:55 <uuidNuni1> paperowl: docker/podman/CI uses alpine a lot
2023-01-03 05:54:18 <uuidNuni1> bare metal server or vps? i don't think so  
2023-01-03 05:55:53 <paperowl> Bare, its my daily driver . Tiny lenovo laptop
2023-01-03 05:56:37 <paperowl> I did the standard alpine install, and all it has is a bare minimal windows manager tmux and xwindows .nothing fancy 
2023-01-03 05:56:41 <uuidNuni1> yeah  i mean big guys not likely to use alpine 
2023-01-03 05:57:19 <paperowl> They would have to...gasps....support their own systems 
2023-01-03 05:57:23 <uuidNuni1> big guys wants stability they don't want new features once installed
2023-01-03 05:57:43 <uuidNuni1> they like it up forever without rebooting
2023-01-03 05:58:34 <paperowl> I guess taking a node offline for a few minute eats into quarterly profits
2023-01-03 05:58:38 <uuidNuni1> but they create vms or containers to build stuff which uses alpine alot
2023-01-03 05:59:12 <paperowl> reason I asked is wonder if any alpine got breached before in commercial use 
2023-01-03 05:59:27 <paperowl> Aka alpine security track record in business
2023-01-03 05:59:45 <uuidNuni1> docker :P
2023-01-03 05:59:57 <uuidNuni1> that's what i can think of
2023-01-03 06:00:26 <uuidNuni1> i use alpine on my VPS anyway  
2023-01-03 06:00:48 <uuidNuni1> i got only a personal web server running. simple
2023-01-03 06:00:59 <uuidNuni1> only 22 80 443 ppen
2023-01-03 06:01:01 <uuidNuni1> open
2023-01-03 06:01:11 <paperowl> Servers are awesome
2023-01-03 06:01:24 <paperowl> Anyone can become a part of the internet 
2023-01-03 10:15:05 <sicelo> < paperowl> Alpine teaches me a ton of stuff I didnt know using a bigger distro  <= like?
2023-01-03 11:49:04 <jhjacobs[m]> <uuidNuni1> "big guys wants stability they..." <- This bullshit is based on what facts?
2023-01-03 11:50:42 <jhjacobs[m]> Let me rephrase that: How did you come to that conclusion?
2023-01-03 11:54:38 <uuidNuni1> interesting, i'm able to chroot to a minirootfs from a GNU/Linux, and run `ls`,  but i can't run ./newroot/bin/busybox ls (i know it's musl). but why former works? 
2023-01-03 11:55:12 <uuidNuni1> jhjacobs[m]: based on RHEL package version changes.
2023-01-03 11:55:56 <jhjacobs[m]> heh
2023-01-03 11:58:01 <ptrc> uuidNuni1: paths to libraries
2023-01-03 11:58:42 <ptrc> when running as ./newroot/bin/busybox, it doesn't know what `ld-musl-${arch}.so.1` is
2023-01-03 11:59:55 <ptrc> there's a chance it would work if you ran it as `LD_LIBRARY_PATH="$PWD/newroot/lib" ./newroot/bin/busybox`, haven't tried it though
2023-01-03 12:01:36 <grawity> LD_LIBRARY_PATH is read by ld.so, so it can't affect how the path *to* ld.so is found (which is hardcoded in the binary) -- you'll want to run it as "/newroot/lib/ld-musl-foo.so.1 /newroot/bin/busybox" (in addition to the LD_ env for the other libraries)
2023-01-03 12:01:57 <ptrc> exactly that ^
2023-01-03 12:02:03 <ptrc> ( tried it in a container right now )
2023-01-03 12:03:02 <uuidNuni1> but why does chroot work?
2023-01-03 12:03:04 <grawity> it's almost exactly like running a script that has #!/bin/sh -- it will run the "host" /bin/sh, not the one from /newroot
2023-01-03 12:03:31 <grawity> a chroot works because chroot changes the meaning of /, and therefore it changes what file the path "/lib/ld-musl-foo.so" resolves to
2023-01-03 12:03:39 <grawity> I mean, literally it's what chroot does
2023-01-03 12:03:45 <uuidNuni1> the chroot is a glibc binary. i mean
2023-01-03 12:04:06 <grawity> yeah but it runs on the host, it doesn't need to pull any libraries from the guest
2023-01-03 12:04:37 <uuidNuni1> but when the arch is different, e.g. chroot from x86_64 to a aarch64 minirootfs... it won't work, right?
2023-01-03 12:04:50 <grawity> it won't because the kernel won't know how to run an aarch64 binary
2023-01-03 12:04:51 <ikke>  unless you have binfmt setup
2023-01-03 12:05:28 <grawity> not sure if binfmt interpreters are resolved against the current root or the original one
2023-01-03 12:05:39 <grawity> never tried the qemu binfmt thing
2023-01-03 12:05:47 <uuidNuni1> yeah, binfmt works i know. 
2023-01-03 12:06:07 <uuidNuni1> i am just curious about glibc chooot into musl rootfs 
2023-01-03 12:06:26 <uuidNuni1> ok. i am gonna install differently. this is interesting : https://blog.ari-web.xyz/b/how-to-manually-install-alpine-linux-on-any-linux-distribution/
2023-01-03 12:06:53 <grawity> after chroot tells the kernel to remap /newroot to /, it doesn't really load any libraries from within the musl rootfs -- it just straight goes to exec the specified binary
2023-01-03 12:07:28 <grawity> chroot won't crash because anything it needs has already been loaded (from the host) by that point
2023-01-03 12:07:37 <uuidNuni1> grawity: got it. that's wonderful. 
2023-01-03 12:07:49 <grawity> and the new binary will work because the / (and therefore /lib) has been remapped to the right location
2023-01-03 12:26:28 <uuidNuni1> OK: 7 MiB in 15 packages, i really like it
2023-01-03 12:28:35 <uuidNuni1> it would be nice if ash has full documents
2023-01-03 13:42:19 <uuidNuni1> there are runlevels like boot default sysinit, how do i know which one to add to for a service?
2023-01-03 13:52:13 <Shiz> uuidNuni1: it's usually default unless you have a very specific reason not to
2023-01-03 13:52:48 <Shiz> see also https://wiki.alpinelinux.org/wiki/OpenRC#Available_Runlevels
2023-01-03 14:14:24 <EF> Afternoon all
2023-01-03 14:34:02 <EF> There appears to be something very odd with dnsmasq/networking/busybox in Alpine.  Setting bind-interfaces or listing an/some interface/except-interface/address in dnsmasq's config with dhcp enableds stops dns respnding on any interface
2023-01-03 14:41:28 <minimal> EF: insufficient information given to explain the issue
2023-01-03 14:51:35 <EF> I'm digging still but a quick Void install appears to be suffering the same fate.  dig times out even when the server is on the localhost
2023-01-03 15:13:34 <uuidNuni1> there are so many KOPT_ vars, where were they defined? https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in
2023-01-03 15:15:10 <uuidNuni1> EF why not read man dnsmasq
2023-01-03 15:15:28 <ptrc> uuidNuni1: this loop right here: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/351dd25e/initramfs-init.in#L352
2023-01-03 15:15:32 <ptrc> it reads from kernel parameters
2023-01-03 15:16:45 <uuidNuni1> ptrc: i see, thanks a lot
2023-01-03 15:17:31 <uuidNuni1> looks like the option modules= is only needed for moodloop
2023-01-03 15:19:15 <ikke> No, it also loads kernel modules to be able to mount your file system for example
2023-01-03 15:19:56 <uuidNuni1> isn't it automatically? when mount /dev/xyz  / 
2023-01-03 15:19:59 <minimal> uuidNunil: using chroot is a typical way to create an Alpine disk image - that's exactly what I use for my own such script, together with binfmt/qemu to create images for a different arch (i.e aarch64 on a x86_64 machine)
2023-01-03 15:20:46 <ikke> uuidNuni1: it needs to find the mountpoint (root device) you specified, and for that to work the kernel module already needs to be loaded
2023-01-03 15:21:41 <uuidNuni1> ikke: that's the device driver, e.g. usb, nvme, emmc, not stuff like filesystem 
2023-01-03 15:21:55 <ikke> sorry, yes
2023-01-03 15:22:22 <uuidNuni1> usually, i have configured mkinitfs.conf for device driver for usb and hard drive
2023-01-03 15:22:23 <ikke> device drivers, not filesystem drivers
2023-01-03 15:22:54 <minimal> uuidNuni1: Busybox's mount doesn't autoload a filesystem module, the "full fat" mount program does
2023-01-03 15:23:24 <uuidNuni1> well, that, is surprising minimal 
2023-01-03 15:24:36 <minimal> uuidNuni1: well I'm thinking of "mount -t auto" behaviour
2023-01-03 15:25:22 <ikke> there is a list of pre-defined filesystems, and bb mount just tries to mount it one by one until it finds one that works or it gives up
2023-01-03 15:26:32 <uuidNuni1> sudo busybox mount -vvv -f /dev/sda3 /mnt/
2023-01-03 15:26:37 <uuidNuni1> mount: would do mount('/dev/sda3','/mnt/','ext3',0x00008000,'(null)')
2023-01-03 15:26:39 <uuidNuni1> lol
2023-01-03 15:26:50 <uuidNuni1> it is ext4 actually 
2023-01-03 15:26:58 <uuidNuni1> but busybox thinks it is ext3
2023-01-03 15:26:59 <minimal> uuidNuni1l: generally for Alpine you'd have "module=ext4" (or whichever) defined in your bootloader's cmdline list
2023-01-03 15:27:12 <minimal> oops modules=
2023-01-03 15:27:50 <ikke> That's what started this discussion
2023-01-03 15:28:34 <uuidNuni1> sudo busybox mount -vvv -f -t ext4 /dev/sda3 /mnt/ : mount: would do mount('/dev/sda3','/mnt/','ext4',0x00008000,'(null)')
2023-01-03 15:28:48 <uuidNuni1> yeah, i think rootfstype= or modules= is needed 
2023-01-03 15:29:21 <ikke> rootfstype is appended to modules
2023-01-03 15:29:49 <minimal> yeah, I just remembered that, lol
2023-01-03 15:29:57 <ikke> modprobe -a $(echo "$KOPT_modules $rootfstype" | tr ',' ' ' ) loop squashfs simpledrm 2> /dev/null
2023-01-03 15:30:54 <uuidNuni1> so in my case, rootfstype= is enough?
2023-01-03 15:31:39 <minimal> unless your rootfs is on NVME where you're also need "modules=nvme" (from memory)
2023-01-03 15:32:27 <ikke> and if you use lvm, you also need lvm, and on and on
2023-01-03 15:33:09 <uuidNuni1> yeah, i see. i'll add my device drivers to modules= and also add rootfstype=ext4. my device is USB+EMMC, USB is dwc3/dwc2 MMC is dwmmc 
2023-01-03 15:33:40 <uuidNuni1> this way i don't need to create a file in features.d
2023-01-03 15:34:06 <ikke> The module needs to be present in the initramfs, if it isn't, you cannot load it
2023-01-03 15:34:30 <minimal> uuidNuni1: features.d files are used to INCLUDE modules in the initramfs
2023-01-03 15:34:32 <ikke> which is why you have an initramfs in the first place
2023-01-03 15:34:38 <uuidNuni1> well. then, i still need to add modules to mkinitfs.conf 
2023-01-03 15:34:42 <ikke> yes
2023-01-03 15:34:50 <minimal> having "modules=XYZ" will have no effect if XYZ are not part of the initramfs file
2023-01-03 15:34:59 <uuidNuni1> got it 
2023-01-03 15:35:51 <minimal> if you run "mkinitramfs -l" that does not rebuild the initramfs but rather shows you what would be placed in the initramfs, based on the current config, oif it was rebuilt
2023-01-03 15:36:08 <uuidNuni1> yeah
2023-01-03 15:36:16 <uuidNuni1> i did it before, 
2023-01-03 15:36:59 <uuidNuni1> created a file called n1.modules in /etc/mkinitfs/features.d/ to include the USB and EMMC drivers, and then it will boot fine 
2023-01-03 15:37:31 <ikke> You can also make an MR here to have the feature available by default
2023-01-03 15:37:31 <minimal> unidNuni1: and add "n1" to features list in /etc/mkinitfs/mkinitfs.conf
2023-01-03 15:37:34 <ikke> https://gitlab.alpinelinux.org/alpine/mkinitfs/
2023-01-03 15:37:56 <uuidNuni1> minimal: sure. 
2023-01-03 15:38:23 <uuidNuni1> once @mps enable these options, i'll redo the alpine installation on my Amlogic box : https://gitlab.alpinelinux.org/alpine/aports/-/issues/14512
2023-01-03 15:39:31 <uuidNuni1> ikke: that drivers are specific to my devices, i don't think it is needed to make a PR/MR. it's not generic drivers 
2023-01-03 15:40:42 <uuidNuni1> it's like rpi.modules, odriod.modules ... there are so many SoCs ... 
2023-01-03 15:40:49 <uuidNuni1> to busy to make MR/PR
2023-01-03 15:43:20 <uuidNuni1> ARM SoCs are fun :) digging into device tree and making a working kernel, that's fund. it helps learn stuff
2023-01-03 15:43:26 <uuidNuni1> x86_64 is boring
2023-01-03 15:50:50 <uuidNuni1> for /etc/init.d/fsck to work, do i set cmdline rw or ro?
2023-01-03 15:53:52 <EF> uuidNuni1: certainly have read manual.  Spent days trying to work this out.  
2023-01-03 15:54:54 <uuidNuni1> EF dnsmasq should say something like bind addr effect dns... 
2023-01-03 15:56:57 <uuidNuni1> EF if you bind to a private addr, i don't think 127.0.0.1:53 will work 
2023-01-03 15:58:49 <minimal> uuidNuni1: from memory Linux kernel defaults to "ro" if not specified
2023-01-03 15:59:36 <uuidNuni1> dnsmasq man : Note that if no --in‐terface option is given, but --listen-address is, dnsmasq will not automatically listen on the loopback interface.  To achieve this, its IP address,  127.0.0.1,  must  be  explicitly given as a --listen-address option.
2023-01-03 16:05:19 <uuidNuni1> minimal: i'll test. archlinux has a fsck hook which requires rw cmdline.
2023-01-03 16:06:33 <minimal> uuidNuni1: are you referring to fsck of rootfs or other fses?
2023-01-03 16:10:11 <uuidNuni1> a quick note : https://0x0.st/oRW7.txt
2023-01-03 16:10:15 <uuidNuni1> i am prepared. 
2023-01-03 16:10:32 <uuidNuni1> minimal: fsck for all partition
2023-01-03 16:10:40 <uuidNuni1> on boot. 
2023-01-03 16:13:19 <minimal> uuidNuni1: the "ro" / "rw" on cmdline only refer to the rootfs
2023-01-03 16:13:39 <uuidNuni1> right. 
2023-01-03 16:14:19 <uuidNuni1> minimal: can you check my `rc-update add`, if you find anything wrong? 
2023-01-03 16:15:04 <minimal> uuidNuni1: re your commands line, you could have looked at my script for an detailed example of how to do that ;-)
2023-01-03 16:15:19 <uuidNuni1> there are init.d/{sysfs,procfs} do i need to enable them?
2023-01-03 16:19:22 <uuidNuni1> minimal: can i have your link? 
2023-01-03 16:20:10 <uuidNuni1> 200M fully working basic server, that's fantastic
2023-01-03 16:21:23 <uuidNuni1> currently it is 65M without kernel. 
2023-01-03 16:21:32 <uuidNuni1> i'm waiting for the kernel :)
2023-01-03 16:22:47 <minimal> uuidNuni1: https://github.com/dermotbradley/create-alpine-disk-image
2023-01-03 16:25:06 <minimal> uuidNuni1: you can do smaller than that with a little tuning
2023-01-03 16:28:00 <minimal> uuidNuni1: I don't enable either sysfs or procfs init.d server - sysfs is autostarted as a dep of another init.d service
2023-01-03 16:30:10 <uuidNuni1> minimal: your script is very detailed and thoughtful
2023-01-03 16:31:17 <minimal> uuidNuni1: it's a script creator, give it appropriate options and it spits out a 2nd script tailored to do what you specified
2023-01-03 16:31:29 <uuidNuni1> i did a grep : grep -R add_init_d_service create-alpine-disk-image
2023-01-03 16:31:33 <uuidNuni1> very helpful 
2023-01-03 16:37:19 <uuidNuni1> sudo dua alp : 23.79 MB alp
2023-01-03 16:37:27 <uuidNuni1> i removed vim 
2023-01-03 16:53:30 <uuidNuni1> Executing mkinitfs-3.7.0-r0.trigger
2023-01-03 16:53:43 <uuidNuni1> where is this file? i can't find it using find 
2023-01-03 16:54:12 <uuidNuni1> i want to regen the initramfs the way mkinitfs.trigger does 
2023-01-03 16:54:48 <minimal> uuidNuni1: it is part of the mkinitfs packaging
2023-01-03 16:54:59 <uuidNuni1> is it on the disk?
2023-01-03 16:55:10 <uuidNuni1> or in some tar ball
2023-01-03 16:55:38 <minimal> https://git.alpinelinux.org/aports/tree/main/mkinitfs/mkinitfs.trigger
2023-01-03 16:56:00 <uuidNuni1> /lib/apk/db/scripts.tar: binary file matches
2023-01-03 16:56:53 <dancesWithCycle> Hi, anyone mind sharing with me how I build a C programm for Alpine --arch aacrh64 on my Debian host so that I can run it on my PostmarketOS host?
2023-01-03 16:58:15 <uuidNuni1> minimal: yeah, but where is it? 
2023-01-03 16:59:28 <minimal> dunno where apk keeps it locally. There are references to triggers in /lib/apk/db/triggers but not the actual script
2023-01-03 17:00:49 <uuidNuni1> minimal: yeah, found it : tar -t -f /lib/apk/db/scripts.tar 
2023-01-03 17:01:53 <minimal> if you're inside a chroot then you should only need to run "mkinitfs" with no options
2023-01-03 17:02:16 <uuidNuni1> no, uname -r shows host kernel version
2023-01-03 17:02:24 <uuidNuni1> i need to append kernel version 
2023-01-03 17:02:48 <dancesWithCycle> How do I setup chroot for Alpine?
2023-01-03 17:05:50 <minimal> uuidNuni1: ah, I do it my setting up mkinitfs config files appropriately before installing the kernel (which then runs the trigger)
2023-01-03 17:11:13 <dancesWithCycle> Is building a C program different from  building an *.apk? I am wondering how the other *.apk's are build for Alpine and PostmarketOS? Can't I just follow the same way?
2023-01-03 17:11:40 <ikke>  
2023-01-03 17:14:30 <uuidNuni1> dancesWithCycle: make a C program is to create a binary then you can move it to somwhere else and run it, make a apk is to create a package which includes your c program and other files then you can install via apk
2023-01-03 17:15:34 <uuidNuni1> for *.apk read : https://wiki.alpinelinux.org/wiki/APKBUILD_Reference#$pkgname.pre-install
2023-01-03 17:15:53 <uuidNuni1> for c program, you may want to make it statically, so it runs everywhere 
2023-01-03 17:39:03 <EF> So the dnsmasq issue appears to be dnsmasq related. Using udhcpd instead for dhcp and dnsmasq for only dns.
2023-01-03 18:03:32 <uuidNuni1> EF: udhcpd does not handle reconnection on carrier lost. use dhcpcd instead 
2023-01-03 18:09:07 <minimal> uuid~Numi1: he is talking about DHCP *servers*, not clients
2023-01-03 18:13:24 <EF> indeed.  Spent two days trying to work out what dnsmasq kills internet traffic when dhcp enabled in dnsmasq also
2023-01-03 18:16:19 <minimal> EF: you still haven't posted any detailed information to enable anyone to help you with the issue
2023-01-03 18:18:03 <blt[m]> Am testing linux-lts from edge (6.1.2) I notice very high temperature on all CPU core and fan are running high but CPU is at 1%, htop is not showing any usage, nor Intel GPU tools. Any idea where to look first ?
2023-01-03 18:19:26 <blt[m]> (Nothing in dmesg)
2023-01-03 18:19:51 <ikke> maybe your cores are not scaling down?
2023-01-03 18:22:02 <blt[m]> ikke: am taking a look at  /sys/devices/system/cpu/cpufreq/policy
2023-01-03 18:23:30 <blt[m]> scaling_cur_freq is showing about 2300002 where my base is 1200000 and max is 4400000
2023-01-03 18:26:02 <uuidNuni1> blt[m]: you need fancontrol 
2023-01-03 18:26:15 <uuidNuni1> it's pwm related
2023-01-03 18:26:55 <blt[m]> but my fan are running because my laptop is very hot
2023-01-03 18:27:26 <blt[m]> I am taking a look to cpupower to see if it is a freqency related issue on 6.1.2
2023-01-03 18:27:55 <blt[m]> I had no issue with 5.15
2023-01-03 18:29:36 <uuidNuni1> blt[m]: better run a watch on the current freq
2023-01-03 18:29:48 <uuidNuni1> as root not sudo.
2023-01-03 18:30:22 <uuidNuni1> watch -n1 cat ...
2023-01-03 18:32:59 <blt[m]> cpupower frequency-set -g powersave is not changing anything
2023-01-03 18:33:04 <blt[m]> current CPU frequency: 2.80 GHz it seems to be the issue
2023-01-03 18:33:22 <blt[m]> as the base is 400MHz and my computer is at 1% of CPU
2023-01-03 18:36:25 <uuidNuni1> blt[m]: cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver
2023-01-03 18:37:18 <blt[m]> uuidNuni1: intel_pstate
2023-01-03 18:37:19 <ikke> blt[m]: what is the scaling_driver?
2023-01-03 18:37:22 <ikke> oh lol
2023-01-03 18:38:30 <blt[m]> hum I see intel_cstate in my lsmod but not intel_pstate
2023-01-03 18:38:37 <ikke> me neither
2023-01-03 18:38:44 <uuidNuni1> blt[m]: what is your cpu 
2023-01-03 18:38:54 <ikke> But my driver is intel_cpufreq
2023-01-03 18:38:57 <blt[m]> Alder Lake
2023-01-03 18:39:02 <blt[m]> Intel 12
2023-01-03 18:39:16 <ikke> but my laptop is still on linux 5.15
2023-01-03 18:39:29 <blt[m]> yep on 5.15 I dont have any issue
2023-01-03 18:39:46 <ikke> Interesting to see what scaling_driver is used for you on 5.15
2023-01-03 18:40:01 <blt[m]> I can roll back on 5.15 to check
2023-01-03 18:40:47 <blt[m]> am Downgrading linux-lts am back in 1 min after a clean reboot
2023-01-03 18:41:16 <uuidNuni1> blt[m]: try this on cmdline intel_pstate=disable see if it makes a diff?
2023-01-03 18:42:03 <scorpion2185[m]1> how can I use osk-sdl on alpine?
2023-01-03 18:42:47 <uuidNuni1> watch cat /sys/devices/system/cpu/cpu[0-9]*/cpufreq/scaling_cur_freq
2023-01-03 18:43:20 <ikke> scorpion2185[m]1: you'd have to ask the pmos devs
2023-01-03 18:43:47 <blt[m]> interesting....... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/fkGTBKAjROifmuuRkcVnDTVl>)
2023-01-03 18:43:53 <blt[m]> still intel_pstate
2023-01-03 18:44:06 <blt[m]> but now everything is OK, no more high temp
2023-01-03 18:44:10 <ikke> And the frequency?
2023-01-03 18:45:07 <blt[m]> going back low :... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JQSvFDMARubAzSwwKGZAeQBs>)
2023-01-03 18:45:15 <blt[m]> Package id 0:  +53.0°C  (high = +100.0°C, crit = +100.0°C)... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/yZxwXUCfdhoZlkRUiGocxfCA>)
2023-01-03 18:45:20 <blt[m]> everything back to normal
2023-01-03 18:45:29 <blt[m]> so there is definitely something on 6.1.2
2023-01-03 18:46:58 <blt[m]> the only kernel parameter am using for my i915 is intel_idle.max_cstate=1
2023-01-03 18:46:59 <blt[m]> perhaps it could be the cause? I can remove to test
2023-01-03 18:47:15 <ikke> worth a try
2023-01-03 18:47:26 <blt[m]> am removing it and upgrading to 6.1.2 to see
2023-01-03 18:47:45 <blt[m]> I know I have some parameters in my UEFI for the states as well but will try like that
2023-01-03 18:49:25 <scorpion2185[m]1> ikke: can alpine devs add it as "native" package too?
2023-01-03 18:50:10 <ikke> Someone interested certainly can
2023-01-03 18:51:07 <psykose_> it's device specific and requires initrd integration, the hell would it be doing in the alpine repos
2023-01-03 18:53:21 <blt[m]> ha! it was definitely the issue ! the temperature is now around 44° and the scaling are now
2023-01-03 18:53:31 <blt[m]> 1309325... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JPXluvHerUZroxBVtLtpnRkA>)
2023-01-03 18:53:42 <scorpion2185[m]1> device specific?
2023-01-03 18:53:48 <blt[m]> it seems that it is even better on power saving!
2023-01-03 18:54:04 <scorpion2185[m]1> e.g. on debian you can install it and enable it
2023-01-03 18:54:34 <scorpion2185[m]1> ah you need the "drivers" or something for the touch
2023-01-03 18:57:10 <uuidNuni1> blt[m]: 2.1G is still high for idle compared to 600Mhz on 5.15
2023-01-03 18:59:23 <blt[m]> uuidNuni1: I agree ! cpupower frequency-set -g powersave doesn't seem to change a lot, I can see from the watch that it is sometimes going down but not for so long, any idea on how to scale down/force ?
2023-01-03 19:00:00 <uuidNuni1> i use schedutil
2023-01-03 19:00:04 <blt[m]> my CPUIs are almost idle, 1% max
2023-01-03 19:01:36 <blt[m]> it is like it doesnt reach the lower state on my case
2023-01-03 19:03:54 <blt[m]> cpupower frequency-set -g performance doesn't seem to impact either
2023-01-03 19:04:08 <blt[m]> in both case they are at 2.1G
2023-01-03 19:06:07 <blt[m]> https://www.tomshardware.com/news/intel-thread-director-coming-to-linux-5-18 maybe related
2023-01-03 19:16:31 <scorpion2185[m]1> 
ACTION uploaded an image:  < https://matrix.org/_matrix/media/v3/download/matrix.org/jsMxFlIKyAuuDOaIGjwIGrsB/2023-01-03_20-14-23.png >

2023-01-03 19:18:26 <scorpion2185[m]1> not sure to how do that
2023-01-03 19:26:58 <scorpion2185[m]1> (WebKitWebProcess:2): Gdk-WARNING **: 1: Settings portal not found: Failed to execute child process “dbus-launch” (No such file or directory)
2023-01-03 19:26:58 <scorpion2185[m]1> (WebKitWebProcess:2): GLib-GIO-CRITICAL **: 6: g_dbus_connection_get_unique_name: assertion 'G_IS_DBUS_CONNECTION (connection)' failed
2023-01-03 19:27:01 <blt[m]> uuidNuni1:  I see almost half are going back low, but still higher than expected and every other seconds all threads are back to 2.1G, which is not very power efficient it is supposed to reach 400M... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/EBehmoxTdfnStEvfwVcYDgyx>)
2023-01-03 19:32:08 <blt[m]> uuidNuni1: maybe not due to the first article but more this one https://www.phoronix.com/news/Linux-5.17-P-State-ADL-Mobile, I have to see if I can set power instead of balance power
2023-01-03 20:09:41 <scorpion2185[m]1> >basically you need a custom init script in the initfs to run osk-sdl on the appropriate luks device, and then you need to build an initramfs with that custom init script in it
2023-01-03 21:19:48 <frag> anyone using bluetooth speaker? seemingly suprisingly easy with pulseaudio, but not great connection/quality (cutting in and out and changing sound quality)
2023-01-03 21:22:35 <paperowl> only bluetooth keyboard which has no issues. But i did have a separate speaker issue which was fixed by going to pipewire from pulseaudio
2023-01-03 21:22:47 <psykose> all my bluetooth stuff has worked on pipewire for years
2023-01-03 21:23:02 <psykose> weak signal (read: went behind X walls) cause it to cut out of course, but
2023-01-03 21:23:47 <Ermine> My airpods failed to connect to my laptop last time I tried it.
2023-01-03 21:24:19 <psykose> you of all people having air pods is suprising
2023-01-03 21:25:12 <Ermine> Well, that was ill-thought purchase
2023-01-03 21:25:40 <psykose> but afaik they should also work, maybe you should debug that at some point
2023-01-03 21:26:19 <Ermine> Sometime
2023-01-03 21:26:44 <Ermine> But i915 on this laptop has more priority
2023-01-03 21:28:57 <psykose> ye olde ancient i915g or
2023-01-03 21:30:27 <Ermine> no, I mean debugging this kernel module.
2023-01-04 02:52:06 <Armote[m]> is it possible to use alpine apk such that everything is built from source locally? similar to gentoo
2023-01-04 02:53:59 <Armote[m]> perhaps a local apk build server but one that only compiles packages on demand?
2023-01-04 02:56:12 <psykose> it's just a package manager, anything is possible
2023-01-04 02:58:00 <q66> at zombocom
2023-01-04 02:58:40 <q66> the only limit is yourself
2023-01-04 08:01:42 <uuidNuni1> $MOCK openrc boot
2023-01-04 08:01:47 <uuidNuni1> what is MOCK?
2023-01-04 08:06:32 <ikke> deja-vu
2023-01-04 08:06:55 <ikke> Ermine asked the same question yesterday (in #alpine-devel)
2023-01-04 08:07:10 <ikke> It's used to prevent those commands to run during testing
2023-01-04 08:07:22 <ikke> (or to run a mock version of those commands)
2023-01-04 08:11:56 <uuidNuni1> so it will not run?
2023-01-04 08:12:34 <ikke> If MOCK is empty, it will just run the command
2023-01-04 08:12:52 <ikke> if it's not empty, it will run whatever MOCK is set
2023-01-04 08:14:01 <uuidNuni1> i chroot install alpine based on minirootfs. it does not boot... 
2023-01-04 08:14:17 <ikke> minirootfs is not meant to boot
2023-01-04 08:14:31 <uuidNuni1> well i install stuff and enabled servicies ...
2023-01-04 08:14:38 <uuidNuni1> i don't know what are missing 
2023-01-04 08:14:43 <ikke> me neither
2023-01-04 08:15:14 <uuidNuni1> but i got a piece of rc.log
2023-01-04 08:15:38 <ikke> uuidNuni1: best way is to install alpine-conf and run setup-alpine
2023-01-04 08:15:40 <uuidNuni1> as i enabled in rc.conf
2023-01-04 08:17:01 <uuidNuni1> https://0x0.st/oRvJ.txt
2023-01-04 08:17:01 <ikke> Do give details
2023-01-04 08:17:49 <uuidNuni1> from the log, do you see usb got mounted?
2023-01-04 08:17:55 <uuidNuni1> usb is the boot device 
2023-01-04 08:18:14 <uuidNuni1> Filesystems couldn't be fixed
2023-01-04 08:18:18 <uuidNuni1> what does that mean 
2023-01-04 08:18:38 <ikke> Not sure
2023-01-04 08:19:12 <uuidNuni1> that message is from fsck
2023-01-04 08:20:27 <uuidNuni1> actually i could just cp a piece of my vm installation over to usb and setup fstab boot etc...
2023-01-04 08:20:35 <uuidNuni1> s/cp/rsync/
2023-01-04 08:20:35 <alpine-meetbot> uuidNuni1 meant to say: actually i could just rsync a piece of my vm installation over to usb and setup fstab boot etc...
2023-01-04 09:27:54 <markand> sometimes on my pi 4 apk del -r hangs for hours
2023-01-04 09:27:57 <markand> like apk del -r xset
2023-01-04 09:29:00 <uniqNuuid> lol. rsync over the vm to the usb, do some bootloader fstab job, it boots 
2023-01-04 09:29:32 <uniqNuuid> but installation on top of minirootfs does not boot, i guess localmount did not run 
2023-01-04 09:32:18 <uniqNuuid> i think it was due to setup-disk was ignored by me 
2023-01-04 10:49:46 <zweik> hello, i am using alpine+sway. does someone know how to setup autostart for syncthing? i tried "exec syncthing-start.desktop" in my sway config but diddnt work. thx
2023-01-04 10:52:36 <Ermine> zweik: did you try symlinking syncthing-start.desktop to /etc/xdg/autostart?
2023-01-04 11:04:46 <zweik> no, i will try this, thx
2023-01-04 11:41:03 <zweik> doesnt work :/ tried /etc/xdg/autostart and ~/.config/autostart also with copy the .desktop file 
2023-01-04 11:41:35 <zweik> but it also doenst work if i try it with firefox.desktop or other desktop files
2023-01-04 11:47:56 <Ermine> in sway config, replace syncthing-start.desktop with what is written after Exec= 
2023-01-04 12:39:27 <Ermine> Updated to 3.17, udevd got removed somehow. Does apk save its log somewhere?
2023-01-04 12:43:26 <Ermine> Also bash in tmux cannot find .bashrc anymore, did anybody stumble upon that?
2023-01-04 12:44:02 <grawity> cannot find, or stopped looking for it?
2023-01-04 12:47:58 <Ermine> grawity: as I've written, it cannot find.
2023-01-04 12:48:26 <grawity> does it give an error message about it?
2023-01-04 12:48:49 <Ermine> Yes
2023-01-04 12:50:18 <grawity> please share that error message with the channel
2023-01-04 12:51:07 <Ermine> -bash: .bashrc: No such file or directory
2023-01-04 12:52:56 <grawity> what's the working directory of the shell when this happens -- your home or something else?
2023-01-04 12:55:10 <Ermine> something else
2023-01-04 12:56:04 <grawity> edit your ~/.bash_profile (or ~/.bash_login or ~/.profile, whichever you have) to source ~/.bashrc instead of just .bashrc
2023-01-04 12:58:14 <Ermine> good catch, thank you
2023-01-04 13:01:04 <chereskata> how can it be done with ash instead?
2023-01-04 13:01:23 <chereskata> (when using gnome-shell on wayland, that ignores .profile)
2023-01-04 13:14:10 <uuidNuniq> chereskata: ln -s .profile .the_profile_gnome_does_not_ignore
2023-01-04 13:15:00 <chereskata> uuidNuniq: how is this mysterious file actually called?
2023-01-04 13:18:23 <uniqNuuid> chereskata: or source your .profile in .bash_profile
2023-01-04 13:18:33 <chereskata> i use ash
2023-01-04 13:18:42 <uniqNuuid> echo 'source ~/.profile' >> ~/.bash_profile
2023-01-04 13:19:14 <chereskata> that will work for sure, but is bash dependend
2023-01-04 13:19:16 <uniqNuuid> well. well. 
2023-01-04 13:19:47 <chereskata> also this should influence only env vars for the bash shell. Desktop wide env variables are still a dead end
2023-01-04 13:20:23 <uniqNuuid> what about /etc/profile? 
2023-01-04 13:20:37 <chereskata> there are some hacks to place them in some gdm config file, but this is SOO unintuitive compared to .profile
2023-01-04 13:21:00 <chereskata> \/etc/profile is used for it atm
2023-01-04 13:21:22 <chereskata> but this would not be as cute, if i had multiple users on my computer
2023-01-04 13:21:55 <chereskata> for example .  ${HOME}/.profile will not work
2023-01-04 13:22:26 <grawity> hmm but /usr/bin/gnome-session deliberately runs $SHELL with the -l option
2023-01-04 13:22:38 <grawity> does ash ignore ~/.profile in that case regardless?
2023-01-04 13:22:51 <grawity> or did that part get patched out from alpine
2023-01-04 13:23:34 <chereskata> Oh neat! I actually did not know about this 
2023-01-04 13:23:42 <chereskata> -I could be a solution
2023-01-04 13:24:11 <grawity> basically it's supposed to work by default :(
2023-01-04 13:24:34 <chereskata> i thought that if a variable is undefined while executing . to profile it would not work 
2023-01-04 13:24:42 <chereskata> so i even did not try it
2023-01-04 13:24:58 <grawity> ...eh?
2023-01-04 13:27:24 <chereskata> i thought that putting . ${HOME}/.profile in /etc/profile would cause an error because ${HOME} is undefinded. By reading man bash /-I i learned, that the variables get passed down to the user session. But i detected my thought error, that suggested that .profile would load after the user session started
2023-01-04 13:28:02 <chereskata> but it would still not load in /etc/profile because ${HOME} is undefined there
2023-01-04 13:28:13 <chereskata> or am i missing smth
2023-01-04 13:28:19 <grawity> no, it is supposed to be defined at that point
2023-01-04 13:28:33 <grawity> $HOME is set by gdm and inherited down by all child processes
2023-01-04 13:28:37 <grawity> it's not per-file or anything
2023-01-04 13:28:49 <chereskata> that was my default assumption
2023-01-04 13:29:09 <grawity> so if the `ash` process has $HOME in its environment, then it'll be available to any sourced script, whether it's ~/.profile or /etc/profile or anything else
2023-01-04 13:29:47 <chereskata> ${HOME} is present when running env
2023-01-04 13:31:13 <chereskata> but .ashrc does not load, neither does .profile. I tried around a month ago to force ash to load its config by writing UNKNOWN_VARNAME=/home/user/.ashrc to /etc/profile
2023-01-04 13:31:46 <chereskata> this var loaded the ashrc file, but did nothing to .profile (as expected)
2023-01-04 13:52:14 <vkrishn> chereskata: how is invoked in your setup ?
2023-01-04 13:52:21 <vkrishn> how is ash ^
2023-01-04 13:52:44 <chereskata> by opening kgx
2023-01-04 13:55:55 <vkrishn> I have not used kgx, but if does not bother you, is it like by just type ash in the terminal ?
2023-01-04 13:56:06 <chereskata> ys
2023-01-04 13:56:08 <chereskata> yes
2023-01-04 13:56:24 <vkrishn> could you try this, before you invoke ash
2023-01-04 13:56:27 <vkrishn> export ENV=/etc/profile
2023-01-04 13:56:30 <vkrishn> ash
2023-01-04 13:58:36 <chereskata> so my finding
2023-01-04 13:59:30 <chereskata> i mixed /etc/profile while describing my problem above. Just replace every /etc/profile mentioned by /etc/environment 
2023-01-04 13:59:45 <vkrishn> yes, should be same
2023-01-04 13:59:55 <chereskata>  /etc/profile does not get loaded in a gnome session by using kgx
2023-01-04 14:00:07 <chereskata> verified by this code:
2023-01-04 14:00:09 <vkrishn> as long as the file is standard source'eable file
2023-01-04 14:00:31 <chereskata> echo ${PS1}
2023-01-04 14:00:41 <chereskata>  \w \$
2023-01-04 14:00:49 <chereskata> export ENV=/etc/profile
2023-01-04 14:00:55 <chereskata> ash
2023-01-04 14:01:09 <chereskata>  \h:\w\$
2023-01-04 14:01:32 <chereskata> => default PS1 till sourcing /etc/profile by hand
2023-01-04 14:04:00 <vkrishn> which env file are you expecting to load with ash ?
2023-01-04 14:06:08 <chereskata> ${HOME}/.ashrc. That needs a loaded ${HOME}/.profile to tell ash to load .ashrc. Probably with ENV=${HOME}/.ashrc 
2023-01-04 14:06:30 <vkrishn> yes, try that
2023-01-04 14:06:44 <chereskata> .profile is ignored on wayland
2023-01-04 14:06:46 <vkrishn> only if $HOME is defined before
2023-01-04 14:08:03 <chereskata> probably ${HOME} is defined by gdm, because i have it when calling env in kgx
2023-01-04 14:09:46 <vkrishn> well I tried in mobile environment, konsole in knoppix, xterm in knoppix - seems to work, maybe your problem is different
2023-01-04 14:09:46 <chereskata> I have an idea
2023-01-04 14:10:00 <chereskata> vkrishn: thank you
2023-01-04 14:10:44 <chereskata> i will try to add a line to /etc/gdm/PostLogin/Default with . ${HOME}/.profile
2023-01-04 14:11:09 <chereskata> https://paste.gnome.org/1qZ1L5JrU
2023-01-04 14:13:39 <chereskata> vkrishn: do you use gdm and gnome there?
2023-01-04 14:13:55 <chereskata> @all thanks for your help
2023-01-04 14:15:44 <vkrishn> kde on knoppix as my main desktop
2023-01-04 14:16:01 <vkrishn> rest are alpinelinux (headless)
2023-01-04 14:16:37 <vkrishn> but I do try tons of linux distros, when free
2023-01-04 14:16:49 <chereskata> cool!
2023-01-04 14:17:17 <chereskata> the problem lies probably in gdm not sourcing .profile in home
2023-01-04 14:37:28 <uniqNuuid> chereskata: as i said.... source your .profile in .bash_profile or whatever gdm likes to source 
2023-01-04 14:37:59 <uniqNuuid> you don't need to install bash... just touch .bash_profile
2023-01-04 14:38:24 <uniqNuuid> and add one line of code to source your .profile in .bash_profile
2023-01-04 14:46:52 <chereskata> never tried that
2023-01-04 15:59:10 <hsv2> I'm fighting since yesterday to get Alpine boot in EFI mode on some computer. On that system, an Alpine installation on a USB pendrive boots in both BIOS and EFI mode, but a similar installation on an internal HDD boots only in BIOS mode. UEFI seems to be unable to load GRUB on HDD for some reason. Somebody may have an idea? more info on my setup: https://bin.disroot.org/?6166e090514f811d#12iBDumub7zuHg
2023-01-04 15:59:10 <hsv2> TeZYyJGgDLEiDRzDK2FFueGEjYXnCp
2023-01-04 16:07:59 <chereskata> uniqNuuid: Hey, your suggestion did not work. But i made it working by configuring gdm to source .profile by default
2023-01-04 16:08:22 <chereskata> i added three lines of code in /etc/gdm/PostLogin/Default
2023-01-04 16:09:10 <blt[m]> does udev takeover the control of powerbutton if I am not using acpid nor elogind ? I would like to hibernate my computer instead of poweroff when pressing power button , is it a good idea to create an udev rule for that ?
2023-01-04 17:02:14 <Ermine> blt[m]: you can install pm-utils and acpid
2023-01-04 17:03:34 <blt[m]> yes got both installed and tried but it seems that udev is faster and take so my idea to work with uded
2023-01-04 17:11:47 <blt[m]> Ermine: to be accurate, the hibernation part is working with pm-hibernate and acpi, but when I resume, everything is OK but when the system is ready it goes straight to poweroff , something is holding the poweroff process until the resume is done...
2023-01-04 17:25:35 <blt[m]> (forgot to mention that doing a pm-hibernate manually is working perfectly, the problem is when I am using the power button)
2023-01-04 17:27:47 <Ermine> blt[m]: Maybe it tries to react to poweroff event after resuming?
2023-01-04 17:28:20 <blt[m]> yes that's what I am trying to understand but don't know how to stop that
2023-01-04 17:28:40 <blt[m]> that's the reason I thought originally creating an udev rule
2023-01-04 17:41:11 <Ermine> Do you have acpid running?
2023-01-04 17:42:36 <blt[m]> yes
2023-01-04 17:43:06 <blt[m]> and again the hibernate is working fine, but when I resume power off kicks off at the end of the resuming
2023-01-04 17:43:10 <blt[m]> * the resuming phase
2023-01-04 17:46:01 <Ermine> You mean, if you do pm-hibernate, after resuming your computers proceeds to powering off?
2023-01-04 17:52:57 <blt[m]> it first resumes and then poweroff
2023-01-04 17:53:52 <Ermine> 
ACTION scratches head

2023-01-04 17:53:55 <blt[m]> I am investingating and I see that udev is rreading a lot of files from /lib/udev/rules.d and found one for power-switch that is perhaps the reason... I am checkig that oneto es
2023-01-04 17:54:01 <blt[m]> s/oneto/one/, s/es/to see/
2023-01-04 18:01:30 <blt[m]> that's what I think is taking over :... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/PoneZurvEwtjugcexEttmrSa>)
2023-01-04 18:01:50 <blt[m]>  * that's what I think is taking over :... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/lKPLsAyLQUSwNljzEhZCwzQR>)
2023-01-04 18:05:11 <Ermine> I'm not really knowledgeable about udev and its rules
2023-01-04 18:05:49 <Ermine> But it doesn't invoke poweroff or something
2023-01-04 18:24:51 <blt[m]> removing the rules is indeed not changing anything 
2023-01-04 18:24:59 <blt[m]> this udev rules seems to only tag
2023-01-04 19:07:33 <minimal> blt[m]: in elogind.conf what is HandlePowerKey set to?
2023-01-04 19:10:40 <scorpion2185[m]1> https://github.com/ReimuNotMoe/ydotool/blob/master/CMakeLists.txt
2023-01-04 19:11:28 <frag> trying to run pipewire, exits with errors (also cant get sound to work at all, help!) https://file.silentsilas.com/N3k.txt
2023-01-04 19:11:51 <scorpion2185[m]1> what are the directory of the cmake
2023-01-04 19:13:05 <scorpion2185[m]1> how can you tell it to don't make systemd files?
2023-01-04 19:13:22 <ikke> you cannot
2023-01-04 19:13:43 <ikke> You can remove them afterwards
2023-01-04 19:13:47 <Ermine> frag: do you use elogind?
2023-01-04 19:13:48 <scorpion2185[m]1> how do you handle this?
2023-01-04 19:14:20 <ikke> rm -r "$pkgdir"/lib/systemd
2023-01-04 19:14:22 <scorpion2185[m]1> build() {... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/IokXYVzvEkDLRyAqWHAaLzvV>)
2023-01-04 19:14:30 <ikke> in package
2023-01-04 19:14:58 <frag> Ermine: Ermine: not that i know of
2023-01-04 19:14:58 <Ermine> frag: if you do, either uninstall elogind or start this service
2023-01-04 19:14:59 <scorpion2185[m]1> got it ,  is build function fine?
2023-01-04 19:15:13 <Ermine> frag: rc-service elogind status
2023-01-04 19:15:31 <frag> Ermine: its "stopped"
2023-01-04 19:15:44 <Ermine> frag: start it
2023-01-04 19:16:09 <ikke> scorpion2185[m]1: no need to specify -j `nproc`
2023-01-04 19:17:03 <scorpion2185[m]1> I copied from the readme
2023-01-04 19:17:16 <ikke> and you can do: cmake -B build
2023-01-04 19:17:22 <ikke> then you don't need to mkdir + cd
2023-01-04 19:17:34 <Ermine> frag: or remove it if you do not intend to use it (e.g. you don't run DEs)
2023-01-04 19:18:00 <ikke> scorpion2185[m]1: and then cmake --build
2023-01-04 19:18:03 <ikke> instead of make
2023-01-04 19:18:16 <ikke> sorry, cmake --build build
2023-01-04 19:18:43 <frag> Ermine: oh i dont have DE ... got a new error after starting elogind : https://file.silentsilas.com/cQg.txt
2023-01-04 19:19:02 <frag> Ermine: i dont think pulseaudio is running btw, dunno if that is a problem
2023-01-04 19:19:22 <Ermine> frag: no need to run PulseAudio if you're using PW
2023-01-04 19:19:39 <Ermine> do you have eudev started?
2023-01-04 19:19:50 <ikke> scorpion2185[m]1: https://tpaste.us/BkNY 
2023-01-04 19:20:30 <ikke> Nowadays we even use Ninja as generator: https://tpaste.us/NOgv 
2023-01-04 19:20:56 <ikke> (then you need to add samurai to makedepends)
2023-01-04 19:20:57 <frag> Ermine: i dont think so 'pgrep eudev' doesnt find anything
2023-01-04 19:21:36 <Ermine> frag: rc-service udev status
2023-01-04 19:21:54 <frag> Ermine: oh, its "started"
2023-01-04 19:22:39 <scorpion2185[m]1> ikke: no pkg folder is created
2023-01-04 19:22:47 <ikke> scorpion2185[m]1: that is done later
2023-01-04 19:23:07 <elly> I didn't know about samurai ikke, thanks for mentioning it
2023-01-04 19:23:29 <ikke> scorpion2185[m]1: https://tpaste.us/x5gW 
2023-01-04 19:23:36 <scorpion2185[m]1> what creates that?
2023-01-04 19:23:48 <ikke> scorpion2185[m]1: check the latest paste
2023-01-04 19:23:55 <Ermine> frag: try udevadm trigger as root
2023-01-04 19:25:29 <scorpion2185[m]1> it install in /usr/local, do you move it manually ?
2023-01-04 19:26:26 <ikke> scorpion2185[m]1: https://tpaste.us/yvek 
2023-01-04 19:28:04 <Ermine> Did anybody see frag's terminal?
2023-01-04 19:28:17 <frag> Ermine: doas udevadm trigger?
2023-01-04 19:28:24 <Ermine> frag: yes
2023-01-04 19:28:37 <frag> Ermine: no errrors or output
2023-01-04 19:28:55 <ikke> scorpion2185[m]1: -DCMAKE_INSTALL_PREFIX=/usr sets the prefix
2023-01-04 19:29:13 <Ermine> frag: now try to run PW again
2023-01-04 19:29:22 <scorpion2185[m]1> I see, but error: /usr/bin/abuild: line 17: -DCMAKE_INSTALL_PREFIX=/usr: not found
2023-01-04 19:29:36 <ikke> Did you also add the \ on the previous line/
2023-01-04 19:31:14 <scorpion2185[m]1> ah no
2023-01-04 19:31:34 <frag> Ermine: more errors https://file.silentsilas.com/g9i.txt
2023-01-04 19:32:04 <scorpion2185[m]1> no need to make a openrc service?
2023-01-04 19:32:29 <frag> Ermine: output got a bit messy btw..
2023-01-04 19:32:32 <ikke> scorpion2185[m]1: If this runs as a service, then it would be good to provide one
2023-01-04 19:32:39 <Ermine> frag: does it go on?
2023-01-04 19:33:30 <scorpion2185[m]1> in a separate package *-openrc ? I need to look an example
2023-01-04 19:33:35 <frag> Ermine: no, now music plays btw :S
2023-01-04 19:33:43 <ikke> scorpion2185[m]1: a subpackage
2023-01-04 19:33:52 <Ermine> 
ACTION scratches head

2023-01-04 19:34:17 <Ermine> frag: do pipewire and wireplumber process run?
2023-01-04 19:34:29 <Ermine> s/do/are
2023-01-04 19:34:29 <alpine-meetbot> Ermine meant to say: frag: are pipewire and wireplumber process run?
2023-01-04 19:34:34 <ikke> scorpion2185[m]1: abuild can split it automatically, you just need to make sure you add $pkgname-openrc to subpackages, and install the service to $pkgdir/etc/init.d/<servicename>
2023-01-04 19:34:43 <Ermine> s/process/processes/ s/run/running/
2023-01-04 19:34:43 <alpine-meetbot> Ermine meant to say: frag: are pipewire and wireplumber processes run?
2023-01-04 19:34:53 <Ermine> s/run/running/
2023-01-04 19:34:53 <alpine-meetbot> Ermine meant to say: frag: are pipewire and wireplumber processes running?
2023-01-04 19:35:54 <frag> Ermine: how do i check? i thought pipewire-launcher exited, but i guess its a daemon? at least 'pgrep wire' finds 3 processes running
2023-01-04 19:36:30 <Ermine> frag: they are put into background by the script
2023-01-04 19:36:47 <frag> ah so TWO pipewire and one wireplumber
2023-01-04 19:37:13 <Ermine> two pipewire? do you mean one of them is pipewire-pulse?
2023-01-04 19:37:27 <scorpion2185[m]1> I see i'll lock lightdm
2023-01-04 19:37:47 <frag> Ermine: https://file.silentsilas.com/IKc.txt
2023-01-04 19:37:54 <scorpion2185[m]1> *look
2023-01-04 19:38:20 <Ermine> frag: kill 'em all and try again, maybe that would help
2023-01-04 19:39:09 <scorpion2185[m]1> so do I make a ydotoold.initd file ?
2023-01-04 19:39:36 <frag> Ermine: no, still two pipewire after killing them and rerunning the command
2023-01-04 19:39:39 <ikke> scorpion2185[m]1: in the package directory, yes
2023-01-04 19:40:23 <Ermine> frag: are you sure you killed them all? pipewire-launcher shouldn't kill them all.
2023-01-04 19:41:15 <frag> Ermine: i killed them with "killall pipewire", quite sure, pgrep wire didnt have any output
2023-01-04 19:42:00 <ikke> scorpion2185[m]1: in the init.d directory it will just have the service name without extension
2023-01-04 19:42:08 <Ermine> 
ACTION scratches head harder

2023-01-04 19:42:45 <scorpion2185[m]1> wait there is one alredy but it doesn't get installed https://github.com/ReimuNotMoe/ydotool/blob/master/Daemon/ydotool.service-openrc.in
2023-01-04 19:42:59 <ikke> scorpion2185[m]1: then you install it manually
2023-01-04 19:43:06 <scorpion2185[m]1> do I copy that in the init.d dir? `install -Dm755 "$srcdir"/lightdm.initd "$pkgdir"/etc/init.d/lightdm`
2023-01-04 19:43:15 <ikke> yes
2023-01-04 19:43:25 <ikke> scorpion2185[m]1: no $srcdir
2023-01-04 19:44:46 <Ermine> frag: did you modify pipewire-laucher by the chance?
2023-01-04 19:45:43 <frag> Ermine: nope, but did change some config
2023-01-04 19:46:07 <Ermine> How did you change it?
2023-01-04 19:46:14 <frag> do you need alsa when you have pw?
2023-01-04 19:47:14 <Ermine> Afaik if something uses alsa directly, you can add pipewire-alsa
2023-01-04 19:47:15 <frag> it was this morning, dont remember well, but tried to disable dbus (i see now it was probably commented out so changing it to false probably didnt do anything :p)
2023-01-04 19:47:36 <frag> Ermine: okok :]
2023-01-04 19:48:28 <frag> also prob some other things
2023-01-04 19:48:51 <frag> but i doubt it would affect two pipewires
2023-01-04 19:49:12 <Ermine> Try to run with default config (and dbus disabled as per wiki)
2023-01-04 19:49:46 <scorpion2185[m]1> the service file looks in /usr/local, do I fix with some command?
2023-01-04 19:50:04 <frag> Ermine: oh 'top' shows the two pipewires, one is '/usr/bin/pipewire -c pipewire-pulse.conf'
2023-01-04 19:50:14 <ikke> scorpion2185[m]1: a patch
2023-01-04 19:50:40 <frag> the other just '/usr/bin/pipewire'
2023-01-04 19:51:09 <blt[m]> <minimal> "blt: in elogind.conf what is..." <- I thought I was not using it but it seems I am, perhaps I installed it for some testing with pipewire... so it seems that it is the issue as HandlePowerKey is default poweroff....
2023-01-04 19:51:32 <blt[m]> I am still not clear on what is taking the precedence between elogind, acpid and udevd
2023-01-04 19:52:22 <blt[m]> (elogind service is stopped on my case btw)
2023-01-04 19:52:32 <Ermine> frag: which alpine version do you run btw?
2023-01-04 19:54:30 <frag> Ermine: 3.16 its on rpi4 btw ... but after removing pipewire-pulse (and pipewire-alsa), kill, restart,  .... i was gonna say i got only one pipewire, but i got two still but no wireplumber
2023-01-04 19:56:04 <frag> ah my mistake, didnt uninstall pipewire-pulse...
2023-01-04 19:56:24 <Ermine> frag: apply this patch for pipewire-launcher: https://gitlab.alpinelinux.org/alpine/aports/-/commit/584be490b6e8904ecd9a78827ebe679ec8fdd5c6#3bd820349b0066b31b8ba75fbe59b4ef2c0b7136
2023-01-04 19:56:37 <Ermine> (or upgrade to 3.17 which includes this)
2023-01-04 19:57:09 <frag> Ermine: but now after apk del pipewire-pulse, i get one pipewire and one wireplumber..!
2023-01-04 19:57:11 <scorpion2185[m]1> i am trying it and it's giving some output but no effect
2023-01-04 19:57:12 <blt[m]> blt[m]: sorry it is started ! (elogind-daemon)
2023-01-04 19:58:25 <frag> (also had to delete all the pulseaudio packages which is fine :)
2023-01-04 19:58:47 <scorpion2185[m]1> ah because I need to click
2023-01-04 19:59:38 <Ermine> frag: pipewire-pulse is needed for compatibility with PulseAudio apps, and that's a lot of apps, not everything works natively on PW
2023-01-04 20:00:19 <frag> Ermine: i mostly use mpv (but also qutebrowser/firefox/telegram-desktop)
2023-01-04 20:00:52 <Ermine> Ah, mpv, firefox and tg can use PW directly afaik
2023-01-04 20:01:46 <scorpion2185[m]1> how do I mkae a patch? https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Patches
2023-01-04 20:02:22 <ikke> scorpion2185[m]1: what I do, in a clean builddir: git init -q && git add . && git commit -qm.
2023-01-04 20:02:29 <ikke> then make your change
2023-01-04 20:02:31 <ikke> then git diff
2023-01-04 20:02:45 <ikke> That you want write to a patch file
2023-01-04 20:03:37 <scorpion2185[m]1> do I clone the repo somewhere else to do this?
2023-01-04 20:04:42 <scorpion2185[m]1> if you need a module uinput, do you something like putting it in a message?
2023-01-04 20:05:03 <frag> will try to log in/out to set the ENV var :>
2023-01-04 20:10:04 <Ermine> frag: wait, does your music work?
2023-01-04 20:10:35 <frag> Ermine: now it works without starting pw :S (i guess alsa)
2023-01-04 20:11:28 <Ermine> I guess it too
2023-01-04 20:11:42 <Ermine> But FF ditched their ALSA backend afaik
2023-01-04 20:16:23 <frag> seems to work in FF both when pipewire is running and not ... how do i know im using pipewire :p
2023-01-04 20:20:34 <psykose> ff alsa still exists
2023-01-04 20:21:16 <Ermine> my bad then
2023-01-04 20:38:55 <ikke> scorpion2185[m]1: there is /etc/modules-load.d where you can dropin a file, though that still requires manual action when you install the package
2023-01-04 22:25:21 <Saijin_Naib> https://www.phoronix.com/news/BusyBox-1.36-Released
2023-01-04 22:27:06 <Ermine> Woah
2023-01-04 22:30:11 <Saijin_Naib> Right? 
2023-01-04 22:30:31 <Ermine> 
ACTION looked in the comments section

2023-01-04 22:30:40 <Ermine> 
ACTION ran away

2023-01-04 22:31:18 <vladh> what is the nsswitch equivalent on Alpine? trying to figure out how Avahi/mDNS installation works
2023-01-04 22:39:56 <Ermine> vladh: you may look at musl-nscd, but I doubt this is what you want
2023-01-04 22:41:43 <psykose> there is no nsswitch and iirc there isn't a way to get that avahi to work? nscd had some plugin stuff but it wasn't the same
2023-01-04 22:41:53 <psykose> i keep forgetting to test that and find out if it's possible
2023-01-04 22:42:45 <psykose> > Also. Who cares if it's written in C? Is this like language racism?
2023-01-04 22:42:52 <psykose> never change phoronix comments. never change
2023-01-04 22:43:20 <vladh> oh, so you're saying mDNS won't work on Alpine? or is there a different way?
2023-01-04 22:44:01 <Ermine> psykose: agreed
2023-01-04 22:44:18 <Ermine> (on phoronix comments)
2023-01-04 22:47:17 <Saijin_Naib> It's never not a mistake to read them, so that's good. Consistent!
2023-01-04 22:47:35 <Saijin_Naib> It's a shame, really, I think the articles are all so well written.
2023-01-04 22:50:58 <psykose> vladh: based on https://github.com/pikhq/musl-nscd/issues/32 probably not, idk how it really works
2023-01-04 22:51:44 <vladh> got it, thanks psykose :)
2023-01-04 22:52:05 <Saijin_Naib> What is the process for asking if a package in Community of a useful, but EOL tool which has a maintained fork of a different name, can be updated/changed over? 
2023-01-04 22:52:26 <psykose> no idea of the process but which one
2023-01-04 22:52:38 <Saijin_Naib> Tootle, the GTK Mastodon client to Toothy, the GTK4 fixup fork 
2023-01-04 22:53:05 <Saijin_Naib> The dev of Tootle has archived it and is done with it 
2023-01-04 22:53:46 <psykose> this https://github.com/GeopJr/Tooth 
2023-01-04 22:53:56 <Saijin_Naib> Toothy supports Mastodon 4.x features and is improved in general
2023-01-04 22:54:04 <Saijin_Naib> Yes, that's it 
2023-01-04 22:54:09 <psykose> it has no tags
2023-01-04 22:54:17 <Saijin_Naib> Nope 😬
2023-01-04 22:54:42 <psykose> nor is it very up to date, but
2023-01-04 22:54:49 <psykose> i guess the current one isn't in good state either :p
2023-01-04 22:54:58 <psykose> this looks better but they have to release it imo
2023-01-04 22:56:07 <psykose> based on https://github.com/GeopJr/Tooth/issues/27#issuecomment-1340899467 the person doesn't think it's ready
2023-01-04 23:03:23 <Saijin_Naib> Hmm 🤔 so keep this on the back burner until dev tags a release? 
2023-01-04 23:05:06 <psykose> sure
2023-01-04 23:14:53 <Saijin_Naib> Thanks! 
2023-01-05 04:13:32 <uniqNuuid> rust is ... too big in size
2023-01-05 04:14:42 <uniqNuuid> i'v built some rust programs... the generated sized is crazy, the final binary size is crazy too
2023-01-05 04:15:50 <uniqNuuid> very nice fearuews in busybox 1.3.6 
2023-01-05 04:16:11 <uniqNuuid> love it. isn't seedrng already in busybox 1.3.5?
2023-01-05 04:20:55 <psykose> backported
2023-01-05 04:21:29 <psykose> what size, my http server is like 600kb
2023-01-05 04:34:09 <uniqNuuid> crond: implement support for setting PATH in crontab files
2023-01-05 04:34:12 <uniqNuuid> nice
2023-01-05 04:34:51 <uniqNuuid> psykose: my http sever is caddy and is 40MB
2023-01-05 04:36:34 <uniqNuuid> to build a rust program, e.g. fd, one needs to download 1GB+ tools, and it generates 300M + codes for each arch, and the binary size is 2M
2023-01-05 04:40:23 <psykose> my 600kb program disagrees
2023-01-05 04:57:47 <dwfreed> caddy is kitchen sink httpd, not surprising it's massive
2023-01-05 05:04:26 <minimal> uniqNuuid: there was a bit of drama between kernel random maintainer and Busybox devs a while back regarding seedrng
2023-01-05 05:47:10 <kode54> figured out my error
2023-01-05 05:47:25 <kode54> it was because I stupidly unpacked a small binary package to root using tar
2023-01-05 05:47:35 <kode54> and it replaced my /usr/lib64 with a folder containing the items
2023-01-05 05:47:47 <kode54> this also rendered my /bin/bash inoperable
2023-01-05 05:48:07 <kode54> and then I proceeded to generate dracut images that way
2023-01-05 05:50:28 <psykose> that is a serious xD moment
2023-01-05 05:50:53 <kode54> good thing I know my way around the installer iso
2023-01-05 05:51:05 <psykose> :)
2023-01-05 05:51:46 <kode54> I had to dig google to figure out what would make a dynamic binary emit "not a dynamic executable"
2023-01-05 05:52:38 <kode54> "oh, the loader can't be found"
2023-01-05 05:53:50 <dwfreed> usually it says no such file or directory, so that's different
2023-01-05 05:54:08 <dwfreed> maybe they improved that recently
2023-01-05 05:54:35 <kode54> that's the error that ldd emits if you scan a dynamic binary whose loader is missing
2023-01-05 05:55:12 <kode54> and dracut uses ldd to find the dependencies of binaries it bundles
2023-01-05 05:55:17 <dwfreed> linux also used to say that
2023-01-05 05:55:21 <kode54> oh, hmm
2023-01-05 05:55:36 <dwfreed> via errno after execve
2023-01-05 05:55:36 <kode54> well, new knowledge acquired, and stupid mistakes to hopefully not make again
2023-01-05 05:55:46 <dwfreed> yeah :)
2023-01-05 07:21:18 <uniqNuuid> dracut is slow, mkinitfs is fast
2023-01-05 07:27:50 <infinite[m]> Dracut support zfs
2023-01-05 08:12:19 <bananicorn> Hello, I would like to log to separate files, but it seems that busybox syslogd does not support any configuration
2023-01-05 08:12:38 <ikke> It does
2023-01-05 08:12:45 <ikke> You just need to create the log file
2023-01-05 08:13:35 <ikke> bananicorn: /etc/syslog.conf
2023-01-05 08:13:48 <bananicorn> Oh I read somewhere that this is not supported
2023-01-05 08:13:59 <bananicorn> Thanks for correcting me;
2023-01-05 08:14:18 <bananicorn> Does this also allow logging to a remote server?
2023-01-05 08:14:30 <ikke> bananicorn: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/busybox/busyboxconfig#L1217
2023-01-05 08:15:04 <bananicorn> Already see the answer, thanks!
2023-01-05 08:15:20 <bananicorn> I assume the syntax is like a traditional syslog.conf file
2023-01-05 08:16:29 <ikke> I think so
2023-01-05 08:17:53 <ikke> I have a simple config, but I did follow the default syntax, which worked
2023-01-05 08:18:50 <bananicorn> Thanks, BusyBox never ceases to amaze me!
2023-01-05 08:19:20 <ikke> That can be both positive as negative :)
2023-01-05 08:19:40 <ikke> although amaze is postivie
2023-01-05 08:19:42 <ikke> positive*
2023-01-05 08:20:04 <bananicorn> Hehe positive in most cases, the documentation or lack thereof can be a bit difficult sometimes though 😅
2023-01-05 08:20:22 <bananicorn> Took me a while to understand how to use busybox sendmail properly
2023-01-05 08:21:46 <bananicorn> But its existence did positively amaze me in that case aswell
2023-01-05 09:26:14 <Apachez> bananicorn: so how do you use busybox sendmail properly then? :)
2023-01-05 09:36:16 <bananicorn> Apachez: Here's how I use it currently https://codeberg.org/unicorn/bbmail.sh
2023-01-05 09:36:44 <bananicorn> I made a little script that sends the outputs of my server's daily/weekly/monthly cron jobs to me via mail
2023-01-05 09:37:07 <bananicorn> It's a terrible script that I wrote when I didn't know much of shell scripting, but I hope it is somewhat understandable :P
2023-01-05 09:37:11 <uniqNuuid> i still don't know why the installation on top of minirootfs does not boot... 
2023-01-05 09:38:28 <uniqNuuid> bananicorn: do you have to setup a mail server?
2023-01-05 09:39:01 <bananicorn> uniqNuuid: Not necessarily, it should be able to log into any SMTP server
2023-01-05 09:39:08 <Apachez> bananicorn: nice and thanks!
2023-01-05 09:39:14 <bananicorn> I run my mail server on a different machine and just supply my login details to the script
2023-01-05 09:39:32 <bananicorn> It can connect via STARTTLS or regular TLS
2023-01-05 09:39:39 <Apachez> uniqNuuid: since its hardcoded you could alter it to send the email directly to the SMTP server where the destination mailbox is located
2023-01-05 09:40:07 <uniqNuuid> yeah. send to itself
2023-01-05 09:40:45 <bananicorn> Haven't tried that but don't see why it shouldn't work :)
2023-01-05 09:41:06 <uniqNuuid> do you need a mail client?
2023-01-05 09:41:28 <bananicorn> Well you need to read the mail you received somehow right?
2023-01-05 09:41:45 <uniqNuuid> i mean the sending part
2023-01-05 09:41:48 <bananicorn> sendmail is the SMTP client
2023-01-05 09:42:11 <bananicorn> It connects directly to the server as an STMP client
2023-01-05 09:42:44 <bananicorn> So nothing else required for sending
2023-01-05 09:43:22 <uniqNuuid> cooll
2023-01-05 09:43:54 <uniqNuuid> most of the time i use telegram bot... via the api + self cooked simple sh script
2023-01-05 09:44:53 <bananicorn> That also works I guess, I just preferred to have it via mail since I have that self-hosted and it's fully FOSS :)
2023-01-05 09:45:01 <uniqNuuid> so basically some code to read the stdin and curl to send the data to telegram's api
2023-01-05 09:45:50 <uniqNuuid> yeah, afterall telegram is not safe to store private info. i agree with you 
2023-01-05 10:41:00 <bananicorn> Also a warning regarding busybox sendmail, I have not yet found another distro whose busybox package includes this utility
2023-01-05 10:41:22 <bananicorn> And the regular sendmail seems to have some different behaviour
2023-01-05 10:42:00 <bananicorn> I still find it wonderful to use on my alpine machines since it doesn't require me to install anything
2023-01-05 10:42:14 <Ermine> bananicorn: shouldn't your MTA package include sendmail tailored to its workings?
2023-01-05 10:42:58 <bananicorn> Ermine: Honestly no idea 🙃
2023-01-05 10:43:10 <bananicorn> Just saw that it is not available or works differently in other distros
2023-01-05 10:43:21 <bananicorn> So I thought I should mention it :)
2023-01-05 10:51:42 <EF> On a minimal Alpine system with no X, appart from /var/log what other folders get written too frequently/a-lot by default? 
2023-01-05 10:55:11 <ikke> /tmp possibly 
2023-01-05 10:57:06 <EF> Ok thanks.  /run perhaps too?  My aim is to remove as many writes from the RPi4's SD card as possible.
2023-01-05 11:02:03 <bananicorn> Unless you do a disk install, this should all be happening in RAM, which is the ideal situation for an SBC in my eyes
2023-01-05 11:04:00 <bananicorn> You can use lbu and create partitions if necessary for persistent storage
2023-01-05 11:27:38 <EF> bananicorn: did sys install but MicroSD cards don't seem to last long here.  Thinking of using a iscsi target to host volatile stuff.
2023-01-05 11:31:02 <bananicorn> It might be easier and more reliable to run from RAM and use another more reliable storage medium for persistent data
2023-01-05 11:32:25 <bananicorn> It would also allow you to store the configuration on the other storage medium
2023-01-05 11:32:52 <bananicorn> Which would make a dead SD completely inconsequential for your data :)
2023-01-05 12:18:54 <Fxzx_mic> busybox 1.36.0 had been released.
2023-01-05 12:19:32 <ikke> nmeum: ^
2023-01-05 13:18:56 <nmeum> ikke: there is an MR already in aports :)
2023-01-05 13:19:43 <nmeum> !42868 still doing some testing and need to fix/debug the x86 segfault
2023-01-05 13:39:21 <Fxzxmic> LOL
2023-01-05 13:39:39 <ikke> nmeum: I'd expected you to be on top of it already :)
2023-01-05 14:01:43 <Fxzxmic> I use my own compiled busybox plus Alpine Linux patches. So there is work again.
2023-01-05 14:32:15 <uniqNuuid> let upgrade the lovely bb
2023-01-05 14:32:31 <uniqNuuid> it's bb that made alpine so small
2023-01-05 14:32:40 <uniqNuuid> oh and openrc
2023-01-05 14:33:13 <uniqNuuid> run from RAM? interesting, is 2GB enough for headless server?
2023-01-05 14:34:01 <elly> definitely
2023-01-05 14:35:03 <uniqNuuid> gonna do some test. that's fun
2023-01-05 14:37:52 <elly> my system right now is sitting at 254MB total RAM used so if I did have 2GB of RAM I would have ample margin still
2023-01-05 14:38:35 <elly> shoutout to Xorg for being more than half of that
2023-01-05 14:42:06 <Fxzxmic> I use the "lovely" bb in WSL1
2023-01-05 14:44:25 <Fxzxmic> That's one of the reasons I used my own compiled busybox
2023-01-05 14:58:24 <Fxzxmic> a repository for my own configuration: https://github.com/fxzxmic/Busybox-for-AlpineLinux-on-WSL-Version-1
2023-01-05 16:53:15 <uniqNuuid> whoever wrote setup-disk is a hardcore
2023-01-05 16:53:58 <minimal> a hardcode what?
2023-01-05 16:54:08 <minimal> s/hardcode/hardcore/
2023-01-05 16:54:08 <alpine-meetbot> minimal meant to say: a hardcore what?
2023-01-05 16:54:57 <uniqNuuid> just a hardcore
2023-01-05 16:55:36 <uniqNuuid> i am reading to findout what i was missing when install from minirootfs 
2023-01-05 17:12:57 <uniqNuuid> i think for arm target, alpine better provide a rootfs, then, people just need to configure the bootloader, fstab, it will boot. currently there is no such thing 
2023-01-05 17:14:00 <uniqNuuid> the tarball for generic arm is not a rootfs, it is something like a "boot me, and then i'll replace you"
2023-01-05 17:15:31 <uniqNuuid> i say a rootfs, is something like what archlinux arm provided : https://archlinuxarm.org/platforms/armv8/generic
2023-01-05 17:16:36 <uniqNuuid> one just format, mount, tar -C , chroot, setup bootloader (usually u-boot), edit fstab... then boot. 
2023-01-05 17:16:42 <uniqNuuid> simple and easy.
2023-01-05 17:16:59 <uniqNuuid> no scripts needed, just cp and vi
2023-01-05 17:18:05 <uniqNuuid> + generate initramfs (in chroot)
2023-01-05 17:26:03 <uniqNuuid> this is how i install alpine on aarch64 right now, what i did: 1. install a basic local vm instance 2. mount a usb to any machine (say lan_box) in LAN network 3. from the vm, rsync --options / root@lan_box:/mnt 4. copy over u-boot extlinux, edit fstab, generate initramfs, reboot
2023-01-05 17:26:58 <uniqNuuid> the vm acts like a "rootfs" i am talking about. 
2023-01-05 17:28:19 <minimal> uniqNuuid: you obviously have missed something when trying to use minirootfs as that method should work
2023-01-05 17:28:45 <uniqNuuid> if needed, regenerate the ssh host key, machine-id ... but that's not a big deal, they can be done after reboot
2023-01-05 17:30:13 <uniqNuuid> minimal: yeah, i think it something in setup-disk
2023-01-05 17:30:38 <uniqNuuid> because i did not do that step, i thought i don't need it. 
2023-01-05 17:30:55 <minimal> uniqNuuid: you don't need to use setup-disk if you are taking care of everything yourself
2023-01-05 17:31:14 <minimal> I don't use setup-disk myself in my scripts
2023-01-05 17:33:05 <uniqNuuid> setup-disk : 1676 lines of sh script! oh my god 
2023-01-05 17:34:48 <Ermine> I looked at it and it supports a lot of configurations: raid stuff, apkovl, encryption, etc
2023-01-05 17:34:55 <uniqNuuid> minimal: your script is even crazier to be honest
2023-01-05 17:35:50 <uniqNuuid> ARM SoC does not need such stuff, just one or two partition. fat+ext4 or ext4+ext4 for /boot and /
2023-01-05 17:36:03 <minimal> uniqNuuid: yes but it is supporting more stuff - you don't need to look at my script, simply run it with the appropriate options and then look at the *smaller* script it creates specific to the option you give it
2023-01-05 17:37:24 <Ermine> setup-disk is common for all images
2023-01-05 17:37:31 <minimal> uniqNuuid: no-one is going to write a different setup-disk script for each different architecture, disk layout, etc
2023-01-05 17:38:40 <uniqNuuid> when you provide a rootfs, no setup-* needed... 
2023-01-05 17:39:17 <uniqNuuid> i have dozen machine... i install one alpine, and tar it, and untar to another machine... 
2023-01-05 17:39:33 <uniqNuuid> or rsync the whole '/' over to another machine 
2023-01-05 17:39:41 <minimal> uniqNuuid: a rootfs that runs on various ARM SBCs that will probably have different storage configuration (eMMC vs SDcard vs USB) and different hardware settings?
2023-01-05 17:39:43 <uniqNuuid> then only need a small modification 
2023-01-05 17:40:42 <uniqNuuid> minimal: it does not matter what partition layout, what disk type... because you got to format the disk first, then rsync over ... and then do fstab stuff. 
2023-01-05 17:41:16 <ikke> or you have a script that does all of that
2023-01-05 17:41:45 <ikke> Doing it manually is anoying and error prone
2023-01-05 17:42:17 <uniqNuuid> the setup-* script is pretty good for x86, that's for sure
2023-01-05 17:42:26 <uniqNuuid> arm soc is a different thing.
2023-01-05 17:42:37 <ikke> Yes, but that's because there is very little standardized
2023-01-05 17:43:48 <uniqNuuid> ls
2023-01-05 17:43:57 <uniqNuuid> sorry wrong pane
2023-01-05 17:45:34 <ikke> .
2023-01-05 17:45:37 <ikke> ..
2023-01-05 17:59:12 <uniqNuuid> or, can i start with a static-apk-tools? 
2023-01-05 17:59:22 <uniqNuuid> that's gonna be interesting 
2023-01-05 18:02:02 <minimal> uniqNuuid: that is what my script uses to create the base alpine chroot
2023-01-05 18:02:25 <uniqNuuid> minimal: lol you may want to man blkid, it has a -o value option, then you don't need sed
2023-01-05 18:02:47 <uniqNuuid> blkid -o value -s UUID /dev/sda1
2023-01-05 18:03:11 <minimal> uniqNuuid: are you talking about Busybox blkid or util-linux blkid though?
2023-01-05 18:03:29 <uniqNuuid> i created a test.sh using your script, and am reading the test.sh 
2023-01-05 18:03:50 <uniqNuuid> minimal: busybox blkid does not have a -s, but your script uses the -s 
2023-01-05 18:04:46 <uniqNuuid> oh, i am wrong 
2023-01-05 18:04:54 <uniqNuuid> it has -s but not documented
2023-01-05 18:05:48 <uniqNuuid> oh, no, it ignores any option.... doas busybox blkid -s -x -aaaaaaa -bababbab  /dev/sda1
2023-01-05 18:05:53 <uniqNuuid> still outputs ... 
2023-01-05 18:06:55 <uniqNuuid> minimal: util-linux has plenty option, and you can use -o value to get the value from -s TYPE
2023-01-05 18:10:34 <uniqNuuid> minimal: you are not minimal at all, look at your script, oh my god. i'm not able to write such code until i die 
2023-01-05 18:14:40 <elly> now I want to see this script
2023-01-05 18:15:14 <Ermine> uuidNuniq: please keep your emotions under control.
2023-01-05 18:17:25 <Ermine> And be *constructive*.
2023-01-05 18:18:11 <uniqNuuid> this api is interesting, https://gitlab.alpinelinux.org/api/v4/projects/5/packages/generic/v2.12.9/"${_host_arch}"/apk.static
2023-01-05 18:27:36 <psykose> there's only two kinds of people
2023-01-05 18:27:41 <psykose> those that write scripts
2023-01-05 18:27:48 <psykose> and those that complain about other people's scripts
2023-01-05 18:27:48 <psykose> :p
2023-01-05 18:29:02 <psykose> elly: iirc it was https://github.com/dermotbradley/create-alpine-disk-image/ or something along those lines
2023-01-05 18:29:18 <psykose> does what it says on the tin
2023-01-05 18:29:25 <uniqNuuid> no, i don't complain, i admire
2023-01-05 18:29:39 <psykose> no worries, just making bad jokes :)
2023-01-05 18:30:16 <elly> wow, this is very legible
2023-01-05 18:31:02 <psykose> yep
2023-01-05 18:32:17 <uniqNuuid> i am confident after reading the generated test.sh... he uses a static apk to do everything.... 
2023-01-05 18:34:56 <uniqNuuid> minimal: i know what went wrong, the minirootfs does not have "alpine-base" pkg by default.... 
2023-01-05 18:35:03 <uniqNuuid> and i did not install that ... 
2023-01-05 18:39:28 <scorpion2185[m]1> only firef.? https://wiki.alpinelinux.org/wiki/Category:Web_Browser
2023-01-05 18:42:12 <minimal> elly: yes it is intended to be legible, and there's more useful log output if you use the "--debug" option
2023-01-05 18:43:26 <minimal> uniqNuuid: yes the minirootfs does not contain alpine-base for a specific reason - that package depends on openrc and the minirootfs is not intended to have an init system
2023-01-05 18:43:51 <uniqNuuid> apline-base is a virtual package. nothing in there, just depends 
2023-01-05 18:46:04 <minimal> uniqNuuid: and it is the depends that would be a "problem" for minirootfs if it was included - that would then drag in the openrc package which is not required/wanted for minirootfs
2023-01-05 18:46:37 <uniqNuuid> i understand that part. i mean, i probably have all the depends installed 
2023-01-05 18:46:52 <uniqNuuid> i don't know, but i'll try your method right now 
2023-01-05 18:47:37 <minimal> with minirootfs as well as adding openrc you would also need to add a kernel package and a bootloader as those are also not expected to me in a minirootfs
2023-01-05 18:48:11 <uniqNuuid> that was taken care of already 
2023-01-05 18:48:44 <uniqNuuid> my notes ; https://0x0.st/oR0-.txt
2023-01-05 18:52:31 <scorpion2185[m]1> ikke:  `git init -q && git add . git commit -qm.`  ? where do I run that ?
2023-01-05 18:52:43 <ikke> scorpion2185[m]1: in the builddir
2023-01-05 18:54:22 <scorpion2185[m]1> src? isn't it deleted when you make a pkg?
2023-01-05 18:54:44 <ikke> yes, it's only temporary
2023-01-05 19:02:34 <scorpion2185[m]1> then how can I do those things? if builddir is src
2023-01-05 19:04:06 <psykose> in prepare()
2023-01-05 19:04:20 <psykose> not sure why you need to do that tho
2023-01-05 19:06:37 <scorpion2185[m]1> a patch , I wonder if can just edit /usr/local to /usr
2023-01-05 19:08:27 <psykose> ah
2023-01-05 19:09:38 <scorpion2185[m]1> https://github.com/ReimuNotMoe/ydotool/blob/master/Daemon/ydotool.service-openrc.in
2023-01-05 19:11:09 <scorpion2185[m]1> with command do I fix /usr/local ?
2023-01-05 19:12:08 <scorpion2185[m]1> with what
2023-01-05 19:14:32 <psykose> the /usr/bin/kill doesn't exist either
2023-01-05 19:16:14 <scorpion2185[m]1> ah yes is that pkill?
2023-01-05 19:17:43 <psykose> no, just make it kill
2023-01-05 19:19:25 <scorpion2185[m]1> sed command?
2023-01-05 19:21:14 <scorpion2185[m]1> (to replace them)
2023-01-05 19:22:14 <donoban> wops, [436779.998943] sway[4448]: segfault at 10 ip 00007f51ef6728b8 sp 00007ffe9f10bc58 error 4 in libwlroots.so.11[7f51ef636000+60000] likely on CPU 2 (core 2, socket 0)
2023-01-05 19:24:07 <donoban> it seems reproducible
2023-01-05 19:24:38 <donoban> [436906.643740] sway[12202]: segfault at 10 ip 00007f85156b48b8 sp 00007ffff46a0fe8 error 4 in libwlroots.so.11[7f8515678000+60000] likely on CPU 5 (core 1, socket 0)
2023-01-05 19:26:36 <donoban> maybe it's just a missversion between my host and a container
2023-01-05 19:29:36 <donoban> upgrading it didn't help
2023-01-05 19:44:19 <wart_> Hi folks.  Did an upgrade to 3.17 and running into an odd bug with pulseaudio --- every once in a while i will lose sound in firefox.  pavucontrol still indicates that it is running fine (bars are hopping up and down), and sometimes (but not always) muting and unmuting fixes it
2023-01-05 19:44:23 <wart_> sometimes i have to restart firefox though
2023-01-05 19:44:35 <wart_> curious if this has been experienced before, thanks in advance
2023-01-05 19:44:56 <wart_> should say it happens in mpv too
2023-01-05 19:46:33 <donoban> wart_: no problem here but I use pulsaudio through pipewire
2023-01-05 19:47:44 <wart_> thanks, yeah not using pipewire but i kind of think it is something about how i am starting things
2023-01-05 19:51:48 <donoban> you could give it a try
2023-01-05 19:52:08 <donoban> it's pretty simple to setup
2023-01-05 20:10:01 <uniqNuuid> minimal: it boots now, installation from minirootfs. I have to "touch /fastboot" to skip fsck. 
2023-01-05 20:11:08 <uniqNuuid> it did not boot previously because fsck error, and returns : Filesystems couldn't be fixed
2023-01-05 20:11:38 <uniqNuuid> and openrc hang. so i read /etc/init.d/fsck code, and found it can be skipped if there is a /fastboot file 
2023-01-05 20:19:57 <uniqNuuid> turns out the /etc/init.d/fsck has some options that busybox does not recognize, it needs util-linux's fsck to work. so the solution is to install util-linux
2023-01-05 20:26:34 <uniqNuuid> apk add cmd:fsck.ext4 cmd:fsck.vfat cmd:fsck, which will install e2fsprogs and dosfstools and util-linux # fsck.ext4 has to be installed, or openrc hang due to /etc/init.d/fsck has some options that busybox fsck does not recognize. 
2023-01-05 20:38:16 <elly> I've never seen this cmd:foo syntax before, is that new?
2023-01-05 20:38:25 <ikke> no
2023-01-05 20:38:28 <Habbie> i learned about it a few years ago at least
2023-01-05 20:38:35 <Habbie> it's a very nice feature that i wish apt had
2023-01-05 20:38:36 <elly> well, today I learned
2023-01-05 20:38:38 <Habbie> (dnf/yum also have it)
2023-01-05 20:38:43 <ikke> It's a provides that abuild automatically adds
2023-01-05 20:39:20 <elly> gotcha, cool :)
2023-01-05 20:40:30 <Habbie> ikke, for everything that the package puts in /bin or /usr/bin?
2023-01-05 20:41:08 <ikke> Habbie: https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L1243
2023-01-05 20:41:39 <Habbie> that's the best possible way to answer that question, thank you for that :)
2023-01-05 20:42:10 <uniqNuuid> i never check, does the setup-alpine method installs those tools e2fsprogs and dosfstools and util-linux by default? 
2023-01-05 20:43:19 <uniqNuuid> busybox fsck does not support ext4 at all, it returns 111 err code, but look at the /etc/init.d/fsck, it mentions busybox....
2023-01-05 20:44:00 <uniqNuuid> /etc/init.d/fsck is owned by openrc-0.45.2-r7
2023-01-05 20:44:43 <psykose> fsck is implemented per f
2023-01-05 20:44:44 <psykose> fs
2023-01-05 20:45:10 <uniqNuuid> is this a bug? many services depend on fsck, but they don't check the where the fsck is from... 
2023-01-05 20:45:11 <psykose> so you need dosfs for the fat fsck, e2fs for the ext fsck, and then nothing else implements fsck so the list ends there (/joke but it's mostly correct)
2023-01-05 20:45:22 <psykose> no
2023-01-05 20:47:00 <uniqNuuid> ah, i see, when normal installation, the setup-disk is used to format the disk ... and then will need e2fsprogs and dosfstools
2023-01-05 20:47:42 <uniqNuuid> because i install on top minirootfs, i skipped setup-disk, and the 2 tools was skipped too... 
2023-01-05 20:48:20 <uniqNuuid> then, here comes the fsck hang... finally, i found the root cause!
2023-01-05 20:49:59 <uniqNuuid> ok ok, i declare (aka typeset on some shell), successfully installed from minirootfs!
2023-01-05 20:51:32 <uniqNuuid> what a journey
2023-01-05 20:57:22 <minimal> uniqNuuid: did you not see in the test.sh script it installed the relevant package(s) for per-fstype fsck?
2023-01-05 21:01:13 <minimal> uniqNuuid: https://git.busybox.net/busybox/tree/e2fsprogs/fsck.c#n32
2023-01-05 21:08:02 <uniqNuuid> minimal: i see. so util-linux{,-misc} is not needed. just the 2 filesystem format tools will be enough. 
2023-01-05 21:09:20 <uniqNuuid> minimal: i read your code, but i did not notice the filesystem related pkgs. my fault. i never thought fsck would not work without them... 
2023-01-05 21:11:17 <minimal> uniqNuuid: it would be unrealistic for a single "fsck" command to be written to understand FAT/ext2/ext3/ext4/xfs/zfs/btrfs/etc and any/all other other filesystem types
2023-01-05 21:12:07 <uniqNuuid> yeah, sounds impossible, 
2023-01-05 21:13:05 <uniqNuuid> minimal: your idea of install and configure mkinitfs first and then install the kernel is very cool. learnt 
2023-01-05 21:16:39 <uniqNuuid> apk add alpine-base mkinitfs e2fsprogs dosfstools # fsck need filesystem tools to work
2023-01-05 21:17:01 <uniqNuuid> that's all for the issue.
2023-01-05 21:19:11 <uniqNuuid> after editing mkinitfs.conf, then apk add linux-edge linux-firmware-none . so total 6 base pkgs. so far so cool. holy f
2023-01-05 21:20:26 <minimal> you only need dosfstools because you're using UEFI (which uses a FAT-formatted ESP partition)
2023-01-05 21:21:43 <minimal> though you're using u-boot's UEFI support right? I seem to remember that it supports some non-standard stuff and so *might* support a ext2 ESP partition, not sure
2023-01-05 21:21:56 <uniqNuuid> no, i have /boot and / both ext4
2023-01-05 21:22:41 <uniqNuuid> my u-boot does not use UEFI.  and my mmc can't be formatted into one part, because it has some important data on the mmc
2023-01-05 21:22:51 <uniqNuuid> i need to avoid some blocks. 
2023-01-05 21:23:02 <minimal> ah, it was the "mount" for /sys/firmware/efi/efivars that gave me that impression
2023-01-05 21:23:18 <uniqNuuid> minimal: there's check 
2023-01-05 21:23:21 <minimal> so why install dosfstools is no FAT filesystems?
2023-01-05 21:24:04 <minimal> s/is/if/
2023-01-05 21:24:04 <alpine-meetbot> minimal meant to say: so why install dosfstools if no FAT filesystems?
2023-01-05 21:24:12 <uniqNuuid> ?
2023-01-05 21:24:16 <uniqNuuid>  /boot can be ext4 or vfat 
2023-01-05 21:24:34 <uniqNuuid> some people liked vfat :) 
2023-01-05 21:24:40 <minimal> but you just said you had /boot and / both ext4
2023-01-05 21:24:57 <uniqNuuid> yeah, i'd like to make a generic script for this device 
2023-01-05 21:26:04 <uniqNuuid> not like a universal one, a script just for phicomm n1
2023-01-05 21:26:51 <uniqNuuid> minimal: https://github.com/laris/Phicomm-N1/blob/master/refs/partition-table-format-of-phicomm-n1.md
2023-01-05 21:27:30 <uniqNuuid> strange device, an old 2015 u-boot loader is on mmc, 
2023-01-05 21:27:50 <minimal> there already seems to be an Alpine setup for this device: https://github.com/yangxuan8282/phicomm-n1/releases/tag/alpine
2023-01-05 21:28:10 <uniqNuuid> but mainline kernel support is very complete. it runs mainline pretty well. 4core A54 max 1.5Ghz 2GB RAM, 8GB MMC
2023-01-05 21:28:26 <uniqNuuid> minimal: that's pretty old setup. 
2023-01-05 21:28:36 <uniqNuuid> it needs so many patches. 
2023-01-05 21:29:04 <uniqNuuid> i chainload u-boot v2022.10 from the mmc u-boot, so i don't need any patch for kernel 
2023-01-05 21:32:09 <Guest370> Hello, all
2023-01-05 21:32:43 <Guest370> Is there any explicit help on getting Apline Linux to install on an IBM z system mainframe, running z/VM?
2023-01-05 21:33:19 <uniqNuuid> minimal: i learnt a lot playing with this device. like a lot... never have thought i'd read setup-* code ... and many others... 
2023-01-05 21:33:52 <uniqNuuid> so it's a very good experience. 
2023-01-05 21:34:18 <ikke> Guest370: There are not a lot of people with experience here
2023-01-05 21:34:47 <uniqNuuid> Guest370: are you talking about super computer?
2023-01-05 21:35:13 <ikke> they are mainframes
2023-01-05 21:35:27 <invoked> just when you think you've seen it all
2023-01-05 21:35:34 <uniqNuuid> how many cores do they have
2023-01-05 21:37:46 <Guest370> it's a z15....
2023-01-05 21:38:27 <Guest370> running IBM's mainframe hypervisor z/VM 7.3
2023-01-05 22:04:31 <ikke> Our s390x machines are hosted and maintained by IBM
2023-01-05 22:25:06 <liliace> hi! im trying to install alpine on my laptop. i don't have an ethernet cable and i need the broadcom-wl module to connect to wireless network. im thinking i need to put [broadcom-wl](https://github.com/antoineco/broadcom-wl) into the installer iso so i can `modprobe wl` during installation
2023-01-05 22:25:10 <liliace> i put the directory for broadcom-wl directly under the top level directory of the iso, and packed the iso back using the command here: https://askubuntu.com/questions/625286/how-to-create-uefi-bootable-iso. but when i boot, it shows the grub command prompt instead of tty1 and i don't know what to do from there. what should i do and am i on the right track so far?
2023-01-05 22:31:08 <satmd> not the solution you are looking for: consider putting that on a usb storage device then you don't have to modify the iso
2023-01-05 22:31:17 <satmd> but maybe much simpler
2023-01-05 22:42:18 <liliace> thanks satmd. that is a good idea :) i only have one usb stick though, so if there's a way to do it without it i'd still like to know
2023-01-05 22:45:12 <satmd> store them on a "recovery partition" on the target system using another iso that has networking
2023-01-06 00:10:30 <bananicorn> Hello :) I have a question about PipeWire and MIDI, I have a little MIDI keyboard  that I would like to use with ardour
2023-01-06 00:11:06 <bananicorn> Pipewire is working fine, I installed pipewire-pulse, -alsa and -jack
2023-01-06 00:11:28 <bananicorn> However I still seem unable to use the jack backend, it simply doesn't recognize my MIDI keyboard
2023-01-06 00:11:43 <bananicorn> I can see it in /dev/snd/by-id/ though
2023-01-06 00:11:58 <bananicorn> I get an error that /dev/snd/seq does not exist
2023-01-06 00:12:29 <bananicorn> Which seems to be the ALSA midi sequencer or something of that sort
2023-01-06 00:12:39 <bananicorn> I am not sure what I am missing here, does anybody have a clue?
2023-01-06 00:14:05 <psykose> modprobe snd adds seq iirc
2023-01-06 00:14:08 <psykose> aside from that no idea
2023-01-06 00:14:51 <bananicorn> I ran modprobe snd but I still have no /dev/snd/seq
2023-01-06 00:18:53 <psykose> it should've loaded snd-seq-midi-event or some shit as a dependency
2023-01-06 00:29:10 <bananicorn> https://paste.debian.net/hidden/1c8cfdad/
2023-01-06 00:29:51 <bananicorn> There are some midi related things there
2023-01-06 00:30:09 <bananicorn> I also see snd_seq_device there
2023-01-06 00:31:43 <bananicorn> But no sign of /dev/snd/seq
2023-01-06 00:42:20 <bananicorn> Is the default kernel compiled with CONFIG_SND_SEQUENCER enabled? I found an issue from the pmOS project, and since that's alpine based I wonder if that could be the cause
2023-01-06 00:42:52 <bananicorn> Unfortunately I am a kernel noob so I don't really know where to look
2023-01-06 00:46:48 <bananicorn> Okay I found the file, does CONFIG_SND_SEQUENCER=m mean enabled or something else? :P
2023-01-06 00:47:05 <bananicorn> Ah I see it means it gets built as a module
2023-01-06 00:49:23 <bananicorn> Okay I just had to run modprobe snd-seq
2023-01-06 00:49:30 <bananicorn> Problem solved!
2023-01-06 00:49:38 <invoked> glad we could help :)
2023-01-06 00:52:14 <bananicorn> Hehe it seems most problems solve themselves as soon as I ask, after fruitlessly debugging for hours
2023-01-06 01:18:44 <elly> rubber duck debugging in action!
2023-01-06 01:35:10 <gctwnl> Wo. It's been  ages since I last used IRC. Anyway, if a binary inside a docker container on Alpine Linux coredumps (and ulimit is unlimited inside the container) where are the coredumps saved?
2023-01-06 01:40:52 <psykose_> cwd
2023-01-06 01:42:38 <dwfreed> depends on the value of kernel.core_pattern sysctl
2023-01-06 01:42:47 <dwfreed> but usual default is cwd
2023-01-06 01:50:47 <bananicorn> elly: Wonderful name for this common phenomenon^^
2023-01-06 01:51:32 <bananicorn> And after reading about it, perhaps I should get myself a rubber duck
2023-01-06 02:26:33 <minimal> as opposed to a USB Rubber Ducky? ;-) https://shop.hak5.org/products/usb-rubber-ducky
2023-01-06 06:22:02 <frag> i think i got bluetooth speaker to work through pipewire now, but it still cuts in and out like every second .. could it be my slow computer/rpi4?
2023-01-06 06:23:20 <frag> do you need a sound card to play through bluetoothspeaker?
2023-01-06 06:29:15 <frag> oh, error: M 07:28:15.887687 spa.bluez5.sink.a2 ../spa/plugins/bluez5/a2dp-sink.c:764:a2dp_on_flush_timeout: error reading timerfd: Resource temporarily unavailable
2023-01-06 07:26:52 <uniqNuuid> minimal : https://0x0.st/oRka.sh . installation from minirootfs, done. 
2023-01-06 10:27:43 <jhjacobs[m]> <uniqNuuid> "minimal : https://0x0.st/oRka.sh..." <- this installs a fully working alpine, yes?
2023-01-06 10:29:16 <uniqNuuid> jhjacobs[m]: yes
2023-01-06 10:29:58 <jhjacobs[m]> nice!
2023-01-06 13:11:41 <Fxzxmic> http://lists.busybox.net/pipermail/busybox/2023-January/090079.html
2023-01-06 13:37:17 <Saklad5> Can you rely on getting security updates in stable APK repositories?
2023-01-06 13:38:03 <ptrc> yes, as long as the repository is supported
2023-01-06 13:38:20 <Saklad5> But other updates only go to edge?
2023-01-06 13:38:27 <ikke> so 3.16/community is no longer supported
2023-01-06 13:38:34 <ptrc> yes
2023-01-06 13:38:37 <ikke> but 3.14/main is still supported and still gets security updates
2023-01-06 13:55:55 <uniqNuuid> i am now running traditional sys mode,  can i switch to diskless mode?
2023-01-06 13:56:44 <uniqNuuid> tried setup-lbu, it add stuff in /etc/fstab, that's all.
2023-01-06 13:56:51 <ptrc> you can use lbu to commit what you currently have and replace the root partition with just an empty one with .apkovl file
2023-01-06 13:57:38 <uniqNuuid> ptrc: in cmdline? 
2023-01-06 13:58:02 <uniqNuuid> what does that look like in kernel cmdline
2023-01-06 13:58:17 <ptrc> i think you can just remove `root=*` and nlplug-findfs will find the apkovl file automatically; not quite sure though
2023-01-06 13:58:56 <uniqNuuid> well, what about the data in '/'
2023-01-06 13:59:49 <uniqNuuid> now i have a file in /media/microsd/n2.apkovl.tar.gz
2023-01-06 14:00:10 <ptrc> you'll get only the stuff you add and commit, anything else would not be available
2023-01-06 14:01:44 <uniqNuuid> yeah i did `lbu ci`, i mean if i remove root=* parameter in cmdline, and how does it know where to load the n2.apkovl.tar.gz and my current data in '/'
2023-01-06 14:02:51 <ptrc> https://github.com/alpinelinux/mkinitfs/blob/6a09876be6/nlplug-findfs/nlplug-findfs.c#L881
2023-01-06 14:03:00 <ptrc> nlplug-findfs can scan filesystems in search of the apkovl file
2023-01-06 14:05:36 <uniqNuuid> so it copy all the data in my '/' to ram? and then load apkovl.tar.gz ?
2023-01-06 14:06:03 <ptrc> no, it unpacks the apkovl to ram
2023-01-06 14:06:23 <ptrc> whatever's in your current / that you won't add with lbu to the apkovl will be lost
2023-01-06 14:07:03 <uniqNuuid> is setup-apkcache needed?
2023-01-06 14:07:53 <ptrc> yes, otherwise it wouldn't be able to boot
2023-01-06 14:08:23 <ptrc> because when booting with apkovl, it only keeps the list of packages - the actual package files are installed from the cached .apk files
2023-01-06 14:09:25 <uniqNuuid> setup-apkcache set the cache location, and after that i run apk cache download 
2023-01-06 14:09:54 <uniqNuuid> it says something wrong.
2023-01-06 14:10:04 <ptrc> what exactly does it say?
2023-01-06 14:17:28 <uniqNuuid> WARNING: Ignoring https://dl-cdn.alpinelinux.org/alpine/v3.17/main: No such file or directory
2023-01-06 14:17:39 <uniqNuuid> ptrc: ^
2023-01-06 14:17:41 <ikke> network issue?
2023-01-06 14:18:18 <uniqNuuid> aha... i run apk update, then it works 
2023-01-06 14:18:20 <uniqNuuid> why?
2023-01-06 14:18:56 <ptrc> it was probably a temporary network issue
2023-01-06 14:19:50 <uniqNuuid> https://0x0.st/oRns.txt
2023-01-06 14:19:54 <uniqNuuid> this is the tree in lbu
2023-01-06 14:20:27 <uniqNuuid> now, i am gonna remove the root=* and reboot
2023-01-06 14:22:01 <uniqNuuid> emmm.. should i rebuilt initramfs? 
2023-01-06 14:22:06 <uniqNuuid> looks like it does not boot
2023-01-06 14:22:59 <uniqNuuid> or maybe i need the apkovl= parameter
2023-01-06 14:33:43 <MehmedArif> Hi
2023-01-06 14:33:49 <MehmedArif> Got a problem setting up alpine-linux aarch64 iso on my amd64 host post-install, is this the right place to discuss?
2023-01-06 14:34:08 <raven523> this is the right place for help with alpine
2023-01-06 14:34:45 <MehmedArif> where would I go for help with qemu, wiki says there is a #qemu channel, but there is #debian-qemu without any activity
2023-01-06 14:35:30 <raven523> there is #qemu on libera.chat
2023-01-06 14:35:52 <raven523> there is a #qemu here on OFTC but it's not in the public channel list
2023-01-06 14:38:21 <MehmedArif> okay, my bad. new to irc, my man. 
2023-01-06 14:38:56 <MehmedArif> I thought all the channels would be listed 
2023-01-06 14:39:47 <uniqNuuid> ptrc: what does the kernel cmdline look like in diskless mode?
2023-01-06 14:39:54 <raven523> https://wiki.qemu.org/Links says their official channel is #qemu on OFTC, but I don't know why it's hidden
2023-01-06 14:44:28 <Saklad5> Does a minimal install of Alpine Linux keep its clock synchronized?
2023-01-06 14:46:40 <raspbeguy> hello, is it possible to make abuild create the package in a custom location like makepkg on archlinux ? I tried with "-P /custom/dir" but it failed because "/custom/dir/alpine/x86_64: No such file or directory"
2023-01-06 15:03:00 <minimal> Saklad5: yes my installs of Alpine do keep their clocks synchronised ;-)
2023-01-06 15:04:03 <ikke> Lol 
2023-01-06 15:04:45 <minimal> Saklad5: as to the general case, if you are referring to NTP sync then you'd need to configured NTP client to run. Are you talking about a diskless or Sys-mode run-from-disk installation?
2023-01-06 15:07:30 <minimal> if it is a Sys-mode installation performed via setup-alpine then that should have run setup-ntp (unless it is one of the "exception cases", i.e. running via LXC, or a KVM-based VM) to prompt you which NTP client software to use for sync
2023-01-06 15:37:23 <jhjacobs[m]> so, just to make sure: its better to learn nftables then awall?
2023-01-06 15:41:44 <scorpion2185[m]1> I can't use epiphany
2023-01-06 15:42:26 <uniqNuuid> if your harware does not have rtc... then rc-update add swclock boot
2023-01-06 15:48:52 <minimal> uniqNuuid: using swclock does not keep a system clock synchronised though, it merely sets the date/time to a potentially more "realistic" data/time, you still need something like NTP to sync time
2023-01-06 15:55:11 <scorpion2185[m]1> what's the script  that implements /init inside initramfs?
2023-01-06 16:02:37 <uniqNuuid> minimal: swclock + ntp
2023-01-06 16:11:17 <Ermine> scorpion2185[m]1: /usr/share/mkinitfs/initramfs-init
2023-01-06 16:13:15 <Saklad5> So it turned out Alpine's virtual image didn't setup anything
2023-01-06 16:13:32 <Saklad5> For time synchronization
2023-01-06 16:13:57 <Saklad5> All my servers were like 20 seconds off
2023-01-06 16:14:07 <Saklad5> Well, you live and learn
2023-01-06 16:14:37 <Saklad5> I'm now running Crony as an NTS client on all of them.
2023-01-06 16:14:41 <Saklad5> Chrony*
2023-01-06 16:16:23 <minimal> Saklad5: when you ran setup-alpine did it not prompt you for which NTP software to use?
2023-01-06 16:16:36 <minimal> Saklad5: do these machines have a RTC?
2023-01-06 16:19:07 <Saklad5> I'm certain it did not prompt me
2023-01-06 16:19:14 <Saklad5> Because I definitely would not have skipped that
2023-01-06 16:20:22 <Saklad5> As for an RTC, I'm not sure: I am running three Alpine servers, two on Vultr and one on Linode
2023-01-06 16:20:59 <uniqNuuid> Saklad5: i use busybox ntp
2023-01-06 16:21:36 <Saklad5> That's actually how I realized I wasn't running anything: I was reading the manual for Busybox
2023-01-06 16:22:01 <Saklad5> I noticed it had NTP, then it hit me: nothing of the sort was running on my servers
2023-01-06 16:22:33 <Saklad5> Ironically, Busybox's implementation is unusable in my mind, by virtue of being, well, NTP
2023-01-06 16:22:35 <uniqNuuid> rc-update add ntpd ...
2023-01-06 16:22:37 <Saklad5> With no security
2023-01-06 16:22:52 <Saklad5> https://www.rfc-editor.org/rfc/rfc8915
2023-01-06 16:23:59 <jhjacobs[m]> hmz, nftables it is
2023-01-06 16:24:31 <scorpion2185[m]1> plymouth-set-default-theme -R
2023-01-06 16:24:31 <scorpion2185[m]1> /usr/share/plymouth/themes//.plymouth does not exist
2023-01-06 16:24:41 <Saklad5> I'm not running a firewall at all: I know every single thing I'm running, so it's not actually doing anything
2023-01-06 16:25:10 <jhjacobs[m]> good for you :-) 
2023-01-06 16:25:17 <jhjacobs[m]> i prefer a firewall anyway ;-)
2023-01-06 16:25:55 <Saklad5> If you're concerned about letting a compromised user/program open ports, that's sensible
2023-01-06 16:26:20 <scorpion2185[m]1> Ermine is this line 7 `splashfile=/.splash-ctrl` the background?
2023-01-06 16:27:39 <Ermine> ?
2023-01-06 16:27:57 <Lord> hey small question : when i do "apk list --upgradable" it says "apk: unrecognized option: upgradable"
2023-01-06 16:28:06 <Lord> but "apk list -u" works
2023-01-06 16:28:13 <Lord> is there a typo or else ?
2023-01-06 16:28:27 <Saklad5> Yes
2023-01-06 16:28:36 <Saklad5> You missed an "e"
2023-01-06 16:28:47 <uniqNuuid> lol
2023-01-06 16:28:53 <uniqNuuid> common english typo
2023-01-06 16:29:00 <minimal> Saklad5: I did mentioned that the setup scripts do check for a couple of situations when not to setup NTP, one of those is when "kvm-clock" is used, i.e. it is a KVM-based VM
2023-01-06 16:29:21 <Lord> o___o
2023-01-06 16:29:22 <Saklad5> Under the assumption it'd be handled anyway?
2023-01-06 16:29:30 <raven523> apk list --help says --upgradable is an option
2023-01-06 16:29:34 <raven523> and --upgradeable
2023-01-06 16:29:35 <scorpion2185[m]1> I cannot   -R, --rebuild-initrd   Rebuild initrd (necessary after changing theme) see message above
2023-01-06 16:29:49 <Lord> "  -u, --upgradable      Consider only upgradable packages"
2023-01-06 16:29:52 <Saklad5> Huh
2023-01-06 16:30:01 <Saklad5> Actually, yeah, both work
2023-01-06 16:30:17 <uniqNuuid> https://0x0.st/oRn5.txt
2023-01-06 16:30:57 <Saklad5> And looking at a dictionary, that isn't quite a typo: both spellings are accepted
2023-01-06 16:31:02 <minimal> Saklad5: the only potential issue with using NTS to be aware of is that as it uses certificates then if your machine's clock is very wrong then there's a possibility (however remote) that a NTS connection won't be possible if the wrong date/time is outside of the certificates valid-after & valid-before ranges lol
2023-01-06 16:31:14 <Saklad5> Yeah, I saw that in the spec
2023-01-06 16:31:42 <Saklad5> I'd prefer things to use DANE, but even that would have that issue
2023-01-06 16:31:52 <Ermine> scorpion2185[m]1: did you install plymouth?
2023-01-06 16:32:28 <Saklad5> I think the odds of the clock being that wrong are much lower than the odds of a connection issue or a malicious attack, so NTS is still a no-brainer in the end
2023-01-06 16:33:58 <elly> you could use tlsdate as well
2023-01-06 16:34:07 <elly> which does have code to deal with your system clock being very wrong when you run it
2023-01-06 16:34:24 <Saklad5> Wouldn't that have the exact same issue with the TLS certificate?
2023-01-06 16:35:11 <elly> tlsdate ignores the cert validity window while bootstrapping the connection, except that the cert must be valid at some point after tlsdate's build timestamp
2023-01-06 16:35:20 <Saklad5> NTS could do that too, no?
2023-01-06 16:35:27 <elly> (so you can't grab some old cert from 1995 to use)
2023-01-06 16:35:33 <elly> it could, I don't know if it does or not
2023-01-06 16:35:40 <elly> my point was that tlsdate already does this without you hacking stuff
2023-01-06 16:35:44 <Saklad5> https://www.rfc-editor.org/rfc/rfc8915#section-8.5
2023-01-06 16:36:00 <Saklad5> I'm not hacking anything, I'm running an NTS client and trusting it follows the advice of the spec
2023-01-06 16:36:05 <elly> ahh
2023-01-06 16:36:35 <minimal> yupe, it's more of an issue for RTC-less devices like Raspberry Pis
2023-01-06 16:37:02 <Saklad5> I'm not a fan of the tendency to solve every problem with HTTP
2023-01-06 16:37:20 <minimal> swclock is a kludge to reduce the "wrongness" window (compared to 1st Jan 1970) lol
2023-01-06 16:38:16 <Saklad5> I'm not a subject-matter expert, and RFC 8915 looks sensible to me. I'm just going to use that. More specifically, I settled on using Chrony
2023-01-06 16:38:19 <elly> Saklad5: me either, but it reliably traverses ~all middleboxes, which cannot be said for NTP :(
2023-01-06 16:39:44 <bananicorn> Hello, I have a bit of a confusing mesa situation going on, I have mesa-dri-classic installed, but it does not seem to exist in repos anymore
2023-01-06 16:40:10 <bananicorn> I am running alpine with an older intel iGPU
2023-01-06 16:40:29 <bananicorn> Did something change here over the last few major versions that I missed?
2023-01-06 16:40:44 <Saklad5> @elly That sort of logic feels like surrender
2023-01-06 16:41:42 <elly> Saklad5: well, here we are :P
2023-01-06 16:41:51 <Saklad5> Anyway, setting up Chrony to use NTS with time.cloudflare.com works fine
2023-01-06 16:42:17 <Saklad5> I'd use the default pool (and maybe set up my servers to join it), but it doesn't support NTS
2023-01-06 16:44:05 <uniqNuuid> Saklad5: i use ip
2023-01-06 16:44:10 <Saklad5> IP?
2023-01-06 16:44:19 <uniqNuuid> timeserver yeah. 
2023-01-06 16:44:25 <uniqNuuid> so there is no DNS issue
2023-01-06 16:44:52 <uniqNuuid> what is wrong with bb ntpd? i don't see any problem
2023-01-06 16:44:54 <Saklad5> I cannot conceive of a reason that I'd care about time synchronization while DNS is broken
2023-01-06 16:45:05 <Saklad5> I can very much conceive of people changing the IP address the server is on
2023-01-06 16:45:37 <uniqNuuid> ip + domain
2023-01-06 16:45:49 <uniqNuuid> domain as backuo
2023-01-06 16:46:51 <Saklad5> There are only two good reasons to use an IP address directly: either you are bootstrapping DNS in some capacity, or you are connecting to something that isn't in the DNS
2023-01-06 16:47:10 <Saklad5> There are other reasons, but they aren't good
2023-01-06 16:47:54 <minimal> Saklad5: chrony also supports getting time from (some) hypervisors via /dev/ptp_kvm, /dev/ptp_vmw, or /dev/ptp_hyperv
2023-01-06 16:47:55 <mu12> hey i have a problem connecting to the internet when i boot the alpine live iso 
2023-01-06 16:48:49 <mu12> the problem is that when i activate usb therting on my phone noting happens 
2023-01-06 16:49:29 <Saklad5> @minimal Hey, one of my servers actually has one of those
2023-01-06 16:49:31 <Saklad5> Thanks!
2023-01-06 16:49:55 <Saklad5> NTS is sort of redundant in that scenario, so I'll see if I can use Busybox there after all
2023-01-06 16:51:04 <minimal> Saklad5: I said Chrony support that, not Busybox ntpd. It is used via a chronyd.conf "refclock PHC" entry
2023-01-06 16:51:57 <Saklad5> I know, I just thought I'd check
2023-01-06 16:52:03 <Saklad5> Doesn't look like it, so Chrony it is
2023-01-06 16:53:00 <minimal> which goes back to uuiqNuuid's question "what is wrong with bb ntpd?", nothing really, it just has limited functionality compared to chrony
2023-01-06 16:54:47 <mu12> is this a problem for everyone or just me 
2023-01-06 16:55:06 <Saklad5> Busybox doesn't have to please everyone, and shouldn't
2023-01-06 16:56:07 <Saklad5> Though there's an argument for removing features that have no use case due to being obsolete
2023-01-06 17:29:40 <uniqNuuid> Saklad5: what features?
2023-01-06 17:35:11 <uniqNuuid> is there a guide to write oneshot daemon? it has start command and then exit. and stop command when OS shutdown. but it is not a long living program. how do i write that in openrc
2023-01-06 18:19:18 <jhjacobs[m]> though i wonder, what futures one needs from NTP other then checking and correcting time :) ive been moving away from chrony to bb ntp
2023-01-06 18:40:48 <scorpion2185[m]1> ErmineYes I installed ALL plymouth packages but doc and dev
2023-01-06 18:51:50 <jhjacobs[m]> Alpine works with plymouth?
2023-01-06 18:54:30 <scorpion2185[m]1> i installed it so I guess that it doesn't use it by default
2023-01-06 19:03:15 <jhjacobs[m]> plymouth is on my list of things to get working
2023-01-06 19:04:05 <jhjacobs[m]> been a bit busy with the year's end, haven't had much time to tinker around
2023-01-06 19:04:11 <jhjacobs[m]> need to learn packaging as well :D
2023-01-06 20:24:16 <scorpion2185[m]1> how can I export MOZ_ENABLE_WAYLAND=1?
2023-01-06 20:24:23 <scorpion2185[m]1> .profile?
2023-01-06 20:24:53 <Ermine> add it to firefox desktop file
2023-01-06 20:26:17 <scorpion2185[m]1> and If i use the command line?
2023-01-06 20:30:16 <Ermine> use text editor which works in terminal
2023-01-06 20:32:37 <scorpion2185[m]1> .profile works fine
2023-01-06 20:33:12 <WhyNotHugo_> I can't get gpg to work with a yubikey. I have rw permission on the device nodes, and the key works for u2f on firefox.
2023-01-06 20:33:15 <WhyNotHugo_> gpg: selecting card failed: No such device
2023-01-06 21:07:29 <WhyNotHugo_> Any known quirks with using smartcards with gpg on alpine?
2023-01-06 21:10:10 <ikke> Not aware of any known issues
2023-01-06 21:12:43 <WhyNotHugo_> `gpg --card-status` works as root.
2023-01-06 21:13:06 <WhyNotHugo_> But udev rules are in place and both /dev/hidraw0 and /dev/hidraw1 are writeable by my current user.
2023-01-06 21:13:47 <elly> strace time?
2023-01-06 21:14:43 <psykose_> strace should indeed show why something fails to access since it works for root
2023-01-06 21:15:24 <WhyNotHugo_> Guess I'll have to strace gpg-agent itself
2023-01-06 21:15:53 <elly> yeah
2023-01-06 21:16:13 <elly> I would guess that gpg-agent is trying to open something other than the hidraw device...
2023-01-06 21:22:44 <WhyNotHugo_> Running gpg-agent --use-standard-socket --daemon forks, so I can´t  continue strace...
2023-01-06 21:22:55 <WhyNotHugo_> Running gpg-agent --use-standard-socket --server doesn´t listen on the usual socket, so gpg starts a new agent.
2023-01-06 21:23:57 <WhyNotHugo_> Oh, strace -f
2023-01-06 21:25:16 <psykose_> you can also start it and use -p to attach
2023-01-06 21:25:49 <psykose_> assuming there isn't some timing you have to hit, doesn't matter where the daemons go really
2023-01-06 21:44:49 <WhyNotHugo_> I'm not sure why, but after a reboot it works.
2023-01-06 21:44:59 <WhyNotHugo_> I wish I hadn´t  spent 20 minutes stracing a dead end :c
2023-01-06 21:56:40 <elly> well, glad it works now
2023-01-06 22:02:58 <WhyNotHugo_> Thanks
2023-01-06 22:22:59 <Ermine> uuidNuniq: redefine start() and stop() functions in your initd
2023-01-06 22:42:33 <WhyNotHugo_> I have `iface wlan0 inet dhcp` in `/etc/network/interfaces`. When booting, the system pauses until wifi is up an dhcp replies. Can I make networking initialise in the background while the rest of the system continues booting?
2023-01-06 22:49:15 <psykose_> https://ptrcnull.me/posts/openrc-async-services/ lost count of how much i've linked this
2023-01-06 23:26:59 <gctwnl> I am building a dovecot/postfix containers based on alpine. This works. Host OS is Ubuntu. Ubuntu mail group is god 8. Alpine gid 8 is mem. Alpine gid 12 is mail. The configs for postfix and dovecot work with the named group mail. Would it be really problematic if I removed group 8 and 12 from Alpine during docker build and recreate group 8 as mail? I kind oif think that is bad.
2023-01-06 23:27:02 <gctwnl> But I was still wondering.
2023-01-07 01:03:56 <frag> 'apk info' lists all packages installed?
2023-01-07 01:07:19 <psykose_> ye
2023-01-07 02:39:52 <WhyNotHugo_> Is the order of `features=` for mkinitfs significant?
2023-01-07 02:41:10 <elly> no, they are all just packed into the initfs I think
2023-01-07 03:22:29 <Armote[m]> if a luks mounted device is accidentally yanked out, often times it remains in /dev/mapper and any attempts to close it fail because it's 'busy' when it no longer exists and even its mount point was succefully unmounted. is there an way to force it? Command failed with code -5 (device already exists or device is busy).
2023-01-07 03:22:54 <Armote[m]> lsof returns nothing
2023-01-07 03:24:57 <psykose_> perhaps has more ideas https://askubuntu.com/questions/429612/device-mapper-remove-ioctl-on-luks-xxxx-failed-device-or-resource-busy
2023-01-07 03:27:08 <Armote[m]> psykose_: lsof hangs for a while then exits having printed nothing
2023-01-07 03:56:53 <Armote[m]> with every instance of unshare (mount) that i closed lsof returned a little faster. after i closed them all the problem was solved. it's sad that the only useful information came from the timing of lsof lol. afterward i found the solution referenced: https://forums.gentoo.org/viewtopic-t-1116050-start-0.html
2023-01-07 04:25:19 <uniqNuuid> ptrc: minimal , after some research, i made it, switch from sys to diskless mode : https://0x0.st/oRhU.txt
2023-01-07 04:43:25 <uniqNuuid> then there is a question, if i made some changes in diskless mode, and lbu commit, and then i switch to sys mode by modifying the kernel parameters, that changes made by lbu won't be in sys mode. 
2023-01-07 04:44:14 <uniqNuuid> i think it needs tar -xz -f apkovl.tar.gz -C /  to make it happen. 
2023-01-07 04:49:35 <minimal> uniqNuuid: why would you switch back and forth between diskless and Sys?
2023-01-07 04:51:52 <uniqNuuid> minimal: in case i have a lots of packages and ram is not enough? 
2023-01-07 04:52:14 <minimal> so use Sys then from the start
2023-01-07 06:43:05 <uniqNuuid> the bb chpasswd is confusing
2023-01-07 06:43:23 <uniqNuuid> chpasswd -e -c sha256 returns help message 
2023-01-07 06:43:35 <uniqNuuid> looks like -e can be used with -c together 
2023-01-07 06:43:42 <uniqNuuid> s/can/can not/g
2023-01-07 06:43:42 <alpine-meetbot> uniqNuuid meant to say: looks like -e can not be used with -c together 
2023-01-07 07:36:06 <uniqNuuid> is bb 1.3.6 on the way?
2023-01-07 07:36:39 <psykose_> busybox 1.3 is from 2007
2023-01-07 07:36:51 <psykose_> (obviously yes)
2023-01-07 07:37:29 <uniqNuuid> s/1.3.6/1.36/g
2023-01-07 07:37:29 <alpine-meetbot> uniqNuuid meant to say: is bb 1.36 on the way?
2023-01-07 07:38:40 <uniqNuuid> diskless mode made /etc/fstab in 600 mode
2023-01-07 07:38:57 <uniqNuuid> non-root can't read. 
2023-01-07 07:39:21 <uniqNuuid> is that a bug or feature
2023-01-07 07:40:22 <psykose_> probably neither
2023-01-07 07:57:50 <wcf> hello
2023-01-07 07:57:54 <uniqNuuid> https://0x0.st/oRFZ.png
2023-01-07 07:58:28 <uniqNuuid> i made /tmp 80% of RAM
2023-01-07 07:58:42 <wcf> NIU B
2023-01-07 07:59:38 <wcf> 这是个字符画吗
2023-01-07 07:59:43 <wcf> 很神奇
2023-01-07 07:59:45 <wcf> 很好
2023-01-07 08:01:38 <uniqNuuid> feel so much better without systemd and glibc!
2023-01-07 08:02:09 <wcf> 我大概无法看懂你说的话
2023-01-07 08:19:07 <psykose> hmm
2023-01-07 08:19:31 <psykose> is there a way to check if something is added to any runlevel other than 'does file exist in /etc/runlevels/'
2023-01-07 08:19:54 <psykose> (to detect if some service is enabled)
2023-01-07 08:21:12 <ikke> rc-status?
2023-01-07 08:22:12 <psykose> how?
2023-01-07 08:22:34 <psykose> if it's rc-status something | grep then it's not really different than find in /etc/runlevels, but i guess
2023-01-07 09:17:15 <ikke> not sure if an improvement, but rc-status has -f ini
2023-01-07 09:18:01 <psykose> hmm, might be
2023-01-07 09:18:54 <psykose> well, doesn't have boot and whatnot, and otherwise is the same as -a
2023-01-07 09:18:56 <ikke> I have a zabbix agent plugin that uses that
2023-01-07 09:19:11 <ikke> -f ini -a
2023-01-07 09:22:41 <psykose> i only now just realised it /is/ the same thing, the -f ini means init format
2023-01-07 09:22:43 <psykose> ini*
2023-01-07 09:22:55 <psykose> smh my head
2023-01-07 09:23:01 <psykose> but yeah that's good to know, thank you
2023-01-07 10:06:45 <WhyNotHugo_> Is there any way to ensure that kms kicks in before the cryptsetup prompt? Currently it's kicking in and then flickers when the resolution changes.
2023-01-07 10:10:07 <frag> upgraded to 3.17, my ash functions in .profile no longer loaded it seems?
2023-01-07 10:10:14 <psykose> i don't actually know what "kms kicks in" is (though i do know what you're referring to), but iiuc it would require some change in mkinitfs
2023-01-07 10:11:03 <psykose> frag: do you still have ENV= set, etc
2023-01-07 10:11:37 <frag> psykose: nope, haha, nice guess!
2023-01-07 10:11:45 <psykose> :)
2023-01-07 10:35:30 <Fxzxmic> Is libnss_* already included in the musl libc?
2023-01-07 10:38:22 <psykose> musl doesn't have any libnss_* support
2023-01-07 11:07:41 <Fxzxmic> Busybox said this in USE_BB_PWD_GRP:If you leave this disabled, busybox will use the system's password and group functions. And if you are using the GNU C library (glibc), you will then need to install the /etc/nsswitch.conf configuration file and the required /lib/libnss_* libraries in order for the password and group functions to work. This generally makes your embedded system quite a bit larger.
2023-01-07 11:08:37 <Fxzxmic> but there isn't any libnss_* in Apline Linux
2023-01-07 11:08:47 <ikke> musl does not use nsswitch
2023-01-07 11:09:53 <ikke> "If you are using the GNU C library"
2023-01-07 11:09:58 <ikke> that does not apply to alpine linux
2023-01-07 11:11:54 <Fxzxmic> Oh, I think that the implementation of libnss_* is included in the musl before.
2023-01-07 11:12:59 <psykose> no, musl has never implemented nsswitch
2023-01-07 11:13:43 <psykose> what that toggle does is define whether the standard user functions that access shadow (*pwent, *pwnam, etc) go through the libc or are implemented in busybox itself
2023-01-07 11:14:02 <psykose> it has nothing to do with nss at all, just those on glibc make use of nss
2023-01-07 11:15:17 <Fxzxmic> Sorry for saying the wrong thing, what I want to say is: I used to think that the implementation of libnss was included in musl.
2023-01-07 11:15:21 <psykose> but since busybox needs a libc to do anything at all, and musl does not have an external nss component, there's no benefit to using the bb implementation
2023-01-07 11:15:51 <Ermine> There's an alternative of nsss, but it requires rebuilding busybox and configuring the nsss daemon
2023-01-07 11:17:36 <Fxzxmic> OK, thanks
2023-01-07 11:22:17 <WhyNotHugo_> Comments are not allowed on /etc/apk/world?
2023-01-07 11:24:27 <psykose> nope, and it gets sorted on like every operation
2023-01-07 11:44:29 <Fxzxmic> In fact, I am rebuilding my own busybox, so that's why I am studying these.
2023-01-07 12:06:15 <Eragon> Hi, I'm new here, is it okay to send a new package on the mailing list for aport ? I've seen the message about the gitlab integration not working so I was wondering if simply sending the patch is enough.
2023-01-07 12:07:09 <psykose> not working means you can't do it
2023-01-07 12:08:00 <psykose> but looking at what you sent, i'll deny it based on the just-forked-other-piece-of-software
2023-01-07 12:11:04 <Eragon> Is forgejo too young ?
2023-01-07 12:11:42 <psykose> let's just say that every time i've seen one of these 'some-people-forked-some-project-over-disagreement' it usually just died after a few months
2023-01-07 12:11:52 <psykose> in this case i have more hope (afaik codeberg will maintain it), but
2023-01-07 12:12:08 <psykose> idk, i'd wait a bit, and then probably replace gitea with it entirely? makes more sense to me
2023-01-07 12:13:06 <psykose> the audacity/tenacity thing was a tad similar and that was quite the experience
2023-01-07 12:14:31 <Eragon> Oh, ok. Could you help me to simply use the package that I made ? I would like to try to replace my gitea with forgejo but I don't understand how I am suposed to use the apk after building it.
2023-01-07 12:19:17 <psykose> hmm
2023-01-07 12:19:33 <psykose> you just install it, if you built it with abuild then you would have a ~/packages/testing
2023-01-07 12:19:50 <psykose> so you can add /home/you/packages/testing to /etc/apk/repositories and it becomes installable
2023-01-07 12:24:59 <Eragon> Oh, I forgot the /testing that's why it wasn't working. Thanks
2023-01-07 12:39:44 <frag> is there a recommended hardware/laptops for alpine (and or just linux in general)?
2023-01-07 12:40:55 <ptrc> as for laptops, you can get basically anything that supports linux natively and alpine should work fine
2023-01-07 12:41:24 <ptrc> generally the only thing to avoid is nvidia graphics, as alpine doesn't support the proprietary driver and nouveau is... fun and quirky
2023-01-07 12:41:38 <ikke> broadcom is less optimal as well
2023-01-07 12:42:25 <frag> ikke: you _should_ have one with broadcom?
2023-01-07 12:42:43 <ikke> no, avoid it :)
2023-01-07 12:42:56 <frag> okok :]
2023-01-07 12:47:41 <psykose> intel wifi, any igpus
2023-01-07 12:47:44 <psykose> not much else to pick
2023-01-07 12:49:22 <ikke> wifi is most of the time easy to replace though
2023-01-07 12:49:44 <ikke> although nowadays with these unibodies, that might no longer be true
2023-01-07 12:49:48 <frag> psykose: should have or should not have?!
2023-01-07 12:49:59 <ikke> intel wifi is fine
2023-01-07 12:50:04 <frag> kinda want a wired net tbh :>
2023-01-07 12:50:21 <frag> but also thin and fanless, jaja..
2023-01-07 12:50:52 <ptrc> ikke: iwlwifi can be a really flaky driver sometimes tbh
2023-01-07 12:51:00 <ptrc> but that's just my experience
2023-01-07 12:51:14 <psykose> frag: should
2023-01-07 12:52:18 <psykose> i mean there's really three vendors isn't there
2023-01-07 12:52:31 <psykose> broadcom is bad enough even their ethernet is unusable sometimes
2023-01-07 12:52:51 <psykose> qualcomm is found in.. nothing
2023-01-07 12:52:54 <psykose> and intel is the rest
2023-01-07 12:52:58 <ikke> realtek?
2023-01-07 12:53:03 <psykose> but yeah if it's standard m.2 wifi key it's easy to replace
2023-01-07 12:53:05 <psykose> ah yes realtek too
2023-01-07 12:55:19 <frag> this good? https://system76.com/laptops/galago
2023-01-07 12:56:22 <psykose> everything from s76 is probably fine as it's linux primary
2023-01-07 12:56:36 <psykose> fwupd can even make it easier to update the firmware and they guarantee it
2023-01-07 12:57:55 <psykose> smh no hidpi tho
2023-01-07 12:57:59 <frag> https://linuxpreloaded.com/
2023-01-07 13:01:21 <frag> isnt fn-key (on thinkpads) where the left-ctrl key should be, very annoying?
2023-01-07 13:01:50 <ikke> yes, though you can switch it in the bios
2023-01-07 13:02:05 <frag> oh!  thats better
2023-01-07 13:02:19 <psykose> don't think you can always do that but probably
2023-01-07 13:02:43 <psykose> ah thinkpad
2023-01-07 13:02:43 <psykose> yes
2023-01-07 13:03:25 <psykose> the keyboard i got has programmable firmware and the ability to merely flash whatever layout i want is quite eye opening
2023-01-07 13:03:51 <frag> psykose: is that a laptop kb?
2023-01-07 13:04:01 <psykose> nah, external, so unrelated
2023-01-07 13:04:08 <psykose> but it makes me realise i wish more things had that
2023-01-07 13:04:23 <psykose> especially laptops that you can't change the keyboard as much physically
2023-01-07 13:05:40 <frag> mhmhm, i have a nyquist split kb with qmk too :] .. but usually use a bt logitech k380 (from the sofa haha)
2023-01-07 13:11:49 <minimal> WhyNotHugo_: add your KMS module to the bootloader's "modules=" list?
2023-01-07 13:18:32 <minimal> uniqNuuid: why would you expect "chpasswd -e -c sha256" to work together? 1 option tells it to use sha256 to encrypt passwords, the other option tells it you are giving it *already* encrypted passwords...
2023-01-07 13:29:45 <WhyNotHugo_> psykose: Hopefully you'll be linking it a bit less in future. https://wiki.alpinelinux.org/wiki/OpenRC#Preventing_slow_services_from_delaying_boot
2023-01-07 13:30:16 <uniqNuuid> minimal: how to use -c ?
2023-01-07 13:30:50 <WhyNotHugo_> minimal: I have kms in my mkinitcpio features. I'm using an UKI, not entirely sure if it interacts with video before KMS kicks in.
2023-01-07 13:32:01 <uniqNuuid> i see. chpasswd -c SHA256 <<<"user:PLAIN_PW"  , chpasswd -e "user:ENCRIPTED_PW"
2023-01-07 13:32:17 <minimal> WhyNotHugo_: mkinitcpio? you mean mkinitfs? if you add the relevant DRM module to the cmdline "modules=" then it will be loaded before LUKS password prompt
2023-01-07 13:34:19 <uniqNuuid> i still don't understand the "modules=" option, since the modules are in initramfs, aren't all of the modules loaded already?
2023-01-07 13:35:14 <minimal> uniqNuuid: modules have to be contained in the initramfs for it to be *possible* to load them, but their presence in initramfs does not mean that will be loaded
2023-01-07 13:35:21 <minimal> s/that/they/
2023-01-07 13:35:21 <alpine-meetbot> minimal meant to say: uniqNuuid: modules have to be contained in the initramfs for it to be *possible* to load them, but their presence in initramfs does not mean they will be loaded
2023-01-07 13:35:45 <psykose> WhyNotHugo_: that's awesome, thanks :)
2023-01-07 13:36:04 <uniqNuuid> that's the reason i ask. minimal 
2023-01-07 13:36:14 <uniqNuuid> 'but their presence in initramfs does not mean they will be loaded'
2023-01-07 13:36:34 <minimal> ask what?
2023-01-07 13:38:11 <uniqNuuid> modprobe -a $(echo "$KOPT_modules $rootfstype" | tr ',' ' ' ) loop squashfs simpledrm
2023-01-07 13:38:32 <uniqNuuid> the source code shows by default if nothing is set in cmdline, 3 modules will be loaded
2023-01-07 13:38:48 <minimal> yes, so what is your question?
2023-01-07 13:39:22 <uniqNuuid> modules are selectively loaded, not blindly loaded
2023-01-07 13:39:37 <minimal> that's a statement, not a question
2023-01-07 13:40:00 <uniqNuuid> that's a statment yes, after looking at the source
2023-01-07 13:40:48 <uniqNuuid> also. /etc/modules get loaded too. 
2023-01-07 13:41:17 <minimal> yes, another statement, I'm still confused as to what your question is
2023-01-07 13:41:35 <uniqNuuid> 'aren't all of the modules loaded already'
2023-01-07 13:41:42 <uniqNuuid> that's was my question. 
2023-01-07 13:41:53 <minimal> and I answered that already saying that no they are not
2023-01-07 13:42:24 <uniqNuuid> yeah got it now. 
2023-01-07 13:42:36 <minimal> otherwise if your initramfs included, for example, ethernet modules for 10 different cards would you expect all 10 to be loaded automatically?
2023-01-07 13:43:15 <uniqNuuid> some distro do it like that
2023-01-07 13:43:36 <uniqNuuid> load all the module in initramfs
2023-01-07 13:43:48 <minimal> load multiple modules for different makes of the same type of hardware?
2023-01-07 13:44:15 <WhyNotHugo_> minimal: Yes, mkinitfs. So do I add `kms` or `amdgpu` to `module=`?
2023-01-07 13:44:39 <minimal> so they might load 10+ network drivers? 20+ sata drivers? 10+ wifi drivers? multiple graphic card drivers? etc
2023-01-07 13:45:10 <uniqNuuid> during initramfs generating, they already sorted the modules needed for the current hardware, archlinux for instance, has a autodetect HOOK
2023-01-07 13:45:46 <minimal> WhyNotHugo_: you add the relevant DRM module for your graphics card, so amdgpu in your case
2023-01-07 13:46:17 <uniqNuuid> minimal: https://0x0.st/oRCy.txt
2023-01-07 13:46:35 <minimal> uniqNuuid: ok, but you're assuming a situation where the modules present in the initramfs have been tailored for a machine machine/machines
2023-01-07 13:47:47 <minimal> uniqNuuid: Alpine's mkinitramfs doesn't do that. For other distros that can do that I assume it is configurable (can be enabled and disabled) as otherwise how would you create a bootable USB stick to be used on multiple different types of machines
2023-01-07 13:48:01 <minimal> ?
2023-01-07 13:51:01 <minimal> uniqNuuid: so autodetect is an *optional* feature of some initramfs creation tools (i.e. dracut from memory)
2023-01-07 13:53:59 <WhyNotHugo_> minimal: archlinux's mkinitcpio also has an "autodetect" option was trims modules not in use in the current system.
2023-01-07 13:54:09 <minimal> uniqNuuid: booster (packaged for Alpine) also supports creating initramfs tailored for particular machines
2023-01-07 13:54:31 <minimal> mkinitfs does not however
2023-01-07 13:54:53 <Apachez> thats why floppyfw could fit kernel + qos + busybox + firewall + whatelse on a 1.44MB floppy while the ubuntu kernel alone 140MB or so in size
2023-01-07 13:55:19 <uniqNuuid> minimal: another question, i included dwc2 dwc3 usb modules in initramfs, but i did not specify them in "modules=", how did that get modprobed during init?
2023-01-07 13:55:59 <psykose> mdev etc
2023-01-07 13:56:43 <psykose> a device manager coldplug discovers devices and loads modules
2023-01-07 13:56:45 <minimal> uniqNuuid: by nlplug-findfs AFAIK
2023-01-07 13:58:18 <uniqNuuid> so "modules=" is used to set modules that can't be probed by mdev or nlplug-findfs? 
2023-01-07 13:58:46 <minimal> psykose: nlplug-findfs does a coldplug
2023-01-07 13:59:05 <ikke> What's a coldplug?
2023-01-07 13:59:11 <uniqNuuid> what does nl stand for in nlplug?
2023-01-07 13:59:16 <ikke> netlink
2023-01-07 13:59:41 <minimal> ikke: one that hasn't reached room temp yet? ;-)
2023-01-07 13:59:42 <uniqNuuid> netlink , fs, sounds not related... 
2023-01-07 14:00:32 <psykose> has nothing to do with networking
2023-01-07 14:01:15 <psykose> it's probably the worst named interface in the entire kernel
2023-01-07 14:01:18 <psykose> nothing worse comes to mind
2023-01-07 14:01:52 <Ermine> It works over socket huh
2023-01-07 14:02:15 <ikke> "What started as more effective means to relay and modify network related information has become a generic kernel and userspace communications fabric via NETLINK_GENERIC."
2023-01-07 14:03:05 <minimal> all I know is that when I worked on the MR to get the initramfs' init to load USB keyboard modules that nlplug-findfs triggered their loading rather than needing an explicit check-and-load to be added, so it appears that nlplug-findfs loads more than just storage-related modules
2023-01-07 14:03:55 <ikke> yeah, it gets info about all kinds of devices
2023-01-07 14:05:14 <minimal> it also resulted in the scenario from Alpine 3.17 onwards that the graphics card specific DRM module is loaded by the initramfs' init (as well as generic simpledrm) whereas in previous releases the graphics card specific DRM module was not loaded until /etc/init.d/hwdrivers ran later in boot
2023-01-07 14:09:35 <WhyNotHugo_> Where can I find some documentation for inittab? I want to figure out what `::once:/sbin/openrc async` mean exactly.
2023-01-07 14:13:19 <minimal> WhyNotHugo_: openrc source code?
2023-01-07 14:14:43 <minimal> if you want to understand the "openrc async" portion, busybox docs/source code for the inittab file
2023-01-07 14:14:58 <Ermine> minimal: alpine uses bb init which launches OpenRC, doesn't it?
2023-01-07 14:15:18 <minimal> Ermine: yes, that's why I mentioned bb as well
2023-01-07 14:15:31 <Ermine> ah, dumb me
2023-01-07 14:20:37 <psykose> this random netlink mention led me to a path of removing a random legacy libnl
2023-01-07 14:20:41 <psykose> sometimes reading is helpful
2023-01-07 14:21:19 <minimal> down the rabbit hole we all go lol
2023-01-07 14:34:22 <vkrishn> could someone pls provide the offline video, links seems broken here, https://alpinelinux.org/conf/ ?
2023-01-07 14:35:26 <ikke> it's kind of gone
2023-01-07 14:36:19 <vkrishn> history now a mystery
2023-01-07 14:42:58 <vkrishn> Ariadne_: ?? ^ 
2023-01-07 15:26:23 <scorpion2185[m]1> is there some log of waht package is removed /installed ?
2023-01-07 15:29:22 <ikke> APKv3 will have that
2023-01-07 15:32:49 <scorpion2185[m]1> I installed firefox and it removed some things like nmcli 😑
2023-01-07 15:34:38 <scorpion2185[m]1> plymouth-set-default-theme -R /usr/share/plymouth/themes//.plymouth does not exist
2023-01-07 15:34:38 <scorpion2185[m]1> I installed all plymouth packages but doc and dev
2023-01-07 15:35:51 <psykose> why would it exist
2023-01-07 15:38:35 <scorpion2185[m]1> I set a theme for it and the -help says that I need to run that
2023-01-07 15:40:05 <scorpion2185[m]1>   -R, --rebuild-initrd   Rebuild initrd (necessary after changing theme)
2023-01-07 15:40:30 <minimal> using invisible theme? ;-)
2023-01-07 15:42:49 <scorpion2185[m]1> the problem is how do I add osk-sdl to init script? theme is not important
2023-01-07 15:55:10 <scorpion2185[m]1> is there a way to set an image when it is booting then?
2023-01-07 16:06:38 <Ermine> Refer any osk-sdl questions to PmOS people. They do develop it.
2023-01-07 16:07:54 <scorpion2185[m]1> alpine devs could add it too as a package
2023-01-07 16:09:28 <ikke> It's still PmOS that develops it
2023-01-07 16:09:39 <ikke> So questions on how to use it should be referred to them
2023-01-07 16:13:39 <scorpion2185[m]1> I asked in osk-sdl room. they told me to use pmos
2023-01-07 16:14:12 <ikke> So it's not meant to be used outside of pmos
2023-01-07 16:14:24 <ikke> So if you want to use it outside of pmos, you have to figure out how to do it yourself
2023-01-07 16:15:47 <psykose> i think i've said that like twice
2023-01-07 16:18:08 <Newbyte> ikke: it's not necessarily specific to pmOS (e.g. PureOS uses osk-sdl too)
2023-01-07 16:19:09 <Newbyte> but yes, getting it working with Alpine's initfs setup is beyond what we support
2023-01-07 17:12:11 <WhyNotHugo_> powerctl indicates that sleeping into the `standby` state is not supported on my setup. Any ideas on how to determine WHY this is not supported?
2023-01-07 17:22:19 <minimal> WhyNotHugo_: what does "cat /sys/power/state" show?
2023-01-07 17:53:17 <WhyNotHugo_> minimal: freeze mem disk
2023-01-07 17:57:00 <WhyNotHugo_> Does osk-sdl use KMS/DRM?
2023-01-07 17:57:48 <minimal> WhyNotHugo_: so there's no standby there which is why powerctl gives you that error
2023-01-07 17:59:13 <WhyNotHugo_> minimal: My question is: why is there no standby?
2023-01-07 17:59:26 <WhyNotHugo_> Is it a hardware limitation? If so, which component?
2023-01-07 17:59:39 <WhyNotHugo_> Or am I missing some software?
2023-01-07 18:03:58 <minimal> WhyNotHugo: that depends on your computer - standby support is optional
2023-01-07 18:04:19 <minimal> either it may not support it or there may be a setting in the BIOS/UEFI that affects it
2023-01-07 18:04:55 <minimal> https://www.kernel.org/doc/html/v5.0/admin-guide/pm/sleep-states.html#standby
2023-01-07 18:07:28 <minimal> WhyNotHugo: try dmesg | grep "PM: " for a list of S-states supported
2023-01-07 18:08:11 <minimal> if S1 is not in the list then it does not support standby (for whatever reason)
2023-01-07 18:24:10 <WhyNotHugo_> ACPI: PM: (supports S0 S3 S4 S5)
2023-01-07 18:24:29 <minimal> there you go, no S1 and therefore no standby
2023-01-07 18:37:46 <Armote[m]> where can the startup scripts for alpine be found? mainly the one that executes mount.
2023-01-07 18:42:46 <scorpion2185[m]1> ikke debian/arch linux use it too, it is not specific to pmos and pmos is alpine.
2023-01-07 18:43:31 <Ermine> Armote[m]: /etc/init.d
2023-01-07 18:46:33 <scorpion2185[m]1> what are the apps that aren't avaiblable on alpine? like proprietary nvidia-driver
2023-01-07 18:49:58 <Ermine> { unavailable apps } = { all apps } - { apps available on alpine }
2023-01-07 18:54:14 <minimal> Ermine: { unavailable apps } = { { all apps } - { apps available on alpine } LIKE nvidia-driver }  ;-)
2023-01-07 18:54:46 <Ermine> minimal: XD
2023-01-07 21:54:29 <WhyNotHugo_> If I want to handle the power button from my user session, I think I need to use the acpid package, right? What starts busybox's acpid? It's not an openrc service, but I see it running.
2023-01-07 22:01:50 <WhyNotHugo_> Okay, so deleting /etc/acpi/PWRF/00000080 results in busybox-acpi ignoring the power button as expected (e.g.: not shutting the system down).
2023-01-07 22:02:11 <WhyNotHugo_> But while sway detects the power button as an input, it emits no event when the button is pressed.
2023-01-07 22:15:25 <bleb> zfs stopped working after upgrade + reboot again
2023-01-07 22:15:29 <bleb> modprobe: ERROR: could not insert 'zfs': Exec format error
2023-01-07 22:15:46 <bleb> any ideas?
2023-01-07 22:24:39 <bleb> huh, apk -U upgrade && reboot has fixed it
2023-01-08 01:19:55 <WhyNotHugo_> Any ideas what optional dependency is required for firefox to show notifications usign the notifiacation-daemon?
2023-01-08 01:20:00 <WhyNotHugo_> The daemon itself is up and running
2023-01-08 01:21:41 <WhyNotHugo_> Ah, it's `libnotify`
2023-01-08 01:22:02 <bl4ckb0ne> a notification client maybe
2023-01-08 03:20:46 <Ecce> Package newbie here does alpines aport system synonymous or based on a ports collection (https://en.m.wikipedia.org/wiki/Ports_collection). Functionally they seem identical to me with apkbuild. Wasnt sure if i was missing anything. Reading up on the ports concept in an article. Thanks
2023-01-08 03:22:16 <Ecce> Correct me if im wrong but this system offers better flexibility than systems like apt?
2023-01-08 03:23:08 <raven523> it doesn't offer the built-in flexibility that FreeBSD Ports and Gentoo's Portage have
2023-01-08 03:23:20 <raven523> you have to manually edit APKBUILDs to change options
2023-01-08 03:23:37 <Ecce> Thanks, is that the only downside?
2023-01-08 03:24:01 <Ecce> The apkbuilds seems straightforward enough the ones i saw 
2023-01-08 03:24:03 <raven523> it also requires building binary packages, which can take up space
2023-01-08 03:25:03 <Ecce> Ok seems to me they kind of mimicked ports but with slight downsides. Appreciate the info
2023-01-08 03:25:06 <raven523> I don't remember how well abuild (the main build script used with aports) can handle building dependencies
2023-01-08 03:25:37 <raven523> the vast majority of alpine users don't use aports directly. Instead they just use apk with the binary repos
2023-01-08 03:25:54 <Ecce> I think it does it automatically if you do abuild -r. But that might just be dling deps
2023-01-08 03:26:40 <Ecce> Makes sense . The article was saying how ports were great and stuff so i wanted to see if we had that
2023-01-08 08:15:08 <uniqNuuid> i'd like to use zstd to generate initramfs permanently, where do i set it and the compress ratio?
2023-01-08 08:21:48 <psykose> mkinitfs.conf gets sourced after the default so i guess in there
2023-01-08 08:22:00 <psykose> initfscomp=gzip is default set and corressponds to -C
2023-01-08 08:22:19 <psykose> and level is -19 with no options
2023-01-08 08:22:37 <psykose> so if you do initfscomp=zstd i guess it would work for zstd -19 and anything else you'd have to patch stuff
2023-01-08 08:22:38 <psykose> lemme try
2023-01-08 08:23:11 <psykose> nope, still gz
2023-01-08 08:23:47 <uniqNuuid> psykose: hmmm. get sourced? 
2023-01-08 08:24:06 <psykose> it's all just shell scripts so sourcing (. file) is the common configuration way
2023-01-08 08:24:20 <psykose> as in what it's doing
2023-01-08 08:24:22 <psykose> doesn't seem to work tho
2023-01-08 08:24:36 <psykose> ah no
2023-01-08 08:25:00 <psykose> it works
2023-01-08 08:25:03 <psykose> i just misspelt it
2023-01-08 08:25:06 <uniqNuuid> i found that chacha20-poly1305@openssh.com on alpine is really slow [2s between LAN ], the is the default ssh cipher... switched to aes128... fast [0.48s], `time ssh LAN_BOX1 exit` 
2023-01-08 08:25:10 <psykose> 34 mb -> 22
2023-01-08 08:25:20 <psykose> just put initfscomp=zstd in /etc/mkinitfs/mkinitfs.conf
2023-01-08 08:25:32 <uniqNuuid> coll!
2023-01-08 08:26:04 <psykose> chacha20 is slower than aes with hwaccel (though iirc that's not a fair comparison, aes256 would be equivalent and you should compare to that)
2023-01-08 08:26:09 <psykose> in any case yeah it would get beaten anyway
2023-01-08 08:26:18 <psykose> but most arm devices don't have aes accel
2023-01-08 08:26:20 <psykose> it's much more even there
2023-01-08 08:27:35 <psykose> but yeah that's nice, i was wondering how to do that for a bit and was lazy
2023-01-08 08:27:41 <psykose> thanks for reminding me, free zstd saving for me
2023-01-08 08:27:50 <psykose> maybe i should go add -T0 so it's a bit faster
2023-01-08 08:29:02 <uniqNuuid> psykose: i have the same box install with archlinux arm, the ssh is fast using default chacha... 
2023-01-08 08:29:09 <uniqNuuid> but on alpine, it is slow, 
2023-01-08 08:29:23 <uniqNuuid> i have 2 of the same hardware. 
2023-01-08 08:29:44 <uniqNuuid> one is on archlinux arm, custom kernel. one is on alpine, linux-edge
2023-01-08 08:30:16 <uniqNuuid> not able to change the compress level? 
2023-01-08 08:31:14 <psykose> you'd have to do more analysis and get some better numbers/comparison and minimise some configs for me to look more into why one is fast and one is slow
2023-01-08 08:31:16 <psykose> no, can't set level
2023-01-08 08:31:20 <psykose> but there's no point to go past 19
2023-01-08 08:31:46 <psykose> i guess for personal use you might be able to use -22 --ultra if you really wanted, but it's not really a good default, though it would be safe for mkinitfs
2023-01-08 08:31:46 <psykose> meh
2023-01-08 08:31:52 <uniqNuuid> -19 is crazy slow, but my initramfs is only 5.2M on gzip with this setup: features="base ext4 keymap usb n1" 
2023-01-08 08:31:56 <uniqNuuid> let me try zstd
2023-01-08 08:32:46 <psykose> slow at compression you mean? sure
2023-01-08 08:32:55 <dwfreed> but compression is once
2023-01-08 08:32:57 <psykose> for me, -19 is 14.8 seconds
2023-01-08 08:33:04 <psykose> but -19 -T0 is 1.9
2023-01-08 08:33:06 <psykose> so i will go add that
2023-01-08 08:33:16 <dwfreed> space is forever, and decompression is every boot
2023-01-08 08:33:25 <psykose> decompression is also still very fast
2023-01-08 08:33:26 <dwfreed> and zstd decompression is incredibly fast
2023-01-08 08:33:28 <psykose> yep
2023-01-08 08:33:37 <uniqNuuid> /boot/initramfs-edge: Zstandard compressed data (v0.8+), Dictionary ID: None : 4.7M 
2023-01-08 08:33:41 <psykose> in any case the -T0 doesn't really lose the speed here
2023-01-08 08:33:45 <psykose> er
2023-01-08 08:33:45 <psykose> ratio
2023-01-08 08:33:54 <psykose> and is still very quick (though no gzip) at -19
2023-01-08 08:34:06 <uniqNuuid> i think -10 is a good one.
2023-01-08 08:34:09 <uniqNuuid> for zstd
2023-01-08 08:34:24 <uniqNuuid> my daily ratio is -10 , it is fast and size is good 
2023-01-08 08:36:09 <psykose> this is really one of those cases where you don't want to trade the speed
2023-01-08 08:36:24 <psykose> but first, really check `time -T0 -19` on the uncompressed initrd
2023-01-08 08:36:26 <psykose> https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/114
2023-01-08 08:36:36 <psykose> time zstd*
2023-01-08 08:36:58 <psykose> funnily it somehow even improved the ratio
2023-01-08 08:36:58 <uniqNuuid> aha , it is the KexAlgorithms that is slow. 
2023-01-08 08:37:02 <uniqNuuid> not the chacha20 
2023-01-08 08:37:08 <psykose> kex doesn't affect throughput
2023-01-08 08:37:11 <psykose> what were you measuring
2023-01-08 08:37:13 <psykose> connection time?
2023-01-08 08:37:49 <uniqNuuid> now i set KexAlgorithms ^curve25519-sha256 in ~/.ssh/config, it is 0.54s now, tested by `time ssh another_box_in_LAN exit`
2023-01-08 08:38:19 <psykose> yeah connection time
2023-01-08 08:38:21 <psykose> very different : )
2023-01-08 08:38:36 <uniqNuuid> if using ssh default which is "sntrup761x25519-sha512@openssh.com", the same test is 1.97s
2023-01-08 08:38:42 <psykose> generally i don't care very much about connection time because i use the ControlMaster stuff for ssh and they stay open
2023-01-08 08:39:10 <uniqNuuid> but on archlinux, no such issue, i don't change the default KexAlgorithms
2023-01-08 08:39:23 <psykose> i've been meaning to optimise my ssh setups and wireguard because they're really slow sometimes, just haven't found the time
2023-01-08 08:44:06 <uniqNuuid> ControlMaster has some issues, like ENV issue
2023-01-08 08:44:32 <uniqNuuid> the ENV does not change... stuff like SSH_CONNECTION, TERM, i did not count... 
2023-01-08 08:55:38 <psykose> did more numbers in the mr above
2023-01-08 08:55:51 <psykose> zstd -9 is much faster than gzip, has a much better ratio, and decompresses much faster
2023-01-08 08:55:57 <psykose> with the only cost being the 100M compression memory use
2023-01-08 08:56:15 <psykose> 5x speed and .8 better ratio
2023-01-08 08:56:20 <psykose> on my hardware at least
2023-01-08 08:56:24 <psykose> should test some arm devices too
2023-01-08 08:57:18 <psykose> hopefully ncopa is as adventurous as me and we could make that the default :p but adding -T0 to -19 is also an ok opt in change for anyone who used that
2023-01-08 08:59:38 <uniqNuuid> yeah, just do "-T0 -9"
2023-01-08 09:00:48 <psykose> nah, that's not really better
2023-01-08 09:00:57 <psykose> they're both very fast, the T0 just doubles the memory use
2023-01-08 09:01:02 <uniqNuuid> that's what i do in /etc/mkinitcpio.conf on archlinux : COMPRESSION_OPTIONS=(-T0 -10) COMPRESSION="zstd"
2023-01-08 09:01:26 <psykose> at -19 you're shaving a ton of time and probably had the memory anyway (it's over 300M). for -9 it's "just 100" which is very safe
2023-01-08 09:01:43 <psykose> making it 200 and shaving milliseconds is not that worth it
2023-01-08 09:02:06 <psykose> of course you can use it (and i do), i'm just referring to mkinitfs only
2023-01-08 09:02:18 <psykose> at the file size it even is you can't use more than 3-4 threads
2023-01-08 09:02:43 <uniqNuuid> -19 does not give much of space save over -9, but it is really slow, even with -T0. 
2023-01-08 09:03:27 <uniqNuuid> maybe provide an option? 
2023-01-08 09:03:42 <uniqNuuid> so users can do whatever they want. 
2023-01-08 09:03:42 <psykose> this isn't about options
2023-01-08 09:03:45 <psykose> this is explicitly not about that
2023-01-08 09:04:18 <psykose> the difference between -9 and -19 here is 3.28 -> 3.72 ratio
2023-01-08 09:04:27 <psykose> personally i'd do that, but again, not the point
2023-01-08 09:04:33 <psykose> the default is special
2023-01-08 09:04:48 <psykose> most people never touch this or even know it exists
2023-01-08 09:05:03 <psykose> so my proposal is indeed zstd -9 for all of mkinitfs
2023-01-08 09:05:24 <psykose> people can configure whatever they want
2023-01-08 09:05:44 <psykose> maybe someone is doing an lrzip+zstd+custom-filters mkinitfs :p
2023-01-08 09:06:19 <uniqNuuid> -T0 has a sister option: --auto-threads={physical,logical} 
2023-01-08 09:10:13 <psykose> i know how it works, what about it
2023-01-08 09:11:04 <uniqNuuid> i looked at your issue, the beganning says: -T0 defaults to hardware cores
2023-01-08 09:11:20 <uniqNuuid> so i mention this in case you missed that one from man pange
2023-01-08 09:11:27 <uniqNuuid> s/pange/page/g
2023-01-08 09:11:27 <alpine-meetbot> uniqNuuid meant to say: so i mention this in case you missed that one from man page
2023-01-08 09:12:41 <psykose> no, just mentioning what it does in case someone read it and wonders how many cores specifically
2023-01-08 09:13:04 <uniqNuuid> aha, the sourcode of /sbin/mkinitfs : case "$initfscomp" in zstd) .... // i was thinking setting that initfscomp="zstd -10" ... that would fail the case test :(
2023-01-08 09:13:40 <uniqNuuid> so. i'll just wait your proposal get merged
2023-01-08 09:15:44 <psykose> i'm already betting $2 the defaults will never change :p
2023-01-08 09:15:52 <uniqNuuid> actually there is a hack after reading the source code : echo 'readonly comp="zstd -10"' >> /etc/mkinitfs/mkinitfs.conf 
2023-01-08 09:16:07 <uniqNuuid> lol. that should work. 
2023-01-08 09:16:15 <uniqNuuid> just a personal hack. 
2023-01-08 09:18:05 <uniqNuuid> and mkinitfs does not have "set -e" so i guess it works. let me try my hack 
2023-01-08 09:19:02 <uniqNuuid> no. it failed : /sbin/mkinitfs: line 304: comp: is read only
2023-01-08 09:19:10 <uniqNuuid> haha. 
2023-01-08 09:21:44 <uniqNuuid> maybe: readonly comp="zstd -10" || true
2023-01-08 09:22:45 <uniqNuuid> no. same 
2023-01-08 09:23:08 <uniqNuuid> never mind. not a big deal. 
2023-01-08 09:25:00 <uniqNuuid> psykose: change initfscomp=gzip to zstd, is dangerous 
2023-01-08 09:25:14 <uniqNuuid> some u-boot does not load zstd initrd.
2023-01-08 09:26:36 <uniqNuuid> most recent u-boot can load gzipped initrd by default. but zstd has to be specificly enabledd in u-boot .config
2023-01-08 09:28:10 <psykose> why would uboot have to load an initfs ever
2023-01-08 09:28:18 <psykose> the kernel can decompress it itself
2023-01-08 09:29:00 <uniqNuuid> i don't know, i remember few days ago, i encountered that issue. and then i rebuild a u-boot with zstd support
2023-01-08 09:29:32 <uniqNuuid> u-boot load vmlinuz and initrd to RAM first. then... handle the job to vmlinuz. 
2023-01-08 09:36:28 <uniqNuuid> psykose: https://0x0.st/o7-U.png
2023-01-08 09:37:52 <psykose> yes, from everything i can find that ramdisk location in memory can be compressed
2023-01-08 09:38:07 <psykose> so all uboot has to do is map the whatever file into the address
2023-01-08 09:38:11 <psykose> why would it decompress it first
2023-01-08 09:38:15 <psykose> this sounds silly
2023-01-08 09:38:46 <uniqNuuid> maybe i am illusional
2023-01-08 09:38:51 <psykose> no, it sounds correct
2023-01-08 09:39:05 <psykose> but this now sounds like "we can't use zstd as a default because uboot is a pile of shit"
2023-01-08 09:39:07 <psykose> oh well
2023-01-08 09:39:24 <uniqNuuid> i did so many tests all kinds of these days, i already forgot what is what. 
2023-01-08 09:40:24 <psykose> as always everyone complains about x86 and then every platform outside x86 looks like whatever is happening there
2023-01-08 09:40:27 <psykose> /snark
2023-01-08 09:41:52 <uniqNuuid> does the dmesg show that kernel decompress initramfs? 
2023-01-08 09:43:01 <uniqNuuid> the straight forward test would be: remove zstd from u-boot, and let it load a zstd initrd, if that went ok.. then i was illusional. 
2023-01-08 09:43:55 <psykose> no, you're probably right :)
2023-01-08 09:46:51 <uniqNuuid> psykose: from the img i provided : echo $((0x7bf13fe9 - 0x7b7e3000))
2023-01-08 09:46:59 <uniqNuuid> the size is 7540713
2023-01-08 09:47:56 <uniqNuuid> and ls -l /boot/initramfs-linux-phicomm-n1.img shows  : 7540713 
2023-01-08 09:48:16 <uniqNuuid> so, u-boot load the initrd as is
2023-01-08 09:48:28 <uniqNuuid> u-boot does not decompress initrd. you are right. 
2023-01-08 09:50:06 <uniqNuuid> but maybe u-boot refuse to load a format that it does not recognize? 
2023-01-08 09:50:12 <uniqNuuid> i don't know. need test. 
2023-01-08 10:44:42 <psykose> ah, so it works without any "zstd support"?
2023-01-08 10:44:42 <psykose> :)
2023-01-08 11:08:00 <uniqNuuid> psykose: that picture was taking a few days ago. i just did a math to show that u-boot does not decompress initrd. it does not mean u-boot can load zsdt img into ram without 'zstd support'
2023-01-08 11:08:25 <psykose> is that a yes or a no
2023-01-08 11:09:12 <uniqNuuid> i think it is a yes. u-boot (without zstd enabled) can load zstd initrd into ram 
2023-01-08 11:09:34 <psykose> no, i legitimately want a yes or a no if you want to go test it
2023-01-08 11:09:39 <psykose> no thinking!
2023-01-08 11:09:42 <psykose> :p
2023-01-08 11:21:36 <uniqNuuid> psykose: aha, zstd decompress support is to decompress vmlinuz. 
2023-01-08 11:21:50 <uniqNuuid> so, it has nothing to do with initramfs. 
2023-01-08 11:21:55 <psykose> wrong
2023-01-08 11:26:14 <uniqNuuid> psykose: grep -i zst ../u-boot-v2022.10/.config : # CONFIG_ZSTD is not set
2023-01-08 11:26:18 <uniqNuuid> i am testing right now 
2023-01-08 11:28:14 <uniqNuuid> done testing. it boots alpine zstd initramfs fine. 
2023-01-08 11:35:07 <psykose> :)
2023-01-08 11:35:23 <psykose> the relevant options are CONFIG_RD_*
2023-01-08 11:36:10 <psykose> it's separate from the self-decompress (which is only a thing on x86 anyway) or zstd support in general or zstd module loading support (which is also usually implemented in the modprobe utility itself too)
2023-01-08 11:36:11 <uniqNuuid> there is no such thing. 
2023-01-08 11:36:23 <uniqNuuid> grep -i config_rd .config returns nothing 
2023-01-08 11:37:27 <psykose> i have no idea what .config is but the kernel has those just fine
2023-01-08 11:38:12 <uniqNuuid> .config is u-boot's config
2023-01-08 11:38:20 <uniqNuuid> ../u-boot-v2022.10/.config
2023-01-08 11:39:10 <psykose> why would uboot's config have a kernel configuration option that lets the kernel decompress ramdisks
2023-01-08 11:39:46 <uniqNuuid> well, aren't we worring about u-boot right now. 
2023-01-08 11:40:10 <psykose> no, i was telling you what thing lets you boot that
2023-01-08 12:19:37 <uniqNuuid> time -p ssh root@localhost exit : on alpine: real 1.86 ; on archlinux : real 0.45
2023-01-08 12:21:35 <uniqNuuid> if "KexAlgorithms ^curve25519-sha256" in .ssh/config, on alpine : real 0.34
2023-01-08 12:24:45 <ikke> The default is sntrup761x25519-sha512@openssh.com?
2023-01-08 12:27:07 <psykose> is that even on the same computer
2023-01-08 12:28:16 <psykose> getting the feeling the 'archlinux' one is a completely different machine so these tests are completely meaningless
2023-01-08 12:28:54 <ikke> plausible
2023-01-08 12:29:28 <ikke> apparently that key exchange algo is quantum hardened
2023-01-08 12:29:28 <uniqNuuid> psykose: same hardware. amlogic s905d
2023-01-08 12:29:45 <uniqNuuid> i have 2 s905d. one on archlinuxarm one on alpine 
2023-01-08 12:30:19 <psykose> missing a part of the benchmark there still
2023-01-08 12:30:20 <psykose> :)
2023-01-08 12:31:37 <uniqNuuid> the default sntrup761x25519-sha512@openssh.com is slow on alpine. 
2023-01-08 12:31:47 <uniqNuuid> i don't know what went wrong.
2023-01-08 12:31:52 <uniqNuuid> what is acct?
2023-01-08 12:33:13 <ikke> process accounting
2023-01-08 12:34:11 <uniqNuuid> openssh depends on openssl? 
2023-01-08 12:34:25 <uniqNuuid> is openssl compiled with sntrup761? 
2023-01-08 12:34:44 <uniqNuuid> how do i list openssl features/components 
2023-01-08 12:35:44 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/openssl/APKBUILD
2023-01-08 12:35:47 <psykose> openssl doesn't implement sntrup761x25519
2023-01-08 12:38:47 <uniqNuuid> so many no...
2023-01-08 12:39:27 <Ermine> Why those ciphers are off?
2023-01-08 12:40:42 <psykose> i'm pretty sure the entire list is redundant to no-weak-ssl-ciphers
2023-01-08 12:41:17 <uniqNuuid> yeah. i see. it is reasonable to disable legacy stuff 
2023-01-08 12:42:33 <uniqNuuid> the openssh-server on alpine is fast thought, connnection from other machine are fast. 
2023-01-08 12:42:57 <uniqNuuid> s/thought/though/
2023-01-08 12:42:57 <alpine-meetbot> uniqNuuid meant to say: the openssh-server on alpine is fast though, connnection from other machine are fast. 
2023-01-08 12:44:10 <Ermine> is sm4 considered weak? Support for this cipher was added in FSCRYPT in linux 6.2.
2023-01-08 12:44:41 <uniqNuuid> sm4 is made in china 
2023-01-08 12:44:53 <uniqNuuid> sm1,2,3,4 are from China. 
2023-01-08 12:46:31 <Ermine> Cool.
2023-01-08 12:49:10 <psykose> sm4 no, though it's quite sussy
2023-01-08 12:52:46 <uniqNuuid> https://github.com/openssl/openssl/blob/19cc035b6c6f2283573d29c7ea7f7d675cf750ce/Configure#L401
2023-01-08 12:54:44 <psykose> someone forgot no-fips smh
2023-01-08 12:55:12 <uniqNuuid> ok, it has nothing to do with openssl. 
2023-01-08 12:55:27 <uniqNuuid> openssl does not have sntrup761
2023-01-08 12:55:58 <ikke> that's what psykose said
2023-01-08 12:56:20 <psykose> i'm known for being wrong, but sometimes i'm even right :)
2023-01-08 12:56:39 <ikke> like a broken clock? :P
2023-01-08 12:56:58 <psykose> just the broken part
2023-01-08 12:58:21 <uniqNuuid> something todo with pam ?
2023-01-08 12:58:54 <Ermine> nothing todo with pam
2023-01-08 12:59:58 <uniqNuuid> i test ssh root@localhost 
2023-01-08 13:00:12 <uniqNuuid> hardware the same. this drives me crazy 
2023-01-08 13:07:30 <psykose> you would have to learn how to profile things to get the answer
2023-01-08 13:09:42 <hsv2> somebody knows how to get X11 or Wayland working on Raspberry PI 1B? They are telling me they don't find any screen/gpu :/
2023-01-08 13:10:23 <uniqNuuid> psykose: which part was not clear
2023-01-08 13:10:57 <psykose> the part where it was not finished because you didn't run all the tests in both places
2023-01-08 13:11:32 <psykose> hsv2: you need some dtoverlay to enable the gpu iirc
2023-01-08 13:11:42 <psykose> as an argument in config.txt, etc etc
2023-01-08 13:12:05 <hsv2> there's no pkg or tool to auto setup that? 😆
2023-01-08 13:12:32 <hsv2> if so i guess i'll just copy over the config from my gentoo on raspb lol
2023-01-08 13:12:33 <psykose> dtoverlay=vc4-kms-v3d unless it didn't work for 1b
2023-01-08 13:14:00 <hsv2> wiki says its for pi 3/4 :/ let's try if it works to on pi 1
2023-01-08 13:17:32 <hsv2> ok, it (slowly) works too on rpi1
2023-01-08 13:17:35 <hsv2> thx
2023-01-08 13:24:25 <uniqNuuid> psykose: https://0x0.st/o7Ho.png
2023-01-08 13:25:10 <psykose> you need far more than one run
2023-01-08 13:25:56 <uniqNuuid> on alpine, curve25519-sha256 is faster than archlinux. 
2023-01-08 13:27:21 <psykose> maybe, or maybe not
2023-01-08 13:28:44 <uniqNuuid> https://0x0.st/o7HH.txt
2023-01-08 13:30:11 <psykose> pog
2023-01-08 13:30:21 <psykose> you're getting it :)
2023-01-08 13:30:25 <psykose> now three more times
2023-01-08 13:30:48 <uniqNuuid> arch : https://0x0.st/o7H8.txt
2023-01-08 13:31:39 <uniqNuuid> you mean 30 times? 
2023-01-08 13:32:03 <uniqNuuid> same, always real 1.85 seconds
2023-01-08 13:32:39 <uniqNuuid> the 2 box has the exactly same hardware. 
2023-01-08 13:33:03 <uniqNuuid> and same ssh config. because i just rsync the config over 
2023-01-08 13:33:52 <psykose> quite the time difference
2023-01-08 13:34:07 <psykose> now you'd have to do some profiling but that's complicated
2023-01-08 13:37:48 <uniqNuuid> https://0x0.st/o7HK.txt alpine curve25519-sha256  is really fast 
2023-01-08 13:40:33 <uniqNuuid> when i do ssh -v, i can feel that it stops at kex for a second 
2023-01-08 13:52:42 <uniqNuuid> i have 2 vps
2023-01-08 13:56:22 <uniqNuuid> same configuration on ssh config and hardware. :  https://0x0.st/o7HT.txt
2023-01-08 13:56:41 <uniqNuuid> it is the oracle ARM vps 
2023-01-08 13:58:07 <uniqNuuid> because the vps is fast, i basically can't feel the diff 
2023-01-08 13:58:40 <uniqNuuid> but on ARM A53, the diff is obvious 
2023-01-08 14:02:35 <psykose> yep
2023-01-08 14:07:22 <uniqNuuid> i'll change all ssh config to use curve25519-sha256 for now. 
2023-01-08 15:08:16 <vkrishn> uuidNuniq: do you use bash to call ssh in alpine setup ?
2023-01-08 16:03:13 <uniqNuuid> vkrishn_: i use bash mainly for the command 'time'
2023-01-08 16:03:23 <uniqNuuid> the time command is different across shells 
2023-01-08 16:09:04 <marc_> hello, when I start pipewire I get this message: https://paste.stratum0.org/?3f85233ae30d42f8#2YjJu9LuUMb7Bky2pdYaW2opW211b9Shp62Tb6LJpCKP
2023-01-08 16:09:04 <marc_> i already have profiled running in the background but is there anything that takes care of this "freedesktop stuff"?
2023-01-08 16:19:15 <vkrishn_> uniqNuuid: only if I could access  https://0x0.st, then could get what's happening new
2023-01-08 16:31:42 <minimal> uniqNuuid: when using chacha20-poly1305 if this is on an Arm machine do you have the "neon" modules loaded? that should give some acceleration
2023-01-08 16:33:34 <minimal> last time I looked one of the related neon modules is not built on Arm for Alpine, but not sure if it is used for poly1305
2023-01-08 16:36:00 <uniqNuuid> minimal: i have neon set to Y
2023-01-08 16:36:30 <minimal> you're using your own built kernel for Alpine?
2023-01-08 16:36:45 <uniqNuuid> oh no, alpine CONFIG_CRYPTO_NHPOLY1305_NEON=m
2023-01-08 16:37:10 <minimal> that's what I said, there are multiple *loadable* neon modules for Alpine
2023-01-08 16:37:15 <uniqNuuid> i don't see neon module in lsmod after some ssh connection via poly1305
2023-01-08 16:37:59 <minimal> there are 2 poly1305 related neon modules and from memory Alpine only build 1 of them and I don't know if the other is important
2023-01-08 16:38:27 <minimal> uniqNuuid: you won't see it in "lsmod" unless something (i.e. you) loaded it
2023-01-08 16:38:30 <uniqNuuid> poly1305 is not an issue, it is the KexAlgorithms=sntrup761x25519-sha512@openssh.com
2023-01-08 16:39:10 <minimal> poly1305 is not an issue yet earlier you said you WERE using it for ssh???
2023-01-08 16:39:16 <uniqNuuid> CONFIG_CRYPTO_NHPOLY1305_NEON=m CONFIG_CRYPTO_POLY1305_NEON=m
2023-01-08 16:39:35 <uniqNuuid> yeah. 1305 is used by default .
2023-01-08 16:39:59 <uniqNuuid> changing 1305 to aes128 is the same. it is KexAlgorithms=sntrup761x25519-sha512@openssh.com that slows down 
2023-01-08 16:40:13 <minimal> plus there is chacha20_neon to load also...
2023-01-08 16:40:57 <uniqNuuid> this is my A53 cpu flags : Flags:               fp asimd evtstrm aes pmull sha1 sha2 crc32 cpuid
2023-01-08 16:41:51 <uniqNuuid> hmm. alpine does not boot from mmc. but it boots from usb. i am investigating 
2023-01-08 16:42:12 <uniqNuuid> shit happens all the time. 
2023-01-08 16:42:54 <uniqNuuid> no new rc.log or dmesg created on mmc boot. so i guess the hardware is not loaded by initramfs. 
2023-01-08 16:43:09 <uniqNuuid> however boot from usb, and it is ok to mount mmc 
2023-01-08 16:44:03 <uniqNuuid> mounting mmc is fine, booting is NOT fine. i can't see from the monitor because my monitor does not turn on after u-boot. 
2023-01-08 16:44:36 <uniqNuuid> the monitor is another issue... due to  CMA got disabled 
2023-01-08 16:44:55 <uniqNuuid> well i don't care about monitor, as i do headless. 
2023-01-08 16:45:04 <minimal> uniqNuuid: it is unclear what your "complaint" is about Alpine and ssh. If you're using stuff for which there is some acceleration for the nload the appropriate modules; if it's something there is no acceleration for then what do you expect?
2023-01-08 16:46:01 <uniqNuuid> minimal: i tested on VPS.... the result is the same. slower ssh connection speed on alpine. 
2023-01-08 16:46:56 <uniqNuuid> https://0x0.st/o7HT.txt minimal 
2023-01-08 16:47:05 <uniqNuuid> this is on Oracle ARM VPS
2023-01-08 16:47:23 <minimal> are you comparing like-with-like? same negotiated options? same (related) kernel modules loaded?
2023-01-08 16:47:55 <minimal> so an Oracle VM should have ARM "ce" support, which is better than neon
2023-01-08 16:48:05 <uniqNuuid> i am saying this KexAlgorithms=sntrup761x25519-sha512@openssh.com 
2023-01-08 16:48:22 <uniqNuuid> which is default ssh client kex
2023-01-08 16:48:45 <minimal> and I'm asking, with the *SAME* SSH negotiated settings? which the same related kernel modules loaded?
2023-01-08 16:49:15 <uniqNuuid> yes. ssh config is rsynced over. 
2023-01-08 16:49:15 <minimal> the effective sshd_config could well be different between Alpine and Arch, I don't know
2023-01-08 16:49:30 <uniqNuuid> kernel config both alpine lts-virt
2023-01-08 16:49:40 <minimal> and the kernel modules/compiled-in drivers?
2023-01-08 16:49:42 <uniqNuuid> yes, i use alpine lts-virt on archlinux. 
2023-01-08 16:50:17 <uniqNuuid> i copy alpine lts-virt config and compiled an archlinux tar.pkg.zst
2023-01-08 16:50:20 <uniqNuuid> kernel
2023-01-08 16:50:24 <minimal> well the obvious point is there are situation when musl is observed to be slower for some stuff than glibc
2023-01-08 16:50:32 <psykose> yawn
2023-01-08 16:50:37 <psykose> we already established it's slower
2023-01-08 16:50:43 <uniqNuuid> i don't think it is musl. 
2023-01-08 16:50:56 <psykose> there is no point to talk in a circle about it for yet another 6 hours
2023-01-08 16:51:05 <minimal> uniqNuuid: so what so you think it is?
2023-01-08 16:51:18 <psykose> the only thing to do from here is bust out the debug symbols and perf profiling
2023-01-08 16:51:24 <uniqNuuid> i don't know. so i changed the KEX to something else. 
2023-01-08 16:52:04 <uniqNuuid> now am using this KexAlgorithms ^curve25519-sha256 in ssh config
2023-01-08 16:52:14 <uniqNuuid> super fast. 
2023-01-08 16:52:38 <uniqNuuid> you get get a list of valid kex : ssh -Q kex
2023-01-08 16:54:01 <uniqNuuid> psykose: i agree, i am not skilled enough to do more investigation, it's beyond my knowledge. so 
2023-01-08 16:54:16 <psykose> you can do one thing
2023-01-08 16:54:23 <psykose> go to main/openssh
2023-01-08 16:54:29 <psykose> add CFLAGS=-O2
2023-01-08 16:54:30 <psykose> rebuild it
2023-01-08 16:54:33 <psykose> ;)
2023-01-08 16:54:54 <psykose> or rather -O2 -flto if you want the same config as arch
2023-01-08 16:56:50 <minimal> uuiqNuuid: it is hard to follow your issue in general as one moment you're talking about Arm SBCs (which is where NEON may be relevant), then you're talking about Oracle Arm servers (where CE may be relevant), and you're testing franken-systems by mixing kernels between distros etc
2023-01-08 16:57:13 <psykose> no, by the end all of it was narrowed down
2023-01-08 16:57:19 <psykose> please don't repeat the same shit for another 6 hours
2023-01-08 16:57:31 <minimal> its the old "triangle of small, secure, fast - pick any 2"
2023-01-08 16:58:12 <psykose> more like "excuses to pick when you can't find the real reason" :p
2023-01-08 16:58:21 <psykose> sometimes you have to say "i don't know"
2023-01-08 16:59:16 <minimal> psykose: well you're already pointed to the "fast" aspect relating to the fact that Alpine isn't generally compiling packages optimised for speed
2023-01-08 17:00:58 <psykose> that is a lot of words to say "i dont know"
2023-01-08 17:01:26 <uniqNuuid> you know what is the fasted one? diffie-hellman-group1-sha1
2023-01-08 17:02:21 <minimal> psykose: more like "if you're not compiling for speed then don't expect the result to be the fastest"
2023-01-08 17:02:25 <uniqNuuid> revert, ^^ it does not connect at all.
2023-01-08 17:02:50 <psykose> the words keep getting longer ;)
2023-01-08 17:03:24 <minimal> psykose: sentences you mean? the words are likely averaging the same length ;-)
2023-01-08 17:03:31 <psykose> heh
2023-01-08 17:03:35 <psykose> :)
2023-01-08 17:04:03 <minimal> if we switch to German then the words will definitely get longer ;-)
2023-01-08 17:05:24 <uniqNuuid> i'm trying to get mmc to boot alpine. 
2023-01-08 17:07:33 <Ermine> minimal: otoh there would be less words
2023-01-08 17:07:51 <minimal> uniqNuuid: that's won't make your SSH connections any faster ;-)
2023-01-08 17:08:54 <minimal> Ermine: "parser error: no spaces found in line" ;-)
2023-01-08 17:10:11 <uniqNuuid> minimal: SSH is fixed by changing to another kex. nothing is gonna be processed further 
2023-01-08 17:10:56 <uniqNuuid> i really like apline. i mean i install alpine from a 7M minirootfs. and it takes 3mins to get a fully working OS using a script. 
2023-01-08 17:12:03 <uniqNuuid> i did similar script to bootstrap debian using debootstrap... it takes forever. only you know how excellent alpine is after you tried enough distro 
2023-01-08 17:14:27 <minimal> that was a joke reply, note the smiley at the end, I obviously understood the 2 things were unconnected
2023-01-08 17:30:51 <vkrishn> silly question,  two "oracle ARM vps" runnning simultaneously is always same ?
2023-01-08 17:31:27 <vkrishn> ;)
2023-01-08 17:34:55 <minimal> is always the same what?
2023-01-08 17:42:50 <vkrishn> its ram usage, cpu(GHz), response time, cache, ram-free ....
2023-01-08 17:43:53 <minimal> vkrishn: virtually yes ;-)
2023-01-08 17:48:53 <vkrishn> why not just throw this problem to https://www.phoronix.com, who have ton loads of experience in setting clean equivalent benchmarkable setup, then might even treat us back with good logs
2023-01-08 17:49:09 <vkrishn> they might^
2023-01-08 17:58:59 <valerius> vkrishn, they do not seem to be interested in this distro
2023-01-08 17:59:05 <valerius> I've brought it up to them before :p
2023-01-08 18:09:14 <vkrishn> I doubt they would do it for free, unless its requested from a famous company
2023-01-08 18:13:10 <invoked> this review of alpine linux brought to you by carls jr, where if it doesn't get all over the place, it doesn't belong in your face
2023-01-08 18:17:59 <nikolashachka[m]> Hello guys.  Help me please.  I'm already tired of figuring it out.  How to make local backups on a flash drive?  When calling lbu commit, it throws an error.
2023-01-08 18:18:08 <nikolashachka[m]> 
ACTION uploaded an image: (128KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/olzJKSrfZcOlKEBPWbNSpqIw/IMG_20230108_231459.jpg >

2023-01-08 18:18:11 <nikolashachka[m]> 
ACTION uploaded an image: (112KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/fanKCMnILUfqJlFYseAbNLhl/IMG_20230108_214707.jpg >

2023-01-08 18:19:35 <nikolashachka[m]> I'm new and I didn't see how it should work at all. Thanks for the help.
2023-01-08 18:20:04 <ikke> nikolashachka[m]: what does mount output?
2023-01-08 18:20:29 <ikke> nikolashachka[m]: did you by chance just dd the iso on the usb disk btw?
2023-01-08 18:20:54 <ikke> /dev/sdb is iso9660, so that's read-only
2023-01-08 18:21:23 <nikolashachka[m]> 
ACTION uploaded an image: (124KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/qDEookhDfftYAOClLcFvFSyi/IMG_20230108_232058.jpg >

2023-01-08 18:21:24 <ikke> /dev/sdb2 is vfat, is that where it should write it?
2023-01-08 18:22:36 <ikke> nikolashachka[m]: and ls -l /dev/usbdisk
2023-01-08 18:25:39 <nikolashachka[m]> 
ACTION uploaded an image: (40KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/RjLPWqdasjGqjOHtViRwxJQe/IMG_20230108_232516.jpg >

2023-01-08 18:25:40 <nikolashachka[m]> 
ACTION uploaded an image: (145KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/zydgRdFkRpSPuIgjEhzZWJFp/IMG_20230108_232403.jpg >

2023-01-08 18:27:58 <nikolashachka[m]> The goal is to make a bootable flash drive with the programs I need to work from RAM and, at the end of the session, make a backup back to the flash drive
2023-01-08 18:28:44 <ikke> Easiest is to use setup-bootable to create a flash drive that is writable
2023-01-08 18:29:02 <ikke> try: rc-service modloop stop
2023-01-08 18:29:09 <ikke> umount /dev/sdb1
2023-01-08 18:29:30 <ikke> or rather: umount /dev/sdb
2023-01-08 18:34:25 <nikolashachka[m]> 
ACTION uploaded an image: (91KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/VBrOvlttyCmERwjrwNIoWhRf/IMG_20230108_233340.jpg >

2023-01-08 18:37:48 <ikke> That usb has a very non-standard setup, /dev/sdb was mounted, so it's a valid partition, but /dev/sdb1 and /dev/sdb2 exist as well
2023-01-08 18:38:05 <nikolashachka[m]> <ikke> "nikolashachka: did you by chance..." <- Yes, I burned the iso from the official site with the help of DD and did nothing else.  The only thing I checked is whether the operating system really works from RAM, I just took out the flash drive from the computer.  It turns out that now I need to perform some setup-bootable settings, but what exactly do I need to enter in order for the lbu commit command to work?  Sorry for
2023-01-08 18:38:05 <nikolashachka[m]> broken English.
2023-01-08 18:38:32 <ikke> I'd recommend doing setup-bootable https://dl-cdn.alpinelinux.org/alpine/v3.17/releases/x86_64/alpine-standard-3.17.0_rc4-x86_64.iso /dev/sdb
2023-01-08 18:38:45 <ikke> (or whatever image you used)
2023-01-08 18:41:18 <nikolashachka[m]> ikke: Perhaps sdb2 is the bootloader?
2023-01-08 18:42:02 <uniqNuuid> can i copy a module from another distro to alpine's modules folder and load it into initramfs? 
2023-01-08 18:42:48 <ikke> uniqNuuid: no, the kernel usually rejects modules that were not compiled for it
2023-01-08 18:47:21 <nikolashachka[m]> <ikke> "I'd recommend doing setup-..." <- Okay, I'll try this iso.  I'll also take some time to learn the setup-bootable setup, thanks!
2023-01-08 18:48:09 <uniqNuuid> ikke: is there a work around?
2023-01-08 18:48:31 <ikke> uniqNuuid: build the module for the running kernel
2023-01-08 18:49:43 <uniqNuuid> ikke: download the sourcecode and make menuconfig and make modules?  
2023-01-08 18:50:55 <ikke> uniqNuuid: is it an out-of-tree module?
2023-01-08 18:51:51 <uniqNuuid> no, it is a module in the kernel but alpine did not enable it .
2023-01-08 18:52:07 <uniqNuuid> nvmem_meson_efuse : drivers/nvmem/nvmem_meson_efuse
2023-01-08 18:53:46 <ikke> I don't have experience with that, so no idea how to do that
2023-01-08 19:15:42 <Emru> Hey, is there a chance that wkhtmltopdf will be back in the repo for newer releases?
2023-01-08 19:18:07 <psykose> no, it's an eol project that relies on a no longer updated web browser
2023-01-08 19:18:09 <psykose> https://wkhtmltopdf.org/status.html
2023-01-08 19:28:06 <psykose> ikke: was there a way to do something like '"xyz" in pkg.Makedepends' with apkgquery
2023-01-08 19:28:12 <psykose> not sure how to cast PackageSpecs to anything
2023-01-08 19:31:04 <Emru> psykose: thanks, I didn't knew
2023-01-08 19:31:40 <psykose> if it's to convert html to pdf afaik there's like 20 options
2023-01-08 19:33:30 <ikke> psykose: you can use map to map it to a list of strings
2023-01-08 19:33:36 <psykose> o
2023-01-08 19:35:09 <ikke> '"curl-dev" in map(pkg.Makedepends, {.Pkgname})'
2023-01-08 19:35:46 <scorpion2185[m]1> can you use icecat on alpine?
2023-01-08 19:39:44 <ikke> You'd have to compile / package it
2023-01-08 19:40:02 <psykose> works, thanks
2023-01-08 19:40:27 <ikke> last stable release is almost 4 years ago
2023-01-08 19:40:37 <ikke> not sure if that's good for a browers
2023-01-08 19:40:40 <ikke> browser
2023-01-08 19:40:56 <ikke> preview release is more recent
2023-01-08 19:41:03 <WhyNotHugo_> I'd like to delete the option listed in this sections -> https://wiki.alpinelinux.org/wiki/Initramfs_init#initramfs_init_cmdline_options_(work_in_progress)
2023-01-08 19:41:13 <WhyNotHugo_> It's outdated, and it's easier to point people to the man page.
2023-01-08 19:41:25 <WhyNotHugo_> There's no point keeping a duplicate (outdated) list.
2023-01-08 19:41:30 <WhyNotHugo_> Any issues with just deleting it?
2023-01-08 19:41:37 <psykose> option?
2023-01-08 19:41:41 <psykose> you mean the section?
2023-01-08 19:41:54 <WhyNotHugo_> *options
2023-01-08 19:41:58 <psykose> go for it
2023-01-08 19:42:15 <psykose> i noticed you've edited a bunch of entries, nice work btw :)
2023-01-08 19:43:31 <scorpion2185[m]1> * ikke https://packages.guix.gnu.org/packages/icecat/102.6.0-guix0-preview1/
2023-01-08 19:44:30 <scorpion2185[m]1> https://packages.guix.gnu.org/packages/icecat/102.6.0-guix0-preview1/  will it compile on alpine /musl?
2023-01-08 19:44:44 <psykose> it's just a firefox skin, so sure
2023-01-08 19:45:07 <psykose> and given that it's literally a firefox skin, i see no point to actually package it
2023-01-08 19:45:15 <psykose> but you can probably build it using the existing builds easily
2023-01-08 19:45:16 <psykose> i guess
2023-01-08 19:45:24 <psykose> unless they messed that up somehow
2023-01-08 19:46:49 <WhyNotHugo_> psykose: Thanks, just leaving some notes of what I find for the next person (who might even be me).
2023-01-08 19:47:58 <psykose> yeah, i keep forgetting to take some too and then i'm indeed the next person missing them every time..
2023-01-08 19:48:06 <scorpion2185[m]1> psykose >. Its main advantage is an ethical one: it is entirely free software.
2023-01-08 19:48:26 <psykose> it's firefox with a reskin and 4 lines of code changed
2023-01-08 19:48:49 <psykose> (not literally but it's not exactly a fork and it's not it's own browser)
2023-01-08 19:48:59 <Ermine> so liberating
2023-01-08 19:49:09 <scorpion2185[m]1> what about pale moon?
2023-01-08 19:49:49 <Ermine> Build it manually. It won't be packaged in Alpine.
2023-01-08 19:50:12 <psykose> like literally it's just https://git.savannah.gnu.org/cgit/gnuzilla.git/tree/data/patches?id=4a87716686104266a9cccc2d83cc249e312f3673
2023-01-08 19:50:16 <psykose> and then some js toggles
2023-01-08 19:50:18 <psykose> and a skin
2023-01-08 19:50:23 <psykose> wow they hid the feedback button
2023-01-08 19:50:27 <psykose> much fork very wow
2023-01-08 19:51:03 <psykose> nothing against to whoever uses it of course, but it's just firefox
2023-01-08 19:51:20 <scorpion2185[m]1> ah I see thanks. well midori looks good
2023-01-08 19:51:45 <Ermine> psykose: it's GNU level of caring about freedom and purity and whatnot I guess
2023-01-08 19:51:55 <psykose> midori has been abandoned for 3 years
2023-01-08 19:52:23 <rvense> i have an ARM device (rockchip rk3288-based) that runs a buildroot linux but i don't have the sources. would it be feasible to take the kernel and device tree and graft an alpine system on it? i've been using a chroot but i'd like to just boot to it
2023-01-08 19:52:49 <psykose> ah or
2023-01-08 19:53:08 <psykose> no it is, nvm
2023-01-08 19:53:12 <psykose> rvense: yeah, should be
2023-01-08 19:53:28 <psykose> if you know how the boot processx works it shouldn't be very hard
2023-01-08 19:54:57 <rvense> i don't know if i know it well enough :)
2023-01-08 19:55:31 <scorpion2185[m]1> I can't use epyphany , is it jut me?
2023-01-08 19:55:59 <psykose> what error do you get
2023-01-08 19:56:09 <psykose> rvense: then i can't say it won't leave the system unbootable :p
2023-01-08 19:57:23 <scorpion2185[m]1> (WebKitWebProcess:2): Gdk-WARNING **: 20:56:55.162: Settings portal not found: Failed to execute child process “dbus-launch” (No such file or directory)
2023-01-08 19:57:23 <scorpion2185[m]1> (WebKitWebProcess:2): GLib-GIO-CRITICAL **: 20:56:55.650: g_dbus_connection_get_unique_name: assertion 'G_IS_DBUS_CONNECTION (connection)' failed
2023-01-08 19:57:23 <scorpion2185[m]1> (epiphany:9892): epiphany-WARNING **: 20:56:55.657: Web process crashed
2023-01-08 19:57:39 <ikke> I get the latter 2 as well
2023-01-08 19:58:25 <scorpion2185[m]1> I disabled dbus on PW-wireplumber
2023-01-08 19:58:26 <Ermine> scorpion2185[m]1: install dbus-x11
2023-01-08 19:58:38 <psykose> echo $DBUS_SESSION_BUS_ADDRESS
2023-01-08 19:59:07 <psykose> you never need dbus-x11
2023-01-08 19:59:52 <psykose> (the fallback just tries to invoke it when dbus is broken)
2023-01-08 20:00:07 <Ermine> psykose: either DE launches dbus session or user should do it manually with dbus-launch afaik
2023-01-08 20:00:40 <scorpion2185[m]1> can I prevent apk to update repos when I install things?
2023-01-08 20:00:45 <psykose> no
2023-01-08 20:01:07 <psykose> well
2023-01-08 20:01:09 <psykose> -U 999999
2023-01-08 20:01:11 <psykose> probably
2023-01-08 20:01:19 <scorpion2185[m]1> checking if I have dbus taking eternity to update repo
2023-01-08 20:01:39 <psykose> that is an echo
2023-01-08 20:01:43 <psykose> as i posted
2023-01-08 20:02:47 <scorpion2185[m]1> no output from echo $DBUS_SESSION_BUS_ADDRESS
2023-01-08 20:02:57 <scorpion2185[m]1> what's -U 999999?
2023-01-08 20:03:16 <scorpion2185[m]1> dbus-x11 was installed already
2023-01-08 20:03:37 <Ermine> U sure?
2023-01-08 20:03:40 <psykose> if there's no output then you don't have running dbus
2023-01-08 20:03:48 <psykose> you have to start it with your session
2023-01-08 20:03:51 <psykose> depends how you start it
2023-01-08 20:04:17 <scorpion2185[m]1> service dbus status
2023-01-08 20:04:18 <scorpion2185[m]1>  * status: started
2023-01-08 20:04:18 <psykose> with a DE it's started for you (or well should be), with a wayland compositor invoked or startx or whatever you can dbus-start-session -- $thing
2023-01-08 20:04:24 <psykose> that's not a user bus
2023-01-08 20:04:39 <psykose> not relevant
2023-01-08 20:04:53 <psykose> er, dbus-run-session
2023-01-08 20:05:14 <scorpion2185[m]1> dbus-start-session -- river  ?
2023-01-08 20:05:23 <Ermine> yes
2023-01-08 20:05:31 <psykose> run
2023-01-08 20:05:33 <psykose> not start
2023-01-08 20:06:36 <ikke> And $thing would be your window manager for example?
2023-01-08 20:06:45 <psykose> ye
2023-01-08 20:07:08 <Ermine> thing=river in this case
2023-01-08 20:07:40 <scorpion2185[m]1> export MOZ_ENABLE_WAYLAND=1
2023-01-08 20:07:40 <scorpion2185[m]1> these are ignored because of missing dbus?
2023-01-08 20:07:40 <scorpion2185[m]1> export XDG_SESSION_TYPE=wayland
2023-01-08 20:08:31 <Ermine> XDG_SESSION_TYPE should be set by compositor
2023-01-08 20:09:26 <Ermine> only firefox cares for MOZ_ENABLE_WAYLAND=1
2023-01-08 20:15:30 <scorpion2185[m]1> now it works
2023-01-08 20:15:32 <psykose> firefox defaults to wayland with the session type and/or detecting the socket nowadays anyway
2023-01-08 20:16:01 <scorpion2185[m]1> It was running on xwayland
2023-01-08 20:17:21 <Ermine> afaik it defaulted to xwayland not so far ago
2023-01-08 20:19:41 <scorpion2185[m]1> the vars were ingnored because of wrong file
2023-01-08 20:20:37 <scorpion2185[m]1> any hope for hibernation?
2023-01-08 20:25:41 <scorpion2185[m]1> why can't I?
2023-01-08 20:26:10 <Ermine> install acpid and pm-utils
2023-01-08 20:27:27 <scorpion2185[m]1> can I use eloginds instead?
2023-01-08 20:27:41 <Ermine> You can
2023-01-08 20:28:09 <scorpion2185[m]1> it doesn't work
2023-01-08 20:28:36 <Ermine> elogind or acpid+pm-utils?
2023-01-08 20:29:10 <scorpion2185[m]1> elogind I installed pm-utils too. I have this error there:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/BJOApAgzTJNMCiObpRoHJCBK>)
2023-01-08 20:29:48 <scorpion2185[m]1> ash: writer ... wiht pm-utils
2023-01-08 20:31:58 <Ermine> try loginctl hibernate
2023-01-08 20:32:35 <psykose> Ermine: did you know there's a gajim update :p
2023-01-08 20:32:54 <Ermine> Yeah, I've got a mail
2023-01-08 20:35:14 <Ermine> woah, major bump
2023-01-08 20:41:31 <scorpion2185[m]1> Ermine: tried that too
2023-01-08 20:43:12 <Ermine> and what did it say?
2023-01-08 20:45:24 <scorpion2185[m]1> i tried that first , IIRC no such device again
2023-01-08 20:47:07 <Ermine> maybe some kernel module is missing
2023-01-08 20:48:32 <WhyNotHugo_> scorpion2185[m]1: You can use dbus-run-session which is shipped by the dbus package itself
2023-01-08 20:48:49 <scorpion2185[m]1>  Ermine  linux-lts, what should I try?
2023-01-08 20:49:01 <scorpion2185[m]1> WhyNotHugo_: yes thanks they told me that already
2023-01-08 20:53:29 <Ermine> scorpion2185[m]1: 1) find out which device do you have, 2) find out which module is responsible for power management, 3) load it
2023-01-08 20:53:46 <Ermine> Also check if you have swap partition set up
2023-01-08 21:11:41 <scorpion2185[m]1> swap is set
2023-01-08 21:11:58 <scorpion2185[m]1> what device is the CPU?
2023-01-08 21:12:24 <Ermine> laptop if your device is laptop, or cpu/motherboard I gues
2023-01-08 21:12:28 <Ermine> guess
2023-01-08 21:13:37 <scorpion2185[m]1> tablet chuwi ubook
2023-01-08 21:13:59 <scorpion2185[m]1> intel N4100
2023-01-08 21:16:05 <scorpion2185[m]1> optional kbd 
2023-01-08 21:16:42 <Ermine> Idk then
2023-01-08 21:18:09 <scorpion2185[m]1> loginctl suspend works
2023-01-08 21:18:55 <Ermine> Maybe hibernation is not supported by your hardware?
2023-01-08 21:19:36 <ikke> dmesg | grep "PM: "
2023-01-08 21:24:23 <scorpion2185[m]1> it works on that hardware on another linux I don't have it here right now I will check it
2023-01-08 21:59:37 <Saijin_Naib> Might be a missing pinctl module. Had the same issue on my N3450 platform before I made a MR to enable it in linux-lts
2023-01-08 22:00:48 <Saijin_Naib> I ended up diffing the loaded modules between Alpine and Manjaro to help isolate the issue
2023-01-08 22:07:53 <uniqNuuid> i got mmc booting, the reason was that i blacklisted drm module. and somehow drm effects emmc, i still don't know why... the heck 
2023-01-09 02:18:43 <Quill> Hello, I'm getting persistent GDBus errors when I use flatpaks after I transitioned from Gnome to Sway, and I'm not sure what to do. I have xdg-desktop-portal and multiple of its backends (wlr, kde, gtk, and gnome), but I keep hitting the same 'failed to call portal.'
2023-01-09 02:20:09 <Quill> For example, when I enter "flatpak run --command=xdg-open org.gnome.Platform https://example.com", I get "failed to call portal: GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface "org.freedesktop.portal.OpenURI" on object at path /org/freedesktop/portal/desktop
2023-01-09 02:21:56 <Quill> in my ~/.profile I set XDG_CURRENT_DESKTOP=sway and XDG_SESSION_DESKTOP=sway
2023-01-09 02:25:29 <Quill> %n Quillith
2023-01-09 02:28:32 <Quillith> test
2023-01-09 02:38:29 <Saklad5> What's a good way of configuring Alpine Linux to only route traffic over a specific IP address?
2023-01-09 02:39:05 <Saklad5> A default route, or something similar
2023-01-09 02:39:25 <Saklad5> I'm confident such a thing exists, but I'm not sure how to specify it
2023-01-09 02:42:55 <dwfreed> ip route add default via <address here> dev <interface here>
2023-01-09 02:43:25 <Saklad5> Thanks. What's the best way to persist that across reboots?
2023-01-09 02:43:37 <Saklad5> Is it possible to do so in /etc/network/interfaces?
2023-01-09 02:43:44 <dwfreed> yes
2023-01-09 02:44:01 <dwfreed> it's just a gateway
2023-01-09 02:44:03 <Saklad5> I can't for the life of me find a manpage for that file
2023-01-09 02:44:44 <Saklad5> Do you happen to know where I could find the documentation for it?
2023-01-09 02:44:49 <dwfreed> https://manpages.debian.org/bullseye/ifupdown/interfaces.5.en.html
2023-01-09 02:45:06 <Saklad5> That applies to Alpine Linux 3.17?
2023-01-09 02:45:31 <dwfreed> considering the path, I would imagine so, at least at a basic level
2023-01-09 02:46:32 <Saklad5> I'm sort of nervous about messing with a file that will cut off my access to the server if I get it wrong
2023-01-09 02:46:57 <Saklad5> I guess I can always just make a backup and restore it from scratch if that happens?
2023-01-09 02:46:58 <dwfreed> that's why you have backups
2023-01-09 02:47:02 <dwfreed> https://wiki.alpinelinux.org/wiki/Configure_Networking#IPv4_Static_Address_Configuration
2023-01-09 02:47:30 <dwfreed> a quick review of this section indicates the syntax is identical to the debian version of this file, at least for this basic level of configuration
2023-01-09 03:01:58 <psykose> what that file accepts also depends on having ifupdown-ng or not
2023-01-09 03:02:31 <psykose> Quillith: just to double check, you are in a dbus session right (echo $DBUS_SESSION_BUS_ADDRESS)
2023-01-09 03:08:57 <Saklad5> Is there a way to say that a machine should use a SLAAC address over a static address?
2023-01-09 03:12:21 <psykose> like slaac with static fallback or
2023-01-09 03:13:56 <Saklad5> I want the machine to use a single SLAAC address for pretty much everything, but I also want certain software to listen on static IPv6 addresses
2023-01-09 03:14:36 <Saklad5> Yes, I could just set the address SLAAC would produce statically, but that feels like the wrong approach
2023-01-09 03:17:14 <psykose> the lazy solution is to configure slaac then add extra statics in a post-up hook manually
2023-01-09 03:17:17 <psykose> i don't know of a better one
2023-01-09 03:17:32 <Saklad5> That would make the SLAAC address preferred?
2023-01-09 03:18:13 <psykose> preferred in what sense
2023-01-09 03:18:39 <Saklad5> Software not explicitly configured to do otherwise would bind ports on that address when connecting to remote servers
2023-01-09 03:19:26 <psykose> ah, then no
2023-01-09 03:25:18 <dwfreed> if you add the secondary addresses with preferred_lft 0, then the SLAAC address would be preferred
2023-01-09 03:26:09 <Saklad5> Ooh, thanks
2023-01-09 03:26:21 <Saklad5> …Where did you find this?
2023-01-09 03:29:01 <Saklad5> I want to make sure I understand it myself, before using it
2023-01-09 03:29:26 <Quillith> psykose:  echo $DBUS_SESSION_BUS_ADDRESS gave the following output-> unix:path=/tmp/dbus-RswsNfcRNM,guid=d171d6bac5d5f49e360ab6ec63bb3853
2023-01-09 03:29:47 <psykose> looks fine, so idk
2023-01-09 03:31:24 <Quillith> When I ran a session in gnome it worked fine, so I have to think I'm missing some sort of setting or environment variable. idk, Appreciate you taking a look though,have a happy NY
2023-01-09 03:40:41 <dwfreed> ip-address manpage
2023-01-09 03:40:51 <dwfreed> preferred_lft LFT
2023-01-09 03:40:58 <dwfreed> the preferred lifetime of this address; see section 5.5.4 of RFC 4862. When it expires, the address is no longer used for new outgoing connections. Defaults to forever.
2023-01-09 08:56:43 <uniqNuuid> it is hard to debug sometimes : rc-update -qq add mount-ro shutdown # do not add this, it cause mmc not boot sometimes
2023-01-09 08:57:00 <uniqNuuid> this is the thing that stops shutdown? and mmc does not boot sometimes.. 
2023-01-09 08:58:20 <uniqNuuid> now i removed it , mmc always boots. no matter "reboot" or "poweroff then cold boot"
2023-01-09 08:59:22 <uniqNuuid> ohh. no. reboot failed again..... 
2023-01-09 08:59:33 <uniqNuuid> revert my message
2023-01-09 09:10:20 <uniqNuuid> in what cases a "reboot becomes poweroff" ?
2023-01-09 09:11:11 <psykose> bad stuff happened cases
2023-01-09 09:14:43 <uniqNuuid> the LED on my device indcate something: when i do 'poweroff', led turns of. when i do 'reboot' and when it failed to reboot, the led is always on. something stuck at shutdown when 'reboot'
2023-01-09 09:15:33 <uniqNuuid> is there an option to force kill everything? so nothing stops shutdown? 
2023-01-09 09:16:27 <uniqNuuid> btw. udev is much faster than mdev on startup time 
2023-01-09 09:18:15 <uniqNuuid> my rc.conf is set to this : https://0x0.st/o7Zv.txt
2023-01-09 09:18:24 <uniqNuuid> i am reading the rc.log to find anything useful 
2023-01-09 09:18:53 <uniqNuuid> when it is on usb boot : reboot, poweroff, coldboot always always work. 
2023-01-09 09:19:33 <uniqNuuid> when it is on mmc : poweroff always work. but reboot or coldboot sometimes does not work... this is really weird 
2023-01-09 09:21:57 <uniqNuuid> and my rc-update list : https://0x0.st/o7Z3.txt
2023-01-09 09:38:58 <uniqNuuid> what is rfkill service ? wifi and bluetooth related?
2023-01-09 09:41:53 <uniqNuuid> aha, is it the acpid that was not enabled? my arm device does not have ACPI at all. 
2023-01-09 09:42:34 <Forza> Hi) (
2023-01-09 09:42:46 <Forza> Hi :) 
2023-01-09 09:44:59 <Forza> I have an annoying bug where my server is rebooting spuriously. Sometimes two weeks past and sometimes just a day. I enabled remote syslog on it and this time I cought a kernel BUG error. But it did not leave a stack trace, but the last message before was from the iscsi server, so maybe it is related. I am using 3.17 with linux-lts 5.15.85-0-lts
2023-01-09 09:45:58 <Forza> I am thinking to upgrade the kernel to Edge which has 6.1.3. would this cause issues, or do I need to move the whole system to edge? 
2023-01-09 09:47:44 <psykose> just the kernel only theoretically should be fine
2023-01-09 09:48:10 <psykose> i'm not going to tell you it's supported but it won't give you dependency conflicts :p
2023-01-09 09:48:42 <Forza> I know it isn't supported, but this issue has been eluding me for 6 months
2023-01-09 09:48:45 <Forza> =(
2023-01-09 09:49:00 <psykose> go for it
2023-01-09 09:49:58 <Forza> Where is the path /var/cache/apk stored? 
2023-01-09 09:51:18 <uniqNuuid> looks like it is the acpid was not enabled. 
2023-01-09 09:54:38 <psykose> stored?
2023-01-09 10:05:09 <ikke> A symlink from /etc/apk/cache 
2023-01-09 10:10:16 <frag> anyone have experience with OKdo x Radxa ROCK 4 SE ? how is it next to rpi4?
2023-01-09 10:21:31 <Forza> ikke: thanks. 
2023-01-09 10:21:42 <Forza> I wonder if this is my problem https://patchwork.kernel.org/project/target-devel/patch/20220613014321.107045-1-yang.lee@linux.alibaba.com/ 
2023-01-09 10:22:15 <Forza> The server did reboot as I booted a windows server which connects to the iscsi target 
2023-01-09 10:23:18 <Forza> BTW, is it OK to share the apk cache directory with other alpine hosts? 
2023-01-09 10:31:31 <Forza> Hmm seems I was on Alpine 3.16. so I upgraded to 3.17 first and rebooted. Now it fails to boot. Seems to get stuck when bring up services. Network scripts failed to bring up interfaces
2023-01-09 10:46:00 <Forza> Hm. Is there a way to prevent networking scripts from running during boot? 
2023-01-09 10:47:13 <Forza> I was able to boot using serial console and see that it is stuck on enabling bond1. It does not accept ctrl-c
2023-01-09 11:25:39 <Forza> Ended up with blacklist=ice to block the network card from being loaded, now it boots 
2023-01-09 11:27:37 <Forza> Maybe it works better with kernel 6.1.3
2023-01-09 11:28:46 <Forza> Nope...  Lock up when network scripts tries to bring up bond1 :(
2023-01-09 11:56:15 <JanJ[m]> so, is there a good replacement on Alpine for virtualbox? 
2023-01-09 11:57:54 <JanJ[m]> i know, there's kvm and qemu, im just hoping for something GUI'ish 
2023-01-09 11:58:06 <ikke> depends on your need, but qemu / libvirt could be
2023-01-09 11:58:29 <ikke> never used it, but libvirt should have a GUI
2023-01-09 11:59:14 <ikke> virt-manager
2023-01-09 11:59:52 <JanJ[m]> ok thanks :)
2023-01-09 12:00:55 <bananicorn> JanJ[m]: There is also gnome-boxes, its GUI might be a bit easier to use than virt-manager :)
2023-01-09 12:01:50 <JanJ[m]> i tried that on Arch, had so much strugles with it.. i droped the whole gnome stack and moved away to kde
2023-01-09 12:01:51 <JanJ[m]> :D
2023-01-09 12:02:10 <JanJ[m]> maybe i should just "bite the bullet" and go full CLI with that
2023-01-09 12:02:21 <Habbie> i like virt-manager
2023-01-09 12:02:42 <Habbie> and if you have virt-manager, you have libvirt, and thus you also have virsh
2023-01-09 12:03:04 <JanJ[m]> :)
2023-01-09 12:03:24 <bananicorn> I had a more stable experience with virt-manager too, just wanted to mention boxes in case you don't know it yet^^
2023-01-09 12:05:19 <bananicorn> These days I just use plain qemu since I don't need a lot of complicated things
2023-01-09 12:05:38 <JanJ[m]> some day i'll get there too ;-)
2023-01-09 12:06:43 <bananicorn> I wrote a little article in case you are interested in how to work with qemu alone
2023-01-09 12:06:45 <bananicorn> https://regrow.earth/blog/2022-10-21_qemu_virtualbox.html
2023-01-09 12:07:28 <bananicorn> For properly managing VMs libvirt/virsh/virt-manager will be more capable though :)
2023-01-09 12:07:44 <Forza> JanJ[m]: libvirt with vit-manager is rather good 
2023-01-09 12:08:32 <Forza> How do I prevent an interface from being automatically brought up during boot, but still can be brought up manually? 
2023-01-09 12:09:01 <JanJ[m]> can't you set it tomanual in /etc/network/interfaces 
2023-01-09 12:09:02 <JanJ[m]> ?
2023-01-09 12:09:12 <Forza> Hm I guess
2023-01-09 12:09:50 <JanJ[m]> https://www.baeldung.com/linux/network-interface-configure
2023-01-09 12:09:55 <Forza> Seems init.d/networking is using ifquery --auto, which is looking only for 'auto' entries 
2023-01-09 12:10:36 <JanJ[m]> so i would disable the auto bond1 line
2023-01-09 12:10:48 <JanJ[m]> had to read back a bit :)
2023-01-09 12:18:36 <Forza> So it worked when I set it to manual 
2023-01-09 12:19:54 <JanJ[m]> Hm
2023-01-09 12:20:31 <Forza> But when I do ifup bond1 the command freezes. In dmesg I have one line "bonding: bond1 is being created..." and nothing more
2023-01-09 12:21:00 <JanJ[m]> Sounds like one of the interfaces is not ready
2023-01-09 12:29:23 <Fxzxmic> There is something wrong with my id command, id -n outputs the help content instead of the correct result, and id -u output is normal.
2023-01-09 12:29:51 <grawity> what's the correct result of id -n?
2023-01-09 12:30:18 <Fxzxmic> username
2023-01-09 12:30:35 <Fxzxmic> like: root
2023-01-09 12:30:47 <grawity> -n never prints that, -u -n does
2023-01-09 12:31:15 <grawity> i.e. -n only works as a modifier for -u/-g/-G
2023-01-09 12:32:32 <bananicorn> Is there a reason why the vim package includes the gvim.desktop file? It does not work without the gvim package, which also includes the gvim.desktop file, so I don't quite understand
2023-01-09 12:32:43 <Fxzxmic> Oh, I see. Ha ha
2023-01-09 12:33:13 <bananicorn> It basically produces a broken .desktop file for vim users that dont have gvim
2023-01-09 12:33:25 <JanJ[m]> But both interfaces are up?
2023-01-09 12:36:20 <Forza> Yes. According to the switch too
2023-01-09 12:36:52 <Forza> Though I can't tell in Alpine because my console is waiting for the command to return 
2023-01-09 12:37:20 <Fxzxmic> The id command instruction is too misleading.
2023-01-09 12:37:47 <JanJ[m]> Can you open an extra console?
2023-01-09 12:38:04 <ptrc> Fxzxmic: that's busybox's fault, both coreutils and the actual posix spec (https://pubs.opengroup.org/onlinepubs/9699919799/utilities/id.html) are quite clear
2023-01-09 12:39:05 <Forza> JanJ[m]: yes but I am silly. I have neofetch showing machine interfaces ip during login, and that stops too
2023-01-09 12:39:56 <Forza> I cycled the port on the switch and I got new lines in dmesg:
2023-01-09 12:40:31 <Forza> 2023-01-09 13:22:50 info kernel: ice 0000:01:00.0 ice0: NIC Link is Down
2023-01-09 12:40:32 <Forza> 2023-01-09 13:22:51 info kernel: ice 0000:01:00.0 ice0: NIC Link is up 25 Gbps Full Duplex, Requested FEC: RS-FEC, Negotiated FEC: RS-FEC, Autoneg Advertised: Off, Autoneg Negotiated: False, Flow Control: None
2023-01-09 12:40:32 <Forza> 2023-01-09 13:22:50 info kernel: ice 0000:01:00.1 ice1: NIC Link is Down
2023-01-09 12:40:32 <Forza> 2023-01-09 13:22:51 info kernel: ice 0000:01:00.1 ice1: NIC Link is up 25 Gbps Full Duplex, Requested FEC: RS-FEC, Negotiated FEC: RS-FEC, Autoneg Advertised: Off, Autoneg Negotiated: False, Flow Control: None
2023-01-09 12:40:45 <Forza> Something is off with the bonding driver perhaps
2023-01-09 12:41:50 <Forza> Maybe I should switch to teaming driver instead of the bonding driver? https://documentation.suse.com/smart/linux/html/reference-network-bonding-teaming-comparison/index.html
2023-01-09 12:43:39 <JanJ[m]> hmmz, link is down usually means cable or other end?
2023-01-09 12:43:58 <JanJ[m]> oh, link is up too
2023-01-09 12:44:01 <JanJ[m]> i need new glasses
2023-01-09 12:44:31 <Forza> :D
2023-01-09 12:44:40 <Forza> Ifup script timed out. I am back in console now
2023-01-09 12:44:48 <Forza> I'll disable neofetch for now
2023-01-09 12:47:34 <Forza> Ifconfig bond1 down hangs too
2023-01-09 12:49:01 <Forza> So definitely a kernel bug. "grep.  /sys/class/net/bond1/*" hangs on one of the files "dormant" 
2023-01-09 12:49:06 <jh81> did you configure the bond itself correctly?
2023-01-09 12:50:48 <Forza> I think so. I have two bonds. Bond0 is OK and bond1 hangs when I try to enable it
2023-01-09 12:51:19 <Forza> Bond0 is two 1gb Intel cards with the igb driver and bond1 is a dual-port 25gb Intel card using the ice driver
2023-01-09 12:52:10 <JanJ[m]> hmz
2023-01-09 12:52:23 <Forza> Ethtool ice0 hangs now too
2023-01-09 12:52:32 <Forza> I'll reboot again and stest anew
2023-01-09 12:52:59 <JanJ[m]> hmz, then we must wait for others, as i'm not very knowledable in regards to bonding on linux
2023-01-09 12:53:42 <JanJ[m]> well, there goes my other personality :|
2023-01-09 12:55:46 <Forza> I'll try the nic in normal mode without bonding
2023-01-09 12:55:58 <Forza> Maybe the card is broken or the firmware is crap
2023-01-09 12:56:05 <JanJ[m]> good idea :)
2023-01-09 12:56:11 <Forza> It's brand new though. 
2023-01-09 12:56:23 <JanJ[m]> maybe its a driver specific issue?
2023-01-09 13:00:48 <Forza> Yea. Seems ethtool ice0 works after reboot before I attempt creating a bond
2023-01-09 13:07:01 <Forza> So it works setting an  IP on ice0 and it is reachable from the network. 
2023-01-09 13:08:19 <Forza> Either a bonding driver issue or a ice driver issue. How annoying. =(
2023-01-09 13:08:23 <JanJ[m]> hmz
2023-01-09 13:16:06 <Forza> https://pkgs.alpinelinux.org/package/edge/community/x86_64/libteam
2023-01-09 13:16:49 <Forza> Maybe try this in the future. First I will see if I can get the Intel drivers compiles and working instead of using the Linux kernel native ones. Perhaps that would be better. 
2023-01-09 14:34:51 <frag> omg, press alt while selecting text in link (at least in qutebrowser) lets you select it instead of clicking it ...
2023-01-09 14:36:07 <aron> frag: omg, it works! thanks
2023-01-09 15:37:41 <JanJ[m]> libvirt works like a charm :)
2023-01-09 15:47:24 <scorpion2185[m]1> ikke  `dmesg | grep "PM: "`  ?
2023-01-09 15:47:32 <scorpion2185[m]1> no output
2023-01-09 15:51:17 <scorpion2185[m]1> <Saijin_Naib> "Might be a missing pinctl module..." <- is linux-lts different from other distro kernel?
2023-01-09 15:58:23 <Habbie> JanJ[m], great :)
2023-01-09 15:59:42 <JanJ[m]> yeah, now the only problem left to figure out is why my KDE desktop shows a disconnected wired network icon, instead of a nice connected wireless icon :)
2023-01-09 16:01:16 <pj[m]> networkmanager + ncsi?
2023-01-09 16:01:32 <JanJ[m]> ncsi?
2023-01-09 16:01:40 <JanJ[m]> Goes to Google
2023-01-09 16:02:02 <pj[m]> add 'microsoft' to that
2023-01-09 16:02:38 <JanJ[m]> i dont have anything microsoft?
2023-01-09 16:02:38 <pj[m]> :P
2023-01-09 16:04:45 <pj[m]> I'm not sure if that's how KDE does it, but networkmanager has `connectivity` setting
2023-01-09 16:04:59 <pj[m]> which checks for existence of an URL
2023-01-09 16:05:18 <pj[m]> on Windows, it's called NCSI
2023-01-09 16:05:23 <JanJ[m]> ahh :)
2023-01-09 16:05:46 <JanJ[m]> well, its more a visual nuissance then a real problem, but still :)
2023-01-09 16:18:45 <Saijin_Naib> <scorpion2185[m]1> "is linux-lts different from..." <- Alpine has its own kernel configuration that doesn't necessarily track upstream kernel config fully.
2023-01-09 16:24:33 <ikke> Any distro has their own config
2023-01-09 17:17:29 <Ermine> JanJ[m]: if you have your network managed by anything but NetworkManager, you may want to remove plasma-nm so that widget disappears
2023-01-09 17:19:01 <JanJ[m]> Ermine: Well, i would prefer to have plasma-nm to manage the network :) as it stands, i need to echo data into some wpa_supplicant file every time i want to change networks
2023-01-09 17:19:38 <Ermine> JanJ[m]: use NetworkManager then.
2023-01-09 17:20:51 <JanJ[m]> I think i tried that, but then i would have no network connection at all :) i need to look into that again it appears
2023-01-09 17:26:07 <uniqNuuid> is vim broken on edge?
2023-01-09 17:26:27 <Ermine> wdym
2023-01-09 17:26:40 <uniqNuuid> vim --clean , and then :q
2023-01-09 17:26:59 <uniqNuuid> it will prompt to save file
2023-01-09 17:27:23 <uniqNuuid> looks like it write stuff already
2023-01-09 17:27:56 <uniqNuuid> this happens on edge repo
2023-01-09 17:42:16 <uniqNuuid> never mind . deleted .viminfo
2023-01-09 17:42:19 <uniqNuuid> fixed 
2023-01-09 18:22:16 <uniqNuuid> aha ,it is tmux broken, exit tmux, vim is fine 
2023-01-09 18:29:32 <scorpion2185[m]1> liske no ouput from `dmesg | grep "PM: "`
2023-01-09 18:29:49 <uniqNuuid> i don't know what is broken, so downgrade to 3.17 
2023-01-09 18:32:51 <Ermine> scorpion2185[m]1: did you mean to tag ikke ? 
2023-01-09 18:33:06 <scorpion2185[m]1> ah yes sorry
2023-01-09 18:33:16 <scorpion2185[m]1> ikke^
2023-01-09 18:39:24 <scorpion2185[m]1> how is the workflow of patching when you package?
2023-01-09 18:47:17 <scorpion2185[m]1> in prepare I have `git init -q && git add . git commit -qm.` , then can I patch from there with sed?
2023-01-09 19:03:17 <ikke> scorpion2185[m]1: the git init part you only do once locally
2023-01-09 19:03:28 <ikke> It's just a way to generate the patch file
2023-01-09 19:05:39 <ikke> so you'd do `abuild unpack`, then cd to $builddir, then run git init -q && git add . git commit -qm., make the changes you want to do
2023-01-09 19:05:49 <ikke> then git diff >../../<filename>.patch
2023-01-09 19:07:34 <scorpion2185[m]1> ah
2023-01-09 19:42:56 <uniqNuuid> is it possible to use another init on alpine?
2023-01-09 19:43:03 <uniqNuuid> like openrc init,
2023-01-09 19:43:16 <Ermine> Yes
2023-01-09 19:44:30 <Ermine> As for openrc init, you need to build one manually
2023-01-09 19:44:45 <uniqNuuid> what other options do we have 
2023-01-09 19:44:55 <Ermine> sinit
2023-01-09 19:45:54 <Ermine> dumb-init
2023-01-09 19:46:03 <uniqNuuid> i don't find the pkgs for sinit
2023-01-09 19:46:21 <ikke> We do not officically support another init atm
2023-01-09 19:46:24 <Ermine> Well that means it isn't here
2023-01-09 19:46:33 <ikke> (does not mean you cannot do it)
2023-01-09 19:47:16 <uniqNuuid> the busybox "reboot" command sometimes does not reboot my box. 
2023-01-09 19:47:29 <uniqNuuid> 20% percentage fail 
2023-01-09 19:48:28 <uniqNuuid> s/percentage/chance/g
2023-01-09 19:48:28 <alpine-meetbot> uniqNuuid meant to say: 20% chance fail 
2023-01-09 19:49:31 <uniqNuuid> for i in {1..30} ; do echo "the $i rebooting test" ; ssh host "doas reboot" ; sleep 60 ; done
2023-01-09 19:50:02 <Habbie> the question is, is that actually 'the fifth reboot works' or 'by that time the first reboot finally kicked in'?
2023-01-09 19:50:08 <uniqNuuid> not always reboot my box, sometimes stuck there... for no reason. 
2023-01-09 19:50:42 <uniqNuuid> once it stucks. the only way is to cold boot. 
2023-01-09 19:52:16 <uniqNuuid> i tried to wait for ... 30mins... no luck. my box is headless. 
2023-01-09 19:52:43 <uniqNuuid> no display. hmm. and i did not setup serial console 
2023-01-09 19:53:39 <uniqNuuid> so i don't know what is going on... that sucks 
2023-01-09 19:55:15 <uniqNuuid> after a cold poweron, i looked at the rc.log and dmesg.... everything is OK
2023-01-09 20:23:20 <Saijin_Naib> > Any distro has their own config
2023-01-09 20:23:21 <Saijin_Naib> Well, yes, that's what I was trying to say 
2023-01-09 20:29:59 <hsv2> how can i list the files belonging to a package?
2023-01-09 20:30:12 <hsv2> looks like apk doesnt do that
2023-01-09 20:30:40 <ikke> apk info -L <pkg>
2023-01-09 20:30:52 <ikke> (package needs to be installed)
2023-01-09 20:31:40 <hsv2> ups, i was wrong. thanks this is what i need, for installed pkgs
2023-01-09 22:40:15 <WhyNotHugo_> How can I get rid of /etc/acpi/PWRF/00000080 and avoid it being installed again after a package update?
2023-01-09 22:40:28 <WhyNotHugo_> I deleted the file yesterday but after an update it's there again.
2023-01-09 22:41:03 <WhyNotHugo_> (I'm handling the power button via my compositor already)
2023-01-09 22:43:45 <zcrayfish> WhyNotHugo_: Personally I wouldn't get rid of it, just comment out the "poweroff" line inside it. When apk tries to reinstall it later, it will should just create an .apk-new file when it sees that you modified it
2023-01-09 23:49:14 <minimal> WhyNotHugo_: that file is provided by the busybox-openrc package
2023-01-09 23:50:59 <minimal> if you delete it then the next time an updated busybox-openrc is installed it will reappear
2023-01-10 01:44:33 <Quill> Hey @psykose, I think I found a solution for the dbus flatpak issue I brought up the other day. I added a Flatpak section to the Sway wiki essentially saying to add the different xdg-desktop-portal frontends to an autostart section in your Sway config file. That seemed to have resolved the issue for me for now. I'm neither a dev nor an expert in this, so I thought I'd let you know in case there's more going on here and it isn't appropriate for the 
2023-01-10 01:44:33 <Quill> wiki. Thanks
2023-01-10 01:45:05 <psykose> i thought they just automatically started
2023-01-10 01:45:06 <psykose> weird
2023-01-10 01:46:39 <Quill> that's what I thought too, but when I remove them from the config file I get a ton of gdbus errors where no interface exists on the path. no clue why this is lol
2023-01-10 01:47:38 <psykose> i don't manually start them but they do indeed all start running
2023-01-10 01:47:51 <psykose> i guess if you ps aux | grep xdg- they're not there if you don't start them?
2023-01-10 01:48:21 <Quill> let me exit a session and try that, brb
2023-01-10 01:54:30 <Quill> I took out the autostarts and tested a flatpak, which gave me the errors. I ran that command and got xdg-document-portal xdg-permission-store xdg-desktop-portal 
2023-01-10 01:54:57 <Quill> and grep xdg-
2023-01-10 01:55:39 <psykose> no -gtk -wlr?
2023-01-10 01:55:46 <Quill> none
2023-01-10 01:55:51 <psykose> and -gnome
2023-01-10 01:55:53 <psykose> weird
2023-01-10 01:55:57 <psykose> but i guess starting it works
2023-01-10 01:56:14 <Quill> no frontends at all, which is weird because I thought the main portal would take care of it?
2023-01-10 01:56:34 <psykose> not sure tbh
2023-01-10 01:56:47 <Quill> I couched the wiki section in "may"s and "might need to"s because apparently this isn't needed across all users hlol
2023-01-10 01:57:44 <Quill> anyway, appreciate your time. Have a good one
2023-01-10 02:01:12 <zpiro> Quill! Noooooo........ I thought I could just be old, like older than dirt and off my lawn.
2023-01-10 02:05:25 <zpiro> Quillith: let's try to couch the grammar interest and documentation engagemnt!
2023-01-10 02:05:42 <Quillith> What do you mean?
2023-01-10 02:06:10 <zpiro> "I couched the wiki section in "may"s and "might need to"s because apparently this isn't needed across all users hlol"
2023-01-10 02:07:37 <Quillith> Oh I just meant that because I'm not sure if this is a universal issue I didn't want the entry to claim it was necessary for 100% of cases. If it would be better I can move the section to "troubleshooting" 
2023-01-10 02:08:03 <zpiro> But, have anyone warned you? The first question, and the only question is whether there is console interface?
2023-01-10 02:08:39 <psykose> what?
2023-01-10 02:08:41 <Quillith> I'm sorry, I'm not sure I understand what you mean
2023-01-10 02:09:34 <zpiro> console being understood as a terminal, and a termional being understood as a keyboard interaction. and terminal being the async of the fact that a console is a terminal with a serial  and deserialisation connection to a processor or operating system.
2023-01-10 02:09:48 <psykose> please don't spam meaningless garbage
2023-01-10 02:10:59 <zpiro> psykose: define terminal and console for me if you please.
2023-01-10 02:11:18 <psykose> troll some other channel and not this one
2023-01-10 02:11:26 <psykose> last warning
2023-01-10 02:13:16 <zpiro> psykose: does it pay well to demand that others make sense and understand what is going on the challen as an irc operator?
2023-01-10 07:24:53 <smallville7123> how do i install qemu
2023-01-10 07:26:19 <psykose> apk add qemu? :p
2023-01-10 07:26:29 <psykose> it's split into a bunch of subpackages so it depends on what you want to do 
2023-01-10 07:27:04 <smallville7123> qemu (no such package)
2023-01-10 07:27:27 <psykose> that sounds like you don't have the community repository in /etc/apk/repositories
2023-01-10 07:27:30 <smallville7123> i currently have `xen-qemu` but im not sure if its the right qemu
2023-01-10 07:28:00 <psykose> no, that's just for xen
2023-01-10 07:28:25 <smallville7123> ok
2023-01-10 07:29:15 <smallville7123> o.o removing an apk actually removes ALL dependancies of that apk
2023-01-10 07:29:45 <smallville7123> on others it just removes only the specified apk, ingoring its dependancies unless explicitly removed as well
2023-01-10 07:29:51 <smallville7123> package*
2023-01-10 07:30:08 <smallville7123> (eg apt, packman, others)
2023-01-10 07:35:00 <smallville7123> i want to use qemu to emulate an AMD cpu
2023-01-10 07:35:34 <infinite[m]> smallville7123: easy
2023-01-10 07:36:07 <infinite[m]> using qemu-system-x86_64 -cpu help
2023-01-10 07:40:00 <psykose> apk del doesn't remove packages, that's why :)
2023-01-10 07:40:15 <smallville7123> o.o
2023-01-10 07:40:43 <smallville7123> unable to find CPU model 'x86 EPYC-v1'
2023-01-10 07:40:59 <smallville7123> even though it appears in --cpu help
2023-01-10 07:41:12 <ikke> smallville7123: https://ariadne.space/2021/04/25/why-apk-tools-is-different-than-other-package-managers/ 
2023-01-10 07:41:47 <psykose> no x86
2023-01-10 07:41:51 <psykose> just EPYC-v1
2023-01-10 07:42:13 <smallville7123> yea
2023-01-10 07:42:20 <psykose> the x86 is just signifying the type
2023-01-10 07:42:57 <smallville7123> seems phenom-v1 is closest to intel, as it only has CPUID.800000001H:EDX.fxsr-opt unsupported
2023-01-10 07:43:16 <smallville7123> tho i get gtk initialization failed
2023-01-10 07:44:01 <psykose> you have to install the subpackage of whatever you use
2023-01-10 07:44:15 <psykose> qemu-ui-gtk is the gtk frontend, etc
2023-01-10 07:45:32 <smallville7123> welp i cannot exit qemu and it has nothing to boot
2023-01-10 07:46:01 <smallville7123> 
ACTION reboots alpine

2023-01-10 07:46:32 <smallville7123> can alpine use the current host filesystem?
2023-01-10 07:46:35 <smallville7123> qemu*
2023-01-10 07:46:47 <smallville7123> eg qemu --cpu $CPU bash
2023-01-10 07:47:04 <smallville7123> to run bash using a specific CPU
2023-01-10 07:56:54 <smallville7123> as im currently running alpine under virtualbox
2023-01-10 08:00:24 <psykose> just use qemu outside virtualbox, why is this so nested
2023-01-10 08:00:34 <psykose> but sure, qemu-user can use QEMU_CPU iirc
2023-01-10 08:00:43 <psykose> you could google for it i don't remember specifically :)
2023-01-10 09:15:27 <smallville7123> how do i create an image to install alpine to
2023-01-10 09:15:41 <smallville7123> with qemu
2023-01-10 09:15:51 <smallville7123> as i get
2023-01-10 09:15:59 <smallville7123> fd0 ( 0.0 GB )
2023-01-10 09:16:06 <smallville7123> sr1 ( 1.1 GB )
2023-01-10 09:16:59 <psykose> like for a persistent vm disk? qemu-img can make them
2023-01-10 09:17:12 <smallville7123> how would i do so
2023-01-10 09:18:12 <psykose> then you have to pass the image to the vm with e.g. -drive file=vm.img,if=virtio
2023-01-10 09:18:18 <psykose> i dunno, plenty of stuff you can find online
2023-01-10 09:18:19 <smallville7123> as i doubt qemu can use an existing virtual box vdi to boot alpine
2023-01-10 09:18:35 <psykose> https://en.wikibooks.org/wiki/QEMU/Images#Creating_an_image
2023-01-10 09:19:03 <psykose> i don't know anything about virtualbox
2023-01-10 09:22:35 <smallville7123> seems like i can convert it to img
2023-01-10 09:22:41 <smallville7123> qcow2 *
2023-01-10 09:44:35 <WhyNotHugo_> zcrayfish, minimal: Thanks
2023-01-10 09:45:15 <WhyNotHugo_> Trying to install a game with wine (probably an unusual thing on Alpine), and I'm getting an error which I suspect related to a missing (optional?) dependency. Any ideas how to figure out if this is the case?
2023-01-10 09:45:17 <WhyNotHugo_> https://www.reddit.com/r/linux_gaming/comments/107plb1/failed_to_load_ntdlldll_error_c0000135/
2023-01-10 09:45:17 <alpine-meetbot> [reddit] failed to load ntdll.dll error c0000135 ('tech support' flair) (self.linux_gaming) | 0 points (0414.0%) | 2 comments | Posted by WhyNotHugo | Created at 2023-01-09 - 20:50:38UTC
2023-01-10 09:46:28 <psykose> it's 32-bit
2023-01-10 09:46:45 <psykose> before like 1-2 versions of wine you'd just get a different failure
2023-01-10 09:47:03 <psykose> the newest versions handle more 32-bit stuff or something, but you still can't run anything 32-bit on 64-only
2023-01-10 09:47:15 <psykose> you need a 32-bit chroot to open that with the 32-bit wine
2023-01-10 09:47:29 <psykose> there's no multilib on musl to facilitate the side-by-side wine+wine64 to work with those
2023-01-10 09:47:39 <psykose> so, only things that are 64-bit windows executables work
2023-01-10 09:47:49 <psykose> which isn't much software :)
2023-01-10 09:48:11 <psykose> even for software that is 64 the installer is 32 half the time, but these days anything new is 64
2023-01-10 09:48:29 <psykose> but making a chroot and sharing some directories does work
2023-01-10 09:50:45 <psykose> personally i would probably flatpak it :p
2023-01-10 09:51:08 <psykose> unless wine gets full 32-bit without host multilib libraries this will always be a pain
2023-01-10 09:51:25 <WhyNotHugo_> Oh, the installer itself is 32bit? Shit.
2023-01-10 09:51:32 <psykose> that error means it is yeah
2023-01-10 09:52:10 <psykose> there's a chance that the output is 64 and you don't need a chroot after installing, but i wouldn't count on it
2023-01-10 09:52:38 <WhyNotHugo_> The output is 64bit, I distinctly remember this game being 64bit.
2023-01-10 09:52:40 <psykose> hm, actually no, the wineprefix would be 32 and you can't mix
2023-01-10 09:52:53 <psykose> not the game, the launcher itself after installing since you need it to open the game
2023-01-10 09:52:57 <WhyNotHugo_> So if the game is 64bit, and the installer is 32bit... what can I do?
2023-01-10 09:53:13 <WhyNotHugo_> It seems I'd need different prefixes?
2023-01-10 09:53:34 <psykose> either a bunch of hacks i don't know of or just flatpakking it (steam has bnet afaik, lutris has a flatpak to install bnet, ..)
2023-01-10 09:53:39 <WhyNotHugo_> Flatpak won't work, some dependencies are outdated and there's no hardware acceleration: https://gitlab.com/freedesktop-sdk/freedesktop-sdk/-/issues/1532
2023-01-10 09:53:54 <psykose> you're on rdna3?
2023-01-10 09:54:07 <psykose> it would, you can use mesa-git
2023-01-10 09:54:16 <psykose> FLATPAK_GL_DRIVERS=mesa-git flatpak run
2023-01-10 09:54:33 <WhyNotHugo_> That's exactly what I'm doing. Steam still reports llvm14.
2023-01-10 09:54:36 <psykose> install mesa-git flatpak for the runtime version, one for 64 and one for 32
2023-01-10 09:54:47 <psykose> the mesa-git is still llvm14
2023-01-10 09:54:53 <psykose> hm
2023-01-10 09:55:01 <psykose> not very convenient
2023-01-10 09:55:20 <psykose> the perils of a niche distro and having hardware released yesterday..
2023-01-10 09:55:42 <WhyNotHugo_> I origianally got an older card but it didn't fit the case so I returned it... and the newer vidfeo card was cheaper :P
2023-01-10 09:56:08 <psykose> all i can think of is updating and building the flatpak images yourself with the new versions, if that even works
2023-01-10 09:56:10 <psykose> and i have never done it
2023-01-10 09:58:34 <WhyNotHugo_> Or I can wait and get back to work I guess. :P
2023-01-10 09:59:10 <psykose> :)
2023-01-10 09:59:11 <WhyNotHugo_> But I take it that wine on alpine would be a pain since there's no multilib.
2023-01-10 10:01:01 <Ermine> Or wine finally converts its stuff to PE
2023-01-10 10:04:19 <psykose> the reason that error changed and looks like that is because they have done most of it :p
2023-01-10 10:04:26 <psykose> it almost looks like it works
2023-01-10 10:04:32 <psykose> before you got 1 line of output and it just failed
2023-01-10 10:05:17 <psykose> does PE conversion remove host 32-bit multilib requirement for e.g. the random libraries in use?
2023-01-10 10:05:47 <psykose> WhyNotHugo_: yeah, it's basically only useful if everything is 64
2023-01-10 10:06:02 <psykose> the mod installer i've used for instance is 64 and runs fine
2023-01-10 10:09:11 <Ermine> psykose: afaik no if something uses .so libraries. But if everything starts to refer only to other PEs, then it would work without multilib.
2023-01-10 10:09:36 <psykose> what is the "something" in this case?
2023-01-10 10:09:50 <psykose> the windows exes obviously don't use any .so's from linux
2023-01-10 10:10:03 <psykose> and all of the wine components are linked to the host libraries for e.g. opengl
2023-01-10 10:11:16 <Ermine> Well, I'm not that knowledgeable
2023-01-10 10:11:55 <psykose> :) that's ok
2023-01-10 10:12:09 <psykose> i don't know how it works at all either and have been wondering for a while
2023-01-10 10:32:52 <hsv2> how do i config directory creation in /var/run? i did some change to run yggdrasil as a normal user but i cant figure out how to create automatically a dir in /var/run so its user can create a socket
2023-01-10 10:34:45 <psykose> put a checkpath in start_pre of the init.d is normally how you'd do it
2023-01-10 10:34:55 <psykose> also it's /run not /var/run technically :p
2023-01-10 10:35:39 <hsv2> well, yggdrasil has /var/run by default.
2023-01-10 10:35:42 <hsv2> thx
2023-01-10 10:36:34 <psykose> same place
2023-01-10 10:36:41 <smallville7123> how do i get qemu to display the alpine systemd startup stuff
2023-01-10 10:37:02 <psykose> the what
2023-01-10 10:38:00 <smallville7123> the OpenRC stuff
2023-01-10 10:38:12 <smallville7123> * Loading hardware drivers ...
2023-01-10 10:38:21 <smallville7123> * Checking local filesystems
2023-01-10 10:38:52 <smallville7123> * Remounting root filesystem read/write ...
2023-01-10 10:38:55 <smallville7123> stuff like thar
2023-01-10 10:39:02 <EtherMan>  as if it did it, but it still stays unmanaged. I've tried with ifupdown=managed and unmanaged but neither has any effect. :/
2023-01-10 10:39:02 <EtherMan> Hey. I'm trying to get networkmanager to work in alpine in a VM but I cannot for the life of me get it to work. If I run 'nmcli g s', I get that state is disconnected, with connectivity unknown. If I manually configure the network with ip commands and run 'nmcli n connectivity check', then I get that connectivity is full, but state stays disconnected. All network interfaces stays unmanaged. Running 'nmcli d set eth0 managed yes' and it acts
2023-01-10 10:39:36 <psykose> did you disable networking first and remove the interfaces from /etc/apk/interfaces
2023-01-10 10:39:42 <psykose> smallville7123: should just be visible as it boots
2023-01-10 10:40:01 <JanJ[m]> oh.. networkmanager.. how the horrors have haunted me!
2023-01-10 10:40:21 <smallville7123> in qemu it isnt]
2023-01-10 10:40:23 <smallville7123> in qemu it isnt
2023-01-10 10:40:41 <EtherMan> etc/apk? I've removed it from /etc/network/interfaces and yes networking and wpa_supplicant are disabled in accordance with the install instructions on the wiki.
2023-01-10 10:41:08 <EtherMan> I have no interfaces in /etc/apk though :/
2023-01-10 10:41:27 <JanJ[m]> i think he ment /etc/network/interfaces ;-)
2023-01-10 10:46:17 <smallville7123> what is that stuff called
2023-01-10 10:47:44 <psykose> you can try `nmcli g logging level debug` and reading some slightly more verbose logs from /var/log/messages
2023-01-10 10:47:57 <psykose> as you try the other commands
2023-01-10 10:49:20 <smallville7123> nmcli: drive with bus=0, unit=0 (index=0) exists
2023-01-10 10:50:23 <pj[m]> nm is unfortunately stupid and broken, on certain machines it just doesn't want to manage my interfaces
2023-01-10 10:52:09 <Ermine> smallville7123: set rc_verbose=yes in /etc/rc.conf
2023-01-10 10:52:16 <psykose> i started it and whatnot and reproduced the same thing as you
2023-01-10 10:52:36 <EtherMan> psykose: With that, then setting managed indeed returns success, even though it's clearly not actually becoming managed :/
2023-01-10 10:52:41 <psykose> then i disabled networking, added the example configuration, restarted networkmanager, and then it worked
2023-01-10 10:53:27 <psykose> maybe you forgot to restart it after the configuration?
2023-01-10 10:53:58 <EtherMan> Naa. Even restarted the VM multiple times.
2023-01-10 10:54:32 <psykose> hmm
2023-01-10 10:54:34 <psykose> sadly no idea then
2023-01-10 10:54:38 <psykose> it's not very debuggable :(
2023-01-10 10:55:23 <Ermine> Maybe you can try connman
2023-01-10 10:57:29 <smallville7123> using graphical mode makes it appear
2023-01-10 10:58:39 <EtherMan> Ermine: connman has nowhere near the features of nm and specifically lacing almost every one that I want.
2023-01-10 10:59:01 <Ermine> Well ok
2023-01-10 11:02:39 <smallville7123> it seems the boot stuff only shows in VGA
2023-01-10 11:05:07 <Ermine> And where do you want it to be
2023-01-10 11:05:34 <EtherMan> Wait... This looks pretty critical... "DBus not available: set HAVE_DBUS in src/config.h"... The error comes from dnsmasq but it's nm trying to constantly start dnsmasq :/
2023-01-10 11:05:58 <Ermine> Did you start dbus?
2023-01-10 11:06:17 <EtherMan> I don't even know what it is
2023-01-10 11:06:29 <Ermine> rc-service dbus start
2023-01-10 11:06:59 <EtherMan> "has already been started"
2023-01-10 11:12:22 <EtherMan> Hm.. dnsmasq service is up and running and nm still wants to start another instance, which can't because adress already in use... ofc...
2023-01-10 11:12:32 <Ermine> Install dnsmasq-dnssec-dbus then
2023-01-10 11:15:06 <psykose> it relies on finding dnsmasq in dbus apparently
2023-01-10 11:15:17 <psykose> not sure why that breaks everything else (since that is definitely optional)
2023-01-10 11:17:14 <EtherMan> Well, that made is stop complaining about that at least... Sadly, it made no difference for getting nm to actually work >_<
2023-01-10 11:28:35 <JanJ[m]> nm sucks :( 
2023-01-10 11:31:46 <JanJ[m]> i have a wifi connection.. but plasma shows only a disconnected wired icon :|
2023-01-10 11:32:01 <JanJ[m]> and for the life of god i can't get it to work
2023-01-10 11:33:01 <satmd> is your wifi interface named eno* or eth*?
2023-01-10 11:33:15 <satmd> then its heuristics might just be stupidly assuming that's always wired
2023-01-10 11:33:17 <JanJ[m]> nope, its named wlan0
2023-01-10 11:33:44 <JanJ[m]> ifconfig shows it just fine :) its a KDE thing i think
2023-01-10 11:34:56 <JanJ[m]> i use networkmanager and wpa_supplicant, installed plasma-nm, lots of other plasma/network related packages
2023-01-10 11:34:57 <JanJ[m]> but im still missing something, somewhere
2023-01-10 11:41:14 <JanJ[m]> funny thing, nmtui shows everything correctly
2023-01-10 11:42:39 <EtherMan> JanJ[m]: nm and wpa_supplicant are mutually exclusive things are they not?
2023-01-10 11:42:56 <psykose> no
2023-01-10 11:43:20 <EtherMan> Hm. Pretty much all guides I find explicitly says to disable wpa_supplicant to use nm
2023-01-10 11:43:26 <JanJ[m]> oh
2023-01-10 11:43:32 <JanJ[m]> well, i can try that too :)
2023-01-10 11:43:48 <JanJ[m]> i had iwd before, so someone suggested switching to wpa_supplicant
2023-01-10 11:43:50 <JanJ[m]> brb
2023-01-10 11:43:53 <psykose> they mean to disable any wpa_supplicant service
2023-01-10 11:43:58 <psykose> networkmanager manages it itself
2023-01-10 11:44:04 <EtherMan> aaah
2023-01-10 11:44:05 <psykose> but it invokes wpa_supplicant for you
2023-01-10 11:44:06 <psykose> or iwd, etc
2023-01-10 11:44:30 <satmd> the guides tell you, so that you don't run two copies of wpa_supplicant :)
2023-01-10 11:44:39 <psykose> e.g. https://superuser.com/a/1338190
2023-01-10 11:45:08 <JanJ[m]> Well im rebooting now, fingers crosses
2023-01-10 11:45:59 <JanJ[m]> I still need to manually mount the nvme partition as /sysroot so im a bit slow
2023-01-10 11:46:15 <psykose> heh, how come
2023-01-10 11:46:28 <hsv2> btw is it possible to set some kind of apk post install hook for a specific package? i would need to reapply cap_net_admin on yggdrasil binary on pkg update…
2023-01-10 11:46:43 <JanJ[m]> but no, that did nothing to solve the wifi problem :|
2023-01-10 11:47:08 <JanJ[m]> psykose, i think because i use refind as efi/bootmanager :)
2023-01-10 11:47:45 <psykose> that doesn't sound like it would break it
2023-01-10 11:48:05 <psykose> i can boot using e.g. just a direct motherboard efi load on nvme fine without any workarounds
2023-01-10 11:48:27 <JanJ[m]> once i fixed these hair tearing problems, i can remove arch and put syslinux on it, im sure that fixes the mounting issue
2023-01-10 11:48:34 <ikke> hsv2: a package can define triggers which get invoked if predefined files/patterns change
2023-01-10 11:48:59 <ikke> hsv2: afaik there is no way to do it without having a package for it
2023-01-10 11:49:11 <hsv2> oh, ok :/
2023-01-10 11:50:31 <hsv2> btw... is it only because upstream do like that or why yggdrasil runs by default as admin instead of using caps like i'm doing?
2023-01-10 11:50:53 <psykose> you could probably update the package init.d in aports to use a user + cap_net_admin as 750 with chgrp for the binary to the user
2023-01-10 11:51:54 <psykose> the systemd unit uses caps https://github.com/yggdrasil-network/yggdrasil-go/blob/develop/contrib/systemd/yggdrasil.service
2023-01-10 11:58:09 <hsv2> oh right... well... i'm a noob with caps, i just found now how to set them on file but to pass them this way at execution... i'll have to dig some doc. what do you mean psykose with 750? setting the binary to be owned by root:yggdrasil for instance?
2023-01-10 11:58:33 <psykose> openrc has a way to pass them
2023-01-10 11:58:43 <psykose> so actually all you have to do is modify the init.d
2023-01-10 11:59:28 <psykose> capabilities="^cap_net_admin"
2023-01-10 11:59:32 <JanJ[m]> WEELLLL.. removing wpa_supplicant from the boot runlevel means i have no wifi AT ALL.. it seems networkmanager is the problem?
2023-01-10 11:59:49 <psykose> sounds like it :)
2023-01-10 11:59:58 <psykose> you need networkmanager-wifi for it to configure wifi, and then some config
2023-01-10 12:02:52 <JanJ[m]> Yeah, i have that :)
2023-01-10 12:02:56 <JanJ[m]> Followed the wiki to the letter
2023-01-10 12:03:20 <hsv2> psykose: great, thanks
2023-01-10 12:19:58 <JanJ[m]> sigh i guess i'll have to live with the fact that i can only use alpine on our servers :(
2023-01-10 12:46:37 <JanJ[m]> Hmz, i suppose i could live with nmtui for the time beeing
2023-01-10 12:50:53 <psykose> you should probably open an issue for it just in case someone can figure it out :)
2023-01-10 13:05:36 <uniqNuuid> looks like chrony does not correct the time, it just reports the time is wrong. have to restart chronyd to correct time. tested by setting a wrong time : date 01100000
2023-01-10 13:05:56 <uniqNuuid> am i missing something
2023-01-10 13:08:51 <psykose> don't think so, i never figured out how it's meant to update either
2023-01-10 13:10:49 <uniqNuuid> hmmm. same here. what about busybox ntpd, i don't think it has this function
2023-01-10 13:11:03 <EF> Chrony doesn't do big changes automagically unless configured to do so but you can force it.
2023-01-10 13:11:09 <uniqNuuid> and opentpd i checked the doc, they don't have such function either 
2023-01-10 13:11:55 <EtherMan> Most ntp services only do minor corrections apart from the first sync. It's supposed to be kept in sync. The local clock is supposed to keep it enough in check in between the checkins
2023-01-10 13:22:38 <EF> uniqNuuid: "chronyc -a makestep" or similar if you need to force a big time jump but it'll not work with chronyd running as well. 
2023-01-10 13:31:18 <uniqNuuid> EF: wonder if it can automatically do it. 
2023-01-10 13:32:04 <uniqNuuid> the thing am concerned is that, when there is no network on boot. then what will happen for these ntp daemons 
2023-01-10 13:32:21 <uniqNuuid> will it sync the time when the network is up?
2023-01-10 13:44:32 <EtherMan> Hmm... Since Networkmanager seems to REALLY not want to cooperate... Is there some other system that handles using dedicated routing tables for different interfaces? So that say all routes relating to eth0 is in table 10 and NOT in main?
2023-01-10 13:45:39 <satmd> coming from gentoo, using wpa_supplicant and openrc there, I know that I can configure routes via /etc/conf.d/net
2023-01-10 13:46:01 <satmd> I do not know wether that applies to alpine in the same way
2023-01-10 13:46:26 <psykose> that sounds like netifrc
2023-01-10 13:47:13 <EtherMan> It's not about configuring routes... I could simply do a startup script for that... It's about certain routes NOT being configured... Such as I do NOT want a 10.250.0.0/24 route in the main table for the interface, which is automatically configured by ifupdown as an example with seemingly no way to avoid it.
2023-01-10 13:48:37 <satmd> in the conf.d-file I can write "a.b.c.d/e table 10" to achieve exactly that
2023-01-10 13:48:45 <EtherMan> psykose: Sounds interesting. I'll have a look. Thanks
2023-01-10 13:49:16 <EtherMan> satmd: That would not stop the route in main that is automatically added by adding the ip to the interface.
2023-01-10 13:49:36 <psykose> i meant the conf.d/net sounds like netifrc, not the solution to the issue
2023-01-10 13:49:41 <psykose> though perhaps it does support that, idk
2023-01-10 13:49:43 <satmd> you are not involving dhcp in this, does it?
2023-01-10 13:50:22 <satmd> if yes, stop dhcp from deploying routes. if no, configure ifupdown to use a /32 subnet instead
2023-01-10 13:51:40 <EtherMan> No dhcp involved no.
2023-01-10 13:52:02 <EtherMan> And the subnet is /24... And I don't want a /32 in the main table either.
2023-01-10 13:53:27 <satmd> configure the interface to something invalid in the main table then
2023-01-10 13:54:16 <satmd> if an interface has an IP address, it will always also have a /32 route for itself
2023-01-10 13:54:20 <EtherMan> That would just result in a broken interface though... I still want a working interface. I just want the routes in a different table that I can place a source rule on.
2023-01-10 13:54:24 <satmd> because it's basically loopback
2023-01-10 13:54:52 <EtherMan> I'm not saying to not have rules... I said I want those rules to be in a different table...
2023-01-10 13:56:26 <uniqNuuid> today is a big day, i fixed the " alpine sometimes does not reboot " issue, https://0x0.st/o7Mi.txt
2023-01-10 13:57:01 <satmd> I do think that if you set ifupdown to use /32 will maybe not look like what you intended, but will actually do what you try to achieve. /32 will not add any route beside its loopback. You then can add your route /24 to table 10 without having it in table main.
2023-01-10 13:57:07 <uniqNuuid> with a tons of boot args : cpufreq.default_governor=schedutil quiet loglevel=3 console=ttyAML0,115200n8 console=tty0 no_console_suspend net.ifnames=0 rootfstype=ext4
2023-01-10 13:57:09 <uniqNuuid> modules=meson-gx-mmc,dwc3-meson-g12a,dwc2,dwc3,mdio-mux-mmioreg,mdio-mux,dwmac-meson8b,dwmac-generic,stmmac-platform,ext4,i2c-mux,i2c-meson,spi-gpio,i2c-gpio,clk-pwm,clk-scpi,scpi-cpufreq,meson-gxl,mdio-mux-meson-g12a,fixed module_blacklist=mc,rfkill,soundcore,lima,meson_ir,drm,rc_core,dw_hdmi_i2s_audio,industrialio zswap.enabled=0 acpi=off debugfs=off numa=off
2023-01-10 13:57:14 <satmd> even if it sounds weird, please give it a try
2023-01-10 14:07:54 <EF> uniqNuuid: you can probably make chronyd wait for a network interface to be up before starting or maybe start it later in the boot process, i.e. default not boot
2023-01-10 14:11:46 <stevenvb[m]> Hello , I am trying to compile lapce (https://lapce.dev/) on alpine ( a rust aplication) but it seems to fail to build openssl-sys/openssl-src-rs crate because the crate checks only on aarch64-unknown-linux-musl and the one for alpine is set to aarch64-alpine-linux-musl.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/QXvJgxsMIGWQpftzTBbbWJXh>)
2023-01-10 14:15:48 <psykose> export OPENSSL_NO_VENDOR=1
2023-01-10 14:15:55 <psykose> apk add openssl-dev
2023-01-10 14:16:47 <stevenvb[m]> my mean question for this channel is why the aarch64-alpine-linux-musl is used and not aarch64-unkown-linux-musl? Is this expected ?
2023-01-10 14:17:06 <stevenvb[m]> Ill try that thanks. openssl-dev was already added.
2023-01-10 14:17:23 <stevenvb[m]> main*
2023-01-10 14:20:49 <psykose> because it's a different triple
2023-01-10 14:20:58 <psykose> and the build fails because the rust ecosystem sucks
2023-01-10 14:22:18 <psykose> and the rustup toolchain fails with __aarch64_ldadd4_acq_rel missing because it's broken
2023-01-10 14:24:18 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/GPukswWKnqwaRSdGbzoEOpNX>)
2023-01-10 14:27:03 <stevenvb[m]> psykose: ah ok thanks then i know its not package related thanks!
2023-01-10 14:27:31 <psykose> you can probably get it to work with something like
2023-01-10 14:28:05 <psykose> RUSTFLAGS="-Clink-arg=-lgcc -Ctarget-feature=-crt-static"
2023-01-10 14:28:09 <psykose> (for the rustup toolchain)
2023-01-10 14:28:32 <psykose> personally i never use them though
2023-01-10 14:28:49 <psykose> or maybe it was -lgcc_s
2023-01-10 14:29:26 <stevenvb[m]> Indeed i rather also not sue the static crt linking. And i would like to use the alpine rust toolchain, also since making a package later will be easier.
2023-01-10 14:29:36 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/rOQXokuCRFUNFeStAsRJKbXR>)
2023-01-10 14:29:39 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/blGaMdERvHTphTeEEGOcLRKp>)
2023-01-10 14:29:43 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JjlXuPSXCySGKmpHxthgiMQn>)
2023-01-10 14:29:46 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/tpOOZnveUsLzXVGDZcwrGYYq>)
2023-01-10 14:29:48 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/nLhAQfBsmgnSZoPwgrUYUUbA>)
2023-01-10 14:29:50 <zewyykrerossiya[m]> Ukraine, US and Europe is invading Russia.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/ePysvVDVxzSJMbBHbqKsZghC>)
2023-01-10 14:30:25 <stevenvb[m]> any admin to block this spam around?
2023-01-10 14:31:00 <lopid> looking it up
2023-01-10 14:32:46 <psykose> lapce is already packaged
2023-01-10 14:33:07 <stevenvb[m]> hmm indeed, it just does not start here on a pinebook pro
2023-01-10 14:33:47 <psykose> time to figure out why :)
2023-01-10 14:33:54 <psykose> whadya see
2023-01-10 14:38:35 <minimal> uniqNuuid: did you try looking at the chronyd.conf manpage? i.e. "initstepslew"
2023-01-10 14:39:05 <stevenvb[m]> (gdb) r
2023-01-10 14:39:05 <stevenvb[m]> Starting program: /usr/bin/lapce 
2023-01-10 14:39:05 <stevenvb[m]> Program received signal SIGILL, Illegal instruction.
2023-01-10 14:40:47 <stevenvb[m]> 0x0000fffff7ab3b70 in ?? () from /lib/libcrypto.so.3
2023-01-10 14:41:34 <stevenvb[m]> how to install debug symbols in alpine?
2023-01-10 14:42:36 <psykose> cont
2023-01-10 14:42:49 <psykose> not the real error
2023-01-10 14:43:17 <psykose> or well
2023-01-10 14:43:32 <psykose> $ ulimit -c unlimited
2023-01-10 14:43:36 <psykose> $ lapce
2023-01-10 14:43:46 <psykose> then when it crashes you'll have a core in ./core
2023-01-10 14:43:57 <psykose> (unless kernel.core_pattern is set)
2023-01-10 14:44:04 <psykose> then you can inspect that
2023-01-10 14:46:36 <stevenvb[m]> wow 36 threads in that coredump :)
2023-01-10 14:47:02 <stevenvb[m]> (gdb) bt... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/ktVmefshlWHYKKvDfOdmdUrQ>)
2023-01-10 14:47:39 <psykose> you need musl-dbg too
2023-01-10 14:47:50 <psykose> but it looks like a random abort from rust
2023-01-10 14:48:11 <psykose> you'll need to actually build it with debug symbols yourself in the apkbuild most likely
2023-01-10 14:48:14 <uniqNuuid> i could make a cron to restart ntpd every 10mins 
2023-01-10 14:48:16 <uniqNuuid> lol 
2023-01-10 14:48:54 <uniqNuuid> minimal: TL;DR
2023-01-10 14:49:08 <uniqNuuid> chrony is too complicated
2023-01-10 14:49:27 <stevenvb[m]> ok thanks for the hints ll try that later today.
2023-01-10 15:17:15 <uniqNuuid> CONFIG_CRYPTO_DEV_AMLOGIC_GXL is not set
2023-01-10 15:20:25 <WhyNotHugo_> There's no support for an encrypted swap in mkinitfs/nlplug-findfs, is there?
2023-01-10 15:21:02 <WhyNotHugo_> I'm thinking of hacking some support. IIRC, the kernel keeps the decryption key for the main partition in its keyring, so it should be pretty straightforward to mount a swap partition that's LUKS-ecnrypted with the same key.
2023-01-10 15:22:59 <minimal> WhyNotHugo_: nlplug-findfs is used to only mount the rootfs, other fs are mounted by openrc init.d script. Now yes swap is not a fs but typically it is also handled by init.d script
2023-01-10 15:23:44 <WhyNotHugo_> The tricky thing about the swap is that it needs to be unlocked pretty early in case we're resuming from suspension.
2023-01-10 15:23:51 <WhyNotHugo_> *from suspend-to-disk
2023-01-10 15:24:11 <WhyNotHugo_> openrc runs a bit too late.
2023-01-10 15:26:29 <minimal> WhyNotHugo_: indeed, however does that work? whilst Alpine does not currently use Secure (UEFI) Boot, that does prevent,via lockdown, suspend to disk AFAIK
2023-01-10 15:27:01 <minimal> I'm not sure about the current situation without Secure Boot if lockdown is enabled
2023-01-10 15:27:38 <pj[m]> <stevenvb[m]> "hmm indeed, it just does not..." <- what do you mean by "not start"?
2023-01-10 15:27:45 <psykose> which lockdown
2023-01-10 15:27:51 <pj[m]> does pinebook pro support OpenGL 3?
2023-01-10 15:28:00 <minimal> psykose: kernel lockdown
2023-01-10 15:28:26 <psykose> pj[m]: not by default afaik
2023-01-10 15:28:29 <psykose> needs PAN_MESA_DEBUG="gl3"
2023-01-10 15:28:34 <minimal> not sure if Alpine enables it
2023-01-10 15:28:41 <pj[m]> lapce won't run without opengl3
2023-01-10 15:28:57 <psykose> probably the issue then
2023-01-10 15:29:04 <minimal> psykose: https://man7.org/linux/man-pages/man7/kernel_lockdown.7.html
2023-01-10 15:29:12 <pj[m]> there is some magic envvar for software accel opengl3
2023-01-10 15:29:20 <psykose> also why does the lapce thing have random forks with like 1 commit added to the tree sitter grammars xD
2023-01-10 15:29:27 <psykose> that env var enabled real gl3
2023-01-10 15:29:30 <psykose> it's just not complete
2023-01-10 15:29:45 <psykose> you can force using llvmpipe, sure
2023-01-10 15:30:07 <pj[m]> LIBGL_ALWAYS_SOFTWARE=1
2023-01-10 15:30:11 <psykose> if you enjoy editing at 3fps :)
2023-01-10 15:30:30 <pj[m]> I mean
2023-01-10 15:30:33 <pj[m]> some people do that
2023-01-10 15:31:48 <pj[m]> psykose: probably because ts grammars are compiled in
2023-01-10 15:31:54 <psykose> not that part
2023-01-10 15:31:59 <psykose> the random fork part
2023-01-10 15:32:11 <pj[m]> you mean forks of lapce/lapce repo
2023-01-10 15:32:17 <pj[m]> or our forks in lapce org
2023-01-10 15:32:58 <psykose> e.g. panekj/tree-sitter-dockerfile
2023-01-10 15:33:13 <pj[m]> because grammars are not added to bindings
2023-01-10 15:33:29 <pj[m]> not grammars, queries
2023-01-10 15:33:36 <pj[m]> *.scm stuff
2023-01-10 15:34:04 <psykose> i'm not sure i understand
2023-01-10 15:34:24 <psykose> it's the same code
2023-01-10 15:34:28 <psykose> just 6 commits behind
2023-01-10 15:34:47 <psykose> aj
2023-01-10 15:34:48 <psykose> i see
2023-01-10 15:34:54 <psykose> hmm
2023-01-10 15:34:58 <psykose> why is that not normally set
2023-01-10 15:35:03 <minimal> WhyNotHugo_: ah, Alpine kernel doesn't have lockdown enabled so that's a non-issue
2023-01-10 15:35:19 <pj[m]> rust bindings can expose highlight/indents/comment/etc queries
2023-01-10 15:35:25 <pj[m]> so it's usually us forking some grammar repo, adding it (commit) and PR'ing to upstream while we use our fork in lapce
2023-01-10 15:35:37 <pj[m]> idk, people forget, some people don't want
2023-01-10 15:35:52 <pj[m]> it also depends on tree-sitter implementation
2023-01-10 15:36:02 <pj[m]> so helix and neovim they all keep own queries
2023-01-10 15:36:06 <WhyNotHugo_> minimal: Not sure how lockdown is related here.
2023-01-10 15:36:06 <psykose> yeah but like.. if it worked, why would they comment it out
2023-01-10 15:36:08 <psykose> weird shit :)
2023-01-10 15:36:16 <pj[m]> it's commented out by default
2023-01-10 15:36:20 <pj[m]> from tree-sitter-cli
2023-01-10 15:36:58 <pj[m]> since queries are not required part of grammar
2023-01-10 15:38:10 <psykose> helix just writes them separately lol
2023-01-10 15:41:46 <pj[m]> yes
2023-01-10 15:41:53 <pj[m]> as I said earlier
2023-01-10 16:12:35 <JanJ[m]> <psykose> "you should probably open an..." <- This is true :) but where do i file an issue?
2023-01-10 16:12:41 <psykose> aports
2023-01-10 16:26:47 <JanJ[m]> https://gitlab.alpinelinux.org/alpine/aports ?
2023-01-10 16:27:27 <ikke> yes
2023-01-10 16:27:36 <JanJ[m]> thanks, just making sure :)
2023-01-10 16:27:48 <JanJ[m]> why not github? out of curiousity?
2023-01-10 16:28:28 <ikke> As an opensource project we like to use opensource software
2023-01-10 16:28:38 <JanJ[m]> fair enough :)
2023-01-10 16:29:42 <psykose> it was github before
2023-01-10 16:29:45 <psykose> and even redmine
2023-01-10 16:29:57 <psykose> from ruby, to ruby, to ruby
2023-01-10 16:29:58 <psykose> :)
2023-01-10 16:30:16 <JanJ[m]> i'm learning Rust, soon i'll need something like this myself :)
2023-01-10 16:30:47 <JanJ[m]> i hear a lot of "hate" since microsoft took over github, so i'm not sure which way to go :)
2023-01-10 16:31:43 <psykose> my repositories are very safe on my hard drive
2023-01-10 16:31:53 <psykose> :p
2023-01-10 16:32:13 <JanJ[m]> that makes collaborating a bit harder, i would imagine ;-)
2023-01-10 16:32:36 <psykose> i don't maintain any projects
2023-01-10 16:32:41 <psykose> as for contributing i use anything
2023-01-10 16:32:43 <psykose> except email
2023-01-10 16:39:36 <JanJ[m]> well, me neither, but hopefully in the future i can contribute to oss projects :)
2023-01-10 16:40:53 <JanJ[m]> after diner i'm going to remove arch from my laptop, mount the home partition to alpine, and be done with it :) 
2023-01-10 16:58:56 <JanJ[m]> Well.. did that fail miserable 🙈
2023-01-10 17:02:01 <donoban> what failed?
2023-01-10 17:02:57 <JanJ[m]> My own brains :( i removed refind, installed syslinux, but somehow im left without bootsble system 🙈
2023-01-10 17:03:26 <JanJ[m]> I better go eat first before i mess up even more
2023-01-10 17:06:30 <JanJ[m]> Plus side: one learns a lot about Linux in general 🙈
2023-01-10 17:09:07 <donoban> :)
2023-01-10 17:09:59 <JanJ[m]> Some day, Alpine will have a liveCD and an installer ;-) somehow
2023-01-10 17:25:38 <elly> the alpine image is already a live USB image and already has an installer
2023-01-10 17:26:31 <ikke> indeed
2023-01-10 17:26:55 <JanJ[m]> Fair enough :) i ment GUI wise :)
2023-01-10 17:29:34 <elly> that would make the live environment quite a bit larger I think
2023-01-10 17:30:50 <JanJ[m]> True :) and i understand its not what Alpine was created for in the first place
2023-01-10 17:33:58 <JanJ[m]> But i do believe there’s a place for Alpine in the desktop world too :) especially because of its small footprint, its security features etc
2023-01-10 17:34:07 <elly> indeed
2023-01-10 17:34:13 <elly> I'm using it on my laptop right now
2023-01-10 17:34:33 <JanJ[m]> Me too :-)
2023-01-10 17:34:43 <elly> one of the difficulties of graphical installers is choosing which graphical environment to use, though - there are many and people have strong opinions between them
2023-01-10 17:34:57 <JanJ[m]> True :)
2023-01-10 17:35:45 <JanJ[m]> But i would argue, that other distro’s have spinoffs with their prefered desktop environments, so why shouldn’t we ,/)
2023-01-10 17:36:33 <elly> because we think it's a bad idea :P
2023-01-10 17:36:51 <elly> (having a preferred desktop environment, and maintaining a distro with it)
2023-01-10 17:39:00 <JanJ[m]> But who said you should do that ;-)
2023-01-10 17:39:37 <psykose> you, saying it should be added? :p
2023-01-10 17:41:00 <JanJ[m]> True, but im not saying YOU should do that ;-)
2023-01-10 17:41:36 <elly> ah
2023-01-10 17:42:27 <psykose> i'm a developer, who else would maintain it
2023-01-10 17:42:28 <JanJ[m]> Some day, i know enough of Alpine to fix that myself 😂
2023-01-10 17:43:53 <elly> this sounds like a fun and educational project
2023-01-10 17:44:05 <psykose> it would be
2023-01-10 17:44:16 <JanJ[m]> I think so too :)
2023-01-10 17:45:40 <JanJ[m]> Especially since i can run Alpine on: my server, my desktop, my mobile
2023-01-10 17:46:00 <JanJ[m]> Its like Apple, only better
2023-01-10 17:59:05 <donoban> who wants a gui installer
2023-01-10 17:59:22 <donoban> what alpine needs is a ncurses based :P
2023-01-10 18:04:45 <JanJ[m]> 😂😂
2023-01-10 18:10:50 <elly> combining the accessibility of a GUI with the modern conveniences of a terminal :P
2023-01-10 18:12:08 <invoked> guis are usually less accessible than tuis
2023-01-10 18:13:35 <JanJ[m]> If its easier to make possible, then its a good start :)
2023-01-10 18:13:40 <hsv2> psykose: where is the openrc capabilities think documented? i dont see it in openrc-run
2023-01-10 18:14:00 <psykose> start-stop-daemon --help
2023-01-10 18:14:27 <elly> invoked: that isn't my experience - if you try navigating a curses app with a screenreader it's often very hard
2023-01-10 18:14:37 <invoked> i'm not talking about curses
2023-01-10 18:14:41 <elly> oh, I was
2023-01-10 18:14:45 <invoked> plain text is accessible
2023-01-10 18:14:48 <elly> replying to the "ncurses based" comment above
2023-01-10 18:14:54 <hsv2> thanks again psykose 
2023-01-10 18:15:22 <invoked> guis just introduce a lot of chrome and positional problems which lead to further complexity when they need screen readers wrapped around them
2023-01-10 18:15:32 <invoked> all because people need things to look pretty
2023-01-10 18:17:43 <JanJ[m]> Well, if i have to spend 8 hours a day looking at something, then yes it has to be pretty
2023-01-10 18:18:09 <JanJ[m]> And since computers arent humans, i prefer beauty over personality
2023-01-10 18:19:23 <elly> this might be a great discussion to take to the -offtopic channel
2023-01-10 18:19:32 <elly> before it turns into "does beauty have value or not?", IRC edition
2023-01-10 18:20:52 <JanJ[m]> ;-)
2023-01-10 18:22:52 <minimal> beauty is in the eye of the beer holder? ;-)
2023-01-10 18:23:11 <JanJ[m]> Im not sure it should be a discussion anyway :-) it’ll be something never ending :)
2023-01-10 19:36:27 <vkrishn> I don't think al devs should take the burden of maintaining the spin-offs, in the way they are intended to run, its the task of the person who starts it. it should be considered a separate distro, based on AL. Pkg maintainers only need to maintain pkgs.
2023-01-10 19:38:25 <uniqNuuid> stopsig="9" set in /etc/init.d/daemon, but the log shows openrc was sending 15
2023-01-10 19:40:16 <uniqNuuid> i am using supervisor="supervise-daemon"
2023-01-10 19:43:23 <JanJ[m]> <vkrishn> "I don't think al devs should..." <- Well, thats what i was trying to say :)
2023-01-10 19:48:00 <uniqNuuid> a bug : https://github.com/OpenRC/openrc/issues/230
2023-01-10 19:49:59 <psykose> it's stopsig=KILL
2023-01-10 19:50:24 <psykose> or rather, it's nothing, never use that
2023-01-10 19:50:26 <psykose> wtf are you doing
2023-01-10 19:51:11 <psykose> ah and s-d doesn't accept it either
2023-01-10 19:54:26 <uniqNuuid> aha i found it supervise_daemon_args="--signal SIGKILL"
2023-01-10 19:55:25 <uniqNuuid> psykose: i tried stopsig="9|-9|KILL|SIGKILL" non works.
2023-01-10 19:55:46 <uniqNuuid> but ^^ above works for upervise_daemon
2023-01-10 19:55:58 <uniqNuuid> s/upervise_daemon/supervise_daemon
2023-01-10 19:55:58 <alpine-meetbot> uniqNuuid meant to say: but ^^ above works for supervise_daemon
2023-01-10 19:56:21 <psykose> yeah and it's beyond idiotic
2023-01-10 19:56:28 <uniqNuuid> why?
2023-01-10 19:56:46 <uniqNuuid> i fount it is pretty slow on stop
2023-01-10 19:57:02 <uniqNuuid> it's my daemon, and i think KILL is fine. 
2023-01-10 19:57:07 <uniqNuuid> not global
2023-01-10 19:59:07 <uniqNuuid> ... KILL is not faster than TERM.. my fault 
2023-01-10 20:32:34 <JanJ[m]> Alpine seems hellbend on scaring me away today
2023-01-10 20:35:24 <psykose> something something mountains
2023-01-10 20:40:36 <JanJ[m]> It keeps throwing “missing bootable device” at me. Tried refind, tried syslinux, tried grub
2023-01-10 20:41:38 <JanJ[m]> i should not have removed the arch partition :X
2023-01-10 20:42:09 <JanJ[m]> oh well, time for bed! Another day another worry
2023-01-10 21:10:50 <uniqNuuid> lol ,zstd -19 takes 100s and makes a 14M .zst, while -10 takes 7s to make a  15M .zst
2023-01-10 21:11:32 <uniqNuuid> raw dir is 43M 
2023-01-10 21:12:35 <uniqNuuid> psykose: alpine really needs to get rid of -19 
2023-01-10 21:12:44 <psykose> ?
2023-01-10 21:13:23 <dwfreed> it's one of those things that runs rarely
2023-01-10 21:13:24 <uniqNuuid> the mkinitfs script 
2023-01-10 21:14:53 <psykose> TIL i can just look at 3 numbers in front of me and make strict global declarations on what should be done
2023-01-10 21:14:59 <psykose> guess my life got so much easier
2023-01-10 21:15:13 <dwfreed> 4 numbers!
2023-01-10 21:15:18 <dwfreed> :D
2023-01-10 21:15:20 <psykose> this message took 300ms ping
2023-01-10 21:15:27 <psykose> dwfreed: you must rewrite the oftc ircd in rust
2023-01-10 21:15:29 <psykose> it's too slow
2023-01-10 21:15:38 <psykose> critical issue
2023-01-10 21:15:42 <dwfreed> it is actually pretty terrible
2023-01-10 21:16:00 <psykose> :)
2023-01-10 21:16:01 <dwfreed> the EU servers handle as many as 10,000 users on a single thread
2023-01-10 21:16:17 <dwfreed> including TLS
2023-01-10 21:16:30 <psykose> that seems not completely terrible
2023-01-10 21:16:46 <dwfreed> the codebase is horrible to work with
2023-01-10 21:16:53 <psykose> yea
2023-01-10 21:17:18 <dwfreed> for a long time we had a broken epoll test in configure, so it was using poll
2023-01-10 21:17:34 <dwfreed> syscall times went down tremendously with that fix
2023-01-10 21:17:42 <psykose> :)
2023-01-10 21:17:45 <psykose> but hey.. free upgrade
2023-01-10 21:17:55 <psykose> if you let things be slow you get to enjoy them being fast again :p
2023-01-10 21:18:22 <dwfreed> https://github.com/oftc/oftc-hybrid/commit/bfb7649c9fce4a60c86a55bfe8b0079308b4ef2f
2023-01-10 21:19:26 <ikke> heh
2023-01-10 22:35:36 <JanJ[m]> Hmz
2023-01-10 22:50:24 <WhyNotHugo_> I guess I need to subscribe to releases upstream for aports I submit... Or make a bot that generates patches when that happens.
2023-01-10 22:51:48 <psykose> you can add them to release-monitoring.org/
2023-01-10 23:18:49 <WhyNotHugo_> psykose: Intersting. Does it have things like webhooks?
2023-01-10 23:19:03 <psykose> don't think so, how come
2023-01-10 23:19:27 <psykose> it feeds the flagged page once a day
2023-01-10 23:22:15 <WhyNotHugo_> I don't have an OpenID account. It killed my server years ago, I though that had died out :/
2023-01-10 23:28:07 <psykose> i have no idea if that even works
2023-01-10 23:28:13 <psykose> i use the fedora one
2023-01-10 23:43:33 <endigma> what is "an OpenID" account
2023-01-10 23:43:37 <endigma> isn't OIDC like a standard
2023-01-11 00:02:54 <WhyNotHugo_> When do packages merged into aports/master get pushed to testing?
2023-01-11 00:03:06 <WhyNotHugo_> Does it just take a while to build and propagate to mirrors, or is there a schedule?
2023-01-11 00:06:30 <psykose> after build
2023-01-11 00:06:46 <psykose> 15min cron
2023-01-11 00:34:55 <stevenvb[m]> <pj[m]> "what do you mean by "not start"?" <- no display and aborted by coredump, but yes probably opengl 3 check, maybe i can add a check and at least a terminal warning ?
2023-01-11 00:50:12 <psykose> did you try  PAN_MESA_DEBUG="gl3"
2023-01-11 00:53:00 <stevenvb[m]> psykose: will try tomorrow , but at least a notification should be nice or maybe fallback to software rendering by default
2023-01-11 01:28:15 <WhyNotHugo_> Some udev rules have `TAG+="uaccess"`. Does this work with eudev? It's intended to give access to "users that are physically present"; how is this determined?
2023-01-11 01:33:58 <psykose> i don't think it does
2023-01-11 01:37:52 <WhyNotHugo_> I've been discussing this with superd/upstream -> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/43063 ...
2023-01-11 01:38:56 <WhyNotHugo_> And I'm playing with the idea of a split package, where waybar-superd has `install_if=waybar superd`, etc.
2023-01-11 01:40:50 <WhyNotHugo_> Any thoughts on this?
2023-01-11 01:48:31 <psykose> no positive ones
2023-01-11 01:48:34 <psykose> hmm, now i'm confused
2023-01-11 01:48:52 <psykose> adding uaccess before broke some stuff, but i can't find a single string for "uaccess" in the eudev codebase
2023-01-11 01:48:59 <psykose> and i can't imagine what else would even care
2023-01-11 01:49:01 <psykose> how weird
2023-01-11 01:50:33 <psykose> i guess stuff like elogind is related to it
2023-01-11 01:55:43 <WhyNotHugo_> 70-u2f.rules from libfido2 has a bunch of uaccess
2023-01-11 01:56:03 <WhyNotHugo_> Also 42-logitech-unify-permissions.rules from solaar
2023-01-11 01:56:42 <WhyNotHugo_> Okay, so I'll just ignore the existence of uaccess in general, that's easier :P
2023-01-11 02:00:29 <WhyNotHugo_> Ah, the only thing that systemd does is apply an ACL on nodes with this tag: https://github.com/systemd/systemd/blob/main/src/udev/udev-builtin-uaccess.c
2023-01-11 02:06:31 <psykose> no i mean something needs to do something with it to have meaning
2023-01-11 02:06:39 <psykose> i would assume eudev did but it does not
2023-01-11 02:06:46 <psykose> so.. why would something break :p
2023-01-11 05:01:18 <uniqNuuid> echo "$ALPINE_REPO"
2023-01-11 05:01:44 <uniqNuuid> the bootarg "alpine_repo" only support one url?
2023-01-11 06:02:12 <psykose> ye
2023-01-11 07:01:24 <uniqNuuid> well, rc-service stop is slow. it takes 1s to stop
2023-01-11 07:01:41 <uniqNuuid> i have like 5 services, so it takes 5s to stop them all 
2023-01-11 07:02:02 <uniqNuuid> any idea how to make it faster?
2023-01-11 07:11:05 <uniqNuuid> turned debug on, actually it is the supervise-daemon that is slow stopping the service. 
2023-01-11 07:12:17 <uniqNuuid> i think i'm get used to systemd's speed. 
2023-01-11 07:13:32 <uniqNuuid> it's blazing fast, these days, i created openrc version of my systemd services. and try to run my 1000 lines of script on openrc
2023-01-11 07:20:37 <uniqNuuid> hmm, i did a test, using rc-service mydaemon start. and then find the 2 pids, one is supervise-daemon's and one is the program's . then i do: kill -15 pid1 pid2 , it takes 0.016 total
2023-01-11 07:20:45 <uniqNuuid> that's wow, fast. 
2023-01-11 07:24:05 <uniqNuuid> i am gonna redo my scripts to not depend on systemd or openrc, just start them traditionally : command args , and kill them using kill.  well, then there is no process monitor. 
2023-01-11 07:44:14 <psykose> what
2023-01-11 11:07:45 <donoban> what takes 0.016? the kill execution or the pid1,pid2 termination?
2023-01-11 11:22:28 <uniqNuuid> donoban: psykose : code: https://0x0.st/o7jx.sh . result : https://0x0.st/o7j4.txt
2023-01-11 11:23:02 <psykose> what about it
2023-01-11 11:24:43 <psykose> optimising the device.. powering off 5 seconds faster is quite the complete waste of time
2023-01-11 11:24:48 <uniqNuuid> if we start or stop multiple services in a script the first one is always fast, but after that, it becomes super slow
2023-01-11 11:25:01 <psykose> and that would be even if it came without any drawback whatsoever
2023-01-11 11:26:23 <uniqNuuid> what do you man 
2023-01-11 11:26:29 <uniqNuuid> s/man/mean
2023-01-11 11:26:29 <alpine-meetbot> uniqNuuid meant to say: what do you mean 
2023-01-11 11:27:02 <ikke> https://xkcd.com/1205/
2023-01-11 11:27:02 <alpine-meetbot> [xkcd] Is It Worth the Time? | Alt-text: Don't forget the time you spend finding the chart to look up what you save. And the time spent reading this reminder about the time spent. And the time trying to figure out if either of those actually make sense. Remember, every second counts toward your life total, including these right now.
2023-01-11 11:27:13 <psykose> i mean that what you're doing doesn't make any sense
2023-01-11 11:27:28 <psykose> there is never a good reason to sigkill daemons to "make shutdown faster"
2023-01-11 11:27:38 <psykose> especially when it's to save like 5 seconds? wtf
2023-01-11 11:27:42 <psykose> you have spent entire hours on this
2023-01-11 11:27:51 <psykose> you will make back this time in like 50 years :p
2023-01-11 11:28:43 <stevenvb[m]> <psykose> "did you try  PAN_MESA_DEBUG="gl3..." <- That indeed makes it work (although the texts are all rendered in black so its not usable),  LIBGL_ALWAYS_SOFTWARE=1 does render correctly . Ill try a to make a patch to see if i can make a notification or maybe fallback to software rendering if that is possible at runtime?
2023-01-11 11:29:00 <psykose> it would be purely an upstream issue
2023-01-11 11:29:07 <psykose> and yeah sounds about right
2023-01-11 11:29:13 <pj[m]> stevenvb[m]: I'm upstream
2023-01-11 11:29:16 <psykose> the gl3 support is incomplete on that gpu so it makes sense
2023-01-11 11:29:23 <psykose> and software being correct too :)
2023-01-11 11:29:25 <psykose> all figured out
2023-01-11 11:30:27 <uniqNuuid> psykose: no, i did not change signal
2023-01-11 11:31:57 <donoban> the first stop took 0.07s, the next 1s... it's probably due some mechanism of synchronization/coherence check
2023-01-11 11:32:03 <uniqNuuid> it is supervise-daemon performance issue. if you think it is not an issue ... i don't know what to say. performance should be consistent stopping each service. 
2023-01-11 11:32:20 <donoban> probably it has that 1s wait somewhere hardcoded, and looks a reasonable wait for waiting a process to stop
2023-01-11 11:33:02 <uniqNuuid> wait for who? the 5 process are not related to eachother... 
2023-01-11 11:33:50 <psykose> you're not just stopping supervise-daemon
2023-01-11 11:33:57 <uniqNuuid> i mean, if it takes 0.07s to stop the first nc, then it should take the same time to stop the rest. 
2023-01-11 11:33:57 <psykose> you're doing rc-service service stop
2023-01-11 11:34:56 <donoban> maybe it just looks the access to rc-service after doing a stop, and there is a stop success fired 1s later
2023-01-11 11:35:14 <donoban> but if you want to know exactly what happens look at the source code or ask open-rc devs
2023-01-11 11:35:16 <uniqNuuid> so? what should i use to stop the service if not using  rc-service ? 
2023-01-11 11:35:26 <alpine-meetbot> donoban meant to say: maybe it just locks the access to rc-service after doing a stop, and there is a stop success fired 1s later
2023-01-11 11:35:26 <donoban> s/looks/locks
2023-01-11 11:38:25 <stevenvb[m]> pj I  see , but i just want to try something to see I can make it work, would you prefer auto switching to software renderer (with or without notification on boot) or just a message when launched via terminal? The last one seems easier. Issue is probably that we cannot do give a dialog to chose software rendering in case opengl 3 is not supported?
2023-01-11 11:39:02 <psykose> i would explain why your reasoning doesn't make sense (for the 10th time in a row) but discussing this is mostly a waste of time
2023-01-11 11:39:23 <psykose> optimise your 5 second shutdown all you wish i guess :)
2023-01-11 11:39:30 <donoban> uniqNuuid: just for curiositty, could you add a 'sleep 1' after each rc-service stop?
2023-01-11 11:40:29 <uniqNuuid> psykose: why do people use zstd?
2023-01-11 11:41:00 <psykose> to compress things, how come
2023-01-11 11:41:16 <uniqNuuid> it's a bug. not that i spend hours to save 5s... it's unfair to say so 
2023-01-11 11:42:09 <uniqNuuid> psykose: why do people use zstd when there is gzip/zip/bzip already 
2023-01-11 11:42:33 <psykose> no idea, you should ask them i think
2023-01-11 11:43:23 <ikke> zstd has properties (reaad trade-offs) that people like
2023-01-11 11:44:01 <psykose> https://img.ayaya.dev/ADPsvRok2G1q
2023-01-11 11:44:05 <psykose> sure do be waiting one second
2023-01-11 11:46:34 <psykose> i can however reproduce it on my aarch64 system :p
2023-01-11 11:46:47 <psykose> and being precisely 1 second makes me think it's some synchronisation-level thing
2023-01-11 11:47:13 <Ermine> syncs across supervise-daemons?
2023-01-11 11:47:32 <psykose> don't think so
2023-01-11 11:47:45 <psykose> but in any case that would take much more investigation to get anywhere
2023-01-11 11:53:39 <psykose> hmm, every machine except mine reproduces it
2023-01-11 11:53:41 <psykose> funny
2023-01-11 11:54:28 <psykose> if you feel like it's something that actually needs fixing then open an openrc issue
2023-01-11 11:54:45 <uniqNuuid> psykose: only on aarch64?
2023-01-11 11:54:50 <psykose> nah
2023-01-11 11:54:53 <psykose> anything
2023-01-11 11:55:54 <uniqNuuid> i tested on x86_64, same 
2023-01-11 12:00:54 <psykose> the specific wait is just in the waitpid() for either handle_signal or waiting for child, but eh
2023-01-11 12:01:13 <psykose> i'm sure you can get your unix hat on and figure out the semantics
2023-01-11 12:54:03 <nikolashachka[m]> Hello guys.  I can't install any bootloader in diskless mode.  It turned out to install only syslinux but at 50% (it starts on one computer, but not on the other).  Now I have a fat32 128M partition, I would like to install grub2 on it, how can I do this?
2023-01-11 12:54:22 <nikolashachka[m]> 
ACTION uploaded an image: (127KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/hzfOwEbgdoHQTyKXErkaLTIU/IMG_20230111_175050.jpg >

2023-01-11 12:58:11 <nikolashachka[m]> I follow this instruction: https://wiki.alpinelinux.org/wiki/Bootloaders.  But for syslinux I don't understand the extlinux --install /boot line.  boot should be mounted to the right partition on the disk, but it doesn't work for me?  For Grub, it also does not turn out to be visible in the attached photo, I am installing Alpine for the first time.
2023-01-11 13:04:31 <nikolashachka[m]> Despite the fact that both computers have bios (no UEFI).  But on one computer the Linux lts window opens and on the other it doesn't.
2023-01-11 13:05:40 <uniqNuuid> nikolashachka[m]: mount failed...
2023-01-11 13:08:06 <nikolashachka[m]> It feels like syslinux is not fully installed, I just transferred mbr.bin using DD, but I didn't do extlinux --install /boot because I couldn't mount /boot, I create /boot in /mnt, am I doing it right  ?
2023-01-11 13:10:11 <uniqNuuid> szzzzzzzzzzzzzxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy
2023-01-11 13:10:17 <uniqNuuid> sorry. my cat. 
2023-01-11 13:10:27 <nikolashachka[m]> 
ACTION uploaded an image: (135KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/RJyAZuRSJWsQAZrSBCuEitsi/IMG_20230111_180857.jpg >

2023-01-11 13:10:28 <nikolashachka[m]> blkid doesn't show sdc1 partition at all
2023-01-11 13:10:34 <nikolashachka[m]> And I want to install any bootloader on this partition
2023-01-11 13:10:52 <uniqNuuid> nikolashachka[m]: why not just boot an ISO and run setup-alpine ... it make life easier 
2023-01-11 13:10:55 <nikolashachka[m]> Hahahaha
2023-01-11 13:11:37 <Piraty> uniqNuuid: re xz: this is somewhat popular text: https://www.nongnu.org/lzip/xz_inadequate.html
2023-01-11 13:11:55 <uniqNuuid> nikolashachka[m]: did you format sdc1 ?
2023-01-11 13:12:34 <nikolashachka[m]> uniqNuuid: I don't want to install it to disk, I want to use it in diskless mode and work from RAM, logging changes with lbu commit.
2023-01-11 13:13:38 <uniqNuuid> nikolashachka[m]: diskless mode : 1. you must have to have a kernel, initramfs for bootloader to find. 2. you must have apkovl.tar.gz 
2023-01-11 13:13:54 <uniqNuuid> diskless mode != without any disk. 
2023-01-11 13:14:30 <nikolashachka[m]> uniqNuuid: Try to re-format sdc1 to fat32?  Then the question is how is it more correct to do this through mkfs.vfat or fdisk?
2023-01-11 13:14:59 <uniqNuuid> fdisk does not format partitions. 
2023-01-11 13:15:08 <uniqNuuid> you need mkfs.vfat 
2023-01-11 13:17:06 <nikolashachka[m]> uniqNuuid: But I like that I can eject the Flash and the operating system continues to work, but the truth can break the lbu after that (
2023-01-11 13:25:15 <nikolashachka[m]> Indeed, after repeated mkfs.vfat, it turned out to mount the disk partition!  I'm glad, thanks.  I decided to try installing syslinux again, but a new error appeared, what's wrong?
2023-01-11 13:25:25 <nikolashachka[m]> 
ACTION uploaded an image: (126KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/QsmfoXZvNaBTQPfdYGtuKkGV/IMG_20230111_182329.jpg >

2023-01-11 13:27:24 <psykose> inb4 has to be /mnt/boot not mnt/boot
2023-01-11 13:27:54 <nikolashachka[m]> And although I did not transfer mbr.bin using DD, but will this probably break the file system again?  It turns out you need to install Grub?
2023-01-11 13:30:05 <nikolashachka[m]> psykose: tried it, same thing.
2023-01-11 13:35:18 <nikolashachka[m]> The grub gives an error such as if I have UEFI, although I have a BIOS.  It's not clear to me, should I specify / sdc1 or / sdc in the path?
2023-01-11 13:35:19 <nikolashachka[m]> 
ACTION uploaded an image: (121KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/BTkeVldUxMPBWJBLSdkyKznZ/IMG_20230111_183321.jpg >

2023-01-11 13:36:57 <psykose> you probably need to do --target=i386-pc
2023-01-11 13:40:41 <nikolashachka[m]> Thank you, progress has gone, but again an error)
2023-01-11 13:40:48 <nikolashachka[m]> 
ACTION uploaded an image: (126KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/hCHDmmChXEUAxaqqmQwyQuaI/IMG_20230111_184005.jpg >

2023-01-11 13:41:35 <psykose> forgot that one
2023-01-11 13:41:55 <nikolashachka[m]> If I just specify /sdc instead of /sdc1, will Grub be able to determine for itself that it needs to be installed on the sdc1 partition?
2023-01-11 13:42:30 <psykose> perhaps apk add eudev; rc-service udev-trigger start; rc-service udev-settle start; then try again
2023-01-11 13:42:34 <telmich> what is your partition layour, nikolashachka ?
2023-01-11 13:42:48 <psykose> it was somethin like that the last time i saw someone with that issue
2023-01-11 13:44:17 <JanJ[m]> boot and me arent exactly friends today either
2023-01-11 13:44:28 <JanJ[m]> but at least i got myself back into my alpine install now :D
2023-01-11 13:44:39 <nikolashachka[m]> 
ACTION uploaded an image: (101KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/qOoCoCVATIaglpTIlCUvLYnn/IMG_20230111_184408.jpg >

2023-01-11 13:45:15 <nikolashachka[m]> telmich: Like this
2023-01-11 13:46:00 <nikolashachka[m]> JanJ[m]: :D
2023-01-11 13:49:18 <uniqNuuid> guys, why don't you use setup-alpine?
2023-01-11 13:51:03 <nikolashachka[m]> 
ACTION uploaded an image: (141KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/RnLTyWPIjdgoMjDNuNdDpefI/IMG_20230111_185034.jpg >

2023-01-11 13:51:06 <nikolashachka[m]> Still didn't work
2023-01-11 13:51:40 <psykose> then i remember wrong :)
2023-01-11 13:52:12 <nikolashachka[m]> uniqNuuid: But then it won't work from RAM?
2023-01-11 14:04:02 <JanJ[m]> I have tried Grub, Syslinux, Refind, i just cant get Alpine to boot anymore 🙈
2023-01-11 14:05:02 <nikolashachka[m]> I read that it is possible to install Grub from under chroot, but how to use chroot in Alpine Linux?
2023-01-11 14:05:09 <nikolashachka[m]> 
ACTION uploaded an image: (142KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/tSbmkaRGUNYbyNvkyutyzncH/IMG_20230111_190353.jpg >

2023-01-11 14:05:53 <JanJ[m]> Did you look at the wiki article about the bootloaders?
2023-01-11 14:06:05 <JanJ[m]> I think you need to mount several directories first
2023-01-11 14:06:21 <JanJ[m]> But im having my own issues with booting 🙈
2023-01-11 14:06:24 <JanJ[m]> So i cant be sure
2023-01-11 14:06:37 <uniqNuuid> nikolashachka[m]: setup-alpine supports diskless mode 
2023-01-11 14:07:36 <uniqNuuid> nikolashachka[m]: what is your hardware? 
2023-01-11 14:09:17 <nikolashachka[m]> uniqNuuid: Do you suggest using two flash drives and choosing a flash drive on which to install the system in setup-alpine mode?  It's just that if you do it from the same flash drive with which I install the system (although it works from RAM), but setup-alpine does not prompt me to select a flash drive.
2023-01-11 14:09:53 <uniqNuuid> one flash is enough. 
2023-01-11 14:10:24 <uniqNuuid> you can use any partition of that drive as storage to save apkovl.tar.gz 
2023-01-11 14:10:39 <uniqNuuid> you just need to mount it to  /media/name
2023-01-11 14:11:03 <nikolashachka[m]> uniqNuuid: First computer: Asus X555l.  The second ancient computer: Acer aspire 5315. There is also UEFI but I have not tried it yet.
2023-01-11 14:13:33 <nikolashachka[m]> uniqNuuid: It does not work for me.  But I found a crutch, I take out and insert the flash drive back and dump the / media / sdXY folder and mount it
2023-01-11 14:13:38 <uniqNuuid> just boot with ISO and use setup-alpine, when it ask questions about which mode, you don't use 'sys' 
2023-01-11 14:13:54 <uniqNuuid> anything other than 'sys', is diskless 
2023-01-11 14:15:31 <nikolashachka[m]> uniqNuuid: Thanks, I'll try again.
2023-01-11 14:17:35 <uniqNuuid> nikolashachka[m]: my note : https://0x0.st/oRhU.txt : switch from sys mode to diskless mode
2023-01-11 14:17:41 <uniqNuuid> manual :)
2023-01-11 14:18:03 <uniqNuuid> a fresh note i made few days ago while experiment with diskless mode. 
2023-01-11 14:25:28 <uniqNuuid> psykose: ssd is the same : code https://0x0.st/o7eY.sh result : https://0x0.st/o7eE.txt
2023-01-11 14:25:35 <uniqNuuid> i think it is an openrc thing. 
2023-01-11 14:26:23 <pj[m]> <JanJ[m]> "I have tried Grub, Syslinux..." <- nvme drive?
2023-01-11 14:28:28 <satmd> 
ACTION has given up on classical non-uefi bios with mbr, because BIOS/MBR is just emulated in recent firmware and one is better of dropping the legacy. I'm not saying that UEFI/FAT32/X509/... makes me really happy.

2023-01-11 14:29:30 <uniqNuuid> uboot is the best thing ever. 
2023-01-11 14:29:41 <pj[m]> windows is the best thing ever.
2023-01-11 14:30:07 <uniqNuuid> well, speaking of windows, i'd recommend Mac. 
2023-01-11 14:30:39 <satmd> windows is great for downloading ISOs of any real OS :P
2023-01-11 14:30:44 <pj[m]> You might not believe this, but I'm actually typing from macOS
2023-01-11 14:31:16 <uniqNuuid> and ARM Mac is the best of best. 
2023-01-11 14:31:26 <uniqNuuid> which is what i am using as desktop. 
2023-01-11 14:31:36 <uniqNuuid> i never used linux as desktop. 
2023-01-11 14:31:46 <uniqNuuid> but linux is super useful for servers. 
2023-01-11 14:32:19 <elly> alpine works pretty well as a laptop OS too :)
2023-01-11 14:32:46 <uniqNuuid> my alpine on arm a53 uses 48M memory out of 1936M
2023-01-11 14:37:09 <elly> that part's nice, indeed
2023-01-11 14:39:44 <JanJ[m]> HAH! after medling with refind and syslinux most of today, im finally back in my alpine laptop :)
2023-01-11 14:39:51 <JanJ[m]> now i need to get rid of refind
2023-01-11 14:39:52 <elly> excellent
2023-01-11 14:40:08 <elly> if you have a blog somewhere I encourage you to write up your meddling somewhere it will get indexed, so others may find it :)
2023-01-11 14:40:17 <JanJ[m]> good idea
2023-01-11 14:40:33 <JanJ[m]> i should start blogging indeed
2023-01-11 14:42:39 <JanJ[m]> i still have a domain laying around that i do nothing with, might as well put it to good use :-)
2023-01-11 14:43:51 <elly> writing about your experiences, and especially how you overcome problems, is also a good way of contributing to the community
2023-01-11 14:43:54 <elly> plus it's fun :)
2023-01-11 14:44:24 <uniqNuuid> JanJ[m]: reEFInd is bloated.
2023-01-11 14:45:02 <JanJ[m]> that may be, but it got the job done
2023-01-11 14:45:39 <elly> I know that feeling of installing some huge piece of software to do one task and then uninstalling it
2023-01-11 14:45:44 <elly> me and libreoffice like 4-5 times a year
2023-01-11 14:56:10 <nangel> elly:  I have a bunch of KVM containers for exactly that reason.
2023-01-11 14:57:43 <elly> yeah
2023-01-11 14:57:53 <elly> I have an ubuntu vm I use for that kind of thing :P
2023-01-11 14:58:01 <nangel> +1
2023-01-11 15:09:57 <uniqNuuid> i never want to go back to debian/ubuntu after alpine. elly 
2023-01-11 15:10:29 <uniqNuuid> they are such bloated and slow piece of xx
2023-01-11 15:55:54 <scorpion2185[m]1> is there something like df ?
2023-01-11 15:56:01 <scorpion2185[m]1> I tried to search df command
2023-01-11 15:56:55 <minimal> df is like df, in fact it is virtually indistinguishable from it
2023-01-11 15:57:57 <ikke> Busybox has df, but there is also coreutils 
2023-01-11 15:58:05 <scorpion2185[m]1> so no way 100% impossible to use nvidia driver blob?
2023-01-11 15:58:55 <scorpion2185[m]1> ah yes https://pkgs.alpinelinux.org/contents?file=%2Fusr%2Fbin%2Fdf&path=&name=&branch=edge
2023-01-11 16:00:21 <scorpion2185[m]1> it is just in /bin/df
2023-01-11 16:00:35 <elly> nvidia does ship binary drivers, I guess you could download and use them?
2023-01-11 16:01:53 <scorpion2185[m]1> i was told that they don't work cause of musl
2023-01-11 16:02:55 <elly> oh right, I forgot the userspace part, sorry
2023-01-11 16:03:02 <elly> yeah, there is no way they will work I don't think
2023-01-11 16:08:20 <JanJ[m]> networkmanager frustrates me to no end
2023-01-11 16:08:30 <ptrc> elly: one *could* make it work, but it would require way more effort than just meddling with gcompat (which is more or less what i did https://twitter.com/ptrcnull/status/1575597520574906369)
2023-01-11 16:09:10 <elly> ptrc: how would you go about doing that?
2023-01-11 16:12:29 <ptrc> hmm, not sure tbh; i got stuck on nvidia userspace libraries returning "initial-exec TLS resolves to dynamic definition" on load
2023-01-11 16:21:38 <scorpion2185[m]1> gcompact?
2023-01-11 16:22:39 <scorpion2185[m]1> ptrc did they work fine? only on x11?
2023-01-11 16:23:25 <ptrc> no, they didn't work fine
2023-01-11 16:23:46 <ptrc> and gcompat is a library that adds functions from glibc that are missing in musl
2023-01-11 16:25:17 <scorpion2185[m]1> I have no experience with that, is it fixable?
2023-01-11 16:25:47 <scorpion2185[m]1> and with every update do you need to fix it again?
2023-01-11 16:26:25 <elly> it doesn't *sound* fixable
2023-01-11 16:29:19 <scorpion2185[m]1> if that  is way easier, it  looks really hard to do then
2023-01-11 16:31:18 <donoban> ptrc: I'm not sure if that kind of error can be bypassed with LD_PRELOAD
2023-01-11 16:31:33 <ptrc> not with LD_PRELOAD, not really
2023-01-11 16:41:17 <Ermine> minimal: re df: XD
2023-01-11 16:54:21 <JanJ[m]> anyone with syslinux knowledge around?
2023-01-11 16:54:43 <JanJ[m]> because right now i have copied vmlinuz-lts and initramfs-lts to the 
2023-01-11 16:55:02 <JanJ[m]> ESP partition, but i would prefer them to stay on the alpine partition in the normaal /boot directory?
2023-01-11 16:56:22 <JanJ[m]> hmz, it seems unpossible to do :(
2023-01-11 17:01:53 <vkrishn> I am not in touch with wiki.a.o, but fished some QAs, and modified it to best of my understanding
2023-01-11 17:01:59 <vkrishn> could someone pls see if its worth adding to https://wiki.alpinelinux.org/wiki/Alpine_Linux:FAQ
2023-01-11 17:02:03 <vkrishn> new ones are here, https://tpaste.us/a0ll 
2023-01-11 17:12:09 <ikke> vkrishn: "What are preferred tool/utils for doing things in Alpinelinux?" why not suggesting https://pkgs.alpinelinux.org?
2023-01-11 17:15:18 <vkrishn> ikke: ok for link, but mentioning /main or /community should be better
2023-01-11 17:18:36 <vkrishn> .. /main over^
2023-01-11 17:22:10 <vkrishn> eg there maybe replacement for wpa_supplicant, but pkg is in community
2023-01-11 17:23:32 <ikke> vkrishn: maybe refer to https://docs.alpinelinux.org/user-handbook/0.1a/Working/apk.html?
2023-01-11 17:26:36 <vkrishn> how to get which repo from 'apk search' ?
2023-01-11 17:27:26 <vkrishn> yes, news docs are nicer
2023-01-11 17:27:31 <vkrishn> new
2023-01-11 17:28:00 <ikke> apk search only lists packages
2023-01-11 17:28:11 <ikke> you could use apk policy <package> to find from what repo it would be installed
2023-01-11 17:29:01 <vkrishn> yes, thanks, I will modify that paragraph
2023-01-11 17:34:06 <vkrishn> ikke: https://tpaste.us/qaDB
2023-01-11 17:46:35 <clius[m]> Hello, when I boot the live iso I have no internet connection even after running setup-interfaces. Does anyone else also have this problem?
2023-01-11 17:48:31 <elly> are you on a laptop with a wifi card and no ethernet?
2023-01-11 17:50:28 <clius[m]> No on a desktop pc with ethernet and wifi
2023-01-11 17:51:53 <elly> hm! what does ifconfig -a tell you?
2023-01-11 17:54:06 <clius[m]> Ah I just found out that my router just doesn't have internet. But thanks for the answer anyway.
2023-01-11 17:54:56 <elly> oh :P
2023-01-11 18:04:46 <JanJ[m]> if only all our solutions where that simple :)
2023-01-11 18:14:08 <vkrishn> most new users find al, and have a ureka moment like \o/
2023-01-11 18:14:38 <JanJ[m]> well.. ive used up all my eureka moment for this year :|
2023-01-11 18:14:43 <vkrishn> and don't realize that AL similicity !=== easyness
2023-01-11 18:15:28 <JanJ[m]> i think Alpine can be both simple and easy :) it just depends on what you're used to
2023-01-11 18:15:53 <vkrishn> easyness largely depends on the docs
2023-01-11 18:16:44 <vkrishn> simplicity comes from developers supplied means/methods
2023-01-11 18:16:51 <JanJ[m]> :) true
2023-01-11 18:17:26 <minimal> JanJ[m]: why are you using Syslinux with UEFI? It doesn't support it very well, that's why Alpine's setup-disk uses grub-efi for UEFI systems
2023-01-11 18:18:11 <JanJ[m]> well.. because grub failed me :) and i dont have the time to format the drive and start all over again
2023-01-11 18:18:42 <JanJ[m]> but yeah, i'll have to look into grub some more, some day
2023-01-11 18:19:18 <minimal> JanJ[m]: I really wouldn't recommend Syslinux for UEFI, I got it to work in the past but it is messy/has limitations
2023-01-11 18:19:51 <JanJ[m]> well, i currently have refind, which boots syslinux which boots alpine
2023-01-11 18:21:35 <minimal> JanJ[m]: and then you find the kernel/initramfs ESP vs /boot problem.....lol
2023-01-11 18:21:58 <JanJ[m]> hm?
2023-01-11 18:22:27 <minimal> "i would prefer them to stay on the alpine partition in the normaal /boot directory"
2023-01-11 18:22:52 <minimal> I hacked around this in the past
2023-01-11 18:22:54 <JanJ[m]> yes, well.. if thats not possible with syslinux, then one has to look for alternatives hmm :)
2023-01-11 18:23:25 <JanJ[m]> so maybe in a few weeks i'll try to fix grub again :) but for now im rather tired of messing around with the system
2023-01-11 18:26:54 <JanJ[m]> i managed to get rid of Arch, and somehow i "fixed" the boot problem where i had to manually mount /sysroot, so at least everything boots to SDDM now :)
2023-01-11 18:28:45 <vkrishn> last month I got over my eureka moments with mobile+alpine
2023-01-11 18:29:07 <vkrishn> JanJ[m]: did you know that you can boot an .iso file from mobile phone ?
2023-01-11 18:29:20 <vkrishn> without dd`ing into anything
2023-01-11 18:30:08 <vkrishn> all fight with laptop gone in few seconds, with new ones begun
2023-01-11 18:32:08 <JanJ[m]> i didn't :)
2023-01-11 18:32:19 <vkrishn> now I just download new iso directly into one mobile, attach usb to any machine and boot to al
2023-01-11 18:33:31 <JanJ[m]> i use Ventoy :)
2023-01-11 18:35:42 <minimal> JanJ[m]: I didn't say it wasn't possible, I mentioned I'd hacked around it in the past
2023-01-11 18:35:47 <vkrishn> I think I came across this, but new method was then already working
2023-01-11 18:36:06 <JanJ[m]> well, i dont like hacking much :) 
2023-01-11 18:42:22 <JanJ[m]> at some point i got grub-efi to install, but it only added a "boot to firmware" entry to it, thats when i got fed up :)
2023-01-11 18:43:21 <eletrotupi> I'm having a hard time to figure it out why my cron jobs are not running, on a regular, non-container scenario, simply adding a dummy crontab for a user doesn't work, even if I restart the crond service (I noticed that it outputs that it doesn't have inotify support?) 
2023-01-11 18:43:37 <eletrotupi> Can anyone lend me a hand or some ideas on how to debug it further?
2023-01-11 18:44:49 <JanJ[m]> i think you can have cronjob mail you if anything goes wrong, maybe that helps?
2023-01-11 18:45:36 <JanJ[m]> https://serverfault.com/questions/691033/how-to-send-an-e-mail-after-a-cron-job that should at least get you some info?
2023-01-11 19:31:24 <JanJ[m]> so, not Alpine specific, but is there a way to make VM's in virtmanager automaticly resize when i resize the windows of the vm?
2023-01-11 21:03:23 <PureTryOut> Is it possible to run a one-off command on every boot and shutdown with OpenRC?
2023-01-11 21:05:27 <ikke> The local service starts scripts in /etc/local.d/ on start
2023-01-11 21:06:09 <ikke> And apperently when it's stopped it will execute /etc/init.d/*.stop
2023-01-11 21:06:27 <ikke> Oh, /etc/local.d/*.start on start
2023-01-11 21:06:39 <dwfreed> and /etc/local.d/*.stop on stop (not init.d)
2023-01-11 21:06:50 <ikke> euh, yes, thank you
2023-01-11 21:07:02 <ikke> my fingers assuming what my brain wants to type :P
2023-01-11 21:10:16 <dwfreed> note that they need to be executable, but they don't need to be shell
2023-01-11 21:27:39 <PureTryOut> ikke: oh that's great, thanks!
2023-01-12 00:15:57 <stinter> hello, during the 3.17.1 virtual install, is there a way to do custom disk/partitions layout? I have created a virtual disk for each of / /home /tmp and [swap]
2023-01-12 04:30:22 <oreo> Hello #alpine! I have an issue with SPF that I would like help investigating. I'm getting an incorrect behavior (fail) with Alpine while on Debian I correctly get a pass. I wonder if this is related to DNS resolutions on busybox or musl. More details: https://pastebin.com/wSEuVS0v
2023-01-12 04:46:41 <psykose> 159.53.49.231 is not in that TXT record?
2023-01-12 04:54:17 <psykose> ah i see
2023-01-12 04:54:22 <psykose> the a: is ignored
2023-01-12 04:59:00 <oreo> @psykose, it's not ignored -- there is an issue with the DNS lookup for it
2023-01-12 04:59:23 <psykose> ye
2023-01-12 04:59:26 <oreo> spf_dns.c:52         Debug: DNS[cache] lookup: spf.jpmchase.com A (1)
2023-01-12 04:59:26 <oreo> spf_dns.c:52         Debug: DNS[resolv] lookup: spf.jpmchase.com A (1)
2023-01-12 04:59:26 <oreo> spf_dns_resolv.c:310 Debug: query failed: err = -1  Address not available (4): spf.jpmchase.com
2023-01-12 04:59:56 <oreo> i get proper results when querying with "dig" however
2023-01-12 05:02:22 <psykose> for some reason even dig +notcp works, i first assumed it was the classic empty truncation
2023-01-12 05:02:25 <psykose> but it doesn't seem to be
2023-01-12 05:02:39 <psykose> the only way i can reproduce it is with `busybox nslookup <domain> 1.1.1.1`
2023-01-12 05:03:18 <psykose> any other client works regardless of custom tcp fallback or not
2023-01-12 05:03:31 <psykose> not sure where the bug is this time
2023-01-12 05:04:40 <oreo> @psykose does that mean it's likely busybox? (and i can try to look them in?)
2023-01-12 05:04:45 <psykose> no
2023-01-12 05:05:00 <psykose> the tooling doesn't use busybox haha
2023-01-12 05:09:24 <oreo> @psykose what do you suggest then?
2023-01-12 05:11:41 <psykose> suggest for what
2023-01-12 05:16:31 <oreo> @psykose on where to go from there. you don't consider this a bug?\
2023-01-12 05:16:38 <psykose> it is
2023-01-12 05:17:33 <oreo> so what do you recommend for next step(s)?
2023-01-12 05:18:30 <psykose> as a workaround you can use a local dns cache
2023-01-12 05:25:05 <oreo> @psykose you mean set up a DNS server within the container? how would that help here?
2023-01-12 05:25:46 <psykose> to set up something like dnsmasq and have 127.0.0.1 as the only nameserver /etc/resolv.conf yes
2023-01-12 05:29:48 <oreo> so i have to figure out how to override docker for this
2023-01-12 05:41:42 <oreo> @psykose i just tried to set up dnsmasq (that points to 8.8.8.8) and update /etc/resolv.conf to point to 127.0.0.1. it still doesn't work
2023-01-12 05:43:24 <psykose> works for me
2023-01-12 05:45:29 <oreo> i'm running dnsmasq like this: `dnsmasq --no-daemon --listen-address=127.0.0.1 --server=8.8.8.8`
2023-01-12 05:46:02 <oreo> and /etc/resolv.conf only contains `nameserver 127.0.0.1`
2023-01-12 05:48:10 <oreo> @psykose what am i missing?
2023-01-12 05:57:43 <psykose> i don't know
2023-01-12 05:58:54 <oreo> @psykose would you mind showing how you got it to work?
2023-01-12 05:59:17 <oreo> it does look like it may be an empty truncation, no?
2023-01-12 05:59:28 <psykose> i didn't do anything i just run dnsmasq locally
2023-01-12 05:59:34 <psykose> in funnier news it now fails again anyway
2023-01-12 06:46:17 <oreo> @psykose i think i know why it "worked" for you: you likely first used "dig" which set the dnsmasq cache. when i do it that way it works as well.
2023-01-12 06:46:57 <psykose> aha, you are right
2023-01-12 06:47:01 <psykose> that solves that one
2023-01-12 06:47:06 <uniqNuuid> psykose: dwfreed | uniqNuuid: that is "normal" , dwfreed | if supervise-daemon is too fast at checking if the child has died, then it has to wait a second
2023-01-12 06:47:10 <psykose> and yeah i have no idea why it fails itself
2023-01-12 06:47:10 <oreo> (although it's still truncated, but at least some hosts are returned)
2023-01-12 06:47:30 <uniqNuuid> from #openrc
2023-01-12 06:48:14 <psykose> uuidNuniq: ah, makes sense
2023-01-12 06:48:25 <psykose> oreo: hmm, for me dig doesn't truncate since it uses tcp by default
2023-01-12 06:56:29 <oreo> @psykose, what i mean is that even after feeding the cache, using `busybox resolve` does not show all records
2023-01-12 06:57:05 <psykose> ah that, sure
2023-01-12 07:00:50 <oreo> (which is still a problem when it comes to use-cases like SPF checks)
2023-01-12 08:29:04 <superherointj> Is GCC in Alpine also static?
2023-01-12 08:29:28 <psykose> do you mean is the gcc executable static? no
2023-01-12 08:29:36 <superherointj> Yes.
2023-01-12 09:44:32 <uniqNuuid> psykose: when s6?
2023-01-12 09:49:01 <psykose> as soon as you install it and write the service files i'd guess
2023-01-12 09:49:13 <psykose> there is even an openrc service to run an svscan tree
2023-01-12 10:05:38 <uniqNuuid> openrc is too slow. 
2023-01-12 10:25:19 <psykose> there's an easy way to power off in 0.2 seconds
2023-01-12 10:25:23 <psykose> you pull out the power cable
2023-01-12 10:25:28 <psykose> there, i fixed all your problems
2023-01-12 10:25:53 <Ermine> ping
2023-01-12 10:25:58 <psykose> pong
2023-01-12 10:26:04 <Ermine> thank you
2023-01-12 10:26:44 <ikke> 5 seconds latency
2023-01-12 10:27:34 <Ermine> slooow
2023-01-12 10:29:20 <psykose> must be an openrc irc client
2023-01-12 10:30:04 <Ermine> s6-ircclient when
2023-01-12 11:25:56 <donoban> 
ACTION is looking for shutdown benchmarks -_-

2023-01-12 12:11:46 <psykose> oreo: i did some more testing with the resolver in musl master (will be in 1.2.4, tcp fallback) and it works there, so eventually it'll work i guess :)
2023-01-12 12:16:08 <Ermine> we need bootchart but for shutdowns
2023-01-12 12:16:46 <lucidiot> unbootchart(tm)
2023-01-12 12:23:09 <oreo> @psykose, does that mean that musl 1.2.4 supports TCP DNS?
2023-01-12 12:23:14 <psykose> yes
2023-01-12 12:39:39 <oreo> @psykose, ah that's pretty good news! when do you think a stable alpine release will have it?
2023-01-12 12:39:46 <psykose> next one
2023-01-12 13:01:31 <pj[m]> suckless mindset
2023-01-12 13:13:04 <WhyNotHugo_> The play/pause buttons on my bluetooth headsets don't work. I see this in my /var/log/messages -> https://paste.sr.ht/~whynothugo/abb5a248f7d5cf30bd789438596ccdce8f67cddd
2023-01-12 13:13:18 <WhyNotHugo_> Am I maybe missing an optional dependency?
2023-01-12 13:14:00 <psykose> ["bluez5.dummy-avrcp-player"] = false in wireplumber config
2023-01-12 13:14:30 <psykose> the default is true and it's broken for those headphones
2023-01-12 13:14:40 <psykose> no idea why
2023-01-12 13:18:59 <WhyNotHugo_> Added that to `.config/wirepumber/wireplumber.conf`, restarted wireplumber and reconnected the headphones. Same error.
2023-01-12 13:19:11 <psykose> do you have mpris-proxy running
2023-01-12 13:19:32 <psykose> it's also not .conf
2023-01-12 13:19:57 <psykose> .config/wireplumber/bluetooth.lua.d/60-bluez-config.lua or something
2023-01-12 13:20:02 <psykose> i forget how they're loaded
2023-01-12 13:20:16 <psykose> and inside bluez_monitor.properties = {
2023-01-12 13:20:43 <Ermine> 50-bluez-config.lua
2023-01-12 13:21:20 <WhyNotHugo_> I'm not using mpris-proxy, I just have sway handle the keypress.
2023-01-12 13:21:35 <Ermine> ./config/wireplumber/bluetooth.lua.d/50-bluez-config.lua
2023-01-12 13:26:48 <WhyNotHugo_> Added https://paste.sr.ht/~whynothugo/105097d4881a52ce508f5b78f9e593de30e4b24f to /home/hugo/.config/wireplumber/bluetooth.lua.d/60-bluez-config.lua. Same issue.
2023-01-12 13:27:56 <WhyNotHugo_> Isn't the dummy player required tho?
2023-01-12 13:28:12 <WhyNotHugo_> IIRC, some headphones don't enable play-pause if there's nothing playing, so the player works around this, right?
2023-01-12 13:32:45 <WhyNotHugo_> Any ideas what the "No such file or directory (2)" part is referring to?
2023-01-12 13:36:53 <Ermine> WhyNotHugo_: you can try running the daemon under strace to see which syscall returns ENOENT and look the file referred
2023-01-12 13:37:13 <psykose> uinput_create tries to open /dev/uinput
2023-01-12 13:37:47 <psykose> https://github.com/RadiusNetworks/bluez/blob/master/profiles/audio/avctp.c#L1050
2023-01-12 13:38:15 <psykose> does a mere modprobe uinput fix it
2023-01-12 13:38:20 <WhyNotHugo_> Ah, that makes sense, since the headphones basically have a "keyboard" and send keyboard events.
2023-01-12 13:38:32 <WhyNotHugo_> (I realise this after reading https://www.kernel.org/doc/html/v4.12/input/uinput.html )
2023-01-12 13:38:54 <psykose> idk what the dummy player does or the semantics of this
2023-01-12 13:39:05 <psykose> i just know some wireplumber upgrade broke my audio on the -xm3 version
2023-01-12 13:39:14 <psykose> and i bisected it to the commit that set that from =false to =true
2023-01-12 13:39:14 <psykose> :3
2023-01-12 13:39:20 <psykose> and so that fixes those
2023-01-12 13:39:21 <WhyNotHugo_> That works! Thanks!
2023-01-12 13:39:27 <psykose> and yeah, it has to be 50-config
2023-01-12 13:39:30 <psykose> but if uinput fixes it
2023-01-12 13:39:34 <psykose> i guess we have the real fx
2023-01-12 13:39:49 <psykose> i am not sure what downsides there are to having uinput loaded or why it's not default
2023-01-12 13:39:52 <WhyNotHugo_> Some headphones won't emit events when the play/pause buttons are pressed UNLESS there's an active player running.
2023-01-12 13:40:04 <psykose> mhm
2023-01-12 13:40:13 <WhyNotHugo_> The dummy player basically makes the headphones think that there is a player running and then they send the play/pause events.
2023-01-12 13:40:30 <Ermine> I wonder why uinput module didn't got loaded automatically
2023-01-12 13:43:14 <psykose> idk if anything does
2023-01-12 13:44:04 <WhyNotHugo_> There's nothing triggering it to load.
2023-01-12 13:47:38 <WhyNotHugo_> https://wiki.alpinelinux.org/wiki/PipeWire#Play/Pause_buttons_not_working_on_bluetooth_headphones
2023-01-12 13:54:35 <stinter> hello, i'm using the 3.17.1 virtual install and trying to do a manual disks/partitions/filesystems setup as per https://wiki.alpinelinux.org/wiki/Setting_up_disks_manually#For_%22sys%22_disk_mode_installs . I have everything set up under /mnt . at what point do I run setup-disk? before or after setup-alpine?
2023-01-12 13:54:41 <psykose> WhyNotHugo_: nice. also that page is a mess
2023-01-12 13:54:51 <psykose> most WIP thing ever
2023-01-12 13:55:58 <psykose> stinter: whenever you want, manual partitioning implies you set up the disks, then the setup-disk will set up the install on the target, then you can chroot to it and setup-alpine again inside or whatnot
2023-01-12 13:56:44 <stinter> ah okay, it doesn't mention a chroot or anything. i'll give it a shot
2023-01-12 13:57:19 <psykose> well idk, it's just how i personally do it :)
2023-01-12 13:57:58 <stinter> ah hok
2023-01-12 13:58:26 <psykose> setup-disk just fully copies the files and whatnot, think of it as the same thing as pacstrap is on arch
2023-01-12 13:58:50 <psykose> then you just finish the bootloader setup and you're done, not much else to it aside from personal configuration
2023-01-12 13:59:08 <psykose> i used gentoo/arch for too long so i always do these things myself as they're not that hard
2023-01-12 13:59:35 <psykose> of course, with lvm/luks it's more complicated :)
2023-01-12 14:00:03 <stinter> it's trying to call some syslinux utilities which don't appear to be installed; apk update only has the cdrom repo, can I just toss a correct line for an http(s) repo so I can get syslinux installed (install media)?
2023-01-12 14:02:19 <stinter> toss that line into /etc/apk/repositories that is
2023-01-12 14:03:58 <psykose> ah you should at least run setup-apkrepos and setup-interfaces first before setup-disk
2023-01-12 14:04:13 <psykose> i forgot to mention that you do need to get to the point you can call setup-disk, but it's basically just networking and repos
2023-01-12 14:04:14 <psykose> so yeah
2023-01-12 14:05:19 <stinter> that makes sense. kinda got there myself doing ifconfig eth0 up && udhcp eth0, nice of setup-apkrepos to exist :)
2023-01-12 14:06:29 <stinter> right on. setup-disk worked
2023-01-12 14:10:01 <stinter> tfw you forget your bind mounds for proc / sys and friends :P
2023-01-12 14:10:11 <psykose> :)
2023-01-12 14:10:44 <psykose> i've been meaning to write a chroot helper to just add a base utility
2023-01-12 14:10:53 <stinter> big gentoo user here so i already have a lot of chroot notes... will just borrow those
2023-01-12 14:11:20 <stinter> brb need to reboot
2023-01-12 14:32:34 <JanJ[m]> heh
2023-01-12 14:53:37 <stinter> psykose: ok so i'm in the chroot and it's functioning and all that jazz. did setup-alpine and it's asking me which disks to use. do i just say none because I already set all that up?
2023-01-12 14:53:50 <psykose> yeah
2023-01-12 14:53:52 <psykose> i think
2023-01-12 14:55:30 <stinter> ok. glad i'm in a VM *snapshot*
2023-01-12 15:00:54 <stinter> how does one install the bootloaded? no grub
2023-01-12 15:01:12 <JanJ[m]> normally alpine-setup does that for you
2023-01-12 15:01:55 <JanJ[m]> i have struggled most of yesterday with getting a bootloader to work, Syslinux isnt really good for UEFI/GPT i found out, so one automaticly ends up with grub as it is the easier choice
2023-01-12 15:02:24 <stinter> nod. that's why is set up dos / mbr, didn't want to mess with it
2023-01-12 15:02:39 <JanJ[m]> then you can use syslinux :)
2023-01-12 15:02:52 <JanJ[m]> the wiki has an article about that. hang on
2023-01-12 15:02:56 <stinter> cool thanks
2023-01-12 15:03:08 <JanJ[m]> https://wiki.alpinelinux.org/wiki/Bootloaders
2023-01-12 15:03:11 <stinter> made things a lot harder on myself by insisting on manual partitions/filesystems hehe
2023-01-12 15:03:15 <stinter> thanks!
2023-01-12 15:03:32 <JanJ[m]> thats the easiest method :)
2023-01-12 15:04:39 <stinter> omg a distro after my heart. i used to do digital forensics as a job, knew my stuff, and always told people an mbr is 440 bytes NOT 448 you dolts</blog>
2023-01-12 15:04:46 <stinter> and this has bs=440 :D
2023-01-12 15:04:55 <JanJ[m]> :)
2023-01-12 15:05:22 <JanJ[m]> i'm rather fond of Alpine too :) Though the last two days this has cooled down a bit :) it'll come back!
2023-01-12 15:10:16 <stinter> lmao. i'm a gentoo junkie so i get it. come up with your own workflow and then just follow it in the future and you're good. take LOTS of notes
2023-01-12 15:10:50 <JanJ[m]> yeah.. i need to take notes indeed, and publish them online :) 
2023-01-12 15:11:10 <JanJ[m]> i need a blog anyways, so much to tell, so little public!
2023-01-12 15:11:22 <JanJ[m]> ofcourse i'll host it on Alpine :D
2023-01-12 15:14:43 <stinter> and we're up and running :D
2023-01-12 15:14:54 <JanJ[m]> with syslinux?
2023-01-12 15:15:03 <stinter> p a r t. y? COZ I GOTTA
2023-01-12 15:15:17 <stinter> yeah, that page was spot on
2023-01-12 15:16:04 <JanJ[m]> nice :)
2023-01-12 15:16:11 <stinter> somehow I didn't set a root pass LOL *backs up*
2023-01-12 15:16:29 <JanJ[m]> so then do: doas passwd
2023-01-12 15:16:36 <JanJ[m]> then you can set a root password ;-)
2023-01-12 15:16:52 <stinter> well, I can't now :P good thing I take paranoid-level virtual machine snapshots so i'm good
2023-01-12 15:17:02 <stinter> can just hop back to right before the reboot
2023-01-12 15:17:12 <JanJ[m]> why cant you? didnt create a user accont?
2023-01-12 15:17:29 <stinter> skipped it coz stupid user tricks
2023-01-12 15:17:42 <stinter> just trying to log in locally as root not over the network of course
2023-01-12 15:17:45 <JanJ[m]> :-)
2023-01-12 15:48:48 <WhyNotHugo_> If you want stupid simple bootloading, using an UKI works: https://wiki.alpinelinux.org/wiki/UEFI_Secure_Boot
2023-01-12 15:49:09 <WhyNotHugo_> Just don't enroll your keys in the firmware and don't enable secureboot if you don't want to deal with that hassle and don't care about that aspect.
2023-01-12 15:56:23 <WhyNotHugo_> https://wiki.alpinelinux.org/wiki/Bootloaders#Using_a_UKI_(UEFI_only) Lemme know if there's any issues with this.
2023-01-12 16:01:21 <psykose> i wonder if there's any decent dark mediawiki theme
2023-01-12 16:03:29 <Ermine> I use an extension to get dark themes
2023-01-12 16:05:06 <psykose> sure, but i can just add themes
2023-01-12 16:05:18 <psykose> currently it's just the default https://wiki.alpinelinux.org/wiki/Special:Version
2023-01-12 16:05:57 <minimal> WhyNotHugo_: of course that last doc reference only caters for x86_64 systems ;-)
2023-01-12 16:14:38 <WhyNotHugo_> psykose: https://darkreader.org/ ?
2023-01-12 16:16:13 <WhyNotHugo_> minimal: Good catch.
2023-01-12 16:16:45 <psykose> i did use darkreader but it breaks quite a few websites, and in general it makes things very slow
2023-01-12 16:16:51 <psykose> invert filters are not "performance"
2023-01-12 16:16:58 <psykose> and the static modes are broken for almost everything
2023-01-12 16:17:07 <psykose> i do mean an actual mediawiki themse since i admin the instance :)
2023-01-12 16:18:35 <psykose> this is kinda ok https://www.mediawiki.org/wiki/Skin:Citizen though it's star citizen :p
2023-01-12 16:18:48 <minimal> WhyNotHugo_: I've been meaning to script UKI setup for both x86_64 and aarch64 for a while
2023-01-12 16:40:09 <elly> psykose: adaptive wiki theme when?
2023-01-12 16:40:19 <psykose> if you know of one i'm all ears
2023-01-12 16:40:33 <psykose> (isn't vector adaptive)
2023-01-12 16:40:52 <elly> I don't know of one
2023-01-12 16:40:54 <psykose> ah i didn't make it default :)
2023-01-12 16:41:00 <psykose> https://img.ayaya.dev/tFuiUg8xo1hC
2023-01-12 16:41:03 <psykose> looks like this
2023-01-12 16:43:02 <psykose> or wait, it is
2023-01-12 16:43:41 <psykose> did you mean adaptive to media-prefer-dark?
2023-01-12 16:44:27 <elly> yeah, that
2023-01-12 16:44:41 <Ermine> adaptiveness: https://i.imgur.com/hnF8Oom.jpg
2023-01-12 16:45:32 <elly> yeah, I meant dark-mode-adaptive, not screen-size-adaptive
2023-01-12 16:48:21 <psykose> https://wiki.alpinelinux.org/wiki/Bootloaders?useskin=Citizen
2023-01-12 16:49:12 <psykose> for some reason you have to change from auto to something back to auto to make it respect dark xD
2023-01-12 16:49:15 <psykose> bottom left
2023-01-12 16:49:25 <psykose> looks much more modern i guess
2023-01-12 16:49:31 <psykose> not sure i like the sidebar being gone though
2023-01-12 16:52:30 <Ermine> Maybe it is worth to remove the warning from the main page btw?
2023-01-12 16:53:30 <elly> yeah, it's now 1 year old
2023-01-12 16:53:42 <JanJ[m]> yikes
2023-01-12 16:53:52 <psykose> what's yikes
2023-01-12 16:53:54 <psykose> Ermine: sure
2023-01-12 16:54:15 <psykose> elly: also seems this dark theme sucks with any of the colored sections like <pre> or warnings..
2023-01-12 16:54:23 <psykose> https://img.ayaya.dev/h3mCuX9o1Wx4
2023-01-12 16:54:24 <psykose> lol
2023-01-12 16:54:29 <psykose> https://img.ayaya.dev/q1rg0zJgtV8L
2023-01-12 16:54:38 <Ermine> eh
2023-01-12 16:54:40 <psykose> now that is some a11y approved contrast
2023-01-12 16:54:54 <elly> dislike
2023-01-12 16:54:59 <psykose> yep
2023-01-12 16:55:08 <psykose> anyway if you find anything half decent throw it at me
2023-01-12 16:55:08 <psykose> :)
2023-01-12 16:57:01 <psykose> Ermine: removed
2023-01-12 16:57:18 <elly> I am not a mediawiki knower so I am unlikely to find anything I'm afraid
2023-01-12 16:57:37 <elly> and to be totally honest I do not really like wikis as a way of documenting things
2023-01-12 16:58:52 <psykose> what do you prefer
2023-01-12 16:59:24 <psykose> i guess docs.a.o is what you would prefer :p it's a tremendous effort to write however
2023-01-12 16:59:40 <psykose> personally i don't write many docs so anything goes for me
2023-01-12 17:00:18 <WhyNotHugo_> minimal: The default theme doesn't have a dark variant? :c
2023-01-12 17:00:21 <elly> yeah, I prefer a more structured way of doing documentation
2023-01-12 17:00:41 <psykose> it's better for reading too tbh
2023-01-12 17:00:41 <elly> wikis imo tend to contain a mix of current stuff, outdated stuff, inaccurate stuff, "how I did this" type guides, etc
2023-01-12 17:00:47 <psykose> ye
2023-01-12 17:01:21 <elly> it does require a lot of work though, it's hard to achieve without a librarian / tech writer to do it
2023-01-12 17:02:05 <Ermine> elly: hence good docs require effort, no matter the format
2023-01-12 17:02:36 <elly> yeah, definitely
2023-01-12 17:03:00 <elly> the main virtue of wikis is that they allow for writing docs with little effort, but as you say, that rarely yields good docs
2023-01-12 17:03:24 <WhyNotHugo_> Something like man.alpinelinux.org would have been super useful before I had an actual working setup.
2023-01-12 17:03:31 <WhyNotHugo_> I still find it would be useful for stuff that's not installed.
2023-01-12 17:03:48 <WhyNotHugo_> A wiki makes it easy to contribute, but also makes it easy for content to stagnate :c
2023-01-12 17:04:18 <WhyNotHugo_> The search is terrible; searching for "Bootloader" shows the page with title "Bootloaders" in 8th place.
2023-01-12 17:04:25 <elly> huh, man.alpinelinux.org is an interesting idea
2023-01-12 17:04:52 <JanJ[m]> what would it do?
2023-01-12 17:04:56 <WhyNotHugo_> https://man.archlinux.org/man/environment.d.5
2023-01-12 17:05:09 <JanJ[m]> hmz
2023-01-12 17:05:32 <psykose> there's already a site that imports manpages from alpine
2023-01-12 17:05:35 <psykose> i forgot what it was
2023-01-12 17:05:42 <psykose> but in general this is a weird kind of site
2023-01-12 17:05:48 <psykose> almost none of them are distro specific
2023-01-12 17:05:53 <Ermine> Upstreams tend to publish their man pages online
2023-01-12 17:06:02 <elly> importing man pages is one thing, but I'd be more interested in like, a man page describing setup-alpine or similar
2023-01-12 17:06:42 <Ermine> So I don't think having distro website with manpages does make any sense (besides distro-specific stuff of course)
2023-01-12 17:06:43 <psykose> ah, sure, only the specific ones
2023-01-12 17:07:05 <elly> one could imagine like... installing.7 or something, which is a man page describing how to install alpine
2023-01-12 17:08:25 <Ermine> btw I'm looking for a tool which could generate both manpages and basic html (is it scdoc?)
2023-01-12 17:08:42 <WhyNotHugo_> There's a companion for scdoc that generates html.
2023-01-12 17:08:51 <WhyNotHugo_> https://sr.ht/~bitfehler/scd2html/
2023-01-12 17:09:49 <JanJ[m]> Ermine: but that means, i would have to visit multiple man sites? why not go for simplicity, and have all the man pages under the alpine website anyway?
2023-01-12 17:09:54 <WhyNotHugo_> Or you can use mandoc. This https://gitlab.com/WhyNotHugo/darkman/-/blob/main/Makefile#L10 generates https://darkman.whynothugo.nl/
2023-01-12 17:09:55 <elly> the real advantage, imo, of doing such things as man pages is that you can ship them with the install image :)
2023-01-12 17:11:25 <WhyNotHugo_> scd2html seems to hyperlink URLs, so I'd go with that.
2023-01-12 17:11:42 <WhyNotHugo_> Does the install image include man pages?
2023-01-12 17:11:45 <JanJ[m]> but using man pages.. who still does that? most users these days will get their phones and open google :) 
2023-01-12 17:11:57 <elly> JanJ[m]: disagree :P
2023-01-12 17:12:39 <Ermine> JanJ[m]: what I really would appreciate if google (or some other search engine) could understand requests for man pages and open them directly for upstream
2023-01-12 17:12:49 <JanJ[m]> ;-) i don't. i wholeheartedly agree with myself :-)
2023-01-12 17:13:29 <Ermine> but having them offline is easier. You are just one command away from what you need
2023-01-12 17:14:03 <Ermine> WhyNotHugo_: thank you
2023-01-12 17:14:09 <JanJ[m]> if you know what you're looking for, sure :)
2023-01-12 17:14:31 <Ermine> if you don't, there's apropos
2023-01-12 17:14:46 <JanJ[m]> hm?
2023-01-12 17:15:01 <psykose> i wish apropos indexing was not glacially slow
2023-01-12 17:15:12 <elly> apropos searches the man database poorly
2023-01-12 17:15:16 <psykose> how can it take a fuckin minute to index like 30MB of text?
2023-01-12 17:15:21 <psykose> it should measured in milliseconds
2023-01-12 17:15:34 <psykose> embarassing honestly
2023-01-12 17:15:45 <WhyNotHugo_> I've been thinking of writing a man indexer+search that doesn't suck. Don't get me wound up, I don't have time right now!
2023-01-12 17:15:52 <Ermine> Never had this problem, but... isn't that GNU code?
2023-01-12 17:15:59 <psykose> you would probably write it in Go and i would never use it
2023-01-12 17:16:00 <elly> bsd apropos is also quite slow
2023-01-12 17:16:09 <psykose> but yes, me too
2023-01-12 17:16:10 <elly> I think it is just written in a straightforward way and does not perform that well
2023-01-12 17:16:12 <JanJ[m]> i didnt know about apropos :)
2023-01-12 17:16:50 <JanJ[m]> but then, i think i hardly scratched the surface of linux's capability's, even after all these years
2023-01-12 17:18:10 <Ermine> No worries, I've learned about these tools not so long ago :)
2023-01-12 17:18:53 <JanJ[m]> :)
2023-01-12 17:19:01 <JanJ[m]> so there is hope for me yet!
2023-01-12 17:19:04 <WhyNotHugo_> JanJ[m]: Don't worry, you'll soon realise you weren't missing out on anything.
2023-01-12 17:20:42 <Ermine> psykose: do you dislike go?
2023-01-12 17:21:27 <JanJ[m]> i missed out on a lot :) ive always only used Debian, i'm not a very good distro hopper
2023-01-12 17:23:40 <JanJ[m]> oh well! time for some groceries! 
2023-01-12 17:24:37 <psykose> get me some chips
2023-01-12 17:25:03 <psykose> Ermine: it's a great language for mediocre software
2023-01-12 17:25:11 <psykose> i wish large companies stopped shitting out languages
2023-01-12 17:31:58 <WhyNotHugo_> I liked golang a lot at first. There was a peak after like like 6 months, and after that, the more I used it the more I disliked it.
2023-01-12 17:32:24 <psykose> it has the good parts
2023-01-12 17:34:22 <WhyNotHugo_> I do like Rust a lot, despite its portability issues.
2023-01-12 17:34:24 <WhyNotHugo_> 
ACTION ducks

2023-01-12 17:36:10 <minimal> I like............<hang on, where's my 2013 Trendy Language of the Month Calendar>
2023-01-12 17:39:19 <Newbyte> WhyNotHugo_: portability issues will be gone soon enough
2023-01-12 17:51:26 <psykose> they won't
2023-01-12 17:52:25 <ikke> just decide on a single platform, then all portability issues are gone
2023-01-12 17:52:45 <psykose> portability is generally the least of the issues of anything
2023-01-12 17:57:33 <Hello71> what do you mean, it works on windows, mac, and linux, that's literally all the systems
2023-01-12 17:57:43 <Hello71> what do you mean there's more kinds of linux than ubuntu
2023-01-12 17:57:50 <psykose> :)
2023-01-12 18:01:44 <ikke> What do you mean, you don't run x86_64
2023-01-12 18:03:37 <Ermine> what is that s390x thing
2023-01-12 18:07:43 <psykose> something i wish we would drop :p
2023-01-12 18:14:50 <vkrishn> I started with wikipedia, then switched to pmwiki, https://www.pmwiki.org/wiki/PmWiki/PmWikiPhilosophy
2023-01-12 18:16:33 <vkrishn> runs super nice on mobiles devices
2023-01-12 18:16:44 <Ermine> I tried to ask ibm for free s390x in their community cloud, but they didn't answer...
2023-01-12 18:17:58 <psykose> should have told them the VP at your company needs it for expedited Q3 target planning
2023-01-12 18:18:27 <Ermine> XD
2023-01-12 18:18:47 <elly> the idea of doing "expedited Q3 planning" in January aligns exactly with my experiences working with IBM
2023-01-12 18:21:40 <Ermine> Need to run SAP on your cloud
2023-01-12 18:24:57 <psykose> :)
2023-01-12 18:37:00 <minimal> Ermine: in the past I've run SAP (Hana) on their bare-metal cloud for SAP ;-)
2023-01-12 18:37:32 <minimal> was an interesting contract, monster machine clusters...
2023-01-12 18:38:06 <Ermine> Most enterprise thing ever
2023-01-12 18:38:34 <psykose> 1 query per machine? ;)
2023-01-12 18:40:51 <minimal> well the entry level machines were 512GiB of ram per node and those sort of customers they told them to go use the largest AWS VMs, the baremetal stuff was multi ~TiB RAM per node, huge numbers of cores per node, large numbers of fast PCI/NVME SSDs per node as quite a few notes per cluster, even for customer test systems. Hana is an *in-memory* DB, the storage is only really used when a node boots
2023-01-12 18:41:26 <elly> wow
2023-01-12 18:41:41 <psykose> and it's still slow as shit innit
2023-01-12 18:42:21 <minimal> psykose~: dunno, I just worked on the automation infra to build the clusters, never used Hana myself lol
2023-01-12 18:42:30 <psykose> :)
2023-01-12 18:42:31 <psykose> fun job
2023-01-12 18:42:38 <psykose> hopefully
2023-01-12 18:42:48 <WhyNotHugo_> How did support for s390x some to be?
2023-01-12 18:42:50 <Ermine> psykose: I've read that as 'shit init' and started wondering what are you talking about
2023-01-12 18:43:01 <minimal> though POST on the only Intel large core boxes used to take a *crazy* amount of time - like 40-60 minutes
2023-01-12 18:43:09 <minimal> s/only/older/
2023-01-12 18:43:09 <alpine-meetbot> minimal meant to say: though POST on the older Intel large core boxes used to take a *crazy* amount of time - like 40-60 minutes
2023-01-12 18:43:28 <Ermine> yikes!
2023-01-12 18:43:51 <minimal> yeah the more recent Intel stuff POSTs a lot faster
2023-01-12 18:44:13 <WhyNotHugo_> *come to be
2023-01-12 18:44:17 <psykose> Ermine: clearly openrc is too slow
2023-01-12 18:44:55 <Ermine> psykose: undoubtedly
2023-01-12 18:44:57 <psykose> WhyNotHugo_: one of the developers had a self-built scavenged mainframe
2023-01-12 18:45:05 <psykose> hobby maintained it
2023-01-12 18:45:18 <psykose> afaik
2023-01-12 18:45:25 <Ermine> psykose: Intel moved their boxes post system to s6
2023-01-12 18:45:26 <psykose> it's been there for over 10 years i think
2023-01-12 18:45:47 <psykose> first commit with s390 mention is from 2011
2023-01-12 18:45:59 <ikke> The builders we use are provided and hosted by IBM
2023-01-12 18:46:00 <psykose> bytesex.h fixes for x86_64, sh, ia64 and s390/s390x
2023-01-12 18:46:04 <psykose> byte sex :)
2023-01-12 18:46:12 <Ermine> XD
2023-01-12 18:46:32 <psykose> Ermine: what intel boxes
2023-01-12 18:46:40 <WhyNotHugo_> ikke: Oh, that makes sense... I guess.
2023-01-12 18:46:52 <WhyNotHugo_> Was wondering who had spare mainframes for CI.
2023-01-12 18:47:00 <psykose> they also got rust onto s390x
2023-01-12 18:47:13 <ikke> Alpine is big business with IBM
2023-01-12 18:47:17 <ikke> lots of containers apparently
2023-01-12 18:47:22 <elly> that's kind of cool really
2023-01-12 18:47:25 <psykose> and if you look at s390x/ppc64le support in various open source projects
2023-01-12 18:47:31 <Ermine> psykose: I mean Intel started using s6 to POST their stuff and it went much faster
2023-01-12 18:47:33 <psykose> all of them are basically paid ibm employees
2023-01-12 18:47:59 <psykose> there was a binutils ppc issue and when we reported it one of the binutils maintainers showed up and fixed it
2023-01-12 18:48:01 <psykose> :)
2023-01-12 18:48:19 <psykose> (an ibm employee)
2023-01-12 18:48:21 <psykose> etc etc
2023-01-12 18:48:29 <WhyNotHugo_> That's pretty cool. So they actually contribute back to what they use.
2023-01-12 18:48:51 <psykose> in the minimal-effort "when it breaks they can fix it" kind of way yeah
2023-01-12 18:49:11 <psykose> Ermine: huh, where did you read that
2023-01-12 18:49:25 <Ermine> I tried to joke
2023-01-12 18:49:30 <minimal> psykose: rather than a minimal-effortless way? ;-)
2023-01-12 18:49:34 <psykose> :)
2023-01-12 18:49:42 <psykose> minimal puts in maximal effort
2023-01-12 18:50:01 <Ermine> TALOS people fixing ppc stuff where
2023-01-12 18:51:29 <psykose> i saw a talos developer in dri-devel working on some gpu stuff for ppc
2023-01-12 18:51:32 <psykose> aside from that no idea
2023-01-12 19:00:25 <lucidiot> instead of s6 why don't we switch to s390x, since 390>6
2023-01-12 19:01:10 <Ermine> :)
2023-01-12 19:08:30 <JanJ[m]> hmz
2023-01-12 20:58:56 <frag> are laptop keyboards basically known for not lasting very long (or at least not as long as "external kbs")?
2023-01-12 21:00:15 <psykose> never had that issue personally but probably same tier as most generic $15 keyboards
2023-01-12 21:01:04 <frag> i guess i havent had a problem either, BUT then never used a laptop as my main machine
2023-01-12 21:10:52 <JanJ[m]> my lenovo laptop works just fine
2023-01-12 21:11:03 <JanJ[m]> never had any kb problems to think of
2023-01-12 21:12:25 <minimal> I find it's the hinges that go on laptops :-(
2023-01-12 21:12:59 <minimal> have a couple of Lenovos with dodgy screen hinges
2023-01-12 21:14:24 <psykose> hah yeah
2023-01-12 21:14:30 <psykose> mine is ok but it has a jiggle now
2023-01-12 21:16:14 <JanJ[m]> This is my first Lenovo, we send them out to clients lots. i used to have everything Apple
2023-01-12 21:16:31 <JanJ[m]> now i have a thinkpad with Alpine + KDE <3
2023-01-12 21:16:50 <JanJ[m]> almost perfect!
2023-01-12 21:18:08 <psykose> now for the damned networking
2023-01-12 21:18:24 <JanJ[m]> indeed :D
2023-01-12 21:18:25 <JanJ[m]> but thats worries for next week
2023-01-12 21:21:32 <minimal> psykose: damned networking? so you don't have a firewall, you have a dam (i.e. waterwall)? ;-)
2023-01-12 21:21:58 <psykose> everyone else has a firewall, so i think it's a natural counter ;)
2023-01-12 21:23:09 <ikke> I had one laptop where some keys started to no longer respond
2023-01-12 21:25:07 <minimal> ikke: did this include the "any" key and so the computer was unusable? lol
2023-01-12 21:25:14 <ikke> minimal: yes
2023-01-12 21:25:27 <psykose> what's the any key
2023-01-12 21:25:37 <minimal> "press any key to continue" :-)
2023-01-12 21:25:42 <psykose> gdi
2023-01-12 21:25:45 <psykose> xD
2023-01-12 21:26:47 <minimal> the old jokes are always the best ones
2023-01-12 21:26:50 <ikke> jhehe
2023-01-12 23:06:38 <Quillith> How do you view /bin/ash or the other /bin/sh ? When I try to look at them via nano they appear to be random jibberish
2023-01-12 23:07:07 <psykose> they are compiled binaries
2023-01-12 23:07:14 <psykose> can't view them in a text editor
2023-01-12 23:07:28 <psykose> the source code is in busybox
2023-01-12 23:07:34 <psykose> ash.c probably, amongst others
2023-01-12 23:07:43 <psykose> it's probably as readable as the binary
2023-01-12 23:08:15 <Quillith> Okay, thank you. If that's so then I'm unlikely to have messed with them. I'm hunting for why Foot and Kitty always open with an empty line and have a tendency to create extra random lines while I'm doing things
2023-01-12 23:08:57 <elly> I would look at your shell environment, like your .profile and similar scripts
2023-01-12 23:09:02 <elly> possibly your prompt setting
2023-01-12 23:10:59 <Quillith> I don't see anything in my ~/.profile, and I launched Sway with default config with no difference. I thought it may be my foot.ini but using default config didn't change that either. I'll have to keep looking
2023-01-12 23:11:19 <psykose> can you take a picture by chance of just opening a terminal
2023-01-12 23:11:55 <Quillith> I don't know how to send pictures through irssi, but it looks like if you opened a terminal and hit Enter
2023-01-12 23:12:21 <psykose> hmmmmmmmmmmm
2023-01-12 23:12:41 <Quillith> it also happens when I use keyboard commands that launch programs, like a browser, or if I type 'exit' into a terminal (it makes an empty line in all other open terminals)
2023-01-12 23:12:52 <psykose> wh
2023-01-12 23:12:56 <psykose> that is really weird :D
2023-01-12 23:13:04 <Quillith> weirdly it doesn't submit whatever text was on the current line- it just moves on to a new one
2023-01-12 23:13:39 <Quillith> yeah I have no idea why. I was trying to make foot.ini launch with neofetch when this started, so I thought maybe I'd messed up my ash somehow. ah well
2023-01-12 23:14:07 <psykose> not at all
2023-01-12 23:14:14 <psykose> but something much more bizarre is afoot
2023-01-12 23:14:26 <Quillith> loooool
2023-01-12 23:14:47 <psykose> :)
2023-01-12 23:15:50 <minimal> ghost in the machine?
2023-01-12 23:17:18 <elly> that's "strace your shell" territory imo, to see if it's your shell or the terminal emulator that is acting up
2023-01-12 23:17:42 <Quillith> I don't think it's the terminal emulator, because I installed kitty and the same thing happened on it too
2023-01-12 23:19:17 <elly> do you see the same behavior if you run, like... 'foot sh -c "echo hi && sleep 10"'?
2023-01-12 23:19:56 <Quillith> No, then there's no $ sign just the word hi and the window vanished
2023-01-12 23:20:10 <elly> what if you do: 'foot sh'
2023-01-12 23:21:02 <Quillith> empty line as if you'd started the terminal and then hit Enter
2023-01-12 23:21:06 <elly> O_o
2023-01-12 23:21:13 <Quillith> like, an empty line with the $ sign
2023-01-12 23:21:18 <Quillith> not pure empty
2023-01-12 23:21:24 <elly> so you see: $, then a newline, then $?
2023-01-12 23:21:31 <Quillith> yes
2023-01-12 23:21:34 <elly> what the f
2023-01-12 23:21:43 <elly> is your keyboard ok?
2023-01-12 23:21:48 <Quillith> yeah
2023-01-12 23:22:17 <psykose> do you have ENV set
2023-01-12 23:22:18 <elly> how sure are you of that? :P
2023-01-12 23:22:19 <Quillith> And when I open a new terminal, say with Super+T, every existing terminal gets another line. But pressing Super on its own doesn't do that
2023-01-12 23:22:29 <elly> ok it is *definitely* an ENV problem
2023-01-12 23:22:35 <elly> if you see that behavior
2023-01-12 23:22:45 <Quillith> I have some ENV in my ~/.profile
2023-01-12 23:22:46 <psykose> nah something makes me think that's not it either
2023-01-12 23:22:50 <psykose> what's it point to
2023-01-12 23:23:12 <Quillith> What do you mean point to
2023-01-12 23:23:17 <psykose> what is it set to
2023-01-12 23:23:17 <elly> what is ENV set to
2023-01-12 23:23:52 <Quillith> oh sorry I misunderstood, I have environment variables in my ~/.profile, not ENV itself. I'll echo ENV just in case though
2023-01-12 23:24:10 <psykose> can you also paste the file to somewhere just in case
2023-01-12 23:25:01 <Quillith> I echoed $ENV and got a blank line, but that's the same thing that happens if I echo $jibberish
2023-01-12 23:25:09 <elly> right, so it's unset
2023-01-12 23:25:31 <elly> could you pastebin your ~/.profile here: https://bpa.st/ and then link the paste? I'm very curious
2023-01-12 23:27:14 <elly> it almost sounds like you have some daemon that starts in your .profile that is emitting blank lines to new shells
2023-01-12 23:27:28 <Quillith> https://bpa.st/2KM6A
2023-01-12 23:27:41 <Quillith> There's nothing much in there beyond what I put in to switch to Sway
2023-01-12 23:27:56 <elly> huh, okay
2023-01-12 23:27:56 <elly> yeah
2023-01-12 23:28:10 <elly> wow, I have no idea, good luck
2023-01-12 23:28:12 <Quillith> I mean if that daemon is there I'd like to co-opt it to run neofetch instead of a blank line, which was my original goal
2023-01-12 23:28:47 <elly> it seems like that it is your shell receiving that blank line, but why it's doing so is mystifying
2023-01-12 23:29:00 <Quillith> Yeah I may need to reinstall, no clue how I borked this but I did also install every audio manager under the sun and come in here asking "why can't I hear anything?" so anything's possible
2023-01-12 23:29:45 <elly> it ought to be debuggable; my next step would be something like: foot sh -c 'strace -o /tmp/shell.trace sh', then seeing if your shell is actually reading a newline from its input pty
2023-01-12 23:30:30 <elly> but, I gotta go to a meeting so good luck :)
2023-01-12 23:30:36 <Quillith> Thank you for your help!
2023-01-12 23:32:02 <Quillith> So that sent me into a > and I opened up a few terminals and yes the new lines are continuing in there
2023-01-12 23:32:02 <psykose> open a terminal, then open another
2023-01-12 23:32:13 <psykose> ah, you got it already
2023-01-12 23:32:14 <psykose> hm
2023-01-12 23:32:23 <psykose> can you paste the strace log
2023-01-12 23:34:46 <Quillith> Where would that be? I tried to nano into /tmp/shell.trace but it was a New File
2023-01-13 00:07:08 <psykose> it doesn't output to file, you have to copy it from the terminal :p
2023-01-13 00:07:26 <psykose> or you have to add -o somefile
2023-01-13 00:10:38 <Quillith> there is nothing being outputted
2023-01-13 00:10:50 <Quillith> ~ $ foot sh -c 'strace -o /tmp/shell.trace sh
2023-01-13 00:10:50 <Quillith> > 
2023-01-13 00:10:50 <Quillith> > 
2023-01-13 00:10:50 <Quillith> > 
2023-01-13 00:10:52 <Quillith> > 
2023-01-13 00:12:41 <dwfreed> because it goes in the /tmp/shell.trace file instead
2023-01-13 00:14:38 <Quillith> When I nano to /tmp/shell.trace it says New File
2023-01-13 00:14:42 <elly> no, you are missing the trailing single quote
2023-01-13 00:14:50 <elly> the > prompt is your outer shell telling you "finish typing the command"
2023-01-13 00:15:01 <dwfreed> oh, yeah, missed that
2023-01-13 00:15:10 <elly> you want: foot sh -c 'strace -o /tmp/shell.trace sh'
2023-01-13 00:15:34 <stinter> what's "foot" do?
2023-01-13 00:15:54 <elly> it's a terminal emulator
2023-01-13 00:15:59 <stinter> oh nice
2023-01-13 00:16:08 <Quillith> https://bpa.st/XUWXA
2023-01-13 00:16:39 <Quillith> Thanks for explaining that, I should've noticed the trailing quote
2023-01-13 00:17:02 <elly> that is not strace output
2023-01-13 00:18:09 <Quillith> nano /tmp/shell.trace gives me New File still. I'm not sure where to find the strace output
2023-01-13 00:18:29 <Quillith> Sorry, I just cut-pasted what the terminal made after I entered the command
2023-01-13 00:19:15 <elly> it should have ended up in /tmp/shell.trace
2023-01-13 00:19:16 <elly> odd
2023-01-13 00:19:21 <elly> do you have strace installed?
2023-01-13 00:19:53 <psykose> it does work, i assume you typod something else this time
2023-01-13 00:20:17 <Quillith> No elly was right, I didn't have strace installed. I'm posting the bpa pastebin now that I've installed
2023-01-13 00:39:47 <Quillith> Okay, I got it, thanks for your patience: https://bpa.st/N6HGQ
2023-01-13 00:41:12 <Quillith> What are all those poll POLLIN events? Is that what's weird?
2023-01-13 00:41:15 <psykose> the --sigwinch's cause newlines
2023-01-13 00:41:17 <psykose> hmm
2023-01-13 00:41:23 <psykose> nah, that's normal
2023-01-13 00:41:40 <psykose> oh
2023-01-13 00:41:44 <psykose> i can reproduce this
2023-01-13 00:41:45 <psykose> ................
2023-01-13 00:41:48 <psykose> it's only in ash
2023-01-13 00:41:58 <psykose> the fuck
2023-01-13 00:42:10 <elly> *what*?
2023-01-13 00:42:13 <Quillith> lay it on me, what did I do
2023-01-13 00:42:18 <psykose> it's only for resizes in same ws
2023-01-13 00:42:27 <psykose> if you swap to another workspace and open terminals
2023-01-13 00:42:30 <psykose> nothing happens to the others
2023-01-13 00:42:38 <elly> wait
2023-01-13 00:42:49 <psykose> basically you have an ash, spawn a sway window, it resizes (tiles), and the ash prints a newline
2023-01-13 00:43:18 <psykose> any movement does it
2023-01-13 00:43:24 <psykose> so super+e spam keeps printing newlines
2023-01-13 00:43:25 <psykose> haha
2023-01-13 00:43:27 <psykose> this is so weird
2023-01-13 00:43:44 <Quillith> Oh so it's movement that's spawning these newlines?
2023-01-13 00:43:52 <elly> so, this strace shows that *on standard in* your shell is receiving "\x1b[12;5R"
2023-01-13 00:44:01 <elly> which looks a heck of a lot like a VT100-style escape code
2023-01-13 00:44:07 <psykose> yeah
2023-01-13 00:44:19 <Quillith> I tested making a bunch of new windows on terminal, and yeah, spamming new windows in Workspace 2 doesn't affect the terminals open in Workspace 1
2023-01-13 00:44:45 <psykose> i think resize events notify the shell
2023-01-13 00:44:51 <psykose> and every shell handles this
2023-01-13 00:44:57 <psykose> but ash.. newlines
2023-01-13 00:44:57 <psykose> idk
2023-01-13 00:45:02 <elly> right, but your shell doesn't think it should receive a VT220 command there
2023-01-13 00:45:09 <elly> or it does but it handles it oddly
2023-01-13 00:45:22 <elly> out of curiosity, TERM=xterm ash 
2023-01-13 00:45:26 <elly> does it still happen?
2023-01-13 00:45:42 <psykose> yesa
2023-01-13 00:45:44 <Quillith> Yes
2023-01-13 00:45:50 <elly> wow
2023-01-13 00:45:51 <psykose> it's the same for any terminal
2023-01-13 00:46:03 <elly> wait oh my god it happens for me too in i3
2023-01-13 00:46:05 <psykose> i think there is some terminal spec that notifies that escape or something
2023-01-13 00:46:06 <psykose> yep
2023-01-13 00:46:08 <elly> this absolutely didn't happen before
2023-01-13 00:46:11 <psykose> wait
2023-01-13 00:46:12 <elly> this must be the busybox 1.36 roll
2023-01-13 00:46:14 <psykose> yeah
2023-01-13 00:46:27 <psykose> let me try downgrade
2023-01-13 00:48:18 <psykose> yep
2023-01-13 00:48:19 <psykose> fixed
2023-01-13 00:48:21 <psykose> 1.36
2023-01-13 00:48:29 <psykose> i'll make an issue i guess
2023-01-13 00:48:33 <elly> thanks
2023-01-13 00:48:36 <elly> good find Quillith
2023-01-13 00:48:56 <Quillith> It wasn't me?? Hooray!
2023-01-13 00:49:31 <psykose> no, but you found a bug
2023-01-13 00:49:32 <psykose> :)
2023-01-13 00:49:32 <Quillith> Thank you both so much! omg I was ready to chuck my whole desk into a volcano, man. Super appreciate all your patience with me
2023-01-13 00:49:43 <psykose> and spent time debugging it with us
2023-01-13 00:49:49 <psykose> very much appreciated
2023-01-13 00:50:03 <Quillith> lol well if you ever need a user with minimal knowledge bumping around to test for bugs, just lmk
2023-01-13 00:52:35 <psykose> :)
2023-01-13 00:52:44 <psykose> elly: actually could you write it up
2023-01-13 00:53:11 <psykose> i got sniped by getting signed out of gitlab on browser close into needing 2fa to log in into phone out of battery into android update optimising apps taking 10 minutes
2023-01-13 00:53:31 <psykose> honestly lethal combo needs a nerff
2023-01-13 01:07:09 <elly> sure, one sec
2023-01-13 01:11:32 <elly> I'm pretty sure it was introduced by busybox 12566e7f9b5e5c5d445bc4d36991d134b431dc6c
2023-01-13 01:11:50 <elly> attempted fix for signal handling with pending read() did not account for SIGWINCH
2023-01-13 01:14:39 <elly> we probably want to unship 1.36 though for this
2023-01-13 01:16:26 <elly> busybox bug tracker send me my confirmation email challenge [IMPOSSIBLE]
2023-01-13 01:20:13 <psykose> i mean to aports :p unless you're really confident
2023-01-13 01:20:53 <elly> confident that it's a busybox bug?
2023-01-13 01:23:01 <psykose> yeah and can perhaps fix it
2023-01-13 01:23:04 <psykose> but i guess it's good enough
2023-01-13 01:23:20 <psykose> the alpine busybox maintainer could just also save you the effort
2023-01-13 01:25:43 <elly> filed: https://bugs.busybox.net/show_bug.cgi?id=15256
2023-01-13 01:26:07 <elly> I doubt I can fix it though
2023-01-13 01:27:13 <psykose> yeah that diff is uh, big
2023-01-13 01:27:53 <elly> psykose: am I holding it wrong? `abuild prepare` in aports/community/chromium yields: $ abuild prepare
2023-01-13 01:27:55 <elly> >>> ERROR: chromium: Is $builddir set correctly?
2023-01-13 01:28:26 <psykose> you need to unpack too
2023-01-13 01:29:00 <psykose> i always forget somethin
2023-01-13 01:29:33 <elly> of course, the "download and untar 1.6GB of stuff" step
2023-01-13 01:32:24 <psykose> :)
2023-01-13 01:32:29 <psykose> hey, new xz makes it faster
2023-01-13 01:32:31 <psykose> 5.4+
2023-01-13 01:32:43 <psykose> it's parallel now so it can actually saturate the disk :p
2023-01-13 01:32:59 <psykose> i think it's like a 2.5x speedup for me
2023-01-13 01:33:59 <elly> alright, I now have my build running, time to Wait
2023-01-13 01:59:08 <elly> 25% done, not too bad :)
2023-01-13 02:11:28 <psykose> :)
2023-01-13 03:46:55 <elly> 60% done
2023-01-13 03:46:55 <elly> hum
2023-01-13 03:49:34 <elly> I miss my workstation
2023-01-13 04:02:45 <psykose> aw
2023-01-13 04:02:47 <psykose> :(
2023-01-13 04:13:12 <bl4ckb0ne> anybody experiencing that? https://gitlab.alpinelinux.org/alpine/aports/-/issues/14550
2023-01-13 04:14:05 <elly> it types the string "t(1.13.1)" whenever you open anything?
2023-01-13 04:14:13 <bl4ckb0ne> yep
2023-01-13 04:14:16 <bl4ckb0ne> driving me crazy
2023-01-13 04:14:24 <bl4ckb0ne> happens on 2 different alpine setup
2023-01-13 04:14:26 <elly> what on earth
2023-01-13 04:14:34 <elly> just vim?
2023-01-13 04:14:39 <bl4ckb0ne> yep
2023-01-13 04:15:19 <elly> out of curiosity, `vim -u /dev/null`
2023-01-13 04:15:20 <elly> still see it?
2023-01-13 04:15:28 <bl4ckb0ne> ive been troubleshooting my own .vimrc for the last hours
2023-01-13 04:15:57 <bl4ckb0ne> got the regular output
2023-01-13 04:16:30 <elly> do you still see the weird string with `vim -u /dev/null`?
2023-01-13 04:16:34 <bl4ckb0ne> nope
2023-01-13 04:16:40 <elly> ok, so it is your vimrc then
2023-01-13 04:16:53 <bl4ckb0ne> gah
2023-01-13 04:17:00 <elly> -u /dev/null just makes vim load /dev/null instead of your vimrc
2023-01-13 04:17:56 <psykose> hmm
2023-01-13 04:17:59 <psykose> but the other person also has it
2023-01-13 04:18:07 <bl4ckb0ne> yep
2023-01-13 04:18:09 <psykose> wtf did y'all put in there
2023-01-13 04:18:19 <psykose> first try only the /etc/vimrc one
2023-01-13 04:18:39 <bl4ckb0ne> dont have any
2023-01-13 04:18:43 <psykose> hmmm
2023-01-13 04:19:03 <bl4ckb0ne> apparently its due to foot
2023-01-13 04:19:08 <bl4ckb0ne> doesnt happend on alacritty
2023-01-13 04:19:11 <elly> wait a sec
2023-01-13 04:19:12 <bl4ckb0ne> https://github.com/vim/vim/issues/11766
2023-01-13 04:19:17 <psykose> i use foot and don't get it
2023-01-13 04:19:28 <psykose> o
2023-01-13 04:19:30 <psykose> i got it
2023-01-13 04:19:32 <psykose> hh
2023-01-13 04:19:50 <psykose> ah
2023-01-13 04:19:50 <elly> oh exciting
2023-01-13 04:19:54 <psykose> it's ncurses 6.4
2023-01-13 04:19:57 <psykose> not the only thing it broke
2023-01-13 04:19:58 <psykose> i guess
2023-01-13 04:20:31 <bl4ckb0ne> how
2023-01-13 04:20:57 <psykose> idk
2023-01-13 04:21:00 <psykose> let me try one thing
2023-01-13 04:21:44 <psykose> nnope
2023-01-13 04:22:29 <psykose> do the bracketed paste workarounds in the issue fix it
2023-01-13 04:23:22 <bl4ckb0ne> whihc one
2023-01-13 04:24:32 <psykose> idk
2023-01-13 04:24:40 <psykose> also the person says it's fixed with new vim.. but it's not..
2023-01-13 04:25:04 <psykose> it's fixed with some other TERM of course but
2023-01-13 04:25:53 <bl4ckb0ne> ship it
2023-01-13 04:26:28 <psykose> the version we have is newer than the version they said fixed it
2023-01-13 04:26:39 <psykose> also the gui in the alpine issue is probably not even on foot
2023-01-13 04:26:51 <psykose> how bizarre
2023-01-13 04:36:02 <bl4ckb0ne> opening a new foot term also adds an extra enter
2023-01-13 04:39:11 <psykose> do you use ash
2023-01-13 04:39:24 <psykose> that was found above :p
2023-01-13 04:39:32 <psykose> https://bugs.busybox.net/show_bug.cgi?id=15256
2023-01-13 04:39:52 <bl4ckb0ne> yep
2023-01-13 04:40:00 <bl4ckb0ne> wasnt busybox updated recently?
2023-01-13 04:41:14 <psykose> yeah, it's in .36
2023-01-13 04:41:23 <psykose> bug report is probably correct
2023-01-13 04:41:37 <bl4ckb0ne> uwuuuu
2023-01-13 04:42:49 <psykose> :3
2023-01-13 04:42:56 <psykose> welcome 2 alpine edge
2023-01-13 04:43:11 <psykose> where the bugs are hot and fresh and the software is newer than arch half the time
2023-01-13 04:43:12 <bl4ckb0ne> all i wanted to do was write some code :<
2023-01-13 04:43:38 <psykose> for ash there's no workaround (though it's benign)
2023-01-13 04:43:39 <psykose> for vim uhh
2023-01-13 04:44:10 <bl4ckb0ne> brb gonna package vscode
2023-01-13 04:46:14 <psykose> it is
2023-01-13 04:47:46 <psykose> i guess TERM=xterm-256color is the meh enough workaround
2023-01-13 04:48:08 <psykose> or
2023-01-13 04:48:27 <psykose> apk add foot-extra-terminfo, add term = foot-extra to foot.ini under [main]
2023-01-13 04:48:42 <psykose> i love ncurses
2023-01-13 04:50:06 <bl4ckb0ne> works4me
2023-01-13 04:50:17 <bl4ckb0ne> oh hey almost midnight, time to code
2023-01-13 04:51:58 <elly> :)
2023-01-13 04:52:23 <psykose> almost 6 am
2023-01-13 04:52:26 <psykose> time to code too i think
2023-01-13 05:01:11 <elly> 38,300 targets built of 53,500
2023-01-13 05:01:16 <elly> chromium really is big
2023-01-13 05:06:32 <psykose> :)
2023-01-13 05:06:48 <psykose> if not for the CI and access to 48 cores i would not have touched it
2023-01-13 05:07:13 <psykose> technically also a 160 core one
2023-01-13 05:08:52 <elly> ya, my normal workstation is a 128-core thing, and also has access to something that is effectively super-ccache
2023-01-13 05:09:00 <elly> so clean builds take like, 90 seconds
2023-01-13 06:22:08 <elly> 44,600 targets... you were not wrong about it taking hours
2023-01-13 06:22:32 <psykose> aye
2023-01-13 06:22:57 <psykose> iirc it's actually 50k not 54k if you do just -C out/Release chrome
2023-01-13 06:23:42 <psykose> +1 for chrome_sandbox, the rest is chromedriver/crashpad
2023-01-13 06:29:23 <elly> compilation is occurring
2023-01-13 06:30:38 <psykose> free speed boost LD_PRELOAD another allocator
2023-01-13 08:42:13 <uniqNuuid> what man do you guys use, mandoc or man-db?
2023-01-13 08:42:57 <uniqNuuid> man-db is faster, but too verbose : https://git.alpinelinux.org/aports/tree/community/man-db/man-db.trigger
2023-01-13 08:43:12 <uniqNuuid> can some one add a -q to the trigger?
2023-01-13 08:52:26 <Fxzxmic> what about busybox man?
2023-01-13 08:53:06 <markand> mandoc obviously
2023-01-13 08:53:23 <markand> supports selecting manual page depending on synopsis
2023-01-13 08:53:30 <markand> can export to PDF and various other formats 
2023-01-13 08:53:46 <markand> that's why man snprintf will bring you to the printf(3) manual page
2023-01-13 08:59:45 <uniqNuuid> markand: when you need man -k you need mandoc-apropos, and it is slow 
2023-01-13 09:24:20 <uniqNuuid> psykose: https://gitlab.alpinelinux.org/alpine/aports/-/issues/14555
2023-01-13 09:24:40 <uniqNuuid> i made some MR, but forgot to change pkgrel
2023-01-13 09:26:02 <uniqNuuid> if you add the `-q` , i'll close the useless MRs
2023-01-13 11:31:48 <WhyNotHugo_> When booting from installation media, installing any `linux-firmware-*` package is not possible due to /lib/modules being a symlink to a readonly mount.
2023-01-13 11:32:03 <WhyNotHugo_> I hacked around this during my system setup by mounting a tmpfs on top and copying the files int.
2023-01-13 11:32:22 <WhyNotHugo_> But I believe using overlayfs here would be best: it retains all the advantages of the loopback mount, while allowing installing additional firmware.
2023-01-13 11:32:32 <WhyNotHugo_> In particular, I needed to install firmware to get my network card to work.
2023-01-13 11:35:54 <ikke> WhyNotHugo_: it's already something that is supported 
2023-01-13 12:18:46 <WhyNotHugo_> ikke: What's the supported approach?
2023-01-13 12:21:35 <WhyNotHugo_> I guess I could have overlayfs mounted myself? But it would be handy if install media was set up this way, rather than having to delete the symlink and do it manually.
2023-01-13 15:03:32 <minimal> WhyNotHugo_: don't linux-firmware-* packages install files into /lib/firmware/ not /lib/modules/ ?
2023-01-13 15:10:02 <uniqNuuid> doas mkfs.ext4 /dev/sda1
2023-01-13 15:10:12 <uniqNuuid> The file /dev/sda1 does not exist and no size was specified.
2023-01-13 15:10:42 <uniqNuuid> but lsblk, blkud, fdisk are able to see /dev/sda{1,2}
2023-01-13 15:10:50 <alpine-meetbot> uniqNuuid meant to say: but lsblk, blkid, fdisk are able to see /dev/sda{1,2}
2023-01-13 15:10:50 <uniqNuuid> s/blkud/blkid/
2023-01-13 15:18:01 <EF> uniqNuuid: try it as root not via doas perhaps
2023-01-13 15:20:51 <uniqNuuid> EF: good suggestion, let me try right now 
2023-01-13 15:23:25 <uniqNuuid> EF: same
2023-01-13 15:23:45 <uniqNuuid> after fdisk, the device can't be seen by mkfs.ext4
2023-01-13 15:25:00 <uniqNuuid> /sbin/fdisk is owned by util-linux-misc-2.38.1-r1
2023-01-13 15:29:25 <heat> uniqNuuid, stat /dev/sda1 ?
2023-01-13 15:29:53 <heat> AFAIK lsblk doesn't look at /dev for block devices but sysfs
2023-01-13 15:31:59 <uniqNuuid> stat : can't find sda1 
2023-01-13 15:32:21 <uniqNuuid> i think it is hardware issue. lack drivers or something. 
2023-01-13 15:32:33 <uniqNuuid> i tried parted to do the partition.. same 
2023-01-13 15:32:47 <uniqNuuid> mkfs can't find it after `parted`
2023-01-13 15:33:13 <uniqNuuid> but, if i unplug and plugin... it is ok. 
2023-01-13 15:35:27 <uniqNuuid> dmesg doesn't say anything err. 
2023-01-13 15:36:19 <markand> hmmm, what could be the reason my hostname gets changed even though I have /etc/hostname?
2023-01-13 15:36:33 <markand> can networkmanager do something fancy like that?
2023-01-13 15:37:26 <minimal> markand: DHCP server providing hostname value along with IP address?
2023-01-13 15:37:46 <uniqNuuid> markand: grep -R hostname /etc
2023-01-13 15:37:54 <uniqNuuid> see what is doing to your hostname. 
2023-01-13 15:37:59 <markand> hmmm, maybe, but what's sure is that it gave me the windows hostname 
2023-01-13 15:38:06 <markand> (dual boot)
2023-01-13 15:39:05 <minimal> markand: MAC address is same so DHCP server will typically treat Windows and Linux boots the same
2023-01-13 15:39:23 <minimal> after all it is the same (physical) machine...
2023-01-13 15:39:40 <markand> looks like I have to add hostname-mode=none  in NetworkManager.conf
2023-01-13 15:39:43 <markand> let's see if that helps
2023-01-13 15:43:18 <WhyNotHugo_> minimal: I might have mixed up the paths (typed them from memory). The path into which linux-firmware installs is a symlink to a read-only mount on the installation media.
2023-01-13 15:53:59 <EF> uniqNuuid: what does dmesg say about sda?
2023-01-13 15:54:50 <uniqNuuid> EF: nothing abnormal. all green
2023-01-13 15:55:20 <uniqNuuid> i changed to another machine to do the fdisk and mkfs. all is fine. 
2023-01-13 15:55:49 <uniqNuuid> i think there is some missing modules on my alpine installation. 
2023-01-13 15:56:23 <uniqNuuid> stuff like voltage regulator or something 
2023-01-13 16:40:34 <minimal> eye
2023-01-13 16:40:47 <minimal> oops
2023-01-13 16:47:37 <WhyNotHugo_> The file installed by the telegram-desktop doesn't match upstream: /usr/share/applications/org.telegram.desktop.desktop
2023-01-13 16:47:53 <WhyNotHugo_> In particular, it uses absolute paths for Exec, which circumvents wrapper scripts.
2023-01-13 16:48:02 <WhyNotHugo_> But I can't find anything in the APKBUILD that would result in this change.
2023-01-13 16:48:23 <WhyNotHugo_> Upstream's file has relative paths: /usr/share/applications/org.telegram.desktop.desktop
2023-01-13 16:56:48 <pj[m]> I'm unsure where do you see that
2023-01-13 16:56:51 <pj[m]> Exec=@CMAKE_INSTALL_FULL_BINDIR@/telegram-desktop -- %u
2023-01-13 16:58:11 <pj[m]> https://github.com/telegramdesktop/tdesktop/commit/c77f07239fe5c134824fc869481bfe35c513262e.patch
2023-01-13 17:02:21 <endigma> what the hell
2023-01-13 17:02:31 <endigma> ```lala```
2023-01-13 17:05:37 <JanJ[m]> Hmz
2023-01-13 17:05:48 <pj> 🤨
2023-01-13 19:01:32 <WhyNotHugo_> pj[m]: You're right. I have an old checkout and this seems to have changed recently.
2023-01-13 19:03:10 <pj[m]> 👍
2023-01-13 19:08:29 <WhyNotHugo_> Does apk have hooks?
2023-01-13 19:08:37 <WhyNotHugo_> Or rather, where are there docs for them?
2023-01-13 19:08:50 <WhyNotHugo_> I want to patch desktop files when they get installed. :P
2023-01-13 19:10:20 <ikke> WhyNotHugo_: not user managable hooks
2023-01-13 19:10:29 <ikke> packages can provider triggers
2023-01-13 19:10:58 <WhyNotHugo_> So you're saying that I just have to make a package that provides a trigger? :P
2023-01-13 19:11:05 <pj[m]> yes
2023-01-13 19:11:24 <WhyNotHugo_> That's actually rather tidy. Any good leads on that?
2023-01-13 19:13:00 <ikke> https://wiki.alpinelinux.org/wiki/APKBUILD_Reference#triggers
2023-01-13 19:18:43 <WhyNotHugo_> Thanks!
2023-01-13 19:18:48 <WhyNotHugo_> `man APKBUILD` is also quite clear on this
2023-01-13 21:11:58 <mfaith> someone mentioned Alpine will be switching to S6 for its init system but I couldn't find any progress-trackers/forums/etc talking about it... is this something slanted for the near future or is there even a dev branch where it's implemented?
2023-01-13 21:12:59 <ikke> mfaith: It's waiting development on s6
2023-01-13 21:13:02 <ikke> so nothing near-future
2023-01-13 21:14:45 <Ermine> when it's done
2023-01-13 21:15:45 <mfaith> gotcha... It looked like s6 was getting a big boost a couple years ago but I didn't find any update since.  Any idea if his work there is still being sponsored and progressing?
2023-01-13 21:16:33 <Ermine> This work is still sponsored
2023-01-13 21:16:43 <mfaith> glad to hear it!
2023-01-13 21:17:11 <elly> is there a list somewhere of what remains to be done for it?
2023-01-13 21:18:56 <ikke> I dont think skarnet gave any updates
2023-01-13 21:19:30 <Ermine> basically everything outlined at https://skarnet.com/projects/service-manager.html
2023-01-13 21:19:55 <Ermine> He posts updates from time to time on his channel, #s6
2023-01-13 21:20:32 <mfaith> is #s6 on oftc or libera or somewhere else?
2023-01-13 21:20:39 <Ermine> here, on oftc
2023-01-13 21:20:45 <mfaith> sweet :)
2023-01-13 21:21:33 <Ermine> Anyway, here is a lot of work to be done, so do not expect s6-rc v1 to go out anytime soon
2023-01-13 21:21:48 <elly> ah, okay
2023-01-13 21:23:26 <mfaith> thanks
2023-01-13 21:28:35 <mfaith> https://skarnet.org/poweredby.html mentions using s6 for that site/server but I guess it's reasonable to assume Alpine would need it to be more featureful and stable.
2023-01-13 21:29:30 <mfaith> Is the plan to adopt it for Alpine somewhat-certain (when it's eventually ready)? Do folks find this exciting?
2023-01-13 21:32:19 <invoked> everyone will have different opinions. but i suspect alpine is valued more for what it currently does than what it doesn't do.
2023-01-13 21:33:01 <invoked> there are many distros to choose from.
2023-01-13 21:34:19 <invoked> there's nothing i'm doing that needs s6, but i'm not opposed to the change
2023-01-13 21:35:49 <Ermine> mfaith: I think this planning is better to be done in vicinity of the release
2023-01-13 21:44:05 <mfaith> I think OpenRC will suffice for my needs too. But conceptually s6 sounded exciting (and it would've been nice to just learn that instead as it takes me extra long learning new technologies). Thanks for chiming in on the topic :)
2023-01-13 22:01:58 <elly> is there an actual openrc package, or is it just a specification implemented by busybox somehow?
2023-01-13 22:03:49 <mfaith> https://pkgs.alpinelinux.org/package/edge/main/x86/openrc
2023-01-13 22:05:50 <dwfreed> https://github.com/openrc/openrc/
2023-01-13 22:05:56 <ikke> openrc is a separate project 
2023-01-13 22:06:59 <elly> oh, I see, alpine-base depends on it
2023-01-13 22:07:03 <elly> I was expecting to see it in world
2023-01-14 01:36:35 <elly> I didn't know setup-disks supported ROOT_SIZE, that's useful
2023-01-14 01:39:16 <minimal> elly: use the source Luke! ;-)
2023-01-14 01:52:18 <elly> indeed
2023-01-14 03:08:46 <elly> rats, I installed alpine on my desktop and the graphics card (an nvidia rtx 2070) is too new for nouveau
2023-01-14 03:08:52 <elly> not quite sure what to do about graphics now
2023-01-14 03:09:11 <psykose> probably nothing you can do
2023-01-14 03:09:36 <elly> unfortunate, it worked neatly under ubuntu when I tried that but I suppose ubuntu has the proprietary nvidia drivers
2023-01-14 03:10:03 <psykose> yeah
2023-01-14 03:10:15 <psykose> hmm, isn't that turing/nv160
2023-01-14 03:10:22 <psykose> should be somethin https://nouveau.freedesktop.org/FeatureMatrix.html
2023-01-14 03:11:03 <psykose> but yeah even if it could render the screen i wouldn't really use it
2023-01-14 03:11:18 <psykose> it crashes all the time, can't play any games, etc
2023-01-14 03:11:18 <elly> not sure, the nouveau I have says: it's an unrecognized device
2023-01-14 03:11:30 <elly> not really sure what to do with that then
2023-01-14 03:11:41 <psykose> which kernel? i'd assume maybe 6.1 has more stuff
2023-01-14 03:11:52 <elly> whatever the latest in 3.17 is, I can check in a moment
2023-01-14 03:13:17 <psykose> 5.15
2023-01-14 03:13:41 <elly> is there a good fallback that will do software rendering or something? I do not particularly care to play games on it, but it is otherwise a powerful machine and would be a more fun place to build chromium :P
2023-01-14 03:16:24 <elly> I wonder if the nvidia binary drivers work
2023-01-14 03:16:28 <psykose> they don't
2023-01-14 03:16:35 <elly> tragic
2023-01-14 03:17:00 <elly> maybe I'll use it to run distcc for my laptop them :P
2023-01-14 03:17:01 <psykose> not sure if you can do anything if nouveau doesn't work, i guess xorg-vesa is supposed to work but i'm not sure of any of that
2023-01-14 03:17:25 <psykose> personally i just bought an amd gpu and everything unsurprisingly works
2023-01-14 03:17:37 <elly> there's always that option
2023-01-14 03:17:41 <elly> annoying!
2023-01-14 03:17:45 <psykose> it is!
2023-01-14 03:17:46 <elly> nvidia release actual drivers challenge
2023-01-14 03:17:48 <elly> [impossible]
2023-01-14 03:18:14 <psykose> even if they did "release open source drivers" it would be a challenge to get to work in alpine
2023-01-14 03:18:29 <psykose> we only really support mesa and don't use glvnd so they would have to add them to mesa itself
2023-01-14 03:18:40 <psykose> it's more likely to see the end of the world i think
2023-01-14 03:18:47 <elly> well, hopefully not
2023-01-14 03:20:30 <elly> https://www.phoronix.com/news/Nouveau-Turing-Linux-4.21 hmmm
2023-01-14 03:20:35 <elly> this allegedly *should* work
2023-01-14 03:20:54 <Fxzxmic> Does WSLg GPU rendering work in alpine?
2023-01-14 03:32:42 <psykose> elly: where are you specifically when it doesn't? i'd imagine the firmware etc aren't on the iso
2023-01-14 03:32:51 <psykose> so you have to get a bit farther and reboot
2023-01-14 04:18:39 <elly> psykose: yeah, I installed to disk and installed xf86-video-nouveau and xf86-video-nv
2023-01-14 04:19:22 <psykose> idk how those really work either sadly
2023-01-14 04:20:00 <elly> yeah, me either
2023-01-14 04:20:03 <elly> X is a bit of a mystery
2023-01-14 04:20:12 <elly> I might try debugging later but if nouveau does not support the 2070 then it doesn't
2023-01-14 04:20:14 <psykose> i've also used wayland for way too long at this point
2023-01-14 04:20:18 <elly> it's possible I am instead missing a firmware
2023-01-14 04:20:25 <psykose> all i know is "install the mesa and install the wayland compositor and start it"
2023-01-14 04:20:32 <psykose> and it just works idk
2023-01-14 04:21:01 <psykose> all nvidia is in linux-firmware-nvidia
2023-01-14 04:21:04 <elly> in the paraphrased words of the meme, girls born after 1990 can't configure xorg
2023-01-14 04:21:39 <elly> (nor can I)
2023-01-14 04:23:53 <psykose> honestly true
2023-01-14 04:24:05 <psykose> didn't get any hot chip at the store though :(
2023-01-14 04:28:54 <heat> elly, xf86-video-nv is old garbo you shouldn't use
2023-01-14 04:29:05 <heat> "This xf86-video-nv driver package contains support for the NVIDIA GeForce 8 series"
2023-01-14 04:30:06 <psykose> xf86-video-n(o)v(ideo)
2023-01-14 05:05:07 <elly> suspicious
2023-01-14 05:19:39 <psykose> sus
2023-01-14 06:16:42 <smallville7123> how do i install the man pges
2023-01-14 06:16:46 <smallville7123> how do i install the man pages
2023-01-14 06:19:57 <smallville7123> found it
2023-01-14 10:05:28 <frag> in a loop in a script, would 'command|busybox cat -A' use much/any more resources than without the cat -A?
2023-01-14 10:07:39 <psykose> so just command | busybox? that doesn't do anything
2023-01-14 10:08:20 <psykose> oh i see what you mean
2023-01-14 10:08:20 <frag> psykose: was a bit lazy, meant without |busybox too :]
2023-01-14 10:08:21 <psykose> hmm
2023-01-14 10:08:48 <psykose> personally don't know :p the cost of that is the buffering in the layers
2023-01-14 10:08:59 <psykose> it also of course uses more cpu because each of them do something
2023-01-14 10:09:09 <psykose> i wouldn't say it's very much
2023-01-14 10:09:11 <psykose> but it's not 0
2023-01-14 10:09:57 <frag> mhm, at least 'time' doesnt show much/any difference
2023-01-14 10:23:52 <psykose> time isn't good for measuring that because it's microoptimisation territory
2023-01-14 10:24:01 <psykose> it also can't tell you the total memory usage of everything at once
2023-01-14 10:24:06 <psykose> afaik
2023-01-14 10:24:18 <psykose> cause there's more than one thing there, it's limited to one process on -v or something
2023-01-14 10:24:39 <psykose> in general though, you don't have to worry about it
2023-01-14 10:25:03 <psykose> cat -A is different from just cat (you change the output), so if that's what you want, use it
2023-01-14 10:25:07 <psykose> if not, then remove the | cat
2023-01-14 10:25:08 <psykose> :)
2023-01-14 10:25:56 <psykose> i don't think there's a catastrophic case there where you run out of memory doing that, unless the writing command keeps some internal buffer it fills faster than the cat reads it
2023-01-14 10:26:15 <heat> if you have some huge dataset you could 'time'
2023-01-14 10:26:34 <psykose> and normally with pipe output that would.. not really be a thing, and the entire scenario (more data production than cat can output) is really really rare
2023-01-14 10:27:19 <psykose> the chance off you having a slow terminal and outputting to terminal itself being slow is the more likely one than any of the stuff happening in those pipes :)
2023-01-14 10:27:26 <frag> hmm for some reason the |cat -A variant seem to work better in the actual script :S
2023-01-14 10:27:39 <psykose> (which is an actual thing! some things can be slowed down by shitty terminal throughput)
2023-01-14 10:28:34 <heat> if you really care about raw throughput you could use pv or something to measure it
2023-01-14 10:29:36 <heat> like 'command | cat -A | pv > /dev/null'
2023-01-14 10:30:22 <psykose> but what if pv itself is slow? :p
2023-01-14 10:30:55 <dwfreed> stdio is generally blocking; if your terminal can't keep up (eg, laggy ssh, because a pty is basically a pipe), and the io is on the same thread as the actual work, indeed you will slow things down
2023-01-14 10:31:08 <heat> as far as I've heard, pv has negligible overhead. but i've never measured it myself
2023-01-14 10:32:15 <dwfreed> (and if you set stdio non-blocking, don't forget to set it back, otherwise you'll break your shell)
2023-01-14 10:33:58 <psykose> yep
2023-01-14 10:34:26 <psykose> it's why i pick fast terminals with no features
2023-01-14 10:37:59 <frag> psykose: which do you prefer?
2023-01-14 10:38:06 <psykose> i use foot
2023-01-14 10:38:24 <frag> duno if i even heard bout it :p
2023-01-14 10:38:31 <psykose> i used to use alacritty (which is even faster tbf) but at that point it gets into yagni territory
2023-01-14 10:38:43 <psykose> and the startup time of a gl context is nonzero
2023-01-14 12:38:02 <pj1> frag: wezterm is also nice and fast
2023-01-14 12:45:17 <satmd> how does xterm perform in this comparison? I'm a purist.
2023-01-14 12:47:38 <psykose> i don't think xterm is very pure
2023-01-14 12:48:02 <psykose> but it definitely speaks to what purism means to people that care about purity :p
2023-01-14 12:48:10 <psykose> (also idk didn't measure it, but probably quite slow)
2023-01-14 12:49:08 <psykose> the latest 378 version of xterm doesn't build against ncurses 6.4/musl and i wanted to report a bug
2023-01-14 12:49:19 <psykose> and all i could find was "email me, the author, directly, for bug reports"
2023-01-14 12:49:21 <psykose> yeah ok
2023-01-14 13:18:08 <vkrishn> psykose: single email to report to should not be considered a bug in itself ;-)
2023-01-14 13:18:18 <psykose> where did i say it's a bug
2023-01-14 13:18:24 <vkrishn> :)
2023-01-14 13:18:47 <vkrishn> but definetly an issue
2023-01-14 13:21:02 <vkrishn> psykose: you do have very good typing speed though, couple of days back it peeked almost 101wpm
2023-01-14 13:21:13 <psykose> hmm
2023-01-14 13:21:17 <psykose> where do you get that number from
2023-01-14 13:21:25 <vkrishn> irc logs
2023-01-14 13:21:36 <psykose> i don't think that's very accurate
2023-01-14 13:21:37 <psykose> hehe
2023-01-14 13:22:07 <vkrishn> :), would like to know more, world speed is 220wpm
2023-01-14 13:30:49 <psykose> you can typeracer me if you want
2023-01-14 13:30:51 <psykose> i'd probably lose
2023-01-14 13:31:46 <vkrishn> nope, I am on diet coke, like stuffs
2023-01-14 13:32:40 <vkrishn> I even do my bread in microwave these days
2023-01-14 13:32:52 <psykose> what does that have to do with anything
2023-01-14 13:33:06 <vkrishn> high octane/carb
2023-01-14 13:33:17 <vkrishn> caffeine
2023-01-14 13:33:24 <psykose> what about it
2023-01-14 13:33:57 <vkrishn> nah, thats off-toppic
2023-01-14 13:36:18 <vkrishn> Ariadne: you back, by chance I can download videos of alpinelinux 2021 conf ?
2023-01-14 13:39:05 <vkrishn> psykose: I think mine peeks ~ 60-80wpm
2023-01-14 13:39:19 <psykose> mine is 2slow
2023-01-14 14:28:29 <vkrishn> first github now codeberg.org. cannot access from my ISP
2023-01-14 14:29:02 <psykose> that's very unfortunate indeed
2023-01-14 14:30:59 <vkrishn> guess what ping github.com still shows, 20.207.73.82
2023-01-14 14:31:14 <vkrishn> managed to do a workaround
2023-01-14 14:44:01 <vkrishn> maybe do salsa.debian.org like site atleast for /main pkgs
2023-01-14 14:47:20 <psykose> what do you mean
2023-01-14 14:47:39 <psykose> that's just a gitlab page
2023-01-14 14:48:34 <vkrishn> they don't store the repo ?
2023-01-14 14:48:44 <psykose> huh?
2023-01-14 14:48:46 <vkrishn> I think fedora people also did it
2023-01-14 14:48:54 <pj1> what repo
2023-01-14 14:49:07 <psykose> do you mean like https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/aaudit or
2023-01-14 14:50:58 <Ermine> lol, this package resides entirely in aports tree
2023-01-14 14:51:16 <psykose> i didn't even realise
2023-01-14 14:51:29 <psykose> like https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/alsa-lib then
2023-01-14 14:51:33 <ikke> There are a couple more
2023-01-14 14:51:48 <psykose> if you mean to downstream fork every single package into its own repository like salsa
2023-01-14 14:51:54 <psykose> well, no comment on that one
2023-01-14 14:52:57 <psykose> the sad part isn't that debian does it, it's that there are entire upstreams that add that debian/ directory into their own repositories
2023-01-14 14:53:45 <vkrishn> no, full alpinelinux copy of upstream
2023-01-14 14:54:13 <vkrishn> with al specific branch/tags in it, if needed
2023-01-14 14:54:16 <ikke> tps://distfiles.alpinelinux.org/distfiles/
2023-01-14 14:54:16 <psykose> yeah, that is just idiotic sorry :)
2023-01-14 14:54:17 <ptrc> there's distfiles.alpinelinux.org
2023-01-14 14:54:25 <ptrc> but that's just a cache
2023-01-14 14:54:25 <ikke> https://distfiles.alpinelinux.org/distfiles/
2023-01-14 14:54:35 <ikke> Just the source archives, which is what we need
2023-01-14 14:55:50 <vkrishn> psykose: how so?
2023-01-14 14:56:03 <psykose> literally what is the point
2023-01-14 14:57:19 <vkrishn> al is distro that does not give source, distfiles is not complete is it ?
2023-01-14 14:57:42 <psykose> the source is the distfiles plus the patches in the aports repo
2023-01-14 14:57:46 <vkrishn> atleast git repo will save disk space
2023-01-14 14:57:53 <psykose> you mean use more disk space?
2023-01-14 14:58:11 <vkrishn> source for all back till  3.0 ?
2023-01-14 15:08:50 <vkrishn> found it, pagure.io
2023-01-14 15:20:03 <vkrishn> with local repo, all APKBUILD files can point to AL version
2023-01-14 15:20:36 <vkrishn> wide idea, but not that bad
2023-01-14 15:25:42 <raajuss[m]> Can I use alpine docker as regular Linux ? 
2023-01-14 15:26:07 <ptrc> "alpine docker"?
2023-01-14 15:26:10 <elly> what do you mean by "as regular Linux"?
2023-01-14 15:27:16 <raajuss[m]> elly: As a daily driver desktop
2023-01-14 15:27:27 <elly> you can definitely use Alpine as a daily-use desktop :)
2023-01-14 15:27:31 <elly> but that's not a docker image usually
2023-01-14 15:28:09 <raajuss[m]> elly: I wanna use it with docker volumes on everyday for testing
2023-01-14 15:28:12 <elly> alpine standard or extended images make fine desktop systems if you are willing to configure them a bit to your needs
2023-01-14 15:28:20 <elly> oh, I don't know much about docker, sorry
2023-01-14 15:28:25 <elly> but alpine definitely works well on desktop
2023-01-14 15:28:47 <raajuss[m]> elly: I wanna use by docker
2023-01-14 15:29:14 <raajuss[m]> elly: Alpine is most used docker image
2023-01-14 15:31:22 <elly> yep, I've heard, but I don't use docker myself so I can't help with that specific question
2023-01-14 15:32:30 <psykose> elly: there's actually some electron 22 crashes i can't seem to figure out that might also be related to chromium :p
2023-01-14 15:33:34 <elly> oh yeah? link :)
2023-01-14 15:33:54 <vkrishn> psykose: what's wrong with https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/aaudit ?
2023-01-14 15:33:57 <elly> electron is a whole different situation but at least shares a lot of code
2023-01-14 15:34:06 <psykose> vkrishn: nothing, why
2023-01-14 15:34:16 <psykose> elly: i'd have to reproduce them again and all, etc
2023-01-14 15:34:37 <psykose> if you want i can send you the 22 upgrade patch and some instructions but it's a lot more effort this time :)
2023-01-14 15:34:54 <psykose> also they are on startup which is annoying
2023-01-14 15:35:05 <psykose> and not every app..
2023-01-14 15:35:06 <vkrishn> ^ do you mean like https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/aaudit
2023-01-14 15:35:27 <raajuss[m]> Should an Arch Linux user switch to Alpine Linux ? 
2023-01-14 15:35:54 <elly> "if you want to"
2023-01-14 15:36:01 <elly> they're different distros with different use cases
2023-01-14 15:36:20 <elly> psykose: tbh I don't really want to diagnose electron problems
2023-01-14 15:36:25 <psykose> that's alright :)
2023-01-14 15:38:01 <JanJ[m]> I'm an arch user who switched to Alpine :) rAajusS 
2023-01-14 15:38:01 <JanJ[m]> i dont think its the right choice for everybody yet..
2023-01-14 15:54:25 <raajuss[m]> <JanJ[m]> "i dont think its the right..." <- Can you elaborate please 🥺
2023-01-14 15:55:37 <JanJ[m]> well, there's a lot more work involved in getting a full blown desktop working in Alpine :) 
2023-01-14 15:55:49 <JanJ[m]> or at least, thats how i feel
2023-01-14 15:56:08 <JanJ[m]> is it worth it? well i think it is :) 
2023-01-14 15:57:00 <psykose> if you are new enough to linux that you don't already know of the differences it is probably not for you :p
2023-01-14 15:57:01 <JanJ[m]> but, why the question? what are you trying to answer? :)
2023-01-14 15:57:04 <raajuss[m]> JanJ[m]: I'm interested to switch only for lightweight and performance advantage of Alpine.
2023-01-14 15:57:04 <raajuss[m]> Is it ? 
2023-01-14 15:57:25 <elly> it is absolutely lightweight, but I doubt you will see different performance from Arch
2023-01-14 15:57:37 <JanJ[m]> Arch can be lightweigt, and honestly i havent found a "wow!! this things is much faster!" yet moment :)
2023-01-14 15:57:42 <raajuss[m]> psykose: I'm a 6 year experienced Arch dev 
2023-01-14 15:57:48 <elly> and you would want to check that your hardware, especially gpu accel and wifi, are well supported by alpine first
2023-01-14 15:57:49 <ikke> apk is much faster :P
2023-01-14 15:57:56 <elly> don't be like me who has a desktop with no working gpu :P
2023-01-14 15:58:12 <JanJ[m]> is it? ikke on what do you base that?
2023-01-14 15:58:20 <ikke> Personal experience
2023-01-14 15:58:26 <JanJ[m]> ahh :)
2023-01-14 15:58:27 <ikke> I use both daily
2023-01-14 15:58:44 <JanJ[m]> i stopped using Arch :P 
2023-01-14 15:58:49 <raajuss[m]> ikke: Arch & Alpine ? 
2023-01-14 15:58:52 <ikke> yes
2023-01-14 15:58:53 <elly> apk specifically is the fastest package manager I've ever used
2023-01-14 15:59:04 <psykose> still wish it had parallel fetch
2023-01-14 15:59:06 <psykose> grumble
2023-01-14 15:59:06 <elly> but time you spend managing packages is a relatively small amount of computer time, imo
2023-01-14 15:59:11 <raajuss[m]> JanJ[m]: Probably you love 32 bit 
2023-01-14 15:59:20 <JanJ[m]> ?
2023-01-14 15:59:43 <ikke> elly: true, but sometimes you are kinda force to do a systme upgrade, and then you are waiting for the package manager to finished
2023-01-14 15:59:44 <psykose> it's very fast with cached apks for i.e. package builds however, entire rootbld trees are near instant
2023-01-14 15:59:44 <ichigo[m]> lol
2023-01-14 15:59:45 <raajuss[m]> JanJ[m]: Alpine is x86
2023-01-14 15:59:58 <ichigo[m]> bro i use 512 bit
2023-01-14 16:00:32 <psykose> alpine supports 8 architectures
2023-01-14 16:00:37 <psykose> well i'd call it 7 i guess
2023-01-14 16:00:37 <JanJ[m]> Alpine runs on several platforms
2023-01-14 16:00:42 <psykose> arch supports 1
2023-01-14 16:01:11 <JanJ[m]> though i havent gotten my Risc-V board yet :) cant try it
2023-01-14 16:01:25 <ikke> JanJ[m]: what board will you get?
2023-01-14 16:01:25 <raajuss[m]> psykose: I tried to switch but I don't want lost AUR
2023-01-14 16:01:39 <elly> ikke: it is true
2023-01-14 16:01:48 <JanJ[m]> i'm looking at the siFive one
2023-01-14 16:01:52 <elly> apk being fast is part of why I update packages every day for sure
2023-01-14 16:01:58 <elly> on my debian system I do not do that :P
2023-01-14 16:02:02 <ikke> heh
2023-01-14 16:02:08 <ichigo[m]> yeah, unless you genuinely believe someone would make a 32 bit only operating system these days,
2023-01-14 16:02:22 <ichigo[m]> usually never happens
2023-01-14 16:02:26 <ikke> we have 3 32-bits arches
2023-01-14 16:02:43 <ichigo[m]> im on postmarket
2023-01-14 16:02:49 <psykose> i am not sure what the point of the aur is aside from either nonfree software or absolute garbage
2023-01-14 16:02:57 <JanJ[m]> postmarket is actually the reason why i switched from arch :)
2023-01-14 16:03:02 <ichigo[m]> the only thing i dont really get so far is the limited locale support
2023-01-14 16:03:16 <JanJ[m]> i wanted everything on the same platform
2023-01-14 16:03:33 <JanJ[m]> no more "arch for desktop, debian for server, alpine for docker" whatever
2023-01-14 16:03:41 <JanJ[m]> so far, im succeeding quite happily :)
2023-01-14 16:03:55 <raajuss[m]> JanJ[m]: How you use alpine ? 
2023-01-14 16:04:09 <raajuss[m]> JanJ[m]: What you use there ? 
2023-01-14 16:04:29 <JanJ[m]> with KDE as desktop :) and im switching most of our servers, containerizing most of our software in the process
2023-01-14 16:04:38 <scorpion2185[m]1> too bad for nvidia-driver only
2023-01-14 16:04:51 <JanJ[m]> i'm on a thinkpad with amd gpu
2023-01-14 16:04:53 <ichigo[m]> i mean it pretty much feels like any other linux ive ever used :) the only difference really is getting used to musl
2023-01-14 16:05:04 <ichigo[m]> apk is good
2023-01-14 16:05:06 <JanJ[m]> that indeed, musl could be a problem
2023-01-14 16:05:25 <ichigo[m]> just dont go trying to install android apk's xd
2023-01-14 16:05:40 <JanJ[m]> but most of what i do is browser based anyway, execept for thunderbird :)
2023-01-14 16:05:43 <elly> musl is a plus a lot of the time and a big minus occasionally, ime
2023-01-14 16:05:52 <elly> but mostly a plus :)
2023-01-14 16:06:00 <ichigo[m]> yeah musl is great
2023-01-14 16:06:04 <JanJ[m]> i think musl can be a plus all the time, you just have to be aware of it on forehand :)
2023-01-14 16:06:29 <scorpion2185[m]1> if you don't have a nvidia-card
2023-01-14 16:07:33 <ichigo[m]> im sure it can do what i want to do (changing the language of the date and time in sxmo) but its probably something in sxmo somewhere cuz LC_TIME works in apps but not in sxmo
2023-01-14 16:07:34 <JanJ[m]> :)
2023-01-14 16:07:35 <JanJ[m]> well, you can always volunteer for that :) 
2023-01-14 16:07:42 <elly> JanJ[m]: it's great right up until you have something that was not tested with musl, and then it rapidly gets less great
2023-01-14 16:07:48 <elly> unless you enjoy fixing that kind of problem
2023-01-14 16:07:56 <JanJ[m]> elly: indeed :)
2023-01-14 16:07:57 <ichigo[m]> yeah, if i figure it out in the first place hah
2023-01-14 16:08:26 <scorpion2185[m]1> if you want to change the date/time on the desktpo check the desktop widget hook
2023-01-14 16:08:42 <raajuss[m]> Jan J.: so can I use everything as I do in Arch in Alpine using musl ??? 
2023-01-14 16:08:57 <JanJ[m]> that depends on what you do :)
2023-01-14 16:09:03 <psykose> most likely not
2023-01-14 16:09:07 <JanJ[m]> but i suspect most things, yeah
2023-01-14 16:09:27 <ikke> raajuss[m]: If you use binary-only programs that are compiled only against glibc, chances are it won't run
2023-01-14 16:09:41 <raajuss[m]> JanJ[m]: I use docker, VBox, Python programming, C++, java programming
2023-01-14 16:09:53 <JanJ[m]> vbox wont work, buyt there's virt-manager for that
2023-01-14 16:10:00 <JanJ[m]> docker works
2023-01-14 16:10:23 <JanJ[m]> i dont do python or c++ or java so i cant say yes for sure :)
2023-01-14 16:10:27 <ichigo[m]> pretty much everything i did on debian works on alpine
2023-01-14 16:10:32 <ichigo[m]> even anki now
2023-01-14 16:10:41 <ichigo[m]> since its in the edge repos
2023-01-14 16:11:10 <JanJ[m]> but with Alpine i'm more inclined to make thinks work. much so because i enjoy these chats much more then other distro's
2023-01-14 16:11:43 <JanJ[m]> sometimes dev's can be.. anal.. so to speak. I havent had that feeling inhere yet
2023-01-14 16:12:27 <JanJ[m]> and since i'm learning much more about docker and what it can do, i imagine one could build a container in which C++ and Python and Java could work :)
2023-01-14 16:12:47 <ikke> We try to be friendly here, but it takes effort
2023-01-14 16:12:51 <psykose> i am probably the third most anal one for reference
2023-01-14 16:12:57 <JanJ[m]> i start with a basic debian container, then add rust to it, and vcdode web, and postgresql, and i have my dev environment :)
2023-01-14 16:13:34 <ikke> there are plenty people with opinions™ here
2023-01-14 16:13:43 <JanJ[m]> opinions are good :)
2023-01-14 16:13:44 <psykose> and all of them are wrong
2023-01-14 16:13:54 <psykose> :p
2023-01-14 16:14:00 <JanJ[m]> but not everyone expresses them as friendly as possible :)
2023-01-14 16:14:18 <ichigo[m]> kek trademarked opinions
2023-01-14 16:14:53 <raajuss[m]> JanJ[m]: You can chat with me. I'm very very experienced Arch developer
2023-01-14 16:15:29 <ichigo[m]> psykose: if all opinions are wrong, wouldnt that make this opinion of all opinions being wrong be wrong? and in that case where do we stand now then?
2023-01-14 16:17:09 <JanJ[m]> rAajusS: but i dont need Arch anymore ;-)
2023-01-14 16:17:36 <raajuss[m]> JanJ[m]: But you can compare and I can learn how Alpine works
2023-01-14 16:17:43 <JanJ[m]> true true :)
2023-01-14 16:18:03 <psykose> ichigo[m]: why the hell does your matrix domain link to a page with a bunch of swastikas :p
2023-01-14 16:18:07 <JanJ[m]> I have much to learn still about Alpine, but i like to think i know a bit about the basics :)
2023-01-14 16:18:35 <ichigo[m]> psykose: its my webserver
2023-01-14 16:19:34 <psykose> saying something there pal
2023-01-14 16:20:11 <raajuss[m]> JanJ[m]: So we can chat
2023-01-14 16:20:15 <JanJ[m]> nice :) i'm almost daily online so sure :)
2023-01-14 16:20:22 <raajuss[m]> I like Alpine Linux because it's so so small
2023-01-14 16:20:47 <ichigo[m]> psykose: idk bro, maybe you should learn what the symbol actually means before judging it
2023-01-14 16:21:27 <raajuss[m]> Docker image of Alpine Linux is just 8mb
2023-01-14 16:21:41 <JanJ[m]> yeah, thats why alpine is so widely used in containers :)
2023-01-14 16:21:56 <ichigo[m]> wow, thats way smaller than i would expect
2023-01-14 16:22:17 <raajuss[m]> JanJ[m]: I want to use that same 8mb Linux as a daily driver
2023-01-14 16:22:41 <raajuss[m]> ichigo[m]: Just search docker Alpine image
2023-01-14 16:22:43 <JanJ[m]> weeeeelll.. i'd think you miss out a lot of stuff :)
2023-01-14 16:23:01 <raajuss[m]> JanJ[m]: Yeah that was my question
2023-01-14 16:23:16 <psykose> ichigo[m]: what does it mean, bro
2023-01-14 16:23:27 <ikke> raajuss[m]: that 8mb is not enough to boot
2023-01-14 16:23:33 <ikke> you need a kernel
2023-01-14 16:23:35 <ikke> you need modules
2023-01-14 16:23:37 <JanJ[m]> BUT, currently my main desktop uses 7.1GB
2023-01-14 16:23:38 <ichigo[m]> i used void for the longest time, it used to be pretty similar to alpine
2023-01-14 16:23:38 <ikke> you need firmware
2023-01-14 16:23:42 <ikke> you need a service manager
2023-01-14 16:23:52 <JanJ[m]> i consider 7.1GB very small for a full blown desktop i use as my daily driver :)
2023-01-14 16:23:59 <raajuss[m]> <raajuss[m]> "Can I use alpine docker as..." <- .
2023-01-14 16:24:07 <ikke> answer: no
2023-01-14 16:24:39 <JanJ[m]> docker was never ment for that :)
2023-01-14 16:24:43 <JanJ[m]> i think
2023-01-14 16:24:46 <raajuss[m]> JanJ[m]: I can get Arch Linux smaller than this
2023-01-14 16:24:46 <ikke> correct
2023-01-14 16:25:24 <JanJ[m]> yeah well, for me its not about smaller or bigger :)
2023-01-14 16:25:24 <raajuss[m]> JanJ[m]: Hm but can I use a full Alpine in docker ? 
2023-01-14 16:25:57 <JanJ[m]> sure, just add a whole lot of apk's to your dockerfile.. but why?
2023-01-14 16:26:10 <ikke> It's just dead weight
2023-01-14 16:26:12 <ichigo[m]> psykose: i still dont really know fully, but basically  its an ancient eastern symbol that dates back more than 10,000 years, probably more, its a classic symbol of good fortune, and its commonly used in japanese anime and culture, its generally known also spiritually as "highest potential".
2023-01-14 16:26:14 <JanJ[m]> docker isnt a hypervisor
2023-01-14 16:26:27 <raajuss[m]> JanJ[m]: Then this will be my Linux desktop for everyday use with docker volumes
2023-01-14 16:26:37 <ichigo[m]> its also an official character in the japanese alphabet
2023-01-14 16:26:48 <raajuss[m]> JanJ[m]: But it can persist
2023-01-14 16:27:03 <JanJ[m]> then install alpine on bare metal? and use containers?
2023-01-14 16:27:34 <raajuss[m]> JanJ[m]: No use docker Alpine and use it as bare metal 
2023-01-14 16:27:48 <ikke> again, that makes no sense
2023-01-14 16:27:50 <ichigo[m]> i can understand american confusion due to crazy man using it towards nefarious ends a few hundred years ago
2023-01-14 16:27:52 <JanJ[m]> heh
2023-01-14 16:28:03 <raajuss[m]> JanJ[m]: Why
2023-01-14 16:28:06 <JanJ[m]> what ikke says :)
2023-01-14 16:28:14 <raajuss[m]> raajuss[m]: What's wrong here
2023-01-14 16:28:34 <JanJ[m]> because the alpine container misses a lot of software :) 
2023-01-14 16:28:47 <ikke> A docker container runs a process on the host, it only needs a filesystem with programs and libraries
2023-01-14 16:28:50 <JanJ[m]> Docker expects things like kernels to exist on the bare metal
2023-01-14 16:28:54 <ikke> to boot linux, you need a kernel
2023-01-14 16:29:00 <raajuss[m]> JanJ[m]: What matrix client you using ? 
2023-01-14 16:29:06 <ichigo[m]> youre not a true alpine user until you get a garden of alpine trees
2023-01-14 16:29:08 <ikke> to support your hardware, you need drivers and kernel modules and firmware
2023-01-14 16:29:20 <JanJ[m]> app.element.io ;-) my docker container for matrix isnt ready yet ;-)
2023-01-14 16:29:26 <ikke> all these things are not present in a docker image because you don't need it
2023-01-14 16:29:36 <ichigo[m]> * know fully, im still researching, but basically
2023-01-14 16:29:45 <JanJ[m]> exactly, because docker expects the docker host to take care of that
2023-01-14 16:29:58 <ikke> docker just runs programs on the host
2023-01-14 16:30:06 <JanJ[m]> you need something to run docker on
2023-01-14 16:30:11 <ikke> docker is sparkling chroots
2023-01-14 16:30:22 <raajuss[m]> ikke: Yes on top of Alpine Linux
2023-01-14 16:30:34 <JanJ[m]> yes, but not the full alpine linux
2023-01-14 16:30:42 <ikke> So if you want to run alpine bare-metal, you need the full os iso, not the minirootfs
2023-01-14 16:31:00 <ichigo[m]> docker is nice but ive never used it before
2023-01-14 16:31:02 <ikke> You can try, but your computer will be confused
2023-01-14 16:31:17 <ichigo[m]> usually i just host whatever straight up
2023-01-14 16:31:22 <raajuss[m]> ikke: But that also boots everytime docker runs the program 
2023-01-14 16:31:44 <ikke> docker does not boot anything
2023-01-14 16:32:00 <JanJ[m]> no, docker doesnt "boot" anything :) your computer boots an OS which starts docker which starts a container :)
2023-01-14 16:32:13 <raajuss[m]> ikke: Then how the program starts in alpine? 
2023-01-14 16:32:19 <ikke> raajuss[m]: what program?
2023-01-14 16:32:20 <JanJ[m]> and this container only contains the library's that the program needs
2023-01-14 16:32:36 <raajuss[m]> ikke: Docker container
2023-01-14 16:32:46 <ikke> starting a container with docker is not much different from starting another program in your OS
2023-01-14 16:33:10 <JanJ[m]> a container is like that.. a box that includes a program and its needed library's
2023-01-14 16:34:14 <JanJ[m]> but you still need software to open that box :)
2023-01-14 16:34:17 <raajuss[m]> ikke: So can't I start an Alpine Linux as a program in docker ? 
2023-01-14 16:34:27 <JanJ[m]> heh, i think you need to read on what docker is :) 
2023-01-14 16:34:47 <ikke> Alpine Linux is not an application
2023-01-14 16:34:50 <raajuss[m]> JanJ[m]: I just want to do that
2023-01-14 16:35:03 <ichigo[m]> <raajuss[m]> "I can get Arch Linux smaller..." <- i would consider under 1gb small
2023-01-14 16:35:12 <raajuss[m]> ikke: So it's not possible ? 
2023-01-14 16:35:24 <ichigo[m]> there are some full desktop iso's around 300mb
2023-01-14 16:35:24 <raajuss[m]> ichigo[m]: Not possible
2023-01-14 16:35:37 <ichigo[m]> raajuss[m]: lol
2023-01-14 16:35:42 <ikke> raajuss[m]: you can start an application which runs on an alpine linux image
2023-01-14 16:36:08 <raajuss[m]> ikke: But I want to start Alpine itself
2023-01-14 16:36:14 <ichigo[m]> floppy disks were a thing
2023-01-14 16:36:24 <ikke> raajuss[m]: docker is not a hyper-visor
2023-01-14 16:36:36 <JanJ[m]> thats what i said :)
2023-01-14 16:36:37 <ikke> you'd use something like qemu
2023-01-14 16:36:42 <ikke> JanJ[m]: yes, and I keep repeating it
2023-01-14 16:36:48 <ikke> until it's clear
2023-01-14 16:36:58 <ikke> docker is not a hypervisor
2023-01-14 16:37:02 <JanJ[m]> if you want to start alpine itself, as an os, you need to install it bare metal, or in a hypervisor like vbox
2023-01-14 16:37:13 <raajuss[m]> ikke: But that won't be 8mb Alpine ! 
2023-01-14 16:37:18 <JanJ[m]> no
2023-01-14 16:37:24 <ikke> raajuss[m]: correct
2023-01-14 16:37:35 <ikke> as we said, you cannot boot anythign with the 8MB image
2023-01-14 16:37:37 <ikke> not possible
2023-01-14 16:37:39 <ikke> unrealistic
2023-01-14 16:37:42 <JanJ[m]> because it must contain more then just the basic library's.. it must also contain a kernel
2023-01-14 16:38:01 <JanJ[m]> and a bootloader, and sometimes a graphical environment
2023-01-14 16:38:20 <JanJ[m]> thats just not doable in 8MB.. only in the times of MS-DOS
2023-01-14 16:38:37 <raajuss[m]> 😭😭
2023-01-14 16:38:48 <raajuss[m]> My dreams broken 
2023-01-14 16:39:01 <raajuss[m]> Wherever
2023-01-14 16:39:10 <raajuss[m]> * Whatever
2023-01-14 16:39:24 <JanJ[m]> well.. you sure dream small then, if 8MB was your max size
2023-01-14 16:39:32 <JanJ[m]> :)
2023-01-14 16:39:38 <psykose> your dreams are just not how linux works
2023-01-14 16:39:53 <psykose> it's like asking can i add 1 + 1 to make lasagna
2023-01-14 16:40:39 <raajuss[m]> Is there any docker matrix room or space ? 
2023-01-14 16:41:09 <JanJ[m]> i can dial 1 to order lasagna though, i have them on speeddial
2023-01-14 16:41:38 <ichigo[m]> psykose: 1 half of lasagna + 1 half of lasagna = 1 lasagna
2023-01-14 16:41:56 <raajuss[m]> JanJ[m]: Bro is matrix truly private and encrypted ? 
2023-01-14 16:41:59 <JanJ[m]> i'm sore docker is on matrix as well :) i havent looked for that yet
2023-01-14 16:42:00 <psykose> honestly true
2023-01-14 16:42:13 <JanJ[m]> matrix CAN be truely private and encrypted
2023-01-14 16:42:18 <JanJ[m]> if its set up that way :)
2023-01-14 16:42:49 <ichigo[m]> you can even use matrix offline encrypted
2023-01-14 16:42:50 <raajuss[m]> JanJ[m]: We're using matrix.org home server. So how are we ? 
2023-01-14 16:42:51 <JanJ[m]> this room is unencrypted so everyone can join and talk.. but you can make an encrypted room
2023-01-14 16:43:12 <raajuss[m]> ichigo[m]: ?
2023-01-14 16:43:14 <JanJ[m]> but the matrix talk is more for a private convo :)
2023-01-14 16:43:22 <JanJ[m]> or.. a new room :)
2023-01-14 16:43:40 <raajuss[m]> JanJ[m]: That's why matrix is my fav
2023-01-14 16:43:49 <ichigo[m]> raajuss[m]: its local network type stuff, i personally have no use case, but it is a feature
2023-01-14 16:44:01 <raajuss[m]> Can I use matrix in replacement of Signal ? 
2023-01-14 16:44:28 <JanJ[m]> yes, but you need others to join you..
2023-01-14 16:44:40 <ichigo[m]> probably, but youd need a bridge or convince your signal friends to make an account
2023-01-14 16:44:44 <JanJ[m]> but like i said, wrong room for that :)
2023-01-14 16:45:11 <JanJ[m]> i'm in the dendrite room, dendrita is a matrix server, you might wanna check that :)
2023-01-14 16:45:15 <raajuss[m]> Is it main room for Alpine ? 
2023-01-14 16:45:28 <ptrc> it is, for alpine-related discussions; for anything else there's #alpine-offtopic
2023-01-14 16:45:40 <raajuss[m]> JanJ[m]: How do I know I'm in what ? 
2023-01-14 16:45:56 <ichigo[m]> i wish matrix spaces and rooms were easier to find/ varied
2023-01-14 16:46:10 <JanJ[m]> rAajusS: i'll invite you, hang on :)
2023-01-14 16:48:37 <markand> any idea why firefox keeps starting as safe mode?
2023-01-14 16:48:45 <markand> even though from plain 'firefox' in command line
2023-01-14 16:49:40 <JanJ[m]> problems with your home directory permissions? corrupted profile?\
2023-01-14 16:52:45 <markand> I've tried removing ~/.mozilla and ~/.cache/mozilla, it freshly opened in safe mode again 
2023-01-14 16:53:35 <JanJ[m]> and it doesnt give any hint?
2023-01-14 16:54:08 <markand> hmm, actually I'm not sure that it's in safe mode, maybe it's just the GNOME's app menu title bar that is wrong
2023-01-14 16:54:19 <JanJ[m]> there's also .config/mozilla i think?
2023-01-14 16:54:58 <JanJ[m]> hmm no, there's not
2023-01-14 16:55:35 <markand> I think GNOME is being confused by the two .desktop file I think
2023-01-14 16:55:46 <JanJ[m]> hmm
2023-01-14 16:56:01 <JanJ[m]> well, i dont like gnome anyway, so i couldn't help you there :)
2023-01-14 16:56:07 <markand> removing the /usr/share/applications/firefox-safe.desktop file and GNOME no longer thinks it's safe mode 
2023-01-14 16:56:16 <JanJ[m]> :)
2023-01-14 16:56:36 <psykose> that sounds like a bug in gnome
2023-01-14 16:56:55 <markand> I'm not huge fan either, but when you have a laptop and carrying it over than use dock and such you're better off using a DE that does everything related to screen and such
2023-01-14 16:57:12 <markand> but oneday I'll probably add more dmenu to automate that in dwm :P
2023-01-14 16:57:31 <psykose> it might get confused by either iconname or startupwmclass
2023-01-14 16:57:32 <psykose> i guess
2023-01-14 16:58:33 <psykose> could you take a screenshot
2023-01-14 17:04:32 <JanJ[m]> time for diner!
2023-01-14 17:04:51 <JanJ[m]> with pineapple.. so a bit of alpine :)
2023-01-14 17:05:58 <Ermine> 
ACTION . o O (are there Alpine users in Alps)

2023-01-14 17:59:45 <hsv2> ehrm somebody has experience with LVM snapshots? I have my root on a LV, I took a snapshot of it with lvcreate --snapshot ... did some changed, did a lvconvert --merge, and rebooted (twice). But for some reason the snapshot LV still exists and the origin LV is still in "merging state" (but it looks like that changes are indeed reverted). Am I missing something or what I didn't understand?
2023-01-14 18:03:23 <pj[m]> > <@raajuss:matrix.org> I'm interested to switch only for lightweight and performance advantage of Alpine.
2023-01-14 18:03:24 <pj[m]> > Is it ? 
2023-01-14 18:03:24 <pj[m]> you said docker and then said lightweight, choose one
2023-01-14 18:03:57 <JanJ[m]> Heh
2023-01-14 18:09:36 <bl4ckb0ne> is anybody using alpine other than for docker?
2023-01-14 18:09:45 <ikke>  yes
2023-01-14 18:09:51 <ptrc> yes
2023-01-14 18:12:14 <hsv2> I don't use docker, I have it on raspberry, on a pendrive as a portable system, and now experimenting on a intel server board :P
2023-01-14 18:12:14 <hsv2> > is anybody using alpine other than for docker?
2023-01-14 18:12:33 <hsv2> oh, sorry for the xmpp style quote. i forgot im on irc
2023-01-14 18:12:59 <Ermine> yes
2023-01-14 18:14:23 <mcrute> ikke: I noticed you have an alpine/Go package, are you interested in MRs for that? I have quite a lot of stuff that I'm maintaining for some side projects that would be a good fit
2023-01-14 18:14:34 <mcrute> would port them over and MR them if you're interested
2023-01-14 18:16:02 <ikke> depends a tiny bit on what, but in general, yes. The goal is to have a library with common functionaity related to alpine to avoid having to duplicate it in a bunch of projects
2023-01-14 18:17:01 <mcrute> well I basically was building a "dev proxy" for packages so it's all key management, package parsing and building, index building, stuff like that
2023-01-14 18:17:28 <mcrute> I also built a lightweight abuild to use for packaging internal go projects
2023-01-14 18:17:58 <mcrute> but like I'd rip out the core of that and offer it up to the alpine/Go repo and then just rebuild my tools on top of alpine/Go
2023-01-14 18:18:59 <ikke> Seems there is at least a bit of overloap
2023-01-14 18:19:02 <ikke> overlap
2023-01-14 18:20:01 <mcrute> okay
2023-01-14 18:21:04 <ikke> mcrute: but feel free to make MRs if you think it adds value. 
2023-01-14 18:22:13 <mcrute> I'll do that... might take a few days to refactor stuff
2023-01-14 18:22:17 <ikke> sure
2023-01-14 18:22:27 <mcrute> no point keeping it all in private repos if it can be shared :-)
2023-01-14 19:34:44 <uniqNuuid> alpine edge vim issue is not a vim issue. it is something else
2023-01-14 19:35:11 <uniqNuuid> the vim auto insert random text in to any file it opens 
2023-01-14 19:35:38 <uniqNuuid> i downgraed vim but keep the other pkgs on edge, the issue is the same 
2023-01-14 19:38:29 <psykose> which terminal do you use
2023-01-14 19:39:12 <uniqNuuid> it does not matter. 
2023-01-14 19:39:24 <Ermine> it does
2023-01-14 19:39:25 <uniqNuuid> v3.17 is fine. edge is not. i use iTerm2 
2023-01-14 19:39:58 <psykose> ok, it doesn't matter i guess
2023-01-14 19:40:00 <psykose> no issue to fix
2023-01-14 19:40:01 <psykose> thanks
2023-01-14 19:40:04 <bl4ckb0ne> the t(1.13.1) stuff?
2023-01-14 19:40:15 <bl4ckb0ne> i havent reported it yet
2023-01-14 19:40:46 <uniqNuuid> it deletes text in the beginning of line 
2023-01-14 19:44:26 <uniqNuuid> psykose: i use iTerm2 all the time, all my linux box are headless... 
2023-01-14 19:44:53 <uniqNuuid> i also tried on iPhone's Terminus 
2023-01-14 19:45:00 <uniqNuuid> a ssh client. 
2023-01-14 19:46:01 <bl4ckb0ne> yeah i noticed that it keeps deleted stuff in memlry
2023-01-14 19:46:33 <bl4ckb0ne> something to do with clipboard it seems, you can se the garbage in :registers
2023-01-14 19:46:57 <uniqNuuid> and among all my linux boxes, only alpine 'edge' vim has this issue and only recently .
2023-01-14 19:51:36 <uniqNuuid> bl4ckb0ne: so i got to use busybox vi. it's actually pretty nice 
2023-01-14 19:51:51 <uniqNuuid> and fast too 
2023-01-14 19:52:33 <bl4ckb0ne> i just `export TERM=xterm`
2023-01-14 19:54:18 <uniqNuuid> bl4ckb0ne: on iOS i use xterm-256color 
2023-01-14 19:56:08 <bl4ckb0ne> or that hack in the foot.ini psykose gave
2023-01-14 19:57:50 <psykose> every terminal i try works except foot
2023-01-14 19:57:53 <psykose> real xD moment
2023-01-14 19:58:21 <uniqNuuid> what font.ini hack 
2023-01-14 19:58:29 <ikke> foot.ini
2023-01-14 19:58:45 <uniqNuuid> where is the file located?
2023-01-14 20:01:20 <uniqNuuid> foot terminal?
2023-01-14 20:01:42 <Ermine> man foot.ini says 1) XDG_CONFIG_HOME/foot/foot.ini 2) ~/.config/foot/foot.ini 3) XDG_CONFIG_DIRS/foot/foot.ini
2023-01-14 20:01:44 <Ermine> and yes
2023-01-14 20:26:51 <frag> busybox ls cant sort by name?
2023-01-14 20:27:23 <Ermine> afaik it does it by default
2023-01-14 20:27:42 <frag> Ermine: ah ofc :]
2023-01-14 23:01:38 <bananicorn> I am trying to wrap my head around working with alpine in RAM + lbu right now and have some trouble
2023-01-14 23:02:05 <bananicorn> I have an µSD card with two partitions, the first containing the whole root filesystem, the second one being intended for persistent storage
2023-01-14 23:03:11 <ikke> run-from-ram means that you don't have a root file system on persistent storage
2023-01-14 23:03:17 <bananicorn> After boot, I add the second partition to fstab with the mountpoint /media/persist, then I went through the installation as normal, selecting /media/persist for the apk cache and lbu backups
2023-01-14 23:03:34 <ikke> The root would be tmpfs
2023-01-14 23:03:42 <bananicorn> Sorry, I used the wrong word then, basically the first partition is the one I boot from
2023-01-14 23:03:51 <ikke> ok
2023-01-14 23:04:04 <bananicorn> It just contains the generic ARM tarball
2023-01-14 23:04:29 <bananicorn> After finishing installation, I ran lbu commit -d
2023-01-14 23:04:46 <bananicorn> However on reboot, my changes were gone
2023-01-14 23:05:27 <bananicorn> Is this because I stored the apkovl on the second partition? Is there a way to store the apkovl on the second partition and still be able to have them loaded properly?
2023-01-14 23:05:36 <ikke> bananicorn: probably
2023-01-14 23:07:27 <ikke> I'm not sure how this would work on an arm system, but there is the apkovl kernel cmdline option
2023-01-14 23:08:29 <bananicorn> Interesting, that should also work with the extlinux.conf I assume
2023-01-14 23:09:01 <bananicorn> By the way, is it required to also add the mountpoint directory (/media/persist) to lbu?
2023-01-14 23:09:25 <ikke> I would not think so
2023-01-14 23:09:38 <ikke> the lbu overlay is loaded in initramfs
2023-01-14 23:11:39 <bananicorn> Okay, thanks for the hint about the cmdline option, I will play with that :)
2023-01-15 04:18:00 <qaqland[m]> why alpine doesn't use busybox passwd. It has another exec.
2023-01-15 04:25:03 <elly> qaqland[m]: what do you mean? on my system /usr/bin/passwd links to /bin/bbsuid which is busybox's setuid wrapper
2023-01-15 04:25:34 <elly> passwd can't link to /bin/busybox because it needs to be setuid
2023-01-15 04:26:26 <elly> it comes from the busybox-suid package though
2023-01-15 04:28:33 <qaqland[m]> elly: ohh, I see. 😃
2023-01-15 04:33:13 <qaqland[m]> thank you!
2023-01-15 07:17:54 <spazialecontinentale[m]> hi how usable is alpine linux as a daily driver? ( Coming for fedora )
2023-01-15 07:18:18 <ikke> depends a bit on how you use it and what you expect from it, but it should be fairly usable
2023-01-15 07:19:30 <ikke> we don't have pre-configured desktop environments
2023-01-15 07:19:58 <spazialecontinentale[m]> ikke: Gnome DE on a laptop, i use only free software ( so i should have no problem with musl )
2023-01-15 07:21:51 <ikke> Should be doable
2023-01-15 07:22:31 <ikke> https://wiki.alpinelinux.org/wiki/Gnome
2023-01-15 07:23:24 <spazialecontinentale[m]> thanks. How about stability? 
2023-01-15 07:23:38 <spazialecontinentale[m]> * about stability? It's comparable to arch linux or fedora?
2023-01-15 07:24:09 <ikke> run-time stablity or package stability?
2023-01-15 07:24:40 <spazialecontinentale[m]> System stability in general
2023-01-15 07:24:56 <zcrayfish> Extremely stable.
2023-01-15 07:25:27 <spazialecontinentale[m]> zcrayfish: also on edge?
2023-01-15 07:27:52 <zcrayfish> I'm not using edge
2023-01-15 07:27:55 <zcrayfish> 
ACTION is running 3.17.1

2023-01-15 07:44:40 <ikke> edge may have occasional breakages
2023-01-15 07:45:09 <ikke> edge is more like archlinux testing repos
2023-01-15 07:47:02 <spazialecontinentale[m]> i was searching on the repo and i see no big difference between 3.17 and edge 
2023-01-15 07:47:02 <spazialecontinentale[m]> ex: firefox is 108.0.2 and also gnome is 43 .r0
2023-01-15 07:47:02 <spazialecontinentale[m]> ragazzi mi avete convinto, appena avrò tempo proverò alpine linux
2023-01-15 07:47:17 <spazialecontinentale[m]> s/ragazzi mi avete convinto, appena avrò tempo proverò alpine linux/you guys have convinced me, as soon as I have time I will try alpine linux/
2023-01-15 07:47:17 <alpine-meetbot> spazialecontinentale[m] meant to say: you guys have convinced me, as soon as I have time I will try alpine linux
2023-01-15 07:48:06 <Ermine> 
ACTION . o O (livin on the edge)

2023-01-15 07:56:23 <spazialecontinentale[m]> i can't find a signal ( chat ) package, does it have a different name?
2023-01-15 07:59:27 <ikke> https://pkgs.alpinelinux.org/packages?name=signal-desktop&branch=edge&repo=&arch=&maintainer=
2023-01-15 07:59:43 <ikke> edge/testing
2023-01-15 08:00:00 <spazialecontinentale[m]> ikke: thanks 
2023-01-15 08:00:01 <zcrayfish> (it's also available in flatpak if the edge/testing version doesn't work correctly)
2023-01-15 08:01:31 <spazialecontinentale[m]> zcrayfish: does a musl flatpak repo exist?
2023-01-15 08:02:14 <qaqland[m]> Vscode remote ssh donnot work when you want to connect to an alpine pc
2023-01-15 08:02:46 <qaqland[m]> It's only the problem for me now
2023-01-15 08:03:16 <qaqland[m]> spazialecontinentale[m]: Flatpak works well
2023-01-15 12:49:28 <bananicorn> I am having trouble understanding and using the apkovl kernel cmdline option properly
2023-01-15 12:50:13 <bananicorn> I am booting generic ARM from /dev/mmcblk0p1, I want to store the apkovl on /dev/mmcblk0p2
2023-01-15 12:50:31 <bananicorn> So in /extlinux/extlinux.conf I have this line:
2023-01-15 12:50:32 <bananicorn> APPEND apkovl=mmcblk0p2:ext4 modules=loop,squashfs,sd-mod,usb-storage quiet
2023-01-15 12:51:20 <bananicorn> However this appears to lead to no apkovl being loaded. If I remove the apkovl=... then it loads the apkovl from /dev/mmcblk0p1
2023-01-15 12:51:54 <bananicorn> Am I missing something here?
2023-01-15 12:53:45 <bananicorn> The only way that I was able to force alpine to load the apkovl from mmcblk0p2 is by removing the default apkovl on mmcblk0p1 and NOT using the apkovl kernel option
2023-01-15 12:53:58 <ikke> bananicorn: it needs to point to an actual file
2023-01-15 12:54:00 <ikke> not just a partition
2023-01-15 12:55:04 <bananicorn> Oh thank you for pointing that out, I seem to have missed the fact that the path is not optional
2023-01-15 12:55:30 <bananicorn> ikke: I assume the path is relative to the given devicename?
2023-01-15 12:55:34 <ikke> https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L42
2023-01-15 13:10:45 <bananicorn> Thank you, I could not get the devicename:path syntax to work unfortunately, is it possible that this only works for devices that are not already mounted?
2023-01-15 13:11:35 <bananicorn> If I just specify a regular path, eg. /media/mmcblk0p2/hostname.apkovl.tar.gz then it works
2023-01-15 13:16:56 <ikke> Not sure if it can be done in another way
2023-01-15 14:31:16 <wyk72> hello guys
2023-01-15 14:31:47 <wyk72> I need help with one utility that I'd like to run under Alpine
2023-01-15 14:32:15 <wyk72> it's the virtual machine manager for libvirt
2023-01-15 14:32:25 <wyk72> it used to work out of the box
2023-01-15 14:32:48 <wyk72> with the latest alpint it's not starting because it's missing python libraries 
2023-01-15 14:32:53 <wyk72> and I can't install them in any way
2023-01-15 14:33:15 <wyk72> I get a ton of errors in missing python modules
2023-01-15 14:33:32 <wyk72> installing them via pip3 is broken too
2023-01-15 14:33:34 <psykose> paste the error somewhere
2023-01-15 14:33:43 <wyk72> I am vary saddened by this
2023-01-15 14:33:46 <wyk72> ModuleNotFoundError: No module named 'libxml2' Traceback (most recent call last):   File "/usr/bin/virt-manager", line 6, in <module>     from virtManager import virtmanager   File "/usr/share/virt-manager/virtManager/virtmanager.py", line 19, in <module>     from virtinst import BuildConfig   File "/usr/share/virt-manager/virtinst/__init__.py", line 50, in <module>     from virtinst.domain import *  # pylint: disable=wildcard-import   File "/usr/sha
2023-01-15 14:34:02 <psykose> somewhere being not here
2023-01-15 14:34:13 <psykose> ok, now `apk policy python3 py3-libxml2 virt-manager`
2023-01-15 14:34:25 <wyk72> I have no idea what to do
2023-01-15 14:34:50 <wyk72> imm-i7-btrfs:~# apk policy python3 py3-libxml2 virt-manager python3 policy:   3.11.1-r2:     lib/apk/db/installed     http://alpinelinux.mirror.garr.it/edge/main py3-libxml2 policy:   2.10.3-r2:     lib/apk/db/installed     http://alpinelinux.mirror.garr.it/edge/main virt-manager policy:   4.1.0-r0:     lib/apk/db/installed     http://alpinelinux.mirror.garr.it/edge/community
2023-01-15 14:35:10 <psykose> https://paste.debian.net/
2023-01-15 14:36:35 <insomniac31> Hello, any idea why ibus can no longer find ibus-anthy after upgrade to 3.17? Did anyone run into similar issue?
2023-01-15 14:37:10 <wyk72> https://paste.debian.net/1267280/
2023-01-15 14:39:42 <psykose> command -v virt-manager
2023-01-15 14:40:00 <wyk72> imm-i7-btrfs:~# command -v virt-manager /usr/bin/virt-manager
2023-01-15 14:40:36 <psykose> hmm
2023-01-15 14:40:54 <blt[m]> Hi, I am wondering if it is mandatory to use acpi for managing how power button acts ? my goal is to pm-hibernate when pushing the button but I am not able to do so only by udev, I am able to do pm-hibernate when level of battery reaches a certain % but not able with power button, I am trying : SUBSYSTEMS=="input", ATTRS{name}=="Power Button", RUN+="/usr/sbin/pm-hibernate". doing a udevadm info --attribute-walk --path=$(udevadm info
2023-01-15 14:40:55 <blt[m]> --query=path --name=/dev/input/event5) result is looking at parent device '/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0C:00/input/input6': ATTRS{name}=="Power Button" any idea what is missing ?
2023-01-15 14:41:00 <psykose> command -v python3
2023-01-15 14:41:14 <wyk72> imm-i7-btrfs:~# command -v python3 /usr/bin/python3
2023-01-15 14:41:36 <Hello71> blt[m]: if that worked it would hibernate when power button module is loaded
2023-01-15 14:41:44 <wyk72> imm-i7-btrfs:~# python3 --version Python 3.9.9
2023-01-15 14:41:45 <Hello71> which doesn't sound very useful
2023-01-15 14:42:12 <psykose> no idea why your python is 3.9.9
2023-01-15 14:42:36 <blt[m]> Hello71: hum, I thought this would be based on event5 which is supposed to be the event generated by the power button, any way to do it instead?
2023-01-15 14:42:44 <wyk72> @psykose which version should it be?
2023-01-15 14:42:47 <psykose> 3.11
2023-01-15 14:43:01 <Hello71> udev listens to uevents, not evdev
2023-01-15 14:43:12 <wyk72> it was a 3.17 machine that I upgraded to edge
2023-01-15 14:43:21 <wyk72> can I force something ?
2023-01-15 14:43:21 <psykose> 3.17 is 3.10
2023-01-15 14:44:04 <psykose> you can apk fix --reinstall python3 i guess
2023-01-15 14:44:22 <psykose> if that doesn't do it then the python actually comes from somewhere else
2023-01-15 14:46:06 <wyk72> @psykose that worked! i.e. "apk fix --reinstall python3"
2023-01-15 14:46:20 <wyk72> now the programs opens regularly
2023-01-15 14:46:28 <psykose> you probably ran something that overwrote the system python
2023-01-15 14:46:36 <psykose> ..while still working
2023-01-15 14:46:37 <psykose> shrug
2023-01-15 14:46:43 <blt[m]> Hello71: looking at my process and my rc-update it seems that I am using udev / udevd
2023-01-15 14:47:18 <psykose> even 3.14 has 3.9.16, so having 3.9.9 is bizarre
2023-01-15 14:47:21 <wyk72> @psykose thanks a bazillion, it's 3 hours I'm swearing "why in the world?"
2023-01-15 14:47:25 <psykose> :)
2023-01-15 14:54:07 <Fxzxmic> command -v and type New command learned.
2023-01-15 18:40:36 <scorpion2185[m]1> does apline have xorg-xinput and xorg-wayland ?
2023-01-15 18:40:47 <ptrc> you mean xwayland?
2023-01-15 18:41:34 <scorpion2185[m]1> yes just xinput for the other I guess
2023-01-15 18:42:15 <ptrc> mhm
2023-01-15 18:43:38 <scorpion2185[m]1> Build-time dependency hwdata found: NO (tried pkgconfig and cmake)... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/oDiFYKJmTmGUlxpzPQIImiHp>)
2023-01-15 18:43:46 <scorpion2185[m]1> but maybe it's normal
2023-01-15 18:48:25 <Ermine> Dependencies of the program you are trying to build are written in that program's documentation
2023-01-15 18:48:40 <Ermine> libavutil is from ffmepg-dev
2023-01-15 18:50:24 <scorpion2185[m]1> /usr/lib/gcc/x86_64-alpine-linux-musl/12.2.1/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find -lOpenGL: No such file or directory
2023-01-15 18:50:42 <scorpion2185[m]1> are they? I copied the deps
2023-01-15 18:50:49 <Ermine> why it should be?
2023-01-15 18:51:10 <Ermine> you want -lGL
2023-01-15 18:51:22 <scorpion2185[m]1> collect2: error: ld returned 1 exit status
2023-01-15 18:51:22 <scorpion2185[m]1> make[1]: *** [Makefile:134: release] Error 1
2023-01-15 18:51:22 <scorpion2185[m]1> ninja: subcommand failed
2023-01-15 18:52:15 <scorpion2185[m]1> how do I tell make to use that instead? i used make install
2023-01-15 18:52:28 <Ermine> edit Makefile
2023-01-15 18:53:53 <Ermine> are you trying to build your program? or you building something else? can you provide us full logs and context and not random lines from it?
2023-01-15 18:55:16 <scorpion2185[m]1> https://github.com/hyprwm/Hyprland/blob/main/Makefile
2023-01-15 18:55:33 <scorpion2185[m]1> I built wayland-protocols 1.31 cause I can't install it
2023-01-15 18:56:17 <scorpion2185[m]1> should I post the full log then?
2023-01-15 18:57:38 <Ermine> what prevents you from installing wayland-protocols?
2023-01-15 18:59:30 <Ermine> did you install all the dependencies listed in the documentation?
2023-01-15 19:02:23 <psykose> lmao hyprland
2023-01-15 19:03:20 <Ermine> psykose: yeah, looked at it and lol'd too
2023-01-15 19:03:56 <scorpion2185[m]1> yes I installed them, arch list
2023-01-15 19:04:17 <scorpion2185[m]1> w.-prot. has old version 1.30 it need 1.31
2023-01-15 19:04:38 <scorpion2185[m]1> <scorpion2185[m]1> "does apline have xorg-xinput and..." <- those were missing
2023-01-15 19:05:08 <scorpion2185[m]1> i installed xipunt and xwayland and xwayland-dev
2023-01-15 19:05:17 <scorpion2185[m]1> why lmao?
2023-01-15 19:05:21 <Ermine> it is 1.31 on edge. Docs say that it would have problems on anything not bleeding edge
2023-01-15 19:08:11 <scorpion2185[m]1> ah I tried testing well it's installed now
2023-01-15 19:09:53 <Ermine> now, post please the full build log
2023-01-15 19:11:35 <scorpion2185[m]1> ninja: job failed: : && /usr/bin/c++ -O3 -DNDEBUG  CMakeFiles/Hyprland.dir/src/Compositor.cpp.o CMakeFiles/Hyprland.dir/src/Window.cpp.o CMakeFiles/Hyprland.dir/src/config/ConfigManager.cpp.o CMakeFiles/Hyprland.dir/src/debug/HyprCtl.cpp.o CMakeFiles/Hyprland.dir/src/debug/HyprDebugOverlay.cpp.o CMakeFiles/Hyprland.dir/src/debug... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/efwTQhRDsjcNFUrCZePDZUyu>)
2023-01-15 19:12:51 <psykose> use meson instead
2023-01-15 19:13:00 <Ermine> by full I mean 'full', not 'starting with error message'
2023-01-15 19:13:25 <Ermine> and please use pastebin for logs
2023-01-15 19:14:14 <psykose> that basically is a pastebin
2023-01-15 19:14:32 <scorpion2185[m]1> https://paste.centos.org/view/1ac86f82
2023-01-15 19:14:44 <psykose> if anything i prefer it because it doesn't have a bunch of css
2023-01-15 19:16:38 <elibrokeit> consider not using hyprland at all
2023-01-15 19:16:57 <psykose> don't think it even builds currently
2023-01-15 19:17:17 <scorpion2185[m]1> elibrokeit why?
2023-01-15 19:18:34 <elibrokeit> the developer of hyprland has an issue with crude bodily humor insults, which becomes extremely evident on a frequent basis any time someone suggests that it would be interesting to package the software in a linux distro "but first can it please use a system copy of wlroots"
2023-01-15 19:19:25 <elibrokeit> also https://github.com/hyprwm/Hyprland/issues/302#issuecomment-1174172559
2023-01-15 19:19:36 <elibrokeit> > it does not. It used to like 3 months ago. (Unless Kainoa fucked up the pkgbuild again in which case I'll have him prosecuted for violating the geneva convention)
2023-01-15 19:19:56 <elibrokeit> really charming fellow lol
2023-01-15 19:21:14 <psykose> oh hey i only had to comment out 2 lines of code to build it this time
2023-01-15 19:21:14 <psykose> https://img.ayaya.dev/9tybEX4DZR21
2023-01-15 19:22:02 <psykose> it's definitely one of the wms ever made
2023-01-15 19:22:23 <elibrokeit> anyways I assure you that hyprland is not interested in being packaged in linux distros, unless it bundles whichever dependencies it feels like
2023-01-15 19:22:41 <elibrokeit> and then it ships a global shared library since the dev doesn't know what static libraries are
2023-01-15 19:22:55 <elibrokeit> of random git commits for wlroots :P
2023-01-15 19:23:07 <psykose> now it does default_library=static afaict
2023-01-15 19:23:18 <elibrokeit> oh miraculous
2023-01-15 19:23:38 <psykose> not hard to fix that anyway as someone that spends all day on it, but
2023-01-15 19:23:42 <psykose> yeah the project is not very high quality
2023-01-15 19:23:53 <psykose> if you need help building it perhaps you shouldn't use it :p
2023-01-15 19:24:00 <elibrokeit> well. the correct way to fix it is using system wlroots
2023-01-15 19:24:20 <elibrokeit> unfortunately it is prone to only compiling against arbitrary unstable commits
2023-01-15 19:24:24 <psykose> spends all day build things* , i don't actually use it
2023-01-15 19:24:56 <elibrokeit> which is a bit of a policy violation :P
2023-01-15 19:27:48 <scorpion2185[m]1> did you build hyprland?
2023-01-15 19:28:11 <psykose> yea
2023-01-15 19:30:27 <scorpion2185[m]1> how can I do it?
2023-01-15 19:30:54 <psykose> <psykose> use meson instead
2023-01-15 19:30:57 <scorpion2185[m]1> 2 lines to remove?
2023-01-15 19:31:24 <scorpion2185[m]1> only using meson?
2023-01-15 19:31:24 <elibrokeit> I would like to re-emphasize the possibility of using a decent WM instead
2023-01-15 19:32:57 <psykose> i would also emphasize it but
2023-01-15 19:33:03 <psykose> i don't think it would be listened to
2023-01-15 19:33:09 <psykose> scorpion2185[m]1: yeah you just use meson instead
2023-01-15 19:35:06 <psykose> elibrokeit: lmao
2023-01-15 19:35:08 <psykose> https://github.com/hyprwm/Hyprland/pull/971#issuecomment-1306147178
2023-01-15 19:35:19 <psykose> the contrast of this to jbeich reporting an actual bug
2023-01-15 19:35:23 <psykose> hilarious
2023-01-15 19:36:03 <elibrokeit> yeah there's a lot of stuff like that 
2023-01-15 19:39:48 <scorpion2185[m]1> I can't use the camera it works on Debian
2023-01-15 19:41:11 <scorpion2185[m]1> ninja -C _build install --tags runtime,man
2023-01-15 19:41:11 <scorpion2185[m]1> ninja: entering directory '_build'
2023-01-15 19:41:11 <scorpion2185[m]1> ninja: unknown target '--tags'
2023-01-15 19:42:06 <elibrokeit> this is not the meson documentation being used
2023-01-15 19:42:37 <scorpion2185[m]1> https://wiki.hyprland.org/Getting-Started/Installation/
2023-01-15 19:43:26 <scorpion2185[m]1> >Sway - For showing how 2 do stuff the overkill way
2023-01-15 19:46:20 <scorpion2185[m]1> why is it not the meson doc?
2023-01-15 19:47:04 <psykose> these docs are wrong
2023-01-15 19:47:20 <Ermine> more lmai
2023-01-15 19:47:23 <Ermine> lmao
2023-01-15 19:49:19 <psykose> anyway it's more like `meson setup out && ninja -C out` and so on
2023-01-15 19:49:38 <psykose> but i will reiterate that if you can't figure this out yourself without getting stuck on every step you should not be using this
2023-01-15 19:49:45 <psykose> i have no idea why you want to torture yourself
2023-01-15 19:50:37 <scorpion2185[m]1> what should I use instead? i ran the first 2 commands
2023-01-15 19:51:00 <psykose> any other wm probably, i dunno
2023-01-15 19:51:27 <scorpion2185[m]1> i mean what command instead
2023-01-15 19:52:12 <psykose> sigh
2023-01-15 19:55:22 <scorpion2185[m]1> `meson setup out &amp;&amp; ninja -C out` ?
2023-01-15 19:56:13 <scorpion2185[m]1> are those the first 2?
2023-01-15 19:59:24 <scorpion2185[m]1> bad developer but maybe the wm is good IDK
2023-01-15 20:00:18 <Ermine> good software does not have build issue
2023-01-15 20:00:26 <psykose> not even close to true
2023-01-15 20:00:28 <Ermine> s/issue/issues/
2023-01-15 20:00:28 <alpine-meetbot> Ermine meant to say: good software does not have build issues
2023-01-15 20:01:02 <Ermine> psykose: ?
2023-01-15 20:01:30 <psykose> Ermine: !
2023-01-15 20:47:26 <elibrokeit> psykose: if you focus on the corollary, that "software with build issues is bad software", maybe?
2023-01-15 20:50:14 <psykose> it's more just that people focus way less on build systems than what they produce
2023-01-15 20:51:08 <elibrokeit> :(
2023-01-15 20:51:58 <psykose> :(
2023-01-15 20:52:35 <elibrokeit> code that does not build is only theoretically good code
2023-01-15 22:06:28 <endigma> not compiling means good software? alpine must be a great development platform then
2023-01-15 22:12:37 <elibrokeit> not sure where you got that from
2023-01-15 22:19:37 <psykose> i've read it three times and it still doesn't make sense
2023-01-16 04:05:37 <uniqNuuid> guys, how do you real logs? 
2023-01-16 04:05:53 <uniqNuuid> tail -f /var/log/messages? 
2023-01-16 04:06:16 <uniqNuuid> s/real/read
2023-01-16 04:06:16 <alpine-meetbot> uniqNuuid meant to say: guys, how do you read logs? 
2023-01-16 04:14:32 <ptrc> uniqNuuid: usually yes, tail is sufficient
2023-01-16 04:14:40 <ptrc> sometimes tail | grep (-v) or something
2023-01-16 04:18:54 <Justin-Alpine> If I remove a third party repo, how can I sync the packages it updated with the official Alpine ones?
2023-01-16 04:19:38 <Justin-Alpine> Fedora has distrosync and it downgrades the packages to the "official" ones.
2023-01-16 04:20:19 <ptrc> you can use apk upgrade -a
2023-01-16 04:20:55 <uniqNuuid> ptrc: logread doesn't seem to work, it says no buffer
2023-01-16 04:22:06 <uniqNuuid> busybox has everything... 
2023-01-16 04:22:06 <Justin-Alpine> Thanks ptrc I'll try.
2023-01-16 04:23:03 <Justin-Alpine> Thanks ptrc that appears to be doing what I need.
2023-01-16 04:24:02 <uniqNuuid> oh,, i found it . need to add a -C to syslogd, so that logread can read it 
2023-01-16 06:27:25 <Fxzxmic> Alpine and WSL are a perfect match.
2023-01-16 08:04:52 <JanJ[m]> Is it possible to make Alpine save old kernels and old initramfs files?? I suddenly have a system that no longer works :/ and i dont think the old kernel gets saved anywhere?
2023-01-16 08:13:53 <lopid> *subscribe*
2023-01-16 08:14:56 <lopid> afaik you have to boot from some other medium and reinstall the apk from /var/cache/apk
2023-01-16 08:15:21 <lopid> if you still have it
2023-01-16 08:20:22 <JanJ[m]> Hmmz, im afraid the kernel doesnt seem to be the problem, more so the chipset drivers i think :( god i hate myself sometimes 🙈
2023-01-16 08:24:01 <JanJ[m]> I really should note down apk versions
2023-01-16 08:25:53 <JanJ[m]> And make backups 🙈
2023-01-16 08:37:22 <uniqNuuid> aha, the mkfs usb issue i reported days ago was due to busybox mdev. switched to mdevd, issue gone. 
2023-01-16 08:39:07 <uniqNuuid> JanJ[m]: there is no old kernel. to install older kernel, you probably need to install from a different branch, e.g. v3.16
2023-01-16 08:46:31 <Fxzxmic> busybox 1.36.0 belongs to which branch? v3.18 or v3.17.2?
2023-01-16 08:49:44 <Fxzxmic> 😘
2023-01-16 09:04:14 <ikke> edge
2023-01-16 09:10:46 <Fxzxmic> I mean in the future.
2023-01-16 09:18:39 <ikke> Then 3.18
2023-01-16 09:20:43 <Fxzxmic> OK, got it
2023-01-16 09:33:56 <JanJ[m]> <uniqNuuid> "Jan J.: there is no old kernel..." <- Well, last night’s apk upgrade definatly upgraded the kernel
2023-01-16 09:34:39 <JanJ[m]> In any case, i should make backups
2023-01-16 09:36:16 <lopid> it would indeed be good if you could choose which kernel to boot into, at boot time
2023-01-16 09:37:04 <donoban> I had lts and edge installed and i could choose what to boot
2023-01-16 09:38:07 <lopid> won't those have substantially different package sets?
2023-01-16 09:39:17 <JanJ[m]> <lopid> "afaik you have to boot from some..." <- My var/cache/apk only contains APKINDEX tar files, which in turn dont seem to contain any apk files
2023-01-16 09:42:22 <donoban> lopid: I mean linux-lts and linux-edge, but inux-edge probably is only on edge
2023-01-16 09:42:58 <donoban> wow no, is also on the release version
2023-01-16 09:43:56 <lopid> it would just give some degree of confidence that you can easily boot into the previous kernel should the latest update fail for you for whatever reason
2023-01-16 09:45:16 <lopid> (rather, the latest kernel fail at boot time, not the update procedure itself)
2023-01-16 09:46:21 <donoban> it seems discussed here 
2023-01-16 09:46:23 <donoban> #12157
2023-01-16 09:47:50 <JanJ[m]> Is there a way to downgrade packages?
2023-01-16 09:48:14 <lopid> just apk add them
2023-01-16 09:48:37 <lopid> the old version, from /var/cache/apk :)
2023-01-16 09:49:03 <JanJ[m]> Like i said before, there are no packages in var/cache/apk
2023-01-16 09:49:11 <JanJ[m]> I would have done that if there where
2023-01-16 09:49:33 <lopid> i don't know if each version is kept on a remote repo
2023-01-16 09:50:05 <JanJ[m]> There’s only tar files which dont seem to contain apk’s eitjer :(
2023-01-16 09:50:34 <lopid> apk's are tarball, iirc
2023-01-16 09:51:53 <JanJ[m]> Yes :)
2023-01-16 09:52:08 <JanJ[m]> But none are there
2023-01-16 09:53:05 <lopid> maybe you have something that runs `apk cache clean`
2023-01-16 09:53:51 <JanJ[m]> Unless that runs by default someonehow
2023-01-16 09:54:02 <lopid> thanks, donoban. maybe we will see that in the next version of apk
2023-01-16 09:54:49 <JanJ[m]> My system boots just fine (a lenovo thinkpad laptop) but the mouse/keyboard seems to fail
2023-01-16 09:54:50 <JanJ[m]> So i cant switch to a terminal either
2023-01-16 09:55:27 <lopid> usb? can you try ps2?
2023-01-16 09:57:11 <JanJ[m]> On a thinkpad laptop? No ;-)
2023-01-16 09:57:21 <JanJ[m]> But i think maybe i need to load some drivers
2023-01-16 09:57:33 <JanJ[m]> With modprobe or something
2023-01-16 10:06:17 <JanJ[m]> Chrooted into the thing, upgrading to edge see if that works 🙈 else i’ll just have to do a new install again
2023-01-16 10:07:15 <lopid> did it work before?
2023-01-16 10:07:50 <JanJ[m]> Yep, had a very nice KDE desktop setup
2023-01-16 10:07:55 <JanJ[m]> Very happy with it
2023-01-16 10:08:54 <lopid> so something broke after an update, and something removed the previous version's apk cache?
2023-01-16 10:09:37 <JanJ[m]> If there even was a previous apk cache
2023-01-16 10:09:58 <lopid> and you could tell there's nothing there now because you can ssh in or something?
2023-01-16 10:10:20 <JanJ[m]> I can chroot into it :)
2023-01-16 10:10:47 <lopid> this coffee is taking longer to sink in this morning
2023-01-16 10:10:55 <JanJ[m]> :)
2023-01-16 10:24:05 <JanJ[m]> Heh, even the 3.17 setup seems to have problems with my system :( cant find the wifi networks
2023-01-16 10:58:23 <WhyNotHugo_> What's up with this MR? https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/37951
2023-01-16 10:58:31 <WhyNotHugo_> I don't see any negative feedback for it, but it's been there for months.
2023-01-16 10:58:44 <WhyNotHugo_> (Asking since I was actually about to try and update the same package)
2023-01-16 11:31:06 <pj[m]> <JanJ[m]> "Is there a way to downgrade..." <- unless you keep cache, no
2023-01-16 11:34:30 <JanJ[m]> So i noticed
2023-01-16 11:35:49 <JanJ[m]> Seems i’ll be using arch for the desktop for a while afterall :) 
2023-01-16 11:41:49 <ikke> JanJ[m]: What is the actual issue you are having?
2023-01-16 11:46:42 <JanJ[m]> I’m stupid and impatient :) 
2023-01-16 11:47:20 <JanJ[m]> Yesterday, a bunch of updates came by. Including kernel etc. This morning, the pc booted up just fine, except mouse and keyboard (neither intern nor extern) would work
2023-01-16 11:48:11 <ikke> not sure if it fixes anything, but ncopa just pushed a new update
2023-01-16 11:48:55 <JanJ[m]> who knows :) by now i installed anew, with even more issues 🙈 since wlan fails to work
2023-01-16 11:49:52 <JanJ[m]> its my own fault really, i need to learn to make backups
2023-01-16 11:51:49 <JanJ[m]> In any case, not even 3.16 seems to work now. And i have no clue why :( i did everything the same i setup the first time 
2023-01-16 11:51:58 <JanJ[m]> but nope, no mouse our keyboard
2023-01-16 11:52:38 <ikke> In a graphical environment or not even in a TTY?
2023-01-16 11:53:17 <JanJ[m]> I cant access the tty, because there’s no mouse or keyboard :) i’m going to try and chroot and disable sddm 
2023-01-16 11:55:48 <JanJ[m]> nope. Not in tty either
2023-01-16 11:56:58 <ikke> And these are USB devices?
2023-01-16 11:57:07 <JanJ[m]> It all works well when booting. Then refind starts openrc 0.44 which boots kernel 5.15.86-0-lts
2023-01-16 11:57:09 <JanJ[m]> and then its all gone
2023-01-16 11:57:10 <JanJ[m]> But when i booted the iso, it worked. 
2023-01-16 11:57:13 <JanJ[m]> Oh wait
2023-01-16 11:57:20 <JanJ[m]> my internel keyboard works
2023-01-16 11:58:23 <JanJ[m]> on the tty it works (internal keyboard) on the gui it does not
2023-01-16 11:58:36 <ncopa> does udev run?
2023-01-16 11:58:40 <JanJ[m]> i HAD it working for the past few weeks :)
2023-01-16 11:58:55 <psykose> are you in the input group
2023-01-16 11:59:12 <JanJ[m]> valid questions! Let me reboot and find out
2023-01-16 12:01:49 <JanJ[m]> Yes, udev runs
2023-01-16 12:03:05 <JanJ[m]> i had this shitshow in the past with an external docking station, upgraded to 3.17 and “magically” everything worked
2023-01-16 12:04:06 <JanJ[m]> but since wlan0 wont work with an error 0x8913 ioct error
2023-01-16 12:06:11 <JanJ[m]> Oh well, lunchtime
2023-01-16 12:13:13 <JanJ[m]> nope, upgrading didnt help either
2023-01-16 12:25:56 <frag> does those usb to RJ45 converters work well for connecting to the wired network?
2023-01-16 12:28:09 <JanJ[m]> I could never get them to work
2023-01-16 12:28:23 <JanJ[m]> not even on macos
2023-01-16 12:30:44 <frag> JanJ[m]: :/
2023-01-16 12:32:11 <JanJ[m]>  ut im beginning to think im just not cut out for computing ;-) 
2023-01-16 14:44:59 <WhyNotHugo_> Any package to decompress rar files?
2023-01-16 14:45:04 <WhyNotHugo_> `unrar` is notably not present
2023-01-16 14:45:48 <aron> libarchive can do it
2023-01-16 14:45:59 <elibrokeit> you could try to package unrar I suppose 
2023-01-16 14:47:28 <psykose> unrar was removed due to nonfree
2023-01-16 14:48:02 <psykose> `bsdtar` from libarchive-tools usually works for me 
2023-01-16 14:48:11 <elibrokeit> I'm sure it was, but users can locally do whatever they want
2023-01-16 14:48:35 <psykose> they indeed can :)
2023-01-16 14:49:23 <elibrokeit> libarchive "quite possibly" works -- it doesn't use libunrar obviously, so clean room coded support is a bit hit or miss
2023-01-16 14:49:58 <PureTryOut> TIL there is bsdtar that can unpack rar files 😱
2023-01-16 14:50:53 <elibrokeit> except when it fails, duh 🤣
2023-01-16 14:51:48 <elibrokeit> which is often, but not as often as it would be if it didn't support them at all :D
2023-01-16 14:52:27 <psykose> i rarely ever find rars at all these days
2023-01-16 14:52:40 <psykose> it's mostly a windows fell-off-a-truck tier of stuff that uses it
2023-01-16 14:52:54 <elibrokeit> indeed
2023-01-16 14:53:08 <psykose> saw it everywhere for games/music or whatnot a decade ago
2023-01-16 14:53:12 <elibrokeit> unfortunately there's an ebook format that uses it
2023-01-16 14:53:16 <psykose> aha
2023-01-16 14:54:33 <PureTryOut> I still see it in games stuff
2023-01-16 14:54:38 <PureTryOut> Every custom map for CoD4 for example...
2023-01-16 14:54:40 <elibrokeit> *Comic Book Archive" -- your choices are .cbr (rar file) or .cbz (zip file)
2023-01-16 14:54:53 <elibrokeit> some people consider rar the lesser evil there
2023-01-16 14:55:51 <psykose> WhyNotHugo_: if you have abuild set up i think the easiest way is also to go to community/7zip and remove the DISABLE_RAR and rebuild it
2023-01-16 14:55:59 <psykose> then it also supports it
2023-01-16 14:56:41 <psykose> it's the same unrar code afaik
2023-01-16 14:56:55 <WhyNotHugo_> Does unrar have licensing issues?
2023-01-16 14:57:12 <pj[m]> psykose: everything related to piracy uses rar
2023-01-16 14:57:49 <WhyNotHugo_> For some reason, rar was popular when I was a teen, so anything compressed during around 2005 is rar.
2023-01-16 14:57:50 <pj[m]> WhyNotHugo_: it's nonfree
2023-01-16 14:58:39 <psykose> i don't think making an unpacker cleanroom-style would have licencing issues
2023-01-16 14:58:45 <psykose> but iirc the format is a mess
2023-01-16 14:58:54 <psykose> and nobody really cares to do it when they can do anything else
2023-01-16 14:59:00 <psykose> peazip is seemingly another alternative
2023-01-16 14:59:01 <pj[m]> you can use gnu unrar
2023-01-16 14:59:15 <psykose> what is that
2023-01-16 14:59:31 <WhyNotHugo_> https://en.wikipedia.org/wiki/Unrar#GNU_UnRAR
2023-01-16 14:59:32 <alpine-meetbot> [wikipedia] Unrar - GNU UnRAR | "This free software is a Linux version of UnRAR that is based on an old version of RARLAB's UnRAR with permission from author Eugene Roshal. It is licensed under the GPL. Work ended in 2007. It does not support the RAR3 format."
2023-01-16 14:59:33 <elly> it's packaged as "unrar-free" in many distros, it's a GPL rar decompressor
2023-01-16 15:00:16 <elly> it doesn't work that well these days though I think
2023-01-16 15:00:16 <psykose> ah it doesn't even support rar3
2023-01-16 15:00:29 <psykose> there's even rar5
2023-01-16 15:00:33 <pj[m]> imo, unrar should be proprietary and rar should be open source
2023-01-16 15:00:34 <elly> oh, looks like libarchive can speak rar though?
2023-01-16 15:00:44 <psykose> minus some failures yeah
2023-01-16 15:00:44 <pj[m]> that way they can vendor lockin
2023-01-16 15:00:58 <pj[m]> yes
2023-01-16 15:01:33 <elly> pj[m]: that would be very funny, it'd be like opt-in ransomware
2023-01-16 15:01:43 <elly> "you have compressed your own files, now buy the decompressor from us"
2023-01-16 15:02:12 <WhyNotHugo_> peazip's rar support depends on either winrar on unrar-nonfree being isntalled.
2023-01-16 15:02:16 <psykose> unlucky
2023-01-16 15:02:16 <pj[m]> bsdtar is shipped on windows btw
2023-01-16 15:02:18 <WhyNotHugo_> 7zip worked :)
2023-01-16 15:02:31 <psykose> 7zip just vendors the same unrar code so unsurprising
2023-01-16 15:02:36 <pj[m]> i mean, there's that one winrar key that everyone knows
2023-01-16 15:03:17 <psykose> we need someone with a substance problem and a month of hyperfocus to make unrar-free-v2-real-colorised
2023-01-16 15:03:26 <aron> winrar 30-day trials also works forever :P
2023-01-16 15:05:05 <JanJ[m]> Heh
2023-01-16 15:07:28 <WhyNotHugo_> I've been using Alpine on the desktop for a couple of weeks now and it's very pleasant.
2023-01-16 15:07:42 <WhyNotHugo_> I appreciate how polishes abuild feels :)
2023-01-16 15:07:52 <psykose> usually people say the opposite
2023-01-16 15:08:19 <WhyNotHugo_> usually people are wrong
2023-01-16 15:08:28 <psykose> hah
2023-01-16 15:08:41 <psykose> hmm, did they update that mesa-git thing by now i wonder
2023-01-16 15:09:00 <aron> I use alpine as desktop more than 2 years now
2023-01-16 15:09:28 <WhyNotHugo_> psykose: Flatpak? Nope, needs to be built from source.
2023-01-16 15:09:38 <psykose> highly unfortunate
2023-01-16 15:09:42 <WhyNotHugo_> When I have time to play videogames I'll just build it myself.
2023-01-16 15:09:50 <psykose> there's always time :D
2023-01-16 15:10:03 <elibrokeit> iirc the rar *and* unrar software come with licenses that forbid you to use it to reverse engineer the format and write a competing OSS archiver/unarchiver
2023-01-16 15:10:39 <psykose> how did libarchive write a decompressor
2023-01-16 15:10:49 <elibrokeit> so you probably technically cannot download it and use it to create a rar file without looking how
2023-01-16 15:10:57 <elibrokeit> but you could get rar files from someone else 
2023-01-16 15:11:02 <psykose> if they can't read the doc or reverse enginner it
2023-01-16 15:11:16 <WhyNotHugo_> Yup: https://github.com/pmachapman/unrar/blob/master/license.txt#L13-L21
2023-01-16 15:11:41 <elibrokeit> you just have to never touch the rar/unrar programs with a ten foot pole or have it on your computer at all I guess.
2023-01-16 15:11:51 <elibrokeit> then you haven't agreed to anything :D
2023-01-16 15:12:03 <WhyNotHugo_> it doesn't say I can't feed an AI the source code and ask the AI to generate a new version.
2023-01-16 15:12:13 <WhyNotHugo_> Behold me new AI: `cat`.
2023-01-16 15:12:28 <psykose> ah it doesn't say anything about reverse engineer
2023-01-16 15:12:36 <psykose> i guess indeed you can just start with rars and unpack em
2023-01-16 15:12:37 <elibrokeit> AI source code laundering is illegal even when done for a good purpose :D
2023-01-16 15:12:48 <psykose> curse you roshal
2023-01-16 15:14:54 <JanJ[m]> the good thing about a seperate /home partition is not having to adjust all your settings again :)
2023-01-16 15:22:33 <elibrokeit> psykose: if I ever get interested in archiving/compression technology, I hereby declare I shall begin that career by writing libuglyarchive, the FOSS rar format handler.
2023-01-16 15:24:30 <JanJ[m]> please do! :)
2023-01-16 15:25:07 <JanJ[m]> We can thewn use that software in our own sync server that rivals iCloud's sync schmuck
2023-01-16 15:25:25 <JanJ[m]> possibility's! i see so many possibility's!
2023-01-16 15:25:55 <JanJ[m]> alas! my life's biggest problem.. so many possibility's, so few personality's to make it all work
2023-01-16 15:26:58 <psykose> as i've been saying, you only need multiple substance addictions to make up for the lack of multiple personalities to accomplish all of that :p
2023-01-16 15:27:27 <JanJ[m]> :D i suppose speed should do the trick
2023-01-16 16:02:09 <JanJ[m]> OMG
2023-01-16 16:02:21 <JanJ[m]> Who pushed the new kernel this morning?
2023-01-16 16:04:27 <psykose> same person as always
2023-01-16 16:04:29 <psykose> how come
2023-01-16 16:05:38 <lopid> ies
2023-01-16 16:05:50 <JanJ[m]> Mouse works, keyboard works
2023-01-16 16:05:54 <JanJ[m]> And wifi works
2023-01-16 16:06:12 <psykose> :)
2023-01-16 16:06:18 <psykose> i guess it was either completely unrelated
2023-01-16 16:06:27 <psykose> or 6.1.4 broke on your hardware and .6 fixed it
2023-01-16 16:06:33 <JanJ[m]> That
2023-01-16 16:06:49 <psykose> edge be like that
2023-01-16 16:07:01 <psykose> i use rc kernels half the time so i know what it's like
2023-01-16 16:07:21 <JanJ[m]> But i dont use edge
2023-01-16 16:07:26 <JanJ[m]> I use 3.17
2023-01-16 16:07:28 <psykose> ah
2023-01-16 16:07:32 <psykose> then the 5.15 versions
2023-01-16 16:07:43 <psykose> that's more weird :P
2023-01-16 16:07:51 <JanJ[m]> Im on 5.15.88 now
2023-01-16 16:07:54 <psykose> i wouldn't have expected some minors in there to break usb hardware or something
2023-01-16 16:08:07 <JanJ[m]> Which was pulled in just now
2023-01-16 16:09:02 <psykose> yep
2023-01-16 16:09:08 <minimal> there's not many changes in 5.15.88 and nothing regarding keyboards
2023-01-16 16:09:25 <psykose> i would be it's a bit higher up like usb or whatnot
2023-01-16 16:09:32 <JanJ[m]> That may be, but it fixed whatever issue there was
2023-01-16 16:09:33 <psykose> but working in tty and not in sddm sounds like something else
2023-01-16 16:09:50 <minimal> no usb changes listed at all for that
2023-01-16 16:10:09 <JanJ[m]> I tell you.. yesterday evening alpine updated. This morning: no mouse and keyboard
2023-01-16 16:10:25 <JanJ[m]> And now after the new linux-lts gets installed, mouse and keyboard
2023-01-16 16:10:43 <JanJ[m]> I wont presume to tell what the problem is, i know too little about Alpine for that
2023-01-16 16:29:29 <JanJ[m]> Though sddm boots to a nice black screen now 😂
2023-01-16 16:30:42 <psykose> now that i don't know how to fix :p
2023-01-16 16:34:52 <JanJ[m]> Something new every day :)
2023-01-16 16:37:40 <uniqNuuid> i found a bug on apk 
2023-01-16 16:39:23 <uniqNuuid> apk search `-d NTP` is able to find chrony. but `-d ntp` will not. that means the `apk search` is case sensitive, which , is nonsense 
2023-01-16 16:41:03 <uniqNuuid> let's vote, if i am not the only one think so, i'll fire a bug 
2023-01-16 16:41:12 <uniqNuuid> s/bug/bug report
2023-01-16 16:41:12 <alpine-meetbot> uniqNuuid meant to say: let's vote, if i am not the only one think so, i'll fire a bug report 
2023-01-16 16:42:58 <ikke> Just open a report
2023-01-16 16:44:27 <JanJ[m]> Indeed
2023-01-16 16:47:40 <uniqNuuid> ikke:  done. 
2023-01-16 16:49:45 <psykose> it's clearly documented as being case-sensitive
2023-01-16 16:49:58 <psykose> this is a feature request and not a bug
2023-01-16 16:50:08 <psykose> aside from that i agree
2023-01-16 16:50:19 <psykose> better as insensitive
2023-01-16 16:50:38 <ikke> " A pattern matches if it is a case-sensitive substring of the package name."
2023-01-16 16:51:38 <uniqNuuid> ok. yeah, doc says so. 
2023-01-16 16:52:04 <uniqNuuid> but insensitive is better, god knows how the decription is written. 
2023-01-16 16:53:06 <ikke> uniqNuuid: I moved the issue
2023-01-16 16:53:41 <uniqNuuid> ikke: thanks 
2023-01-16 17:15:21 <JanJ[m]> Slowly im getting there :) alpine boots to sddm, mouse works, keyboard works, but it throws me back to the loginscreen when i press login
2023-01-16 17:15:49 <psykose> should read the logs i guess
2023-01-16 17:15:55 <psykose> i think sddm writes something somewhere
2023-01-16 17:17:26 <JanJ[m]> true
2023-01-16 17:17:36 <JanJ[m]> but i havent gotten to that part yet ;-)
2023-01-16 17:22:12 <psykose> you can swap ttys with ctrl-alt-fkeys and log in on another and whatnot
2023-01-16 17:22:25 <JanJ[m]> i know :)
2023-01-16 17:22:35 <JanJ[m]> thats all working just fine
2023-01-16 17:30:51 <JanJ[m]> I'M BAAHAACK!
2023-01-16 17:31:09 <JanJ[m]> got booted back to sddm because bash wasnt installed
2023-01-16 17:36:20 <psykose> ah
2023-01-16 17:36:33 <psykose> you set your login shell as bash and forgot to install it? or
2023-01-16 17:38:16 <uniqNuuid> probably installed via script which uses sed to modify /etc/passwd but did not install bash
2023-01-16 17:40:37 <uniqNuuid> the first time i chroor install alpine from minirootfs , and forgot to install e2fsprogs, and it did not boot because fcsk failed and openrc exec rc-abort 
2023-01-16 17:42:25 <uniqNuuid> that leads me to look at rc.log and then to look into how fsck was called  
2023-01-16 17:42:48 <uniqNuuid> a good experience 
2023-01-16 17:48:08 <JanJ[m]> i have no clue why it wants bash
2023-01-16 17:48:10 <JanJ[m]> but npw it has it
2023-01-16 17:48:14 <JanJ[m]> its all happy
2023-01-16 17:49:23 <psykose> you should probably find out
2023-01-16 17:49:26 <psykose> and post the logs somewhere
2023-01-16 17:51:11 <JanJ[m]> maybe its because my home partition gets shared with manjaro too
2023-01-16 17:51:40 <JanJ[m]> mind you, i'll keep an extra os as backup for a while
2023-01-16 17:51:44 <psykose> what do you have in ~/.profile
2023-01-16 17:52:28 <psykose> and is your login shell bash
2023-01-16 18:00:06 <frag> is it just me or mpv youtube videos often just render sound now?
2023-01-16 18:00:33 <psykose> render sound?
2023-01-16 18:01:03 <frag> psykose: i only get audio, not video (at least on many videos)
2023-01-16 18:01:19 <psykose> mpv doesn't play youtube videos so it's just whatever yt-dlp pipes to it
2023-01-16 18:01:34 <psykose> so you can figure out why yt-dlp fetches only audio i guess
2023-01-16 18:01:43 <psykose> hmm
2023-01-16 18:01:47 <psykose> show an example video
2023-01-16 18:02:20 <frag> https://www.youtube.com/watch?v=maQH9Qs0u2U
2023-01-16 18:03:10 <frag> hmm once i got video, once not...
2023-01-16 18:03:14 <frag> ok gotta troubleshoot
2023-01-16 18:04:25 <psykose> reproduced too
2023-01-16 18:05:50 <psykose> when it doesn't work it prints Could not read init fragment.
2023-01-16 18:06:04 <psykose> gonna guess it fails to download/pipe it in a certain way
2023-01-16 18:33:35 <scorpion2185[m]1> mpv can play youtube video
2023-01-16 18:34:05 <scorpion2185[m]1> does it use something else to play them?
2023-01-16 18:35:21 <psykose> as already mentioned it does not play youtube video but it invokes yt-dlp which pipes something into it
2023-01-16 18:39:37 <scorpion2185[m]1> removing yt-dlp doesn't remove mpov
2023-01-16 18:39:37 <scorpion2185[m]1> mpv
2023-01-16 18:41:07 <psykose> indeed, why would it
2023-01-16 18:41:32 <scorpion2185[m]1> i removed it and mpv can play video from youtube
2023-01-16 18:41:50 <psykose> i removed it and it cannot
2023-01-16 18:42:02 <psykose> you still have either yt-dlp or youtube-dl in PATH somewhere
2023-01-16 18:42:58 <scorpion2185[m]1> i have that other right
2023-01-16 18:43:35 <scorpion2185[m]1> do people use ash as their shell? no completion no .ashrc by default
2023-01-16 18:44:00 <psykose> on remote boxes sure
2023-01-16 18:44:11 <psykose> i don't have completions in my normal shell either
2023-01-16 18:44:51 <scorpion2185[m]1> what's you shell?
2023-01-16 18:45:00 <psykose> yash
2023-01-16 18:45:44 <scorpion2185[m]1> yet another shell?
2023-01-16 18:46:49 <psykose> that's the name
2023-01-16 18:47:05 <psykose> i don't really recommend it
2023-01-16 18:49:06 <scorpion2185[m]1> the name stands for that
2023-01-16 18:49:55 <scorpion2185[m]1> >Command line completion with predefined completion scripts for more than 100 commands
2023-01-16 18:52:05 <psykose> i don't have any enabled
2023-01-16 19:03:18 <JanJ[m]> <psykose> "what do you have in ~/.profile" <- There is no .profile
2023-01-16 19:03:46 <psykose> sure, that works
2023-01-16 19:03:53 <psykose> what is the login shell
2023-01-16 19:04:49 <JanJ[m]> Bash :| i think
2023-01-16 19:06:53 <JanJ[m]> yup, bin/bash
2023-01-16 19:07:08 <scorpion2185[m]1> #️⃣#️⃣
2023-01-16 19:07:14 <frag> ive been using busybox ash as login shell, works suprisingly well..
2023-01-16 19:07:26 <frag> (btw)
2023-01-16 19:08:08 <JanJ[m]> Nice :) maybe for the future i will try that
2023-01-16 19:08:19 <JanJ[m]> for now i just want it to “just works”
2023-01-16 19:53:40 <JanJ[m]> sigh, now manjaro wont open firefox because the firefox on alpine is newer :D
2023-01-16 19:53:56 <JanJ[m]> today, i do not like my life
2023-01-16 19:54:28 <psykose> hah
2023-01-16 19:54:40 <psykose> yeah mixing that doesn't go too well
2023-01-16 19:54:56 <psykose> one of them will be slightly newer half the time
2023-01-16 19:56:21 <JanJ[m]> i suppose todayis as good as any day to start using firefox sync :|
2023-01-16 19:58:48 <psykose> sadly doesn't work perfectly :p but at least the core stuff will sync fine (bookmarks, history, password stuff, extensions minus their config ..)
2023-01-16 20:00:34 <JanJ[m]> :) i dont use many extentions anyway
2023-01-16 20:06:01 <Uniformbuffer[m]1> hi, sorry for the weird question, i was wondering if Alpine is currently able to run the new amd gpus (rx 7900 series) and i'm trying to get some info about. Tecnically to run them they require to have mesa 22.2+ compiled with llvm 15+ and kernel 6.0+. Alpine stable satisfy all of those (if i pick the linux-edge kernel), so i would like to ask if anyone has already tried
2023-01-16 20:07:27 <minimal> Uniformbuffer[m]: what about firmware? is that packaged in the linux-firmware-radeon (or -amd?) package in 3.17?
2023-01-16 20:08:33 <minimal> s/amd/amdgpu/
2023-01-16 20:08:33 <alpine-meetbot> minimal meant to say: Uniformbuffer[m]: what about firmware? is that packaged in the linux-firmware-radeon (or -amdgpu?) package in 3.17?
2023-01-16 20:09:01 <psykose> they work yeah
2023-01-16 20:09:08 <psykose> someone here has one
2023-01-16 20:12:49 <Uniformbuffer[m]1> indeed linux-firmware is dated as 2022-12-14, that's the day the required firmware for the new gpus got upstreamed. Cool, thanks a lot for the confirmation!
2023-01-16 20:16:00 <psykose> you can always check if you know the codes https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/log/
2023-01-16 20:16:25 <psykose> afaict there's more updates to it since then but probably not a must
2023-01-16 20:30:11 <Uniformbuffer[m]1> yeah, i expect it to have some performance issues during the initial months, but having it working at least is a big relief
2023-01-16 20:32:10 <Uniformbuffer[m]1> also if i really need i can always move to Edge and go back when stable (i know it's not tecnicallysupported, but apk is the only package manager i managed to do it pretty painlessly)
2023-01-16 20:34:37 <psykose> it's not necessarily unsupported to downgrade to stable
2023-01-16 20:34:46 <Uniformbuffer[m]1>  * also if i really need i can always move to Edge and go back when stable. I know it's not tecnically supported, but i already managed to do it with apk pretty painlessly. I tried to do the same thing with other distros, but apk dependency solver is the only reliable enough to do it 
2023-01-16 20:34:56 <psykose> if anything it works fine, just probably have to run apk fix once and remember the ole -a
2023-01-16 20:35:06 <psykose> mixing edge with stable is what isn't
2023-01-16 20:35:27 <psykose> or mixing any releases, aside from that go nuts
2023-01-16 20:36:36 <Uniformbuffer[m]1> psykose: oh so I misunderstood, good to know!
2023-01-17 04:29:02 <uniqNuuid> psykose: can you also fix mandb-trigger on v3.17? thanks 
2023-01-17 05:35:18 <ikke> uniqNuuid: done
2023-01-17 05:42:01 <Saijin_Naib> <Uniformbuffer[m]1> "also if i really need i can..." <- This. Painlessly moved all my machines from Edge back to latest-stable here with three apk commands, no fuss, and quite quickly. About the only thing I had to do manually was prune orphaned packages I compiled myself or were only available in Edge/testing.
2023-01-17 06:03:32 <uniqNuuid> ikke: thanks, will it get rebuilt?
2023-01-17 06:03:49 <ikke> Yes there is a pkgrel bump in there 
2023-01-17 06:04:33 <uniqNuuid> the rebuilt is triggered as long as there is a new version, and get distributed to mirror?
2023-01-17 06:10:24 <ikke> yes
2023-01-17 06:10:39 <ikke> that all happens automatically
2023-01-17 06:50:03 <uniqNuuid> ikke: installed : (1/2) Upgrading man-db (2.10.2-r2 -> 2.10.2-r3)
2023-01-17 06:50:09 <uniqNuuid> thanks a lot. 
2023-01-17 07:03:48 <uniqNuuid> ikke: i still see this mandb: iconv_open ("UTF-8//IGNORE", "ISO-8859-1"): Invalid argument
2023-01-17 07:04:06 <uniqNuuid> but if i do mandb -q /usr/share/man, no such output 
2023-01-17 07:05:14 <uniqNuuid> if i do e.g. : apk add/del dash-doc, i see iconv_open ("UTF-8//IGNORE", "ISO-8859-1"): Invalid argument 
2023-01-17 07:06:24 <uniqNuuid> if i do : doas ./man-db-2.10.2-r3.Q1w0EBcKseg4buJhp6CWsWR3AjJ1E=.trigger . no such `iconv_opem ...` output
2023-01-17 07:06:47 <uniqNuuid> weird. 
2023-01-17 07:07:48 <uniqNuuid> only happens when using apk to add/del doc
2023-01-17 07:08:01 <uniqNuuid> s/doc/xxx-doc/
2023-01-17 07:08:01 <alpine-meetbot> uniqNuuid meant to say: only happens when using apk to add/del xxx-doc
2023-01-17 07:26:50 <pj[m]> <WhyNotHugo_> "usually people are wrong" <- not in that case
2023-01-17 07:59:37 <uniqNuuid> never mind it is a groff-doc issue. ikke 
2023-01-17 09:10:24 <ShawCross> I noticed that alpine's adduser command sets the home directory permissions to o+rx which means anyone can read anyone else's home directory. Isn't this a security issue? On other linux distros the default does not allow anyone to read.
2023-01-17 09:10:56 <lopid> it actually varies over distros and over time
2023-01-17 09:11:31 <lopid> some give you your own group, some just dump you in 'users'
2023-01-17 09:12:00 <ShawCross> I'm used to not allowing others to read but yeah it probably varies. Still I feel this a security issue and the default should be changed.
2023-01-17 09:12:29 <ShawCross> Has that ever been discussed for alpine?
2023-01-17 09:12:48 <lopid> 
ACTION shrugs

2023-01-17 09:12:48 <alpine-meetbot> ¯\_(ツ)_/¯
2023-01-17 09:13:29 <ShawCross> If it hasn't been discussed, where should I start the discussion? An issue? Irc? Mailing list?
2023-01-17 09:13:29 <psykose> it's what busybox adduser does
2023-01-17 09:13:39 <ShawCross> Yes I checked busybox's source code
2023-01-17 09:13:41 <psykose> it also by default does g+s
2023-01-17 09:13:43 <ShawCross> that doesn't make it a good default
2023-01-17 09:13:49 <psykose> so.. the busybox mailing list
2023-01-17 09:14:21 <ShawCross> The problem is that alpine uses adduser by default in many places. So I feel this is an alpine decision.
2023-01-17 09:14:40 <ShawCross> busybox is going to say "if you care about this then change the permission manually or don't use adduser"
2023-01-17 09:15:20 <ShawCross> For example, alpine packages that install services that have home directories in /var/lib use adduser
2023-01-17 09:15:25 <ShawCross> this means those directories are world readable
2023-01-17 09:16:08 <ShawCross> This is obviously a security issue with alpine to me
2023-01-17 09:19:46 <lopid> such packages change those permissions if necessary as part of installation. postfix, mariadb, for example
2023-01-17 09:20:04 <ShawCross> In that case I have an issue to report with a specific package that forgets this.
2023-01-17 09:21:03 <ShawCross> I'm going to edit https://wiki.alpinelinux.org/wiki/Setting_up_a_new_user too to warn users that their home directory is world readable by default.
2023-01-17 09:22:04 <pj[m]> <ShawCross> "I'm used to not allowing..." <- it's not a security issue
2023-01-17 09:22:08 <psykose> your best bet is probably patching it and proposing it in a merge request
2023-01-17 09:22:24 <ShawCross> I agree that what I'm used to does not make it a security issue
2023-01-17 09:22:46 <ShawCross> What most people are used to *can* make it a security issue
2023-01-17 09:23:10 <ShawCross> Sometimes it is still a security issue even if most people are used to to the insecure default
2023-01-17 09:24:34 <ShawCross> If you feel that home directories should be world readable then why does mariadb (example from above) change the permissions? I feel it is a more reasonable default to not allow it. This means allowing requires the extra effort, not disallowing it.
2023-01-17 09:25:09 <pj[m]> Go argue with bb
2023-01-17 09:25:22 <ShawCross> is bb the maintainer?
2023-01-17 09:25:35 <ShawCross> I mean it more in the sense that it should be consistent across alpine
2023-01-17 09:25:44 <pj[m]> It is consistent though
2023-01-17 09:26:01 <ShawCross> lopid said that some packages revoke the world readable permission from their services home directories
2023-01-17 09:26:04 <ShawCross> this is inconsistent
2023-01-17 09:26:40 <ShawCross> alpine chooses to use busybox adduser as the default way to add users. this choice can be changed.
2023-01-17 09:27:05 <lopid> there is a default. the default is applied to X. X is then changed
2023-01-17 09:27:09 <pj[m]> yes, you can propose it
2023-01-17 09:27:27 <pj[m]> certain packages will still change the default
2023-01-17 09:27:36 <ShawCross> where I am getting at: why is the default changed for some packages?
2023-01-17 09:27:46 <pj[m]> ask maintainer
2023-01-17 09:27:54 <ShawCross> presumably the answer is that there is secret information in the home folder
2023-01-17 09:28:04 <lopid> the default isn't changed, that's what makes it a default
2023-01-17 09:28:07 <ShawCross> and i feel this should be the sane default assumption for all home folders
2023-01-17 09:28:20 <pj[m]> most of the time: because it stores data known to have sensitive values
2023-01-17 09:28:55 <ShawCross> right and i feel the same assumption makes sense for normal users.
2023-01-17 09:29:09 <pj[m]> just like packages change owners/permissions in /var{,/log}, /etc, etc.
2023-01-17 09:29:33 <ShawCross> the reason i noticed this in the first place is that a maintainer of a package i use forgot to remove the default world readable permission but the package does store sensitive informatino there
2023-01-17 09:29:41 <ShawCross> if the default was not allowing it, this mistake would not happen
2023-01-17 09:30:02 <ShawCross> i completely agree that in some cases readable makes sense and in some not
2023-01-17 09:30:09 <lopid> one should never rely on defaults to be a certain way if it is that important
2023-01-17 09:30:20 <ShawCross> it's about making it safe by default
2023-01-17 09:30:25 <pj[m]> sensitive information shouldn't be stored in application account home anyway
2023-01-17 09:30:26 <ShawCross> and extra work to make it unsafe (not readable0
2023-01-17 09:30:53 <lopid> it is the equivalent of web developers not specifying bgcolor because they assume everyone is using a white background
2023-01-17 09:31:13 <ShawCross> with this argument everyone should be forced to specify the permissons all the time
2023-01-17 09:31:19 <ShawCross> but this isn't the case either
2023-01-17 09:31:25 <lopid> if it is that important, yes
2023-01-17 09:31:59 <pj[m]> Maybe we should have a warning when package creates account+home but it's done in hook scripts unfortunately
2023-01-17 09:32:03 <ShawCross> i feel it would be an obvious security improvement to the current state to have not readable by world by default 
2023-01-17 09:32:15 <ShawCross> (have to afk for an hour)
2023-01-17 09:32:30 <lopid> it would be a security issue if packages DIDN'T specify permissions on sensitive data
2023-01-17 09:32:36 <pj[m]> as it was already stated you can make patch and propose
2023-01-17 09:33:13 <ShawCross> I don't know where to make a patch to change alpine's default.
2023-01-17 09:33:16 <lopid> after all, a sysadmin can change the default
2023-01-17 09:33:17 <ShawCross> I would like to make one.
2023-01-17 09:33:25 <pj[m]> in bb source code
2023-01-17 09:33:48 <ShawCross> I will do that too but like I said this is a distribution choice.
2023-01-17 09:33:59 <ShawCross> busybox is less likely to change it than a specific distro
2023-01-17 09:34:53 <ShawCross> Alpine could be changed to not use busybox adduser by default.
2023-01-17 09:35:13 <lopid> it wouldn't be alpine
2023-01-17 09:35:19 <ShawCross> alpine makes changes
2023-01-17 09:35:25 <ShawCross> ship of theseus
2023-01-17 09:36:48 <pj[m]> given that Alpine uses busybox everywhere, it will definitely not change adduser
2023-01-17 09:36:57 <pj[m]> if you need different behaviour you can install coreutils
2023-01-17 09:37:16 <pj[m]> or shadow, I don't remember which one has useradd
2023-01-17 09:38:12 <psykose> https://github.com/mirror/busybox/blob/24198f652f10dca5603df7c704263358ca21f5ce/loginutils/adduser.c#L283
2023-01-17 09:38:19 <psykose> make that 750 and propose it
2023-01-17 09:40:22 <pj[m]> so many issues open..
2023-01-17 10:25:54 <uniqNuuid> ShawCross: chmod 700 /home/$user  , this is what i do for myself
2023-01-17 10:28:57 <uniqNuuid> ShawCross: i think /etc/skel can change the default for all adduser command. 
2023-01-17 10:29:33 <ShawCross> that's unlikely because the permission is hardcoded into the adduser source code
2023-01-17 10:29:49 <uniqNuuid> no, if you do adduser --help 
2023-01-17 10:29:56 <uniqNuuid> -k SKEL         Skeleton directory (/etc/skel)
2023-01-17 10:30:06 <ShawCross> I believe the skeleton directory does something else
2023-01-17 10:30:13 <uniqNuuid> there is no /etc/skel by default, but if you create one ... and test yourself 
2023-01-17 10:30:24 <ShawCross> Doesn't it copy the contents of skel into the new home?
2023-01-17 10:30:29 <ShawCross> I don't think it will touch the directory itself
2023-01-17 10:32:24 <uniqNuuid> ShawCross: let me take a look at it's source code. 
2023-01-17 10:32:41 <uniqNuuid> maybe there is a file called /etc/adduser.conf that works 
2023-01-17 10:32:58 <ShawCross> I don't see any config file mentioned in the adduser source code.
2023-01-17 10:44:48 <uniqNuuid> ShawCross: hmmm. you are right
2023-01-17 10:45:16 <uniqNuuid> then... there is a hack. 
2023-01-17 10:45:53 <uniqNuuid> parse /etc/passwd. find all homes, chmod 700 all of them. set a crontab to do it everyday 
2023-01-17 10:46:56 <uniqNuuid> just some checks, if not 700, then chmod 700 
2023-01-17 10:47:15 <uniqNuuid> that will be secure enough, i think
2023-01-17 10:52:15 <ShawCross> for me personally i want to make sure to change permissions immediately when the home directories are created. looking at /etc/passwd to find all the automatic created ones is a good idea.
2023-01-17 10:53:23 <uniqNuuid> ShawCross: for dir in $(cat /etc/passwd | cut -d: -f6 | grep -vE '^/$|nologin') ; do [ -r $dir ] && echo $dir ; done
2023-01-17 10:55:28 <uniqNuuid> besides my own home, i found these world readable home : /var/spool/mail /srv/ftp /srv/http
2023-01-17 10:56:59 <ShawCross> it depends on the package. for some world readable is fine, for others not.
2023-01-17 11:02:18 <uniqNuuid> ShawCross: then grap -E '^user_you_care_about|^user_you_care_about|...' /etc/passwd
2023-01-17 11:18:37 <WhyNotHugo_> For the `darkman` package, the build script relies on `.git` to determine the current version and inject that into the binary.
2023-01-17 11:18:52 <WhyNotHugo_> AFAIK, an APKBUILD can only pull from https, but not git repositories.
2023-01-17 11:19:00 <WhyNotHugo_> So what I'm doing is patching the version via the APKBUILD.
2023-01-17 11:19:40 <WhyNotHugo_> I'm not totally happy with the approach. Any comments or advise on this?
2023-01-17 11:46:08 <psykose> that is the exact correct approach already
2023-01-17 12:00:30 <uniqNuuid> WhyNotHugo_: github offers download zip , e.g. : https://github.com/ClementTsang/bottom/archive/$FULL_COMMIT_ID.zip
2023-01-17 12:01:17 <WhyNotHugo_> uniqNuuid: Those zips don't contain the `.git` directory. So any versioning metadata is missing.
2023-01-17 12:02:26 <uniqNuuid> then use tag? 
2023-01-17 12:03:05 <donoban> WhyNotHugo_: can't you inject it with $pkgver?
2023-01-17 12:16:22 <psykose> or as i would it a bit better, it's an approach that takes one line of shell code
2023-01-17 12:16:31 <psykose> after you type thousands of them a day it's the best one :p
2023-01-17 12:33:33 <WhyNotHugo_> Okay, that's what I did, so looks like it was the right approach :)
2023-01-17 12:33:34 <WhyNotHugo_> https://gitlab.alpinelinux.org/alpine/aports/-/blob/8c1451da9445cc3ec75be41cb123031528d855a3/testing/darkman/APKBUILD#L23
2023-01-17 12:34:25 <psykose> thanks algitbot
2023-01-17 12:34:29 <pj[m]> that's the proper way to do it
2023-01-17 12:34:37 <pj[m]> although not through LDFLAGS
2023-01-17 12:35:06 <psykose> if you want my 1-cent opinion LDFLAGS are c compiler ldflags and should generally never be used for anything else
2023-01-17 12:35:17 <psykose> for prior precedence i usually see GO_LDFLAGS instead in terms of makefiles
2023-01-17 12:35:50 <psykose> so if you changed yours to GO_LDFLAGS+= .. then it just becomes make VERSION=$pkgver instead, nothing else
2023-01-17 12:42:00 <WhyNotHugo_> Given that LDFLAGS don't necessarily work on go, and other distros migth have the same issue, I've changed this upsream.
2023-01-17 12:45:50 <psykose> it depends, generally every distro sets those flags, but i'd imagine ones that have build-style meta-templating perhaps unset "LDFLAGS" for golang based things and instead pass GOFLAGS="-extldflags=\"$LDFLAGS\"" for consistency
2023-01-17 12:45:51 <psykose> :)
2023-01-17 12:45:56 <psykose> we don't do that so it's all over the place
2023-01-17 12:46:41 <psykose> there's also GO_CFLAGS and the like that i think are read by default by go itself but i'm not sure
2023-01-17 12:46:52 <psykose> or was it CGOFLAGS
2023-01-17 12:46:57 <psykose> no clue
2023-01-17 13:19:06 <psykose> oh nice, someone actually sent a patch to busybox for adduser
2023-01-17 13:34:42 <Ermine> neato
2023-01-17 13:40:09 <Ermine> I wonder why anybody needs debian compatibility in busybox
2023-01-17 13:42:32 <psykose> where that
2023-01-17 13:42:43 <Ermine> http://lists.busybox.net/pipermail/busybox/2023-January/090100.html
2023-01-17 13:43:59 <psykose> busybox doesn't read "adduser.conf" so i'm not sure what this is relevant to
2023-01-17 13:44:20 <psykose> indeed  i am confused with the reply
2023-01-17 13:44:21 <psykose> :)
2023-01-17 13:47:21 <Ermine> I guess debian adduser reads that file and thus creates new home dirs with 755 by default. And I guess reply author wants bb adduser to use the same mode. *shrug*
2023-01-17 13:47:33 <qaqland[m]> Is the busybox in alpine Linux a complete one?
2023-01-17 13:48:33 <qaqland[m]> (Ignore my poor English.)
2023-01-17 13:49:09 <qaqland[m]> <psykose> "busybox doesn't read "adduser...." <- Busybox's adduser has a sketch argu
2023-01-17 13:49:33 <psykose> it has a what
2023-01-17 13:49:39 <Ermine> What do you mean? Is it built with everything? Afaik no. Here's the config: https://git.alpinelinux.org/aports/tree/main/busybox/busyboxconfig
2023-01-17 13:49:57 <psykose> it's built with almost everything
2023-01-17 13:50:18 <qaqland[m]> Skeleton*
2023-01-17 13:56:38 <psykose> that doesn't affect home dirs
2023-01-17 14:02:20 <ShawCross> i sent the patch to busybox btw
2023-01-17 14:03:24 <ShawCross> I think tito might just have been explaining the historical context for why the current mode is 755. I'm not sure if he meant to say it has to match debian.
2023-01-17 14:04:35 <Ermine> hysterical raisins
2023-01-17 14:05:48 <ShawCross> i had to google that to get it :)
2023-01-17 14:11:03 <vkrishn> why not start a wikipage, https://wiki.alpinelinux.org/wiki/Post_install_security_hardening_guidelines, then see how things proceed
2023-01-17 14:11:28 <ShawCross> such a page might be useful. i already edited the "user management" page to add a warning about the default permissions.
2023-01-17 14:12:02 <vkrishn> that can later be split into admins+users
2023-01-17 14:20:36 <Fxzxmic> This is how I use Alpine Linux now: https://imgloc.com/i/H4ToF
2023-01-17 14:24:05 <minimal> separate from the 750/755 discussion, it is not clear why setgid is set on home directories
2023-01-17 14:26:15 <minimal> if Tito's "rule" of "because Debian's adduser does things" was to be applied then I see that on Debian in /etc/adduser.conf SETGID_HOME=no
2023-01-17 14:27:12 <minimal> though that file does have a comment that "yes" "..was the default for versions << 3.13 of adduser. Because it has some bad side effects we no longer do this per default"
2023-01-17 14:28:29 <minimal> so Debian changed this behaviour in June 2000
2023-01-17 15:22:11 <ShawCross> minimal: in what git repository does debian keep their /etc/adduser.conf ?
2023-01-17 15:29:23 <minimal> https://sources.debian.org/src/adduser/3.130/debian/
2023-01-17 15:30:08 <minimal> actually it seems SETGID_HOME was deprecated in July 2022. I was checking on an older Debian system.
2023-01-17 15:32:21 <minimal> also in Aug 2022 they stopped asking about DIR_MODE during package install: https://sources.debian.org/src/adduser/3.130/debian/README/#L89
2023-01-17 15:33:03 <minimal> oops, original link should be https://sources.debian.org/src/adduser/3.130/
2023-01-17 15:33:35 <ShawCross> thanks
2023-01-17 15:42:43 <scorpion2185[m]1> can you prevent apk to update repo when you are installing?
2023-01-17 16:04:19 <elly> sure, don't configure any remote mirrors
2023-01-17 16:04:26 <elly> then configure them after install
2023-01-17 16:07:20 <scorpion2185[m]1> when I want to install a package it updates all the repo unless it is has been done recently
2023-01-17 16:07:33 <scorpion2185[m]1> i don't like that
2023-01-17 16:07:48 <scorpion2185[m]1> i used apk add not apk update ; apk add
2023-01-17 16:08:55 <ikke> Only by overriding the cache timeout via cli 
2023-01-17 16:10:47 <scorpion2185[m]1> how can I do that?
2023-01-17 16:11:08 <WhyNotHugo_> Why does /etc/zoneinfo/ have a copy of the zoneinfo file that's already in /usr/share/zoneinfo?
2023-01-17 16:12:54 <elly> normally that's a symlink into /usr/share/zoneinfo instead, no?
2023-01-17 16:13:15 <psykose> i don't have an /etc/zoneinfo at all
2023-01-17 16:24:54 <elagost> I have an /etc/zoneinfo folder. Just has one dir and one file that's my current TZ.
2023-01-17 16:31:23 <WhyNotHugo_> It's just one dir with the timezone in it. It's a copy of the file in /usr/share.
2023-01-17 16:31:33 <WhyNotHugo_> And /etc/localtime is a link to this copy.
2023-01-17 16:32:36 <minimal> In my case /etc/timezone contains the name of the timezone and /etc/localtime is a symlink to /usr/share/zoneinfo/<timezone path>
2023-01-17 16:33:14 <WhyNotHugo_> /etc/timezone != /etc/localtime
2023-01-17 16:33:34 <minimal> I know, I never said it was
2023-01-17 16:33:37 <WhyNotHugo_> Trying to pair a bluetooth mouse fails and all I see is: daemon.err bluetoothd[3488]: src/service.c:service_accept() input-hog profile accept failed for XX:XX:X:XX:XX:XX:XX
2023-01-17 16:33:54 <WhyNotHugo_> minimal: Oh, so your /etc/localtime seems a bit more sane.
2023-01-17 16:51:15 <kit_ty_kate> hi, i just wanted to know. Is the RISCV64 support planned for a stable release (e.g. 3.18) or is it staying in edge indefinitely until it is more stable?
2023-01-17 17:01:56 <psykose> don't think it's any time soon
2023-01-17 17:02:18 <psykose> i think we have it pending on native hardware for a variety of reasons
2023-01-17 17:02:37 <psykose> which given the current look of things to me i'd say is a very long time, so perhaps someone else will decide differently in a near-er future
2023-01-17 17:02:48 <psykose> but definitely not 3.18 or something soon like that i'd say :)
2023-01-17 17:14:16 <WhyNotHugo_> Okay, so bluez reports that the bluetooth mouse is disconnected -- but moving the mouse moves the cursor.
2023-01-17 17:14:37 <WhyNotHugo_> I've powered down the bluetooth controler... and the mouse still moves the cursor. Obviously it's haunted.
2023-01-17 17:18:22 <WhyNotHugo_> Oh, there was a unifying receiver on the usb hub, and the mouse decided it was a good time to connect to that XD
2023-01-17 17:24:05 <kit_ty_kate> psykose: ok, thanks!
2023-01-17 17:29:26 <ikke> kit_ty_kate: One of our criteria is that we run the package tests for an arch, which we don't do at the moment because we emulate rv64
2023-01-17 17:30:08 <ikke> https://gitlab.alpinelinux.org/alpine/tsc/-/issues/24
2023-01-17 17:30:16 <psykose> dare i say said emulation is quite the endless list of issues
2023-01-17 17:45:06 <minimal> WhyNotHugo_: Unifying receiver? There can only be one! ;-)
2023-01-17 18:13:37 <WhyNotHugo_> The receiver is absolutely terrible tho. Mouse movement works abour 30% of the time, and with an absurd lag.
2023-01-17 18:14:37 <dwfreed> move the receiver away from the PC
2023-01-17 18:15:36 <dwfreed> logitech keyboards usually come with a short 3 inch (or so) extender; if you've got one of those, it works *wonders*
2023-01-17 18:16:07 <dwfreed> or put it on a USB hub or something
2023-01-17 18:20:00 <WhyNotHugo_> Huh, odd, pairing with the "Bolt receiver" asked me to click left/right in a very specific sequence.
2023-01-17 18:20:20 <WhyNotHugo_> I've never seen this before -- but I wonder how it "accidentally" connected to the receiver when I was trying to make bluetooth work.
2023-01-17 18:21:06 <WhyNotHugo_> dwfreed: Thanks for the advise; I used a 50cm extnsion I had lying around and it works wonders.
2023-01-17 18:22:49 <dwfreed> yeah, the receivers don't work very well when they're in close proximity to large chunks of grounded metal (presumably it's faraday cage like behavior, which would certainly explain it), like a PC case or an aluminum laptop case
2023-01-17 18:37:15 <WhyNotHugo_> That makes perfect sense.
2023-01-17 18:37:52 <WhyNotHugo_> I also heard they don't work if you plug them next to a USB3 port with a lot of data going through. Apparently it causes interference too.
2023-01-17 18:38:06 <psykose> tomorrow on reddit: musl libc breaks logitech wireless receiver
2023-01-17 18:39:38 <minimal> WhyNotHugo_: yes USB3 is know to cause interference with Wifi (and therefore I assume potentially for Blutooth), that's why external USB3 HDDs and SSDs tend to come with a short cable
2023-01-17 18:39:51 <dwfreed> WhyNotHugo_: yeah, usb3 causes a lot of interference on 2.4 GHz, which is the frequency band that the unifying receivers use; also affects 2.4 GHz wifi and bluetooth
2023-01-17 18:40:20 <dwfreed> the first consumer routers to ship usb3 didn't shield the usb3 ports, caused a lot of problems
2023-01-17 18:41:07 <minimal> it also complicates board design for SBCs like Raspberry Pi etc as the placement of USB ports versus placement of Wifi chip and antenna is important
2023-01-17 18:41:34 <WhyNotHugo_> This is why 5.0Ghz is a thing.
2023-01-17 18:41:55 <dwfreed> not everything supports 5 GHz :P
2023-01-17 18:41:59 <WhyNotHugo_> Huh, this is a "Logi Bolt" receiver, not a  "Unifying receiver". Specs say it just uses BT-LE.
2023-01-17 18:42:21 <dwfreed> I mean, the only difference there is really the protocol
2023-01-17 18:42:37 <dwfreed> the old unifying receivers and bluetooth both use 2.4 GHz
2023-01-17 18:43:01 <scorpion2185[m]1> <ikke> "Only by overriding the cache..." <- can you explain how?
2023-01-17 18:43:02 <WhyNotHugo_> Using solaar to pair the receiver to the mouse made my click left/right in a specific sequence. I wonder if bluez needs to support that.
2023-01-17 18:44:20 <ikke> scorpion2185[m]1: --cache-max-age 10080
2023-01-17 18:46:29 <scorpion2185[m]1> can you put that in apk conf?
2023-01-17 18:46:34 <psykose> no
2023-01-17 18:46:40 <scorpion2185[m]1> are those seconds?
2023-01-17 18:46:43 <psykose> yes
2023-01-17 18:47:22 <ikke> minutes
2023-01-17 18:47:48 <ikke> "--cache-max-age AGE   Maximum AGE (in minutes) for index in cache before it's refreshed"
2023-01-17 18:48:16 <scorpion2185[m]1> so do I append that to apk?
2023-01-17 18:48:35 <ikke> yes, it's a global option
2023-01-17 18:48:43 <psykose> oh huh, actual minutes
2023-01-17 18:48:43 <psykose> large
2023-01-17 18:48:59 <ikke> that's one week :)
2023-01-17 18:49:50 <scorpion2185[m]1> apk --cache-max-age 10080
2023-01-17 18:49:50 <scorpion2185[m]1> ERROR: '--cache-max-age' is not an apk command. See 'apk --help'.
2023-01-17 18:50:16 <psykose> try specifying an actual command
2023-01-17 18:50:29 <psykose> apk --cache-max-age 1234 add ..
2023-01-17 18:51:11 <scorpion2185[m]1> so do i have to use it  everytime?
2023-01-17 18:51:16 <ikke> yes
2023-01-17 18:51:52 <scorpion2185[m]1> i can add it to the alias , thanks
2023-01-17 18:53:30 <scorpion2185[m]1> hey both camera works now, the preview on cheese is all ruined though
2023-01-17 18:53:41 <scorpion2185[m]1> *work
2023-01-17 18:59:21 <scorpion2185[m]1> still no hibernation
2023-01-17 18:59:45 <psykose> don't remember how cheese works but wasn't there a bunch of options to pick the camera format and if you guess wrong it looks like a new age painting
2023-01-17 19:02:03 <scorpion2185[m]1> photos are fine it is just preview. are there other apps to take photo/video?
2023-01-17 19:03:35 <scorpion2185[m]1> >No effects found
2023-01-17 19:07:38 <scorpion2185[m]1> looks like  old tv
2023-01-17 19:09:33 <scorpion2185[m]1> Saijin_Naib did you fix hibernation by looking kernel modules right?
2023-01-17 19:26:30 <Saijin_Naib> <scorpion2185[m]1> ">No effects found" <- Our version is missing a package to provide those filters
2023-01-17 19:27:28 <psykose> which one was it
2023-01-17 19:27:32 <Saijin_Naib> <scorpion2185[m]1> "Saijin_Naib did you fix hibernat..." <- Sort of. I found out that for my platform I needed pinctl_brockton enabled or built as a module in the kernel 
2023-01-17 19:30:33 <scorpion2185[m]1> 
ACTION uploaded an image: (152KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/bGzbJsBmIJjYjTTkboKvQteB/2023-01-17_20-30-13.png >

2023-01-17 19:31:21 <scorpion2185[m]1> other camera has a simila gray scale effect, but only in the preview (there are no effects actually)
2023-01-17 19:31:36 <scorpion2185[m]1> how did you find out?
2023-01-17 19:32:27 <Saijin_Naib> psykose: https://gitlab.gnome.org/GNOME/cheese/-/blob/master/meson.build#L143
2023-01-17 19:32:39 <Saijin_Naib> gnome-video-effects
2023-01-17 19:33:01 <psykose> ah
2023-01-17 19:33:11 <psykose> yeah this thing has no updates in like 3 years haha
2023-01-17 19:34:13 <Saijin_Naib> scorpion2185[m]1: DIFFing loaded modules and kernel configuration between Alpine and Manjaro, then compiling my own kernel to test. Took me about 6mo to learn and figure it all out with help from folks here 
2023-01-17 19:35:13 <Saijin_Naib> psykose: Shouldn't really matter for gstreamer effects, right? I mean, they have existed for like 12 years or more by now they should be pretty solid 
2023-01-17 19:36:24 <Saijin_Naib> scorpion2185[m]1: I had that too before I got all the kernel modules I needed for my platform 
2023-01-17 19:36:29 <psykose> if it supports them sure
2023-01-17 19:36:32 <psykose> idk anything about it
2023-01-17 19:37:14 <Saijin_Naib> psykose: Yep, still a runtime dep. shipping with Cheese on every other distro I've used 
2023-01-17 19:38:35 <Saijin_Naib> Mostly silly, but a few are actually useful for work like axis flip, etc 
2023-01-17 19:38:45 <Saijin_Naib> But even the silly ones have a purpose for some folks 🤷
2023-01-17 19:38:45 <Saijin_Naib> I'm just too old and boring for them myself 
2023-01-17 19:42:37 <scorpion2185[m]1> so maybe some alpine devs could fix that
2023-01-17 19:45:06 <Habbie> does apk in the alpine docker image filter (remove) init.d files from packages?
2023-01-17 19:45:13 <ikke> no
2023-01-17 19:45:32 <Habbie> hmm, then why is my /etc/init.d empty after apk add pdns nginx
2023-01-17 19:45:37 <ikke> these files are always moved to an -openrc subpackage
2023-01-17 19:45:45 <Habbie> ah! that helps, thanks
2023-01-17 19:45:50 <ikke> which only gets pulled in if you have openrc installed
2023-01-17 19:45:57 <Habbie> right
2023-01-17 19:46:02 <ikke> You can still manually add it
2023-01-17 19:46:05 <Habbie> just did
2023-01-17 19:46:11 <ikke> :)
2023-01-17 19:46:20 <Habbie> so kind of like what i asked except i asked very poorly, thanks
2023-01-17 19:46:24 <Habbie> as the docker image ships without openrc
2023-01-17 19:46:56 <ikke> indeed
2023-01-17 19:47:09 <minimal> in many/most cases you don't have an init running in a docker container, so no need for init.d files
2023-01-17 19:47:23 <ikke> https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L2035 if you are curious
2023-01-17 19:47:25 <Habbie> yes, i do agree
2023-01-17 19:47:36 <Habbie> ikke, ack, nice
2023-01-17 19:48:15 <Habbie> the output of 'abuild help' is .. unexpected :)
2023-01-17 19:48:21 <ikke> heh
2023-01-17 19:48:37 <ikke> then you learn that abuild is a general command invoker
2023-01-17 19:48:44 <Habbie> yep
2023-01-17 19:49:34 <JanJ[m]> is anyone using alpine with samba? 
2023-01-17 19:49:53 <JanJ[m]> we have very slow speed, only since a week or so :|
2023-01-17 19:50:16 <aron> JanJ[m]: what do you mean?
2023-01-17 19:51:23 <JanJ[m]> copying files to a sama share is terribly slow, like 30MB/s
2023-01-17 19:52:38 <aron> let me check my speed
2023-01-17 19:52:54 <psykose> 30 mb/s sounds fast
2023-01-17 19:52:58 <JanJ[m]> it seems a windows 11H2 thing.. i am hoping someone could verify this
2023-01-17 19:53:15 <JanJ[m]> we have a gigabit network.. 300MB/s sounds fast ;-)
2023-01-17 19:53:23 <psykose> gigabit is 125
2023-01-17 19:53:37 <psykose> unless you mean 30 megabits and not megabytes
2023-01-17 19:53:37 <aron> I can write it ~95 mb/s
2023-01-17 19:53:40 <psykose> that would be actually slow
2023-01-17 19:53:43 <aron> mbytes/s
2023-01-17 19:54:03 <Saijin_Naib> Yep, with overhead, I get 124MB/s fully saturated link 
2023-01-17 19:54:08 <JanJ[m]> well, i mean slow ;) whatever the bits and bytes are
2023-01-17 19:54:33 <JanJ[m]> well, 124 is still better then 30
2023-01-17 19:54:36 <aron> I think 95 mbytes/s is the max speed with samba's overhead on gigabit lan
2023-01-17 19:54:43 <dewdude> 115
2023-01-17 19:54:59 <Saijin_Naib> <scorpion2185[m]1> "so maybe some alpine devs..." <- Fix what? If they don't have your specific platform they won't know what needs enabling, unfortunately 
2023-01-17 19:55:00 <dewdude> I max my synology out over smb at 115MB/s
2023-01-17 19:55:10 <aron> then my disk is slow
2023-01-17 19:55:22 <aron> anyway, faster than 30
2023-01-17 19:55:30 <JanJ[m]> heh, what use is a gigabit network, if everything is so slow
2023-01-17 19:56:35 <Saijin_Naib> Bit is 1/8th a byte, the more common unit of measure so you need to scale the value 
2023-01-17 19:56:49 <Saijin_Naib> 1,000 megabit is 125megabyte
2023-01-17 19:56:50 <aron> my test was alpine -> samba -> alpine, and source is a 10+ years old hdd
2023-01-17 19:57:47 <JanJ[m]> a lot of windows 11H2 users seem to report slow network stuff, so now at least i can say thats the problem :)
2023-01-17 19:59:27 <minimal> Saijin_Naib: "mega" (MB) is a 1000-based prefix, "mebi" (MiB) is a "1024"-based prefix
2023-01-17 20:00:35 <Saijin_Naib> minimal: Yeah, that's what I did the math on. 
2023-01-17 20:01:25 <minimal> just clarifying, it can get confusing when people start using MB, MiB, Bps, bps, etc interchangeably
2023-01-17 20:01:40 <aron> (minus the tcp/ip+samba overhead)
2023-01-17 20:01:59 <JanJ[m]> well thanks for the clarification ;-) i'll try to remember it
2023-01-17 20:03:01 <JanJ[m]> time to look for another solution if its a limitation of samba/smb protocol
2023-01-17 20:03:15 <JanJ[m]> maybe i should try webdav :)
2023-01-17 20:03:43 <aron> try sshfs :P
2023-01-17 20:03:59 <JanJ[m]> i dont think that works on windows ;-)
2023-01-17 20:04:23 <aron> sure, not ;)
2023-01-17 20:04:27 <aron> maybe with winscp
2023-01-17 20:06:31 <JanJ[m]> the horror ;-)
2023-01-17 20:06:55 <aron> or consider replacing windows with linux
2023-01-17 20:07:14 <aron> :P
2023-01-17 20:07:18 <JanJ[m]> yes, when i'm suicidal i'll propose that to the board of directors
2023-01-17 20:07:44 <Saijin_Naib> <minimal> "just clarifying, it can get..." <- I don't know when to use what 😭
2023-01-17 20:07:44 <Saijin_Naib> Just stick to whatever units are on the tin, haha 
2023-01-17 20:08:46 <Saijin_Naib> <JanJ[m]> "time to look for another..." <- Can you deploy NFS? Windows does have limited support for NFS shares
2023-01-17 20:09:19 <psykose> i've never heard good things about nfs.. but i have heard of it
2023-01-17 20:09:31 <psykose> samba on the other hand i only hear "it doesnt work" :p
2023-01-17 20:09:41 <JanJ[m]> actually, between linux systems it works wonderously :) in my experience
2023-01-17 20:10:20 <aron> in my experience, samba just works well
2023-01-17 20:12:25 <Saijin_Naib> <JanJ[m]> "time to look for another..." <- Should not be that bad. My Linux NAS with SSDs shares via samba to Windows and Linux clients within my gigabit network and gets up to 124MBps to my Linux box and usually around 100 to our Windows boxes.
2023-01-17 20:12:25 <Saijin_Naib> What version of smb are you using, and that bug in 22H2 should have been fixed a few patch Tuesday ago... Don't have the KB number handy, though 
2023-01-17 20:13:13 <JanJ[m]> yeah.. patch tuesdays.. we're behind on those because mostly they break things these days :X
2023-01-17 20:13:50 <JanJ[m]> samba is 4.16, comes with nAlpine
2023-01-17 20:14:12 <JanJ[m]> i cant explain why it worked for a few months, up untill "last week" 
2023-01-17 20:14:35 <JanJ[m]> i think thats when ms pushed out H2 to these "half managed" systems
2023-01-17 20:17:02 <JanJ[m]> hmz, i have a senology nas, then its 90 instead of 30
2023-01-17 20:17:49 <JanJ[m]> well that suuuucks! i really want to get rid of that thing, stick it all to Alpine :)
2023-01-17 20:19:02 <Saijin_Naib> JanJ[m]: Same. Mine is TerraOS, an ancient fork of OpenWRT that is mostly unmaintained... Want Alpine, but can't figure out the ARM platform stuff 
2023-01-17 20:20:00 <JanJ[m]> well, i have a raspberry pi :) im looking for a descent risc-v board to get my hands on :)
2023-01-17 20:52:22 <invoked> nfs was good when there wasn't really any (better) option ... and mounting stuff via fuse wasn't a thing either. nfsv4 came a bit late. the idea of what "nas" means these days is anybody's guess.
2023-01-17 23:11:54 <endigma> unraid!!!
2023-01-17 23:12:28 <endigma> idk, i've seen linux desktop users choose SMB over nfsv4, idk whats even going on there
2023-01-17 23:12:38 <endigma> iscsi too
2023-01-17 23:12:46 <endigma> plus like, webdav i guess
2023-01-17 23:56:41 <Saijin_Naib> Security at all, I guess
2023-01-17 23:56:48 <Saijin_Naib> Even if smb is imperfect 
2023-01-17 23:59:59 <q66> samba is an incredibly crummy codebase
2023-01-18 00:04:20 <psykose> what you don't like a build system that does a full rebuild on install
2023-01-18 00:04:48 <q66> psykose: as a matter of fact it only does a full relink
2023-01-18 00:04:51 <q66> big difference
2023-01-18 00:05:07 <psykose> with lto i think there isn't much difference there, considering it's like 1k links :p
2023-01-18 00:05:20 <q66> also that's far from the worst thing 
2023-01-18 00:05:26 <psykose> yes yes
2023-01-18 00:05:32 <psykose> we know the openldap and friends chain in there
2023-01-18 00:05:35 <q66> even in the build system alone 
2023-01-18 00:06:37 <q66> it's super easy to have a set of configure options for samba that totally break it to the point of raising completely inexplicable waf error messages 
2023-01-18 00:07:16 <invoked> i've got ptsd from working with iscsi. i'd wrap myself in an army blanket if i had one.
2023-01-18 00:08:31 <psykose> best i can offer is a comfy weighted blanket
2023-01-18 00:10:05 <invoked> i hear those are nice. anyway, probably if you build something that relies on those things (samba, nfs, iscsi, et al) my advice would be: go rethink & rebuild it, so that it does not rely on those things
2023-01-18 00:12:17 <invoked> (blobs in a db -- not better)
2023-01-18 00:15:14 <invoked> don't really know how owncloud/nextcloud work, but i assume it's like dropbox, and even without knowing how those work i'd bet it's probably better than the sort of low level problems you'd encounter with eg, open-iscsi
2023-01-18 00:32:27 <psykose> don't think php cloud software a whole layer of abstraction up is going to be any better
2023-01-18 00:35:33 <invoked> sure. i wouldn't make an argument for php, but otherwise i'd say a web api would be less hassle.
2023-01-18 00:36:04 <invoked> no stale file handles, or initiator timeouts that freeze the whole system, etc etc.
2023-01-18 00:45:03 <invoked> (forgot to hit enter) something like dropbox does more at scale than any of these "nas" things. dropbox initially was an aws thing until they essentially built out an aws equivalent on-prem. all this cloud stuff makes me gag, but the pragmatist will go with what actually works.
2023-01-18 00:45:34 <invoked> but still, i think most people have to learn things the hard way  :)
2023-01-18 08:09:15 <JanJ[m]> heh, nextcloud
2023-01-18 09:00:29 <Fxzxmic> busybox website is too slow!
2023-01-18 09:01:00 <dwfreed> and we're supposed to do something about that how?
2023-01-18 09:02:54 <Fxzxmic> I'm just complaining. We have nothing to do with that.
2023-01-18 11:32:33 <JanJ[m]> Well, we could take them
2023-01-18 11:32:33 <JanJ[m]> Over by force! Then onto the kernel site..
2023-01-18 11:32:41 <JanJ[m]> Alpine for president!
2023-01-18 11:36:09 <aron> lol
2023-01-18 12:10:29 <pj[m]> busybox is the last thing you would want to take over
2023-01-18 12:17:40 <Fxzxmic> Is Alpine the most active distribution using Busybox?
2023-01-18 12:18:08 <dwfreed> depends on how you define active
2023-01-18 12:20:28 <JanJ[m]> Actually, busybox would be the first thing i’d take over. The more control you habe over the software you use, the easier it is to steer in the direction you want
2023-01-18 12:20:49 <dwfreed> you're free to fork busybox and make it your own
2023-01-18 12:21:01 <dwfreed> it is, after all, copyleft
2023-01-18 12:22:16 <elagost> is there a way to mirror alpine updates on a local server but not have to mess too much with /etc/apk/world? I used to use dnf-local on Fedora and I enjoyed only downloading most updates once for the multiple systems in my home that ran Fedora.
2023-01-18 12:22:38 <ikke> elagost: setup-apk cache combined with apk cache?
2023-01-18 12:23:02 <psykose> you can set up some form of http proxy stealing stuff
2023-01-18 12:23:10 <psykose> it's cleartext afterall
2023-01-18 12:23:33 <elagost> ikke: a cache might just be what I was looking for. Thanks. 
2023-01-18 12:25:11 <elagost> basically just have to share that via nfs and mount as /var/cache/apk on the client systems, right?
2023-01-18 12:25:18 <ikke> yes
2023-01-18 12:25:49 <elagost> sweet. That's what I was doing with dnf-local. I'll try that today. Thank you for the super quick responses as always.
2023-01-18 13:16:16 <EF> I take it Alpine doesn't have ecryptfs enableed in the kernel? 
2023-01-18 13:18:39 <psykose> it is
2023-01-18 13:18:56 <psykose> you probably didn't reboot after upgrading the kernel so a modprobe fails
2023-01-18 13:20:14 <EF> so if I build the ecrpytfs tools I should be able to decrypt then?  Have got Synology problems! 
2023-01-18 13:21:48 <psykose> probably
2023-01-18 13:21:57 <psykose> ecryptfs-utils also exist
2023-01-18 13:22:22 <EF> thank you. Oh, must have searched poorly. 
2023-01-18 13:29:45 <pj[m]> <JanJ[m]> "Actually, busybox would be the..." <- the only direction I would steer it over would be a furnace, and then make own "busybox"
2023-01-18 13:32:32 <psykose> pjbox wen
2023-01-18 14:07:40 <pj[m]> it's called lazybox, and I've not had time to work on it recently
2023-01-18 14:22:14 <hodapp> what's wrong with busybox?
2023-01-18 14:22:25 <endigma> it never picks up the phone
2023-01-18 14:22:56 <hodapp> I see what you did there
2023-01-18 14:33:26 <qaqland[m]> <Fxzxmic> "Is Alpine the most active..." <- openwrt is also active
2023-01-18 14:36:35 <Fxzxmic> Oh, yes
2023-01-18 15:01:54 <JanJ[m]> :-)
2023-01-18 15:33:22 <WhyNotHugo_> rootlesskit fails for me because it uses invalid arguments for the `ip` command -> https://paste.sr.ht/~whynothugo/4d7b7ff3e3c9ca5ee47e5c5a2e0a18133ce20b3a
2023-01-18 15:33:34 <WhyNotHugo_> This is basically the same command that is used for rootless docker containers.
2023-01-18 15:33:44 <WhyNotHugo_> Should I report this upstream?
2023-01-18 15:34:21 <WhyNotHugo_> I suspect they target gnu-ip ?
2023-01-18 15:35:38 <psykose> what do they execute verbatim
2023-01-18 15:36:58 <ptrc> WhyNotHugo_: isn't iproute2 from kernel, not gnu?
2023-01-18 15:37:52 <WhyNotHugo_> The command they execute is in the paste. I'm not sure what arguments they pass to ip.
2023-01-18 15:38:12 <WhyNotHugo_> Seems like the `ip` from `iproute2` is required
2023-01-18 15:38:14 <psykose> that's what i mean, it's not in the paste
2023-01-18 15:38:33 <psykose> ah
2023-01-18 15:38:36 <psykose> it's all the way to the right
2023-01-18 15:38:51 <psykose> i get trolled by these fancy pastebins in non-raw mode every time
2023-01-18 15:38:52 <psykose> sigh
2023-01-18 15:39:55 <WhyNotHugo_> What's fancy and non-raw about it?
2023-01-18 15:40:12 <elly> pastebins just provide the actual text challenge [IMPOSSIBLE]
2023-01-18 15:41:34 <elly> WhyNotHugo_: it looks like this in w3m: https://elly.town/t/srht-paste.png
2023-01-18 15:41:37 <ptrc> well, busybox doesn't implement `ip tuntap`
2023-01-18 15:41:54 <ptrc> so the easy fix would be to add `iproute2-minimal` or something to rootlesskit depends
2023-01-18 15:42:11 <psykose> what's fancy is it doesn't look like this https://paste.sr.ht/blob/97c4fba4a12a56067fdc209a616947f8e84acf07
2023-01-18 15:42:13 <psykose> :-)
2023-01-18 15:42:17 <WhyNotHugo_> I'm packaging docker-rootless-extras, so I'm including it as a depends there.
2023-01-18 15:42:48 <WhyNotHugo_> psykose: You find wrapped code easier to read?
2023-01-18 15:43:46 <psykose> no, but this isn't code :p
2023-01-18 15:44:36 <psykose> i added iproute2-minimal to rootlesskit
2023-01-18 15:45:15 <WhyNotHugo_> I'll exclude it from docker-rootless-extras.
2023-01-18 15:45:25 <WhyNotHugo_> (which I'm still testing)
2023-01-18 15:46:04 <WhyNotHugo_> docker-rootless-extras requires inserting a few kernel modules to work. Is shipping a file for /etc/modules-load.d/ in the package the accepted approach?
2023-01-18 15:46:26 <psykose> sure
2023-01-18 15:58:06 <JanJ[m]> sooo in regards to my "slow alpine samba server" i got this briljant idea last night! i simply rip the smb.conf from the much quicker synology nas! :D
2023-01-18 16:03:01 <psykose> did it work
2023-01-18 16:05:18 <JanJ[m]> i dont know.. ive been to the dentist for toothache
2023-01-18 16:09:53 <qaqland[m]> Afaik, install wlroots-dev will help to wayland build
2023-01-18 16:11:50 <qaqland[m]> Some repo include a ci/cd file which is helpful
2023-01-18 16:11:50 <minimal> JanJ[m]: do we need to add "may cause toothaches!" warning to the samba package? ;-)
2023-01-18 16:12:40 <JanJ[m]> well.. "causes hairloss" would be nice.. the toothache, thats windows's fault!
2023-01-18 16:14:00 <psykose> once it starts going best to get rid of all of it, right? ;-)
2023-01-18 16:14:47 <JanJ[m]> i must say, it makes me look prettier :)
2023-01-18 16:15:12 <psykose> baldness is fashion
2023-01-18 16:52:48 <hsv2> a fresh installed networkmanager isn't supposed to auto create connections for the ether devices available? I don't usually use it, but I think to remember to create manually the configuration to use dhcp
2023-01-18 16:53:22 <hsv2> remember to not need to create manually*
2023-01-18 16:55:32 <hsv2> oh… ok, if I try to create it it says me the device is strictly unmanaged… was I wrong to delete it from /etc/network/interfaces (?)
2023-01-18 16:58:45 <JanJ[m]> networkmanager.. well there's a topic that brings me toothache
2023-01-18 17:03:03 <hsv2> solved... wiki doesn't mention to configure networkmanager ifupdown plugin to manage those ifaces
2023-01-18 17:03:30 <JanJ[m]> could you edit the wiki then?
2023-01-18 17:03:34 <JanJ[m]> update it? :)
2023-01-18 17:03:43 <JanJ[m]> i'm struggling with networkmanager myself
2023-01-18 17:04:22 <hsv2> or not.... was to quick XD... now the interface appears in connection list but it keep telling me it is unmanaged
2023-01-18 17:04:57 <hsv2> i'll keep notes on the setup if i manage to get it to work
2023-01-18 17:06:23 <JanJ[m]> thank you, god bless!
2023-01-18 17:09:58 <JanJ[m]> i'm stuck with a "LAN not connected" icon in KDE, even though i have wifi connected.. tried everything in the wiki that had anything close to "networkmanager" on it, including some obscure recepie for almond soufle... nothing
2023-01-18 18:17:40 <Saijin_Naib> <scorpion2185[m]1> "so maybe some alpine devs..." <- What exact CPU do you have? 
2023-01-18 18:23:22 <scorpion2185[m]1> Intel Gemini Lake N4100
2023-01-18 18:26:28 <psykose> whoa
2023-01-18 18:26:36 <psykose> is that a real 1ghz cpu
2023-01-18 18:28:54 <scorpion2185[m]1> idk my first 8 core
2023-01-18 18:29:21 <Ermine> psykose: I have Gemini lake N5000, which should run at 1.1GHz but it runs at 2+ GHz in practice
2023-01-18 18:29:38 <psykose> yeah the turbo is 2+
2023-01-18 18:30:06 <psykose> also it's 4 cores not 8
2023-01-18 18:31:11 <scorpion2185[m]1> right my first 4 cores
2023-01-18 18:31:20 <minimal> I've got an AMD C-70 somewhere here with is 2 core, 1000MHz (1333~MHz turbo) ;-)
2023-01-18 18:31:31 <minimal> 9W TDP :-)
2023-01-18 18:32:09 <psykose> loses to an rpi4 :p
2023-01-18 18:33:27 <minimal> haven't tried them in a race but I suspect the AMD would have better disk speed lol
2023-01-18 18:33:51 <minimal> also have a Sempron 3850 APU which is 4 core, 1.3GHz
2023-01-18 18:35:16 <Ermine> I once set an upper bound of 1.1 Ghz and got lags
2023-01-18 18:35:36 <minimal> and a AMD E2-3800 which is 4 core, 1.3GHz
2023-01-18 18:36:20 <minimal> .....collects low power CPUs :-)
2023-01-18 18:40:09 <Ermine> At 2Ghz the place where cpu is placed is always warm. And it starts boiling when I run Zoom!
2023-01-18 18:40:15 <Ermine> x86 is just too hot
2023-01-18 18:44:11 <minimal> "starts boiling"? perfect timing for making a cup of tea/coffee to sip during boring Zoom conferences lol
2023-01-18 18:44:24 <minimal> pun intended ;-)
2023-01-18 18:46:09 <JanJ[m]> Zoom..
2023-01-18 18:47:41 <Ermine> I dislike it but I don't really have a choice.
2023-01-18 18:54:52 <JanJ[m]> Thats the problem most of us have :)
2023-01-18 19:16:43 <scorpion2185[m]1> * Saijin_Naib Intel Gemini
2023-01-18 19:20:04 <psykose> iirc there was some gemini stuff added to the kernel but i don't have complete knowledge of what the platform needs
2023-01-18 19:20:46 <psykose> or why it doesn't work for you :p
2023-01-18 19:21:14 <psykose> and unless you get lucky enough with someone with the same hardware figuring it out, you'd have to figure it out yourself
2023-01-18 19:21:20 <psykose> (and then it would be enabled fine of course)
2023-01-18 19:21:32 <scorpion2185[m]1> is possible to just add all the modules of debian ?
2023-01-18 19:22:15 <scorpion2185[m]1> why these modules aren't included?
2023-01-18 19:23:06 <scorpion2185[m]1> do you think that >musl is the future and glibc is bound for deprecation   ?
2023-01-18 19:29:37 <psykose> most likely not
2023-01-18 19:29:56 <psykose> uhh idk, probably just no desire to compile every module from the kernel so it's more demand based
2023-01-18 19:32:16 <elagost> scorpion2185[m]1: I have a gemini lake refresh processor (N5030) but with me right now, may be similar enough. What was the problem you were having?
2023-01-18 19:34:16 <hsv2> JanJ[m], about your nm issue, did you check log file? Somewhere in NetworkManager.conf man page they show a exemple how to enable trace logging for issue debugging
2023-01-18 19:37:59 <psykose> elagost: it was about hibernate not working, i.e. cat /sys/power/state  not having disk as an option, etc
2023-01-18 19:38:42 <elagost> People still use hibernate, eh? Interesting.
2023-01-18 19:39:46 <psykose> i don't, but i guess it's useful if you're someone that keeps a ton of stuff open but don't have a modern device that stays 50+ hours in sleep
2023-01-18 19:40:04 <psykose> but in linux the latter is always scuffed anyway lol
2023-01-18 19:40:20 <psykose> i've had good luck but if i had a dollar for every "sleep broke my graphics/wifi/.."
2023-01-18 19:40:46 <psykose> regardless if it's as simple as a missing module it would be nice to fix :)
2023-01-18 19:42:37 <Saijin_Naib> <Ermine> "psykose: I have Gemini lake N500..." <- My N3450 is similar, rated 1.10, runs around 2.1 most times. Gotta break the laptop open and put better thermal compounds and see if it doesn't scooch up a bit more 
2023-01-18 19:43:41 <gray_philip[m]> Got 2-3 minutes to spare? 🏹... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/TWJieRTBDSfCuToMsiTjXQop>)
2023-01-18 19:43:44 <elagost> enabled states might be a firmware thing?
2023-01-18 19:49:17 <JanJ[m]> <hsv2> "Jan J., about your nm issue, did..." <- not yet, should do that indeed
2023-01-18 19:49:22 <psykose> Saijin_Naib: if you've never repasted it, and especially if it's been a while, it's worth the time
2023-01-18 19:49:32 <psykose> a few %/degrees of improvement for sure :)
2023-01-18 19:49:36 <psykose> just don't do it on carpet
2023-01-18 19:49:41 <psykose> i destroyed a laptop like that once
2023-01-18 19:49:45 <psykose> :p
2023-01-18 19:49:54 <psykose> mainboard straight into carpet
2023-01-18 19:51:44 <Saijin_Naib> <elagost> "enabled states might be a..." <- For me, it worked elsewhere due to Alpine missing the modules. I'm trying to dig through the kernel config and see if I can tell what they might need 
2023-01-18 19:53:04 <Saijin_Naib> psykose: Never on this laptop, which is almost three years old and very cheap, so likely terrible compound 
2023-01-18 19:58:46 <psykose> :)
2023-01-18 19:59:04 <psykose> yeah can probably gain 10% headroom there if it was thermal bottlenecked
2023-01-18 19:59:12 <psykose> old paste gets real bad on laptops
2023-01-18 20:03:56 <Saijin_Naib> $110 MSRP... Probably doesn't have any 🤣
2023-01-18 20:05:55 <psykose> :)
2023-01-18 21:42:42 <Ermine> psykose: lld binary on Elbrus is 12Gb (from some WIP llvm port I guess)
2023-01-18 21:43:02 <psykose> uhhh
2023-01-18 21:43:21 <psykose> the hell kind of instructions are going on there
2023-01-18 21:43:35 <psykose> full debug llvm without any stripping linked into it too? :p
2023-01-18 21:45:05 <Ermine> Some guy in chat conjectured that yes, debug info is here and it' -O0
2023-01-18 21:46:56 <Ermine> it's
2023-01-19 01:13:12 <elagost> Just checked my gemini lake laptop, it has 'freeze mem disk' in /sys/power/state. I'm using linux-edge on alpine 3.17.
2023-01-19 01:30:52 <minimal> elagost: I think the point was what does "dmesg | grep 'ACPI: (supports" return?
2023-01-19 01:31:12 <minimal> i.e. which S states
2023-01-19 01:35:13 <elagost> right, I get S0 S3 S4 S5
2023-01-19 01:40:06 <minimal> elagost: so S4 is hibernate
2023-01-19 01:42:21 <elagost> so standard alpine kernels support hibernate on gemini lake cpus I guess.
2023-01-19 01:42:31 <minimal> which maps to "disk" in /sys/power/state. The other person didn't see this
2023-01-19 02:38:01 <qaqland[m]> 
ACTION uploaded an image: (15KiB) < https://matrix.org/_matrix/media/v3/download/mozilla.org/e6df610050dc0a02ea6ae829daa2288245a8578b/image.png >

2023-01-19 02:39:12 <psykose> what's up
2023-01-19 02:39:29 <qaqland[m]> user=root is not in the shadow group
2023-01-19 02:39:40 <psykose> so what
2023-01-19 02:40:25 <qaqland[m]> never see this before 🙂
2023-01-19 02:40:38 <psykose> ah
2023-01-19 02:40:45 <psykose> i guess i haven't either
2023-01-19 02:40:52 <psykose> nobody is even in that group
2023-01-19 02:40:53 <psykose> funny
2023-01-19 02:42:21 <qaqland[m]> It's my first time know the user doesn't need to be in the group 
2023-01-19 02:42:41 <Fxzxmic> root is not in the tty group too.
2023-01-19 02:47:02 <dwfreed> on debian, root is only in the root group
2023-01-19 02:52:24 <elly> it doesn't matter what groups root is in generally, since it ignores access checks
2023-01-19 02:52:52 <psykose> root is op
2023-01-19 02:53:05 <elly> nerf root
2023-01-19 02:59:19 <dwfreed> I mean, technically that's not true
2023-01-19 02:59:28 <dwfreed> one can make uid 0 pretty powerless
2023-01-19 04:29:05 <Saijin_Naib> <scorpion2185[m]1> "Intel Gemini Lake N4100" <- Looks like that module is enabled, so it must be something else
2023-01-19 08:46:01 <EF> Not strictly an Alpine thing but trying to recover an ecryptfs mount, have passcode etc, it mounts but when trying to view the folder I get "ls: /data/Reference/5/MyDir: No such file or directory"
2023-01-19 08:46:45 <EF> data is copied from an img, recovered from a Synology drive via ddrescue.  The original files seem okay.
2023-01-19 08:47:12 <EF> Tried in Ubuntu and now in Apline 3.17.  Any suggestions, hints or pointers welcome
2023-01-19 08:47:57 <psykose> do you `ls` as root so you even have permissions to see the folder
2023-01-19 08:48:21 <EF> Yes, as root, chmod 777 and chown root:root as well
2023-01-19 09:04:56 <psykose> something unrelated then
2023-01-19 09:09:27 <EF> it is bizarre as it can clearly see something, followed Synology's recovery guide, but just cannot access anything.  Most weird.
2023-01-19 10:45:03 <WhyNotHugo_> What's the state of this -> https://wiki.alpinelinux.org/wiki/Alpine_Wall
2023-01-19 10:45:16 <WhyNotHugo_> The last update was in 2017. This is obsoelte, right?
2023-01-19 10:45:30 <JanJ[m]> one certainly hopes not
2023-01-19 10:45:34 <JanJ[m]> it still works
2023-01-19 10:46:12 <ikke> awall is still maintained 
2023-01-19 10:55:58 <WhyNotHugo_> Oh, interesting.
2023-01-19 10:56:06 <JanJ[m]> another tooth lost over nothing :D
2023-01-19 10:56:39 <WhyNotHugo_> The article kinda sounds like "this is a plan of something we'll do", and since it had no further updates, I guessed it never happend. Guessed wrong.
2023-01-19 10:57:12 <JanJ[m]> the links below to awal and zero-to-awal are updated in 2021 :)
2023-01-19 11:27:36 <bananicorn> Hello :)
2023-01-19 11:28:42 <bananicorn> Is anyone here using Alpine with PipeWire and realtime audio? I have some confusion about parts of the configuration that I'd appreciate some help with
2023-01-19 11:29:13 <bananicorn> Also running without logind, do rtkit is not an option
2023-01-19 11:30:08 <alpine-meetbot> bananicorn meant to say: Also running without logind, so rtkit is not an option
2023-01-19 11:30:08 <bananicorn> s/do/so/
2023-01-19 11:31:14 <Ermine> Why? You can use rtkit without elogind
2023-01-19 11:33:20 <Ermine> bananicorn: you can use instruction at https://wiki.alpinelinux.org/wiki/PipeWire to get realtime
2023-01-19 11:33:48 <bananicorn> The instructions did not work for me unfortunately, I get an I/O error from rtkit when starting pipewire
2023-01-19 11:35:10 <Ermine> can you please post the logs?
2023-01-19 11:35:26 <bananicorn> Sure, give me a second to change my config again :)
2023-01-19 11:40:51 <Ermine> Also, which version of Alpine do you have?
2023-01-19 11:42:42 <bananicorn> I am running alpine edge, using swaywm with seatd, dbus is running
2023-01-19 11:43:49 <Ermine> bananicorn: do you have both session bus and system bus running?
2023-01-19 11:44:20 <Ermine> bananicorn: also if you want to go rtkit way, you need to have polkit installed
2023-01-19 11:46:21 <bananicorn> Ermine: Yes, dbus-daemon is running and I started sway with dbus-run-session
2023-01-19 11:46:24 <bananicorn> I may not have polkit
2023-01-19 11:47:07 <bananicorn> I have polkit-libs, I assume I need something more than that?
2023-01-19 11:48:32 <bananicorn> But I would actually prefer not installing additional things unless I need to
2023-01-19 11:48:33 <Ermine> bananicorn: yes, there's polkit daemon
2023-01-19 11:49:03 <bananicorn> If I go without rtkit, I saw that I had to adjust the limits.conf values, which I did
2023-01-19 11:49:19 <Ermine> bananicorn: rtkit consults polkit on whether to make given thread realtime
2023-01-19 11:49:22 <bananicorn> But it seems I cannot actually set negative nice values in the configuration
2023-01-19 11:49:34 <bananicorn> configuration of pipewire/wireplumber that is
2023-01-19 11:50:09 <bananicorn> https://paste.debian.net/hidden/7759e565/
2023-01-19 11:50:31 <bananicorn> This is the output of the /usr/libexec/pipewire-launcher
2023-01-19 11:51:25 <psykose> whadya mean you "cannot set negative nice values"
2023-01-19 11:51:41 <bananicorn> It always gives me "could not set nice-level to -11: I/O error"
2023-01-19 11:52:02 <psykose> that's not "you" setting it
2023-01-19 11:52:10 <psykose> what does `ulimit -r` output
2023-01-19 11:52:30 <bananicorn> ulimit -r returns 0
2023-01-19 11:52:42 <psykose> where did you put that limits.conf
2023-01-19 11:53:18 <bananicorn> I added it to /etc/security/limits.conf to be sure
2023-01-19 11:53:48 <psykose> do you have util-linux-login/shadow-login installed and did you relog after editing it
2023-01-19 11:54:31 <bananicorn> It appears I do not have those, do I understand correctly that util-linux-login is preferred?
2023-01-19 11:55:29 <psykose> sure
2023-01-19 11:55:50 <Ermine> bananicorn: limits.conf is PAM thing, and default bb login does not use PAM
2023-01-19 11:56:08 <bananicorn> Alright, I'll install util-linux-login, relog and try again :)
2023-01-19 12:03:18 <bananicorn> Ermine, psykose: Problem solved! Thank you both for your time!
2023-01-19 12:03:41 <bananicorn> I will add a note to the wiki about this :)
2023-01-19 12:05:35 <psykose> i already edited it
2023-01-19 12:06:18 <bananicorn> You're just too good 😄
2023-01-19 12:13:06 <psykose> definitely noit
2023-01-19 12:22:48 <Ermine> There's also a way with portals
2023-01-19 12:25:50 <JanJ[m]> 🙈
2023-01-19 12:58:32 <JanJ[m]> i need to get awall working for a router with Wireguard support! tried that a year ago but didnt succeed
2023-01-19 12:58:52 <ikke> JanJ[m]: where did you get stuck on?]
2023-01-19 12:59:14 <JanJ[m]> i cuoldn't get wireguard to work
2023-01-19 12:59:23 <JanJ[m]> but this time, i'll make notes and document :)\
2023-01-19 12:59:31 <JanJ[m]> back then i wasnt in these channels :)
2023-01-19 12:59:36 <ikke> JanJ[m]: I can also recommend looking at awall-policies
2023-01-19 12:59:49 <JanJ[m]> yeah :)
2023-01-19 12:59:54 <ikke> it's a set of policies that can help you give more structure to your policies
2023-01-19 13:00:12 <ikke> It's not perfect, but it can help
2023-01-19 13:01:14 <JanJ[m]> i'm learning Rust, getting better at it all the time, eventually i want to write a web frontend for awall. json support in rust is ublime!
2023-01-19 13:01:22 <JanJ[m]> so thats a match made in heaven :)
2023-01-19 13:01:46 <ikke> https://tpaste.us/YYJM
2023-01-19 13:01:55 <WhyNotHugo_> Can I drop-in a file in /etc/nftables.d/ so as to allow incomming trffic on port 22?
2023-01-19 13:02:07 <WhyNotHugo_> I'd prfer to avoid editing the default file and putting a drop-in file, if possible.
2023-01-19 13:02:10 <ikke> JanJ[m]: That's the policy I use for a dmvpn hub
2023-01-19 13:02:12 <ikke> err
2023-01-19 13:02:15 <ikke> wg hub
2023-01-19 13:02:17 <JanJ[m]> thanks ikke :) 
2023-01-19 13:02:44 <ikke> adp-wan is defined by awall-policies
2023-01-19 13:03:30 <bananicorn> WhyNotHugo_: Yes, check the bottom of the /etc/nftables.nft file, it loads any .nft files from that directory :)
2023-01-19 13:03:37 <bananicorn> Then just have to reload nftables
2023-01-19 13:03:47 <JanJ[m]> yeah, those adp policies are lacking documentation, so i wasnt sure how to use them :) but by now im good enough with Alpine, i think i can make it work :D
2023-01-19 13:05:30 <WhyNotHugo_> bananicorn: I noticed that. I guess I actually wanted to ask what I should put in this file in order to prepend to an existing chain.
2023-01-19 13:05:55 <WhyNotHugo_> Oh, priority -1 is allowed?
2023-01-19 13:06:53 <bananicorn> Yes, priority can be numeric or those pre-defined strings like "filter"
2023-01-19 13:06:56 <bananicorn> or even "filter - 1"
2023-01-19 13:07:35 <bananicorn> You can either use the command syntax of the cli tool or the syntax that you see with nft list ruleset
2023-01-19 13:07:36 <aron> JanJ[m]: your slow samba issue is solved?
2023-01-19 13:07:45 <JanJ[m]> nope
2023-01-19 13:08:08 <aron> I thought copying that other config file help :)
2023-01-19 13:08:30 <JanJ[m]> and i thought 30mb/s (or mbtis/s im still not sure) was slow.. this morening i got send a pic from a user who got 455kb/s :D
2023-01-19 13:08:38 <JanJ[m]> nah, havent done that yet
2023-01-19 13:08:42 <aron> you could try top and iotop to determine where's the bottleneck
2023-01-19 13:09:16 <aron> on alpine side, ofc
2023-01-19 13:09:30 <JanJ[m]> its an 8th gen HP Micrososerver, im sure that doesnt exactly help either :)
2023-01-19 13:11:01 <ikke> Performance issues are the worst to find
2023-01-19 13:11:30 <psykose> you might've forgotten a -funroll-loops somewhere
2023-01-19 13:11:44 <ikke> yum, funroll-loops
2023-01-19 13:12:10 <q66> funroll loops sounds like some kind of bread or cereal
2023-01-19 13:12:20 <aron> :)
2023-01-19 13:13:28 <psykose> bake me the funroll pastries q66
2023-01-19 13:14:08 <JanJ[m]> actually.. just thinking.. 
2023-01-19 13:14:51 <q66> psykose: what would the recipe be though
2023-01-19 13:15:29 <psykose> you tell me chef
2023-01-19 13:15:43 <q66> i never baked funrolls
2023-01-19 13:16:15 <WhyNotHugo_> I tried this drop-in file with priority -1 and 1, but connections to sshd are always dropped -> https://paste.sr.ht/blob/01020f841a4336d532cf595655b3b4f5e2296b51
2023-01-19 13:16:29 <bananicorn> We'll definitely need flour, water, yeast, a pinch of salt
2023-01-19 13:16:51 <bananicorn> The rest depends on the amount of fun
2023-01-19 13:17:01 <psykose> it's chain input not chain input_sshd pretty suree
2023-01-19 13:17:07 <q66> yeah those things will get you zero level of fun
2023-01-19 13:17:33 <psykose> i bet throwing flour at q66 is a lot of fun
2023-01-19 13:17:42 <q66> sounds like a mess 
2023-01-19 13:17:43 <psykose> source'd
2023-01-19 13:19:05 <bananicorn> WhyNotHugo_ I think what happens in your cases is that with priority -1, writing "accept" just makes it move on to the next chain, which drops packets by default
2023-01-19 13:19:25 <bananicorn> And with priority 1 it never gets to input_sshd because input drops it before
2023-01-19 13:19:38 <bananicorn> So you have to insert the rule in the input chain
2023-01-19 13:19:43 <WhyNotHugo_> psykose: If I name the chain `input`, it nftables complains that the chain already exists. Not sure how I'm suppost to prepend to it.
2023-01-19 13:19:47 <psykose> ah
2023-01-19 13:20:01 <WhyNotHugo_> bananicorn: Can I somehow prepend the rule into the chain via a drop-in file?
2023-01-19 13:20:17 <bananicorn> Hmm, strange, let me examine my setup
2023-01-19 13:20:28 <psykose> the pmaports rules only  do
2023-01-19 13:20:28 <psykose> https://img.ayaya.dev/BCeReKNIOVeP
2023-01-19 13:20:30 <psykose> and nothing else
2023-01-19 13:20:34 <psykose> and i know that works
2023-01-19 13:20:41 <psykose> no prio or anything
2023-01-19 13:21:00 <bananicorn> Yes, you have to leave out the hook/priority if you are just adding to an existing chain
2023-01-19 13:21:04 <bananicorn> That's the problem here
2023-01-19 13:21:22 <bananicorn> Otherwise my file is practically identical with "chain input" and the ssh rule
2023-01-19 13:21:24 <WhyNotHugo_> Ah, yup, that's exactly it!
2023-01-19 13:21:48 <bananicorn> Glad it works :)
2023-01-19 13:21:54 <WhyNotHugo_> yeah, mine now looks like the pmos one -- except that I don't need to duplicaet the comment :P
2023-01-19 13:22:51 <bananicorn> Feel free to message me directly if you have any more trouble, I spent way too much time on nftables recently^^
2023-01-19 13:23:14 <WhyNotHugo_> bananicorn: Thanks, will keep it in mind. Tho all I wanted to do was enable SSH really.
2023-01-19 13:23:30 <bananicorn> Even better :)
2023-01-19 13:27:56 <JanJ[m]> i think i found the problem >:)
2023-01-19 13:30:25 <JanJ[m]> seems urbackup is taking up looooots of resources when 13 workstations starting their backups at the same time
2023-01-19 13:32:50 <psykose> would be weird you concluded it was a recent issue related to specific new version clients then, unless it just happens to be so well timed for them haha
2023-01-19 13:38:12 <JanJ[m]> i related it to the upgrades to Windows 11H2 because there is an actual problem with H2 and SMB copy speeds
2023-01-19 13:38:21 <JanJ[m]> so it seems rather concidal
2023-01-19 13:39:25 <JanJ[m]> in my heart i always knew it wasnt my beloved Alpine <3
2023-01-19 13:42:31 <psykose> sometimes it is :p
2023-01-19 13:53:36 <wej> hello! i have an issue with ffmpeg on alpine. whenever i try to transcode a video with it while using the gpu with vaapi the transcode is extremely slow if i use a scaling filter. any ideas what's going on?
2023-01-19 13:54:23 <wej> when i try the exact same thing with the same hardware on voidlinux, the same transcode runs about 50x faster
2023-01-19 13:55:13 <psykose> do you perchance have a sample commandline and file
2023-01-19 13:56:17 <wej> one moment, i can give you a sample command line
2023-01-19 13:57:31 <wej> ffmpeg -analyzeduration 200M -ss 00:00:45.000 -init_hw_device vaapi=va:/dev/dri/renderD128 -filter_hw_device va -hwaccel vaapi -autorotate 0 -i file:"/path/to/video.mp4" -autoscale 0 -map_metadata -1 -map_chapters -1 -threads 0 -map 0:0 -map 0:1 -map -0:s -codec:v:0 h264_vaapi -rc_mode VBR -b:v 292000 -maxrate 292000 -bufsize 584000 -profile:v:0 high -force_key_frames:0
2023-01-19 13:57:32 <wej> "expr:gte(t,45+n_forced*3)" -vf "scale=320:240,hwupload=derive_device=vaapi" -sws_flags fast_bilinear -sws_dither none -pix_fmt vaapi -codec:a:0 aac -ac 2 -ab 128000 -ar 48000 -af "volume=2" -copyts -avoid_negative_ts disabled -max_muxing_queue_size 2048 -f hls -max_delay 5000000 -hls_time 3 -hls_segment_type mpegts -start_number 15 -hls_segment_filename "/tmp/foo%d.ts"
2023-01-19 13:57:34 <wej> -hls_playlist_type vod -hls_list_size 0 -y "/tmp/foo.m3u8"
2023-01-19 13:57:43 <wej> oops, didn't realize it was this long
2023-01-19 13:58:28 <wej> pretty much any video file will do as far as i can tell, i was using some h264 encoded source file
2023-01-19 14:00:59 <psykose> just running it i get.. 13.5x
2023-01-19 14:01:20 <wej> that's ok, i guess, i was getting 0.3x
2023-01-19 14:02:04 <wej> i was running it on an intel gpu (gen9.5)
2023-01-19 14:02:13 <psykose> now i guess you get to compare libva/mesa/intel-media-driver versions
2023-01-19 14:02:37 <psykose> for reference is it glibc or musl void
2023-01-19 14:03:07 <wej> glibc void
2023-01-19 14:03:22 <wej> libva is 2.16.0 on alpine, 2.17.0 on void
2023-01-19 14:03:59 <wej> i was using libva-intel-driver on both, with both running version 2.4.1
2023-01-19 14:04:39 <psykose> not media? the intel-driver one iirc is the old one
2023-01-19 14:04:52 <wej> yeah, on alpine i tries both, no difference
2023-01-19 14:05:00 <wej> on void i just tried the old one
2023-01-19 14:05:54 <psykose> void also has 4.4 ffmpeg but i'm not sure if that would be relevant
2023-01-19 14:06:07 <psykose> and i guess on alpine you can upgrade to edge for the same 2.17 libva
2023-01-19 14:06:14 <psykose> but i dunno, all i have is guesses
2023-01-19 14:07:07 <psykose> you'd probably have to compare some performance traces which takes a long time and is a lot of work to figure out why one is that much slower
2023-01-19 14:07:29 <wej> thanks for your input, i'll give edge a try, just to be sure
2023-01-19 14:08:00 <psykose> (you can also apk add ffmpeg4 to get the 4.4 one)
2023-01-19 14:08:53 <wej> ah yes, i'll try that
2023-01-19 14:09:18 <psykose> ah, or maybe i had removed the cli
2023-01-19 14:09:35 <psykose> yeah it's libraries only
2023-01-19 14:09:36 <psykose> nevermind
2023-01-19 14:09:43 <wej> oh, yes indeed, no cli tool installed with ffmpeg4
2023-01-19 14:19:42 <handlerug> can apk deal with repos that have different packages signed with different keys?
2023-01-19 14:20:12 <ikke> sure
2023-01-19 14:20:20 <ikke> The name of the key is in the index
2023-01-19 14:20:21 <handlerug> okay, thanks
2023-01-19 14:21:40 <ikke> That means the keyname on /etc/apk/keys should match
2023-01-19 14:21:47 <ikke> s/on/in/
2023-01-19 14:21:47 <alpine-meetbot> ikke meant to say: That means the keyname in /etc/apk/keys should match
2023-01-19 14:24:56 <aron> is there an alpine-uboot-EDGE-aarch64.tar.gz somewhere? so, not the latest numbered release, but one built from the edge?
2023-01-19 15:24:39 <JanJ[m]> Hmmz.. KDE shell suddenly crashed 🙈
2023-01-19 15:28:52 <scorpion2185[m]1> <Saijin_Naib> "Looks like that module is..." <- i edited grub do you use it?
2023-01-19 15:35:28 <Saijin_Naib> <scorpion2185[m]1> "i edited grub do you use it?" <- I do, but I don't need to force anything. Did you diff the modules and kernel config between Alpine and the booted distro that works for your computer? 
2023-01-19 15:36:17 <scorpion2185[m]1> no, can you share waht you added to grub? i have lvm on luks
2023-01-19 15:45:54 <Saijin_Naib> I don't have anything special in my grub config as I stated before 
2023-01-19 15:54:56 <scorpion2185[m]1> I have swap partition of 9gb , ram is 8Gb
2023-01-19 15:55:17 <scorpion2185[m]1> pm-is-supported --hibernate   (not output)
2023-01-19 15:57:20 <scorpion2185[m]1> echo disk > /sys/power/state
2023-01-19 15:57:20 <scorpion2185[m]1> ash: write error: No such device
2023-01-19 16:06:12 <minimal> deja vu
2023-01-19 16:21:17 <wej> psykose: i did a little more testing with my ffmpeg problem. i just tried with ffmpeg 4.4 from alpine 3.15 - no difference, i also tried with alpine edge, also the same. i guess it is not the versions that make the difference then
2023-01-19 17:17:23 <Forza> Hi
2023-01-19 17:17:33 <Forza> Is there a debug kernel for Alpine? 
2023-01-19 17:17:56 <ikke> We do not package one 
2023-01-19 17:18:33 <Forza> Ah ok. I'll compile my own then. 
2023-01-19 17:18:56 <Forza> I have a serious problem with spontaneous reboots on my server. 
2023-01-19 17:19:47 <Forza> I enabled remote syslog and managed to capture a kernel: BUG: NULL pointer dereference, but 
2023-01-19 17:20:07 <Forza> But not what subsystem it happened in or if it was a specific driver 
2023-01-19 17:22:03 <Forza> Exact same issue with 5.15 lts as with 6.1.4 Edge kernel too
2023-01-19 17:32:25 <Forza> Could it be a hardware issue? 
2023-01-19 17:33:26 <ikke> Hard to tell without details
2023-01-19 17:33:40 <ikke> We do have one server that also randomly reboots, but I think it's more frequent that what you describe
2023-01-19 17:33:46 <ikke> and it's aarch64
2023-01-19 18:07:01 <Forza> Sometimes once per day and sometimes every few weeks 
2023-01-19 18:09:55 <Forza> I think I somehow correlates to iscsi. I use this alpine machine as an  iscsi target for some windows servers and some reboots seems to have happened while windows was rebooting or so. But hard to tell 
2023-01-19 18:11:29 <Forza> Can I use Dracut on Alpine? 
2023-01-19 18:15:28 <Forza> I. E. Not use apk to install custom kernels? 
2023-01-19 18:17:06 <JanJ[m]> So ikke Awall is still maintained, but it uses iptables.. but iptables is beeing replaced with nftables no? What does that mean for Awall?
2023-01-19 18:17:19 <donoban> there is a dracut pacakge on testing
2023-01-19 18:18:20 <donoban> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/testing/dracut/README.alpine
2023-01-19 18:59:02 <ikke> JanJ[m]: 21.1.3
2023-01-19 18:59:25 <ikke> err
2023-01-19 18:59:28 <ikke> https://gitlab.alpinelinux.org/alpine/awall/-/issues/9653
2023-01-19 18:59:29 <ikke> that
2023-01-19 19:05:57 <minimal> Forza: "Not use apk to install custom kernels?" - note sure what you mean, mkinitfs, dracut, and booster are tools to create initramfs files
2023-01-19 19:17:20 <invoked> dracut is useful for using stuff like (eg) ignition where you can manipulate disks on boot, but if you're not doing stuff like that, i'm not sure i see the upside
2023-01-19 19:18:40 <invoked> other than you're just used to dracut from rhel (or some other) and want to use that for comfort zone reasons
2023-01-19 19:34:48 <JanJ[m]> clicks link
2023-01-19 19:37:23 <JanJ[m]> well, i gave my 2 cents :)
2023-01-19 19:58:46 <Forza> minimal: I have built a custom alpine kernel on the past, which gave me a apk to install. 
2023-01-19 19:59:23 <minimal> Forza: ok, but you were asking about dracut which is a tool to generate a initramfs file, not a kernel
2023-01-19 20:00:24 <Forza> invoked: yes, I use Dracut on my Gentoo to built initramfs. I used to always build a static kernel, but there was something that needed modules. Can't actually remember really. But I stick with it since
2023-01-19 20:01:14 <JanJ[m]> dracut.. i seem to remember thats also somehow involv3ed into making boot screens
2023-01-19 20:01:18 <Forza> Anyway. I suppose I could simply build a static kernel with no initramfs file. Should work, right? 
2023-01-19 20:01:28 <minimal> on Alpine whenever you install a kernel package then depending on which of mkinitfs/dracut/booster is installed they have a "trigger" that activated to regenerate the initramfs
2023-01-19 20:02:07 <Forza> Dracut is pretty nice. Can add a initramfs environment, logging and boot stuff before init happens
2023-01-19 20:02:11 <minimal> Forza: it is still not clear what you are actually asking.
2023-01-19 20:02:33 <Forza> I want to debug my reboots 
2023-01-19 20:02:51 <minimal> Forza: dracut is ONE of the options in Alpine to do that, mkinitfs is the default option, booster is another alternative
2023-01-19 20:03:11 <Forza> I only seem to get one line I dmesg and that is bug NULL pointer dereference 
2023-01-19 20:03:36 <Forza> :) 
2023-01-19 20:04:20 <Forza> Initramfs is good if I would use the same kernel on multiple systems or change hardware much
2023-01-19 20:04:24 <minimal> I still do not get the connection - initfam is loaded/run by kernel at boot, if you are getting reboots then that is unrelated to initramfs
2023-01-19 20:05:37 <Forza> True
2023-01-19 20:06:20 <Forza> But if I make a custom kernel I need a matching initramfs, or do all need mods built in
2023-01-19 20:06:28 <minimal> "Initramfs is good if I would use the same kernel on multiple systems" - I still unclear what you're trying to say
2023-01-19 20:07:02 <minimal> when you install a kernel (whether custom or not) then the initramfs in (re)generated for that kernel automatically
2023-01-19 20:09:01 <Forza> When if i download the kernel sources and do make install/make modules_install? 
2023-01-19 20:09:13 <Forza> Even if... 
2023-01-19 20:09:38 <minimal> no,, when you "apk add linux-whatever" or "apk -a upgrade" and a new kernel package is installed
2023-01-19 20:10:43 <minimal> when you said custom kernel I assumed you meant creating a APK package for a custom kernel
2023-01-19 20:16:55 <Forza> I would maybe if there was a debug kernel. I think ill just download from kernel.org and build it
2023-01-20 04:48:57 <Asmadeus> hmm, just noticed something "fun": lib sizes on aarch64 exploded with 3.17 ? The most eloquent example I saw can be observed with the iptables package: iptables-1.8.8-r1 (v3.16) is 1.7MB while iptables-1.8.8-r2 (3.17) is 7.5MB. each of the many .so in /usr/lib/xtables/ just went from 8-12KB to 68KB, while 'size' says they contain the same (or very close) amount of code; the culprit seems to be that 
2023-01-20 04:49:03 <Asmadeus> there is now a big hole before .init_array in the sections? (objdump -h)
2023-01-20 04:49:19 <Asmadeus> Does anyone know what caused that, and why the change?
2023-01-20 04:56:12 <dwfreed> probably a change in gcc or binutils
2023-01-20 05:12:41 <Asmadeus> well, yes, downgrading binutils makes 'echo 'int foo() { return 0; }' | gcc -xc - -shared -o foo' go from 68 to 8K; I guess I can have a look at the changelog...
2023-01-20 05:15:47 <Asmadeus> might be related to https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=9833b7757d246f22db4eb24b8e5db7eb5e05b6d9
2023-01-20 05:16:34 <Asmadeus> since aarch64 can have 64K page sizes, .data is aligned to that and binaries grow much bigger
2023-01-20 05:17:38 <Asmadeus> Since I'm using 4K pages this is slightly overdoing it...
2023-01-20 07:39:38 <psykose> Asmadeus: https://gitlab.alpinelinux.org/alpine/aports/-/issues/14126 i had found the same symptoms
2023-01-20 07:39:45 <psykose> i never found the cause in time for the full rebuild though :/
2023-01-20 07:39:57 <psykose> and i think 2.40 binutils now fixed it
2023-01-20 07:41:15 <psykose> hmm, no
2023-01-20 07:41:17 <psykose> it made it worse
2023-01-20 07:43:15 <psykose> and of course rebuilding with lld makes it go from 7.5MB->1.5MB
2023-01-20 07:48:53 <psykose> i guess the minimum size being the max page for the platform is indeed overdoing it as you found
2023-01-20 07:50:02 <Asmadeus> oh! thanks!
2023-01-20 07:50:31 <Asmadeus> Yeah, and for security the mappings don't have to correspond to the pages on disk anyway, the linker could remap to page boundaries on loading
2023-01-20 07:59:29 <psykose> not really sure if it's fixable for 3.17 anymore, but if you find a resolution for binutils i guess at least we can fix the next release
2023-01-20 07:59:36 <psykose> and edge where things rebuild here and there
2023-01-20 07:59:44 <psykose> since even .40 still has this
2023-01-20 07:59:55 <Asmadeus> for 3.17 we could at least rebuild the worst offenders (e.g. iptables, modemmanager I found grew quite a bit)
2023-01-20 08:00:01 <Asmadeus> But yeah, need fixing upstream first
2023-01-20 08:00:18 <psykose> it's really just every binary lol, though yeah anything with tiny executables grew by multiples
2023-01-20 08:00:23 <Asmadeus> I'll have a closer look on monday
2023-01-20 08:00:28 <psykose> sure sure
2023-01-20 08:00:56 <Asmadeus> Yeah, iptables just has a gazillion tiny .so so it's relatively more obvious
2023-01-20 08:01:43 <psykose> triple annoying as most aarch64 platforms are shitty sdcard/emmc platforms with no storage
2023-01-20 08:26:11 <Asmadeus> (yeah, well, guess why I noticed... :D)
2023-01-20 08:29:51 <aron> if kernel hangs after "Starting kernel ..." on an orange pi zero2 (H616, aarch64), is there any option to get some details why this happens?
2023-01-20 08:31:18 <Asmadeus> aron: assuming the console normally works from linux e.g. console=ttyXYZ is set properly and quiet isn't set?
2023-01-20 08:31:47 <aron> I removed quiet, and explicitly set console=ttyS0,115200n8
2023-01-20 08:32:59 <aron> what I'm thinking now: can I change from linux-lts to linux-edge?
2023-01-20 08:33:06 <aron> I mean, on the SD card
2023-01-20 08:33:34 <aron> 3.17's linux-edge is full of enabled sun50i configs, which might be needed
2023-01-20 08:34:19 <aron> like CONFIG_SUN50I_H616_CCU=y -- seems to be critical
2023-01-20 08:43:07 <psykose> should be able to
2023-01-20 08:43:44 <aron> is there a howto for this somewhere?
2023-01-20 08:44:47 <aron> initramfs-lts and modloop-lts seems a bit problematic
2023-01-20 08:46:58 <psykose> you just install the edge kernel and set it as the one to boot in whatever bootloader
2023-01-20 08:47:11 <psykose> ah modloop
2023-01-20 08:47:16 <psykose> probably more steps i guess
2023-01-20 08:48:05 <aron> I can change the content of the SD card on a different arch, since aarch64 does not boot at all (yet)
2023-01-20 08:48:43 <aron> armv7 to be specific, but it runs alpine
2023-01-20 08:50:50 <Asmadeus> with qemu-user you could chroot into the sd card even on a different arch, but for a quick test to see if you get further you could just grab the apk, untar it and overwrite the Image file as a first test
2023-01-20 08:51:02 <Asmadeus> that should get you some logs even if it's missing modules, if that was the problem
2023-01-20 08:51:56 <aron> there is noting to much on the sd card really, just what I got in alpine-3.17-u-boot.tar.gz
2023-01-20 08:52:29 <aron> it will work with qemu-user?
2023-01-20 08:53:07 <aron> or is there a way to generate the content of that tarball with a different kernel?
2023-01-20 08:55:58 <aron> I'm at the very beggining with this sbc, I got a working u-boot, but I was unable to build a working version with aarch64-none-elf-gcc
2023-01-20 08:56:56 <aron> (my build is crashing when u-boot spl starts BL31)
2023-01-20 08:58:56 <ikke> aron: There is also update-kernel, but not sure if you can use it for edge
2023-01-20 09:01:24 <aron> ikke: thanks, after looking into that script, it seems you need a running kernel for that (which I don't have yet)
2023-01-20 09:36:19 <aron> yay! I managed to fix it
2023-01-20 09:36:59 <aron> I just needed to unpack the initramfs and add the modules from linux-edge.apk, then repack it
2023-01-20 09:37:13 <aron> + add the vmlinuz to boot/
2023-01-20 09:38:59 <aron> no networking tho, but at least I can start to hack it
2023-01-20 10:34:15 <donoban> I'm trying to determine why 'pipewire' depends on 'avahi-libs' but only found '/usr/lib/pipewire-0.3/libpipewire-module-zeroconf-discover.so' which is on 'pipewire-zeroconf'
2023-01-20 10:34:31 <donoban> is an expected behaviour or maybe there is a bug on abuild?
2023-01-20 10:35:58 <psykose> libpipewire-module-raop-discover.so depends on avahi-libs
2023-01-20 10:36:12 <psykose> and libpipewire-module-protocol-pulse.so
2023-01-20 10:37:56 <donoban> uhM, let me check I tried to ldd all pipewire so files
2023-01-20 10:38:43 <donoban> where is libpipewire-module-raop-discover.so ? I don't have it on /usr/lib/pipewire-0.3
2023-01-20 10:39:03 <psykose> /usr/lib/pipewire-0.3/libpipewire-module-raop-discover.so
2023-01-20 10:39:59 <donoban> [donoban@localhost][/usr/lib/pipewire-0.3]% ls | grep raop
2023-01-20 10:40:01 <donoban> libpipewire-module-raop-sink.so*
2023-01-20 10:40:21 <donoban> it seems that I miss it
2023-01-20 10:41:48 <donoban> lol, I just rebuilted it without avahi-dev
2023-01-20 10:41:51 <psykose> i'm sure you'll figure it out
2023-01-20 10:42:34 <donoban> sorry
2023-01-20 10:42:42 <donoban> it's just my local version
2023-01-20 10:43:54 <donoban> "Automatically creates RAOP (Airplay) sink devices based on zeroconf information."
2023-01-20 10:44:02 <donoban> shouldn't it be moved to zeroconf subpackage?
2023-01-20 10:46:17 <donoban> my libpipewire-module-protocol-pulse.so is not linked against avahi but I suppose it's due I disabled it on compile
2023-01-20 10:48:29 <psykose> i don't think there is any purpose in moving around like 80kb of files
2023-01-20 10:49:07 <donoban> well, I would like to remove the dependency between pipewire and avahi
2023-01-20 10:49:24 <psykose> no, since the pulse module links the samee
2023-01-20 10:49:30 <psykose> and avahi-libs is another 120kb of files
2023-01-20 10:50:51 <donoban> yeah if there is a dependency with pulse module... end thread :D
2023-01-20 10:51:03 <psykose> ?
2023-01-20 10:51:26 <aron> how can I install a package without its dependencies?
2023-01-20 10:51:31 <psykose> you can't
2023-01-20 10:52:20 <donoban> I mean that I can't do nothing psykose 
2023-01-20 10:52:33 <psykose> then build your own pipewire with everything disabled
2023-01-20 10:52:57 <psykose> this is a distro not your gentoo installation
2023-01-20 10:53:03 <donoban> yes that's an option
2023-01-20 10:54:23 <donoban> well, my idea was not so weird, if the so that causes the dependency could be moved to the relevant subpackage...
2023-01-20 10:54:47 <psykose> it's not the relevant subpackage
2023-01-20 10:54:57 <donoban> just my local state lead me to don't understand waht was happening
2023-01-20 10:55:52 <donoban> yeah, it's all fine now
2023-01-20 10:59:20 <aron> okay, I managed to avoid to download 100 unnecessary package (linux-firmware-*) by installing only 1
2023-01-20 10:59:37 <aron> how can I create a modloop on a running system?
2023-01-20 11:01:46 <psykose> linux-firmware-none satisfies none in that case
2023-01-20 11:02:09 <aron> yeah, I know, but I wanted that 1
2023-01-20 11:07:15 <aron> hmm, it's a squashfs
2023-01-20 11:07:46 <aron> with kernel modules + firmware
2023-01-20 11:07:49 <aron> interesting
2023-01-20 11:12:20 <donoban> archlinux has raop modules on zeroconf subpackage, https://archlinux.org/packages/extra/x86_64/pipewire-zeroconf/files/
2023-01-20 11:15:09 <psykose> that's awesome, they sure saved themselves 100kb of disk space and no dependencies
2023-01-20 11:16:38 <donoban> well, currently alpine saved 56kb with pipewire-zeroconf :D
2023-01-20 11:16:39 <psykose> did the same, but it doesn't remove any avahi libs
2023-01-20 11:17:01 <donoban> yeah I supposed it
2023-01-20 11:18:05 <donoban> I was just wondering if that RAOP modules need zeroconf for work
2023-01-20 11:19:44 <grawity> arch's pipewire-zeroconf depends on avahi, pipewire itself doesn't
2023-01-20 11:20:16 <donoban> grawity: and pipewire-pulse?
2023-01-20 11:21:47 <grawity> separate to avoid conflicts with actual pulseaudio
2023-01-20 11:21:57 <grawity> so that you could have pulseaudio for audio but pipewire for video
2023-01-20 11:22:42 <donoban> in alpine both pipewire and pipewire-pulse depend on avahi
2023-01-20 11:24:08 <psykose> not with those modules moved
2023-01-20 11:24:28 <psykose> but what i mean is that practically nobody uses i.e. pipewire for video with some other sound server
2023-01-20 11:24:52 <psykose> they either have pipewire-pulse and a whole setup like that, or they're not using pipewire at all and use xorg instead
2023-01-20 11:25:06 <psykose> i've never seen someone use pipewire for video on wayland while using some other sound server except years ago
2023-01-20 11:25:42 <psykose> so unless you were doing exactly that, you'll still have pipewire-pulse and the same avahi-libs
2023-01-20 11:26:01 <psykose> rip your 120k diskspace
2023-01-20 11:26:08 <psykose> :D
2023-01-20 11:26:18 <donoban> I don't mind the diskspace
2023-01-20 11:26:45 <donoban> it's just would prefer to don't have avahi :D
2023-01-20 11:27:40 <psykose> the libs themselves are useless
2023-01-20 11:28:23 <psykose> (i.e. no daemons or anything..)
2023-01-20 11:30:28 <donoban> welll, I will just lose 5mins looking what pulse does with it when enabled
2023-01-20 11:30:37 <WhyNotHugo_> Any ideas what's wrong with this command? `date +'YYYY-MM-DD hh:mm'`
2023-01-20 11:30:43 <WhyNotHugo_> It seems to match date --help.
2023-01-20 11:31:01 <ikke> You are missing % signs
2023-01-20 11:31:31 <WhyNotHugo_> I removed them since --help doesn't show them.
2023-01-20 11:31:49 <ikke> What date util are you using
2023-01-20 11:31:59 <WhyNotHugo_> busybox
2023-01-20 11:32:40 <donoban> lol, it seems that pipewire-protocol-pulse module has its own (sub)modules
2023-01-20 11:32:42 <WhyNotHugo_> Huh, yeah, it works with the %
2023-01-20 11:32:42 <aron> try this:
2023-01-20 11:32:43 <aron> date +'%Y-%m-%d %H:%M'
2023-01-20 11:32:59 <aron> works with both date
2023-01-20 11:33:02 <donoban> [donoban@localhost][/tmp/pipewire-0.3.64/src/modules/module-protocol-pulse/modules]% 
2023-01-20 11:33:03 <psykose> the YYYY ones are recognised TIME formats not FMT formats
2023-01-20 11:33:25 <donoban> inside it there is a module-zeroconf
2023-01-20 11:34:01 <psykose> i.e. you can do date +"%y-%m" -d 1990-01-01
2023-01-20 11:34:46 <WhyNotHugo_> psykose: Ah, thanks
2023-01-20 11:34:52 <psykose> they're formats to pass for the -d/-s dates to display/set for parsing
2023-01-20 11:35:29 <WhyNotHugo_> I like busybox date/grep because they force me to write more portable scripts and not gnu-specific scripts.
2023-01-20 11:41:28 <frag> WhyNotHugo_: indeed, another tip is to use to use #!/usr/bin/dash .. im surprise at whats not posix..
2023-01-20 11:48:56 <psykose> if you want an extremely strict posix shell then you can use `yash -o posixlycorrect`
2023-01-20 11:49:16 <psykose> for containers you could install yash-binsh which makes it /bin/sh and when invoked as such the option is set
2023-01-20 11:49:23 <psykose> note though that a lot of alpine is not posix sh :p
2023-01-20 11:49:29 <psykose> i think even mkinitfs was/is broken by that
2023-01-20 11:50:54 <psykose> libtool scripts in builds specify /bin/sh, and guess what :)
2023-01-20 11:50:56 <psykose> (they break)
2023-01-20 11:52:33 <psykose> personally i don't find posix sh portability to be very exciting
2023-01-20 11:53:11 <psykose> 1) because i write a lot of shell script and it would be like 50x harder to not have common extensions available "${} things, local, .."
2023-01-20 11:53:20 <psykose> 2) because for the past 20 years you can just install a better shell
2023-01-20 11:53:22 <psykose> it's free software
2023-01-20 11:53:35 <psykose> i'm a bit tired of "my ultrix can't run something"
2023-01-20 11:53:38 <psykose> just install something better ffs
2023-01-20 11:54:24 <psykose> it's not like you have to pay someone to get a shell that understands `local`
2023-01-20 11:54:53 <psykose> but the moment the posix sh standard is updated to some 2027 version and actually specifies some commonly used stuff, i am all for it as a baseline
2023-01-20 11:55:06 <psykose> the ash list of few extensions are more than fine with me
2023-01-20 12:00:07 <WhyNotHugo_> Yeah, I don't want to be super-strict-posix, I mostly want my scripts to run okay on bsd/busybox and not just gnu.
2023-01-20 12:03:46 <Forza> Hi. 
2023-01-20 12:04:10 <Forza> How can I make mkinitramfs choose the new kernel instead of the old/running kernel's modules?
2023-01-20 12:06:47 <psykose> you mean mkinitfs? it generates it from the installed new kernel only
2023-01-20 12:10:20 <Forza> ya... I see " [ -z "$kernel" ] && kernel=$(uname -r)"
2023-01-20 12:10:58 <Forza> Should I use something else like Dracut?
2023-01-20 12:11:17 <Forza> or I could partch the mkinitfs tool
2023-01-20 12:11:26 <psykose> i'm not sure what issue you have
2023-01-20 12:11:33 <psykose> you didn't say what the problem was :p
2023-01-20 12:11:34 <Forza> I compiled a kernel from source
2023-01-20 12:11:43 <Forza> and installed it, and might want the initrd
2023-01-20 12:12:03 <Forza> but I did compile most needed stuff in the kernel instead of mod, so it might boot anyway
2023-01-20 12:12:24 <psykose> mkinitfs works on custom built kernels fine
2023-01-20 12:12:51 <psykose> it triggers when anything in /usr/share/kernels/ is changed
2023-01-20 12:13:29 <psykose> if you made a new linux-whatever package copying the normal kernel build apkbuild, it would just be a `whatever` folder in there and the mkinitfs trigger would generate you a initramfs-whatever
2023-01-20 12:14:15 <psykose> if you mean you installed something outside of normal apk packages, then uhh
2023-01-20 12:14:33 <Forza> yea
2023-01-20 12:14:55 <psykose> mkinitfs version-name from the name of yours in /lib/modules would work manually
2023-01-20 12:15:05 <psykose> it just matches things in that directory
2023-01-20 12:15:47 <Forza> mkinitfs -l shows it is loading files from ./lib/modules/6.1.4-0-lts/ and not from my kernel
2023-01-20 12:15:57 <Forza> ah
2023-01-20 12:16:00 <psykose> it's not relevant what mkinitfs -l shows
2023-01-20 12:16:52 <psykose> though if you do `mkinitfs -l <yourkernel>` that also works
2023-01-20 12:21:50 <Forza> I didnt know how to specify my own kernel with mkinifs. Now I know, and it works
2023-01-20 12:21:52 <Forza> :D
2023-01-20 12:21:59 <Forza> so thank you
2023-01-20 12:43:33 <Forza> The new kernel seems to run ok. Lets see if I manage to capture more debug stuff if it spontaneously reboots again
2023-01-20 13:17:24 <JanJ[m]> With the new kernel KDE keeps crashing on me :(
2023-01-20 13:30:48 <Forza> :(
2023-01-20 15:53:26 <donoban> JanJ[m]: do you know the crash is kernel related? or is a kernel panic?
2023-01-20 16:05:18 <markand> hmmm, for some reason I have /dev/dri/card1 but not 0 and Xorg complains when trying to start X 
2023-01-20 16:06:34 <markand> GNOME on wayland works fine though
2023-01-20 16:07:09 <donoban> i have also only /dev/dri/card1
2023-01-20 16:09:16 <psykose> as do i
2023-01-20 16:10:12 <JanJ[m]> <donoban> "Jan J.: do you know the crash is..." <- i dont know if its kernel related, thats why i didnt say "the new kernel crashes my kde" 
2023-01-20 16:11:32 <donoban> do you see something on dmesg?
2023-01-20 16:12:00 <JanJ[m]> nope, im trying to figure out where i can see the logs for kde :) if there are any logs at all
2023-01-20 16:14:20 <donoban> do you use Xorg or wayland?
2023-01-20 16:15:28 <JanJ[m]> Xorg, Wayland is too new :P
2023-01-20 16:16:14 <donoban> maybe is something on Xorg log
2023-01-20 16:16:24 <JanJ[m]> though im always willing to switch if that means a stable desktop environment on Alpine :)
2023-01-20 16:17:20 <invoked> not advocating anything: wayland has been much less fiddly for me
2023-01-20 16:17:37 <donoban> I'm using wayland since migrated to alpine on my desktop and I feel that I had 0 problems
2023-01-20 16:18:22 <donoban> but it would be nice if you found the problem
2023-01-20 16:18:24 <donoban> :)
2023-01-20 16:18:39 <JanJ[m]> did you document how to migrate to wayland? :) 
2023-01-20 16:19:25 <donoban> uhM, I don't remeber doing anything special, do you use some login manager?
2023-01-20 16:19:42 <JanJ[m]> sddm
2023-01-20 16:19:59 <JanJ[m]> it shows wayland, but the last time i tried that, the graphics where horrible :)
2023-01-20 16:20:21 <donoban> uh, low resolution?
2023-01-20 16:21:08 <psykose> kde is kinda fiddly with wayland still
2023-01-20 16:21:08 <JanJ[m]> truthfully, i recently added manjaro because of the crashes. i really cant have that when im with clients :) no... more like.. "odd" windows half visible, closed windows that remain "painted" on the screen
2023-01-20 16:21:20 <psykose> it's also qt5 and qt5 is notably more shit on wayland than qt6
2023-01-20 16:21:26 <psykose> i'd give it another year for kde
2023-01-20 16:23:59 <JanJ[m]> :)
2023-01-20 16:25:00 <JanJ[m]> really, i just want to create an Alpine desktop experience :) something i can share with friends and family :)
2023-01-20 16:36:09 <markand> hmmm I wonder why Xorg fails with no screen founds then
2023-01-20 16:38:31 <psykose> post the full xorg error long thing it outputs
2023-01-20 16:38:38 <psykose> into the file
2023-01-20 16:38:39 <psykose> somewhere
2023-01-20 16:39:35 <invoked> psykose: the limited qt5 stuff i do run on wayland seems fine (eg keepassxc)
2023-01-20 16:39:47 <markand> https://bpa.st/5M3AM
2023-01-20 16:40:09 <psykose> invoked: which wm do you use
2023-01-20 16:40:15 <invoked> swaywm mostly
2023-01-20 16:43:41 <elly> markand: you have no video modules installed for X?
2023-01-20 16:43:54 <elly> grep xf86 /etc/apk/world
2023-01-20 16:44:02 <elly> see if you have any xf86-video-*
2023-01-20 16:44:21 <markand> you don't need any since it uses modesettings
2023-01-20 16:44:45 <elly> not even xf86-video-modesetting?
2023-01-20 16:44:49 <elly> I thought it was still needed
2023-01-20 16:44:59 <markand> it's installed 
2023-01-20 16:45:07 <invoked> psykose: i still have the QT_QPA_PLATFORM=wayland and QT_WAYLAND_DISABLE_WINDOWDECORATION=1 in my env, i don't know if that is still needed anymore. apart from that stuff seems ok.
2023-01-20 16:47:25 <psykose> do QT_FONT_DPI=160 or something high like that and move your mouse in like a 2-tier dropdown from a top menu (like a dropdown from the top bar, then a second layer in there and move it up and down and watch what it selects)
2023-01-20 16:47:33 <psykose> does seem to be a bit better than in .5 though
2023-01-20 16:47:43 <psykose> markand: weird, not sure why it tries to open 0 after loading it
2023-01-20 16:48:06 <markand> D:
2023-01-20 16:48:11 <psykose> you can override it with kmsdev=1
2023-01-20 16:48:11 <psykose> i guess
2023-01-20 16:48:15 <psykose> in the options for modesetting
2023-01-20 16:48:20 <psykose> but this seems silly
2023-01-20 16:48:42 <markand> let me try a dirty symlink first
2023-01-20 16:49:13 <psykose> ah
2023-01-20 16:49:15 <psykose> KMSDEVICE=1
2023-01-20 16:49:17 <psykose> it reads env too
2023-01-20 16:49:27 <psykose> and yes
2023-01-20 16:49:31 <psykose> there's no fallback
2023-01-20 16:49:34 <psykose> it tries 0 and nothing elsee
2023-01-20 16:50:08 <psykose> https://gitlab.freedesktop.org/xorg/xserver/-/blob/master/hw/xfree86/drivers/modesetting/driver.c#L236
2023-01-20 16:50:33 <psykose> er, KMSDEVICE=/dev/dri/card1
2023-01-20 16:50:35 <psykose> but yeah
2023-01-20 16:51:26 <markand> oh...
2023-01-20 16:51:39 <markand> it's because I wasn't able to write access /dev/dri/card1 (forgot to add me to video group)
2023-01-20 16:51:48 <markand> if only Xorg could just tell permission denied ._.
2023-01-20 16:51:55 <invoked> psykose: funny you mention that, i'm getting a 4k thinkpad next week. so i'm sure i'll see all the high dpi weirdness i've been insulated from thus far.
2023-01-20 16:52:11 <psykose> you can test it ahead of time like that :)
2023-01-20 16:52:17 <markand> now Xorg starts but gdm still don't log me in in GNOME on Xorg session yet but I'll dig into it
2023-01-20 16:52:27 <psykose> i don't even have a highdpi display really (27" 1440p) but i just like things to be bigger
2023-01-20 16:52:50 <psykose> markand: ah, the classic..
2023-01-20 16:52:53 <psykose> glad you found it
2023-01-20 16:52:55 <markand> :D
2023-01-20 16:53:00 <invoked> yeah. i'm not wild about the 4k but 1440p wasn't an option, because, lenovo
2023-01-20 16:53:09 <psykose> yeah 4k laptops are.. a bit silly
2023-01-20 16:53:31 <psykose> my 1080p 14" laptop display has a higher dpi than my current screen
2023-01-20 16:53:48 <invoked> if i wasn't coming into it cheap i wouldn't bother.
2023-01-20 16:53:49 <psykose> i kinda miss that (text looks nice on high dpi, to the point i prefer not even using a bitmap font)
2023-01-20 16:54:49 <invoked> i'm very used to 1440p, though, everything else i do is on 1440p ips screens, and no joke, i see pixels on any 1080p now.
2023-01-20 16:55:30 <invoked> and this is leaving aside any aspect ratio griping...
2023-01-20 16:55:47 <invoked> but i've adapted to 16:9 so, it is what it is
2023-01-20 16:56:13 <psykose> mhm :)
2023-01-20 16:58:15 <invoked> i specifically wanted a xeon laptop and that narrows things down quite a bit.
2023-01-20 18:01:30 <frag> xmodmap syntax seems so much simpler than xkb (which is impossible to understand!)
2023-01-20 18:41:55 <mbx> hello ^^
2023-01-20 18:53:37 <mbx> rc-status
2023-01-20 18:53:49 <mbx> wrong window ahaha
2023-01-20 18:56:32 <psykose> Runlevel: async
2023-01-20 19:10:51 <mbx> quick question: why is rc-service not responding with the running status? even the verbose flag isn't printing more info.
2023-01-20 19:11:04 <mbx> rc-status*
2023-01-20 19:12:27 <lopid> it isn't on the right?
2023-01-20 19:12:32 <lopid> that's →
2023-01-20 19:13:00 <mbx> nop, below Runlevel: default it only lists service names, but not the running status
2023-01-20 19:13:55 <lopid> it's not a display issue? does $COLUMNS match the display?
2023-01-20 19:14:21 <psykose> they're on the right for me
2023-01-20 19:14:41 <mbx> $COLUMNS is empty, gonna test this out, but I no problems of cutting-off text so far
2023-01-20 19:15:11 <mbx> setting COLUMNS has no effect on the output
2023-01-20 19:15:14 <lopid> COLUMNS may be bash only, tbf
2023-01-20 19:15:30 <mbx> im on ash, within a VM on proxmox
2023-01-20 19:15:45 <mbx> kvm console
2023-01-20 19:16:08 <lopid> indeed, it's not posix. but still
2023-01-20 19:16:48 <lopid> do they appear if you redirect to a file?
2023-01-20 19:17:38 <mbx> yes, they are added to the file when I pipe
2023-01-20 19:18:18 <mbx> okay my bad, this kvm window screwed me
2023-01-20 19:18:32 <mbx> thanks a lot
2023-01-20 19:33:49 <mbx> is it actually enough to install sway, xwayland and seatd for a minimal setup? seatd is running, my user is added to input, video and I don't want elogind. sadly I get an error when I launch sway: backend/wayland/vackend.c:553: Could not connect to remote display: No such file or directory, I've set XDG_SEAT, XDG_SESSION_TYPE, XSG_SESSION_CLASS and WAYLAND_DISPLAY as adviced in the Wiki.
2023-01-20 19:34:13 <mbx> can't find the backend.c in the net
2023-01-20 19:34:54 <mbx> "[wlr] [backend/wayland/backend.c:553]"*
2023-01-20 19:35:49 <mbx> I also set XDR_RUNTIME_DIR forgot to mention that
2023-01-20 19:35:57 <mbx> XDG*
2023-01-20 19:38:11 <invoked> you don't need xwayland unless you have something that needs x.
2023-01-20 19:41:17 <invoked> what url are you following?  i don't set any of those.
2023-01-20 19:42:40 <psykose> you don't need to set any of that except XDG_SESSION_TYPE=wayland i guess
2023-01-20 19:43:10 <psykose> the reason why it fails is probably indeed because you set wayland_display
2023-01-20 19:43:15 <psykose> that is set by the compositor itselff
2023-01-20 19:43:48 <psykose> having it set probably makes wlroots think it's already inside a wayland session and to set up inside an existing one and then it fails to connect to $xdg_runtime_dir/$wayland_display
2023-01-20 19:44:56 <invoked> i don't think MOZ_ENABLE_WAYLAND is even still necessary. (might be, but it doesn't hurt anything so i haven't removed it).  along with QT_QPA_PLATFORM if you need qt5.
2023-01-20 19:45:08 <psykose> yep https://gitlab.freedesktop.org/wlroots/wlroots/-/blob/master/backend/backend.c#L323
2023-01-20 19:46:08 <psykose> even session type is set since libseat in 0.15
2023-01-20 19:46:11 <psykose> so yeah set none of that
2023-01-20 19:46:17 <invoked> MOZ_ENABLE_WAYLAND=1 is for firefox/tor browser, if it's still needed.  QT_QPA_PLATFORM=wayland for qt5, if that's still needed.  i haven't bothered to look. then a couple others for sdl and evas, maybe.
2023-01-20 19:46:25 <psykose> yep
2023-01-20 19:46:29 <mbx> unfortunately, it has less than 500 lines of code, but this is all good to know. Still new to wayland, trying to figure out how to get/set the proper compositor
2023-01-20 19:46:52 <psykose> what does
2023-01-20 19:46:58 <mbx> is sway the compositor?
2023-01-20 19:47:05 <psykose> yes
2023-01-20 19:47:19 <mbx> okay, then it's set to sway
2023-01-20 19:47:26 <psykose> what is set to sway
2023-01-20 19:47:46 <invoked> also if you do need xwayland you need to explicitly enable it in sway's config (see the docs)
2023-01-20 19:47:46 <mbx> XDG_CURRENT_DESKTOP, XDG_SESSION_DESKTOP
2023-01-20 19:48:05 <mbx> thanks, I think I don't need xwayland yet, but I will keep that in mind
2023-01-20 19:48:07 <psykose> sure you can set that, though they don't really do anything
2023-01-20 19:49:06 <mbx> @invoked I follow https://wiki.alpinelinux.org/wiki/Sway and all the links in it
2023-01-20 19:50:16 <invoked> that needs to be fixed up, probably.
2023-01-20 19:51:05 <mbx> @psykose I removed all variables except for XDG_SESSION_TYPE and the error remains
2023-01-20 19:51:32 <psykose> sway -dV &> log.txt and pastebin it somewhere
2023-01-20 19:52:56 <mbx> I forgot to unset them properly, it now says Unable to create backendn any DRM device backend on any GPU
2023-01-20 19:53:46 <psykose> did you install mesa-dri-gallium
2023-01-20 19:54:42 <mbx> yes
2023-01-20 19:54:49 <mbx> https://imgur.com/a/ROoMnrJ
2023-01-20 19:55:26 <mbx> 22.3.3-r0, I'm on edge reps
2023-01-20 19:58:04 <invoked> what card do you have
2023-01-20 19:58:59 <mbx> its a kvm on a proxmox instance, dedicated server with integrated graphics, Intel® Xeon® Processor E3-1230 v6
2023-01-20 20:01:37 <mbx> in proxmox, the graphic card is set on "Default", I could select something else
2023-01-20 20:02:09 <mbx> https://imgur.com/a/uvbGV38
2023-01-20 20:02:43 <psykose> it sounds like https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3167
2023-01-20 20:04:07 <psykose> so yeah play with the vm gpu settings i guess
2023-01-20 20:04:31 <psykose> virtio-gpu should work afaik
2023-01-20 20:04:44 <psykose> i have no idea as i don't use vms to run desktops
2023-01-20 20:08:58 <mbx> okay, i uninstalled xwayland and the error was the same, now I switched to virtio and now it complains that it cannot start the Xwayland binary. I thought xwayland is not necessary O.o
2023-01-20 20:10:03 <psykose> you have to pass xwayland off in the config afaik
2023-01-20 20:20:36 <invoked> xwayland is not necessary to run sway. xwayland is executed by sway, if you've enabled it in your configuration (and restarted) and only if something needs x11.
2023-01-20 20:20:54 <mbx> I definitely progress. After putting xwayland disable in the sway conf I get a new error "execvp failed: No such file or directory001)" and huge log
2023-01-20 20:22:33 <invoked> are you using someone else's config file
2023-01-20 20:22:51 <mbx> no, the default one that came with sway
2023-01-20 20:24:06 <psykose> just paste the log please
2023-01-20 20:24:41 <mbx> yep, I want to. it's huge and I have SSH not setup yet, since this VM is beind a VM
2023-01-20 20:24:48 <mbx> a DHCP*
2023-01-20 20:25:56 <mbx> oh u got tpaste in the repo, that's nice
2023-01-20 20:26:24 <psykose> there's always `| nc termbin.com 9999`
2023-01-20 20:27:32 <mbx> lovely, thanks for the tip
2023-01-20 20:27:33 <mbx> https://termbin.com/q4mf
2023-01-20 20:28:54 <psykose> well, basically just more gpu incompatibility
2023-01-20 20:29:04 <psykose> as denoted by the [ERROR] [wlr] [EGL] command: eglInitialize, error: EGL_NOT_INITIALIZED (0x3001), message: "DRI2: failed to create screen"
2023-01-20 20:29:24 <mbx> F
2023-01-20 20:34:59 <mbx> I got it working
2023-01-20 20:35:34 <mbx> I selected SPICE as GPU in proxmox, lucky
2023-01-20 20:35:47 <mbx> SPICE (qxl)
2023-01-20 20:36:00 <mbx> Thanks so much for your help, very appreciating it
2023-01-20 20:50:41 <psykose> grats
2023-01-20 21:28:12 <spazialecontinentale[m]> hi, anyone use alpine for their homeserver?
2023-01-20 21:28:12 <spazialecontinentale[m]> if so do you use container like podman/docker or normal package through apk?
2023-01-20 21:29:51 <lopid> normal package whenever possible
2023-01-20 21:31:51 <spazialecontinentale[m]> lopid: why? for security?
2023-01-20 21:31:51 <spazialecontinentale[m]> Using apk packages is more inconvenient?
2023-01-20 21:32:04 <lopid> why add bloat?
2023-01-20 21:32:19 <spazialecontinentale[m]> lopid: is podman bloat? 
2023-01-20 21:32:29 <lopid> compared to no padman? yes
2023-01-20 21:32:30 <spazialecontinentale[m]> why is used in server production then?
2023-01-20 21:32:50 <lopid> because of a different use case
2023-01-20 21:33:12 <spazialecontinentale[m]> ( i'm not an expert, i just want to know what technology to use )
2023-01-20 21:33:57 <spazialecontinentale[m]> lopid: so for small/home server is overkill?
2023-01-20 21:34:32 <lopid> it depends what it is. for "normal" stuff, yes
2023-01-20 21:35:27 <spazialecontinentale[m]> lopid: a nextclound sever accessible from outside 
2023-01-20 21:35:32 <lopid> like why would i use vim in docker? ls, mutt… there is no reason for it. for some larger packages, there may be, like nextcloud
2023-01-20 21:37:09 <spazialecontinentale[m]> lopid: > a nextclound sever accessible from outside
2023-01-20 21:37:09 <spazialecontinentale[m]> i'm confused now hahha.
2023-01-20 21:37:20 <spazialecontinentale[m]> s/i/I/
2023-01-20 21:37:20 <alpine-meetbot> spazialecontinentale[m] meant to say: I'm confused now hahha.
2023-01-20 21:38:29 <lopid> it may be self contained and easier to manage
2023-01-20 21:39:07 <spazialecontinentale[m]> does it add a layer of security?
2023-01-20 21:39:12 <lopid> nextcloud, for example, has dozens of native packages available. versus one docker image
2023-01-20 21:39:24 <ikke> docker should not be considered a security layer
2023-01-20 21:41:08 <spazialecontinentale[m]> if docker/podman only add a "simplicity" layer i will try to use the normal package for learning purpose 
2023-01-20 21:42:20 <ikke> docker mostly helps when you start running into different applications needing different dependencies
2023-01-20 21:43:23 <spazialecontinentale[m]> i will run only nextcloud so i should encounter no problem
2023-01-20 21:43:40 <spazialecontinentale[m]> thanks a lot for the help
2023-01-20 21:45:01 <psykose> it can also be considered a security layer if your threat model doesn't include someone targetting you
2023-01-20 21:45:17 <psykose> less automated stuff that gets out of that than otherwise
2023-01-20 21:45:48 <psykose> but meh, indeed it's mostly about convenience depending on what you find easy
2023-01-21 02:09:51 <mbx> I installed libinput but libinput debug-events outputs "debug-events is not installed" ... well that's not a package and it should be installed when installing libinput, it should not?
2023-01-21 02:12:33 <psykose> libinput-tools
2023-01-21 02:12:44 <psykose> https://pkgs.alpinelinux.org/contents?file=libinput-debug-events&path=&name=&branch=edge
2023-01-21 02:13:24 <mbx> thanks, I installed only libinput
2023-01-21 02:15:31 <mbx> yes it works, next time I will use apk list to find similar packages
2023-01-21 10:58:26 <spazialecontinentale[m]> hi, i need ntfy (https://pkgs.alpinelinux.org/package/edge/testing/x86_64/ntfy) but it's on egde. How can i get it without moving to edge?
2023-01-21 11:02:59 <ikke> spazialecontinentale[m]: right now it's only in testing, so that will never be built for stable branches
2023-01-21 11:03:15 <ikke> spazialecontinentale[m]: if you don't want to move to edge, you have to build it yourself on a stable branch
2023-01-21 11:03:50 <spazialecontinentale[m]> how much time a package need to move to stable?
2023-01-21 11:04:12 <ikke> spazialecontinentale[m]: there does not have to be much time in between, but it depends on the maintaainer
2023-01-21 11:04:14 <ikke> maintainer*
2023-01-21 11:06:41 <spazialecontinentale[m]> ikke: can i reuse the APKBUILD?
2023-01-21 11:06:45 <ikke> sure
2023-01-21 11:08:05 <qaqland[m]> can i open edge to install apk and close edge after install
2023-01-21 11:08:33 <ikke> grawity: combining different branches is not supported
2023-01-21 11:08:50 <ikke> you will run into dependency issues
2023-01-21 11:08:57 <ikke> qaqland[m]: ^
2023-01-21 11:09:00 <ikke> sorry, wrong highlight
2023-01-21 11:09:57 <spazialecontinentale[m]> ikke: also can i write to the maintainer to ask for stable release? I don't know if there can be two maintainers of the same package
2023-01-21 11:11:12 <ikke> spazialecontinentale[m]: No, a package has a single maintainer, and it would only be the next stable release that would contain the package. It's unusuall to backport new packages to a stable release
2023-01-21 11:11:45 <spazialecontinentale[m]> ok thanks
2023-01-21 11:15:02 <ikke> spazialecontinentale[m]: and yes, asking the maintainer to move it to community (which will result it being included in the next release) is normal
2023-01-21 11:16:53 <spazialecontinentale[m]> doas modprobe tun
2023-01-21 11:16:53 <spazialecontinentale[m]> modprobe: FATAL: Module tun not found in directory /lib/modules/5.15.86-0-lts
2023-01-21 11:16:53 <spazialecontinentale[m]> sorry i'm new to alpine how can i fix this?
2023-01-21 11:17:25 <ikke> spazialecontinentale[m]: most likely you need to reboot
2023-01-21 11:17:31 <ikke> what does uname -r return?
2023-01-21 11:17:46 <spazialecontinentale[m]> 5.15.86-0-lts
2023-01-21 11:18:00 <ikke> hmm
2023-01-21 11:18:03 <ikke> it's the same
2023-01-21 11:18:13 <ikke> does /lib/modules/5.15.86-0-lts exist?
2023-01-21 11:18:14 <spazialecontinentale[m]> i will try to reboot
2023-01-21 11:18:43 <ikke> (well, it makes sense it's the same, it would always try to look in the folder of the running kernel)
2023-01-21 11:19:51 <spazialecontinentale[m]> ikke: now it work thanks
2023-01-21 11:19:58 <ikke> alright, good
2023-01-21 11:20:23 <ikke> We don't have a good solution yet to keep the running kernel around
2023-01-21 11:20:33 <ikke> (or rather, the modules for the running kernel)
2023-01-21 13:04:26 <Saklad5> Is there a way to edit a flag made on an APK package?
2023-01-21 13:04:55 <Saklad5> There were multiple updates for a piece of software, Dendrite, in quick succession
2023-01-21 13:05:48 <Saklad5> I can't see a way to flag it again either, and waiting for it to be "updated" before flagging it yet again seems like a waste of the maintainer's time
2023-01-21 13:06:48 <ikke> Saklad5: what you do mean with flags?
2023-01-21 13:07:00 <ikke> oh, on pgks.a.o
2023-01-21 13:07:03 <ikke> then no
2023-01-21 13:07:32 <ikke> Most of the time you would check what's the latest version anyway
2023-01-21 13:07:44 <ikke> so no need to update the flag
2023-01-21 13:08:30 <Saklad5> OK
2023-01-21 13:10:14 <Saklad5> Changing the subject, I made an issue over a month ago asking for a minor tweak to a package's configuration, and I haven't gotten any response from the maintainer. Is there a way to at least get confirmation it will be looked at eventually?
2023-01-21 13:10:55 <pj[m]> Nope
2023-01-21 13:11:01 <pj[m]> what's the package and where did you ask
2023-01-21 13:11:21 <Saklad5> https://gitlab.alpinelinux.org/alpine/aports/-/issues/14440
2023-01-21 13:12:25 <pj[m]> You are not going to get a response
2023-01-21 13:12:50 <Saklad5> Is there a problem with what I did?
2023-01-21 13:12:53 <ikke> nop
2023-01-21 13:13:07 <ikke> jirutka seems to be a bit busy lately, so responses are erratic
2023-01-21 13:13:18 <pj[m]> well, you will get a response, but after few months
2023-01-21 13:13:32 <Saklad5> That's fine, I can wait so long as it happens eventually
2023-01-21 13:13:57 <ikke> in any case, it will very likely not happen for 3.17
2023-01-21 13:14:13 <pj[m]> ikke: if by lately you mean a whole past year and now
2023-01-21 13:14:15 <Saklad5> I don't particularly care, I'm not planning to keep anything on it long-term
2023-01-21 13:14:32 <Saklad5> On 3.17, I mean
2023-01-21 13:14:41 <pj[m]> he's usually more active in more important topics
2023-01-21 13:14:50 <Saklad5> Understandable
2023-01-21 13:15:04 <pj[m]> You can open MR so it's there hanging
2023-01-21 13:15:12 <pj[m]> I might eventually ping him on Twitter
2023-01-21 13:15:26 <Saklad5> Ew
2023-01-21 13:15:53 <pj[m]> ¯\_(ツ)_/¯
2023-01-21 13:16:10 <pj[m]> he's also on Discord
2023-01-21 13:16:30 <pj[m]> although never seen him active there
2023-01-21 13:17:12 <Saklad5> So long as this is addressed eventually, I'm satisfied. I can build it myself if I get tired of waiting
2023-01-21 13:18:36 <pj[m]> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/33367
2023-01-21 13:18:37 <pj[m]> one day
2023-01-21 13:19:09 <Saklad5> ???
2023-01-21 13:19:22 <ikke> well, there the pipeline still fails, so makes sense that it's not merged yet
2023-01-21 13:19:52 <pj[m]> I should probably bump it again since there is a new version from this year
2023-01-21 13:20:04 <pj[m]> maybe they fixed tests
2023-01-21 13:20:11 <Saklad5> He's updated the package I want tweaked twice since I opened the issue, so he hasn't abandoned it or anything
2023-01-21 13:20:59 <Saklad5> Though that does make it a little frustrating that he couldn't add two flags to the build configuration while he was working on it
2023-01-21 13:21:34 <Saklad5> Or, you know, acknowledge it at all
2023-01-21 13:23:04 <pj[m]> don't mind it, he's very busy person that does a lot for alpine
2023-01-21 13:23:12 <Saklad5> I can tell
2023-01-21 13:24:01 <Saklad5> Might want to reduce their workload a little, though.
2023-01-21 13:24:24 <Saklad5> Maybe I should look into volunteering as maintainer at some point?
2023-01-21 13:25:37 <panekj> Author	2020	2021	2022	2023	Total
2023-01-21 13:25:37 <panekj> Leo	6065	2925	109	0	9099
2023-01-21 13:25:37 <panekj> psykose	0	188	7629	806	8623
2023-01-21 13:25:41 <panekj> psykos is leo
2023-01-21 13:26:08 <panekj> 🤔
2023-01-21 13:26:13 <Saklad5> It's sort of terrifying
2023-01-21 13:26:17 <Newbyte> what are these numbers I'm looking at?
2023-01-21 13:26:23 <panekj> commits
2023-01-21 13:26:53 <panekj> "commits per author per year"
2023-01-21 13:27:03 <panekj> https://git.alpinelinux.org/aports/stats/?period=y&ofs=10
2023-01-21 13:28:48 <mbx> hi here, I've setup Alpine with Wayland and Sway in a KVM, it's working, but it's damn slow. Is this a common problem with Wayland, will x be faster in a VM?
2023-01-21 13:29:09 <panekj> I'm sure your problem is that you are using Linux desktop in KVM
2023-01-21 13:29:11 <Saklad5> Alpine Linux is very fast on my virtual private servers, which all use KVM
2023-01-21 13:29:23 <Saklad5> All headless, though
2023-01-21 13:29:37 <mbx> panekj: Well Debian LXDE works fast in a VM compared to this
2023-01-21 13:30:20 <mbx> it's really terribly slow, like windows dragging behind and once I launch Firefox it's almost freezing
2023-01-21 13:30:22 <panekj> what host, what guest configuration, what sway config
2023-01-21 13:30:31 <Saklad5> How much RAM did you allocate?
2023-01-21 13:30:42 <mbx> proxmox dedicated server with Xeon CPU
2023-01-21 13:30:48 <mbx> 512MB
2023-01-21 13:30:56 <Saklad5> Not sure that's enough for a GUI
2023-01-21 13:31:12 <Saklad5> If it was thrashing, those are the symptoms I'd expect
2023-01-21 13:31:21 <mbx> ok will try more, let me show you my sway config as well, sec
2023-01-21 13:31:32 <Saklad5> Before you do, just check if it is actually running out
2023-01-21 13:32:23 <mbx> https://termbin.com/tdzh my sway config
2023-01-21 13:32:34 <mbx> oh sorry Saklad, let me check real quick
2023-01-21 13:32:58 <Saklad5> Alpine Linux seems like an odd choice of OS to use a GUI with. Its main strength is being as minimal as it can be without breaking standards
2023-01-21 13:33:23 <Saklad5> GUIs aren't exactly lightweight
2023-01-21 13:33:42 <panekj> I don't see how that relates to using GUI
2023-01-21 13:33:55 <mbx> Saklad5: I know right, I just want to use it to config my pfSense in a browser from the same network without exposing pfSense's web configurator to the world.
2023-01-21 13:34:11 <mbx> so I wanted a light-weight OS that allows me to get a browser
2023-01-21 13:34:20 <WhyNotHugo> fprintd won't work without polkit?
2023-01-21 13:34:53 <mbx> and it gives a great opportunity to test alpine, planning in using it for servers as well
2023-01-21 13:36:18 <mbx> wow Saklad, that's it, just needed more RAM
2023-01-21 13:36:27 <mbx> thanks brother
2023-01-21 13:36:54 <mbx> super happy that all works eventually
2023-01-21 13:37:49 <mbx> one more thing though. I did my research when it comes to using Alpine as a server and I concluded that it's light-weight, but it's slower with compiling incl JIT and the sorts
2023-01-21 13:37:58 <mbx> So python is slower too
2023-01-21 13:38:22 <Saklad5> Compared to…
2023-01-21 13:38:37 <mbx> Unfortunately, I need to mass host a LEMP stack, and my fear is PHP will run slow as well, there is also an issue on alpine how PHP 8 is very slow
2023-01-21 13:39:09 <mbx> compared to glibc and gnu tools instead of musl libc and busybox
2023-01-21 13:39:18 <Saklad5> You could just install those, no?
2023-01-21 13:39:46 <mbx> good point, I could. Not having systemd is still a big pro
2023-01-21 13:40:15 <Saklad5> I do enjoy OpenRC. It's much easier to configure dependencies.
2023-01-21 13:40:41 <psykose> you installed an entire vm to run a web browser in it to access one web page to configure something in another vm.. and that is your concern?
2023-01-21 13:41:02 <Saklad5> Yeah, this seems like a convoluted method of solving your problem
2023-01-21 13:41:30 <mbx> never had a problem with systemd, but the footprint alone makes it worth it. I gonna need to host around 100 linux containers on this server, at least this is my hope, something that wouldn't be possible with systemd I think
2023-01-21 13:41:55 <Saklad5> I always espouse a zero-trust approach: secure things with encryption, not network isolation.
2023-01-21 13:42:44 <mbx> psykose: the VM is also to play with it, I also look forward to install browsh on it for a remote web browser experience. I want to experiment with it. :)
2023-01-21 13:44:17 <Saklad5> Word of advice: if you are running Alpine Linux on a VPS, you can probably use the virtual image instead of standard
2023-01-21 13:44:30 <mbx> Saklad5: agreed, security through obscurity is a bad take, stills feels better this way. I wouldn't want to expose my home router admin panel simply because I have to trust an additional party (manufacture of the router), same goes to pfSense.
2023-01-21 13:44:45 <Saklad5> I feel like you have to trust it regardless
2023-01-21 13:44:50 <mbx> Saklad5: Thanks, I installed the virtual image
2023-01-21 13:46:30 <Saklad5> Also, I managed to go weeks before realizing I hadn't installed an NTP/NTS client on my servers
2023-01-21 13:46:44 <Saklad5> I think it's supposed to prompt you to do that during installation, but that didn't happen for me
2023-01-21 13:47:33 <Saklad5> It's not hard to install chrony and set it up with either the clock provided by the VM host or an NTS server, though
2023-01-21 13:48:00 <Saklad5> All my servers had drifted by like thirty seconds before I fixed it
2023-01-21 13:50:04 <mbx> I think I don't have NTS here as well, but it got the right time, probably from the host
2023-01-21 13:50:13 <Saklad5> Busybox has an NTP client, but it doesn't seem to support "hardware clocks" that VMs can use or NTS. And without NTS, an attacker could easily mess with your server clock
2023-01-21 13:50:40 <Saklad5> It's not that common right now, but if I'm going to set it up manually I'm going to set it up *correctly*
2023-01-21 13:52:19 <mbx> i c, definitely something I should tackle as well later
2023-01-21 13:52:36 <mbx> LXC should be unaffected though
2023-01-21 13:52:47 <mbx> or is it?
2023-01-21 13:52:54 <Saklad5> Unaffected by what?
2023-01-21 13:53:10 <mbx> by the lack of NTS
2023-01-21 13:53:44 <Saklad5> NTS is basically NTP with TLS
2023-01-21 13:53:59 <Saklad5> Except slightly more roundabout than usual, for reasons I am not qualified to discuss
2023-01-21 13:54:26 <Saklad5> Normal NTP is completely vulnerable
2023-01-21 13:54:40 <mbx> is ntpsec = nts?
2023-01-21 13:55:13 <Saklad5> I think that's an implementation.
2023-01-21 13:55:27 <Saklad5> I considered it as well before choosing Chrony
2023-01-21 13:56:00 <Saklad5> NTS is Network Time Security: https://datatracker.ietf.org/doc/rfc8915/
2023-01-21 13:57:08 <mbx> I realise it's an topic in and on itself, so Chrony is your recommend implementation of NTP
2023-01-21 13:57:31 <Saklad5> I am very much a novice here, but after a bit of research yes
2023-01-21 13:57:37 <Saklad5> You should probably do your own
2023-01-21 13:57:50 <mbx> i will thx
2023-01-21 13:58:39 <Saklad5> It'd be nice if Busybox's NTP implementation could use hardware clocks, since my Linode server actually has one. But it doesn't, so I installed Chrony just for that anyway
2023-01-21 14:01:31 <Saklad5> NTS doesn't encrypt the actual timing updates, by the way, since that is obviously public
2023-01-21 14:02:02 <Saklad5> This is one of the rare examples of not needing the confidentiality bit of TLS
2023-01-21 14:04:39 <mbx> interestingly setup-ntp won't create /etc/ntp.conf
2023-01-21 14:04:48 <Saklad5> Anyway, you should definitely make sure your VM host doesn't expose timing data before configuring an NTS server in the guest: if it does, you can probably assume that it is already running one
2023-01-21 14:06:21 <Saklad5> And even if it uses NTP, that's not a concern of yours: if someone tampered with the host clock, you'd probably have issues regardless of the guest clock says
2023-01-21 14:07:17 <Saklad5> And by "NTS server" I meant "NTS client"
2023-01-21 14:07:30 <mbx> my proxmox host uses chrony
2023-01-21 14:07:53 <Saklad5> If it exposes it to the VM, you just need to use that
2023-01-21 14:08:03 <Saklad5> Probably a socket in /dev
2023-01-21 14:08:33 <Saklad5> If it doesn't, and you control the host, figure out if you can rectify that
2023-01-21 14:10:52 <mbx> the time was set right when I installed the VM, doesn't that mean that it's passed to my VM properly already?
2023-01-21 14:11:10 <Saklad5> I would have thought so, but no
2023-01-21 14:11:12 <mbx> not sure what socket to look for, there are a bunch of
2023-01-21 14:11:25 <Saklad5> Does rc-status show something that sets the time?
2023-01-21 14:11:35 <Saklad5> Because if it doesn't, nothing's doing it
2023-01-21 14:11:43 <mbx> no, but I now it does since I run setup-ntp
2023-01-21 14:12:01 <mbx> it wasn't showing anything beforehand
2023-01-21 14:13:10 <mbx> I have to go afk for a few minutes, I will figure this out later, thx for bringing this to my attention :)
2023-01-21 14:14:04 <Saklad5> No problem
2023-01-21 14:45:23 <pj[m]> <mbx> "never had a problem with systemd..." <- why not?
2023-01-21 14:51:33 <spazialecontinentale[m]> doas adduser marco abuild
2023-01-21 14:51:33 <spazialecontinentale[m]> abuild -r
2023-01-21 14:51:33 <spazialecontinentale[m]> abuild-apk: User marco is not a member of group abuild
2023-01-21 14:52:02 <spazialecontinentale[m]>  * doas adduser marco abuild... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/skDVgtMfGNUsxnjYvyrGsmas>)
2023-01-21 14:53:42 <spazialecontinentale[m]>  * doas adduser marco abuild... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/PXiuaJXMixUJifFUGcloxFRM>)
2023-01-21 14:54:17 <kwadroke[m]> you need to logout and back in for your user to get the new group(s)
2023-01-21 14:55:35 <spazialecontinentale[m]> kwadroke[m]: I forgot, thanks!
2023-01-21 18:12:56 <mbx> is there a way to swap two keys on the entire system without building a new keyboard layout? mac keyboards are weird in that it swaps the signals of ^ and <. the browser / proxmox tries to do things right by sending the signal to the qemu container, but Alpine things it deals with a native keyboard, so it swaps to counter for the mac keyboard, thus ^ and < are swapped in my system
2023-01-21 18:13:16 <mbx> this is really only a problem when emulating any linux while using a mac keyboard
2023-01-21 18:13:27 <mbx> it makes working with vim annoying
2023-01-21 18:13:49 <psykose> don't the vm settings let you swap that back first as an input thing
2023-01-21 18:14:07 <mbx> this is proxmox, so I couldn't figure a way
2023-01-21 18:14:58 <mbx> i could also just swap it for an application, but it runs in my browser and I don't want the swap to happen in my browser on every page
2023-01-21 18:15:06 <minimal> mbx: you're earlier discussion about time, NTP, etc - your VM's system clock will typically be initially set by the hypervisor
2023-01-21 18:16:00 <mbx> oh nice, minimal, so I don't have to care about setting up NTP at all? That's neat.
2023-01-21 18:18:20 <minimal> mbx: well the VM time could then drift, so as Proxmox is KVM-based I'd think specifying bothn "clocksource=kvm-clock" on your cmdline and also using chrony with a PHC specified as /dev/ptp_kvm would keep the VM time in line with the host clock - of course you'd still need to ensure the host's time is accurate (i.e. running a NTP client to sync it with elsewhere)
2023-01-21 18:20:24 <mbx> ah yes right, I remember Saklad5 mentioned NTP hardware clock isn't supported by busybox, so I initialize crony. I will dig into that later, thanks
2023-01-21 18:22:22 <mbx> in regards of my keyboard issue, I think i hit up nativefier with my proxmox instance and configure karabiner on my macbook to swap the keys for my electron app
2023-01-21 18:23:01 <mbx> the swapping should happen prior entering the VM, since it would behave correctly on bare metal with real keyboard hardware
2023-01-21 20:43:56 <Saijin_Naib> https://matrix.to/#/!NqiBUfFeQgPvAYuiKG:matrix.org/$16743067289430RvDiE:libera.chat?via=matrix.org&via=libera.chat&via=envs.net
2023-01-21 20:44:41 <Saijin_Naib> @psykose: I am experiencing the same issue, unsure of how to narrow down root cause.
2023-01-21 20:44:41 <Saijin_Naib> File a bug with Alpine, or is this possible upstream XFCE issue? 
2023-01-21 20:44:53 <psykose> i can't view that
2023-01-21 20:45:49 <ikke> me neither
2023-01-21 20:46:28 <Saijin_Naib> Good to know that permalink function works 🤣
2023-01-21 20:46:29 <Saijin_Naib> hey, guys, I spotted a strange error in xfce4-screensaver... AND I assume, it has to do with this file: */home/username/.xscreensaver...* in the line 8: lockTimeout.... when I let my screen locked for more then 30mins, I cannot login anymore.... the dialog simpy does not appear... the screen stays black and the mousepointer moves around flickering when using the mouse ..... does that soud familiar ? 
2023-01-21 20:46:56 <Saijin_Naib> 100% repro rate locally on latest-stable here. 
2023-01-21 20:48:58 <psykose> does it sound like https://gitlab.alpinelinux.org/alpine/aports/-/issues/13037
2023-01-21 20:49:06 <psykose> which was.. well, patched
2023-01-21 20:50:25 <Saijin_Naib> Yeah, same behavior, and running latest-stable with all updates, so no chance of packages that old 
2023-01-21 20:51:45 <psykose> i.e. same thing in the logs?
2023-01-21 20:52:36 <psykose> if so, not sure why it's still an issue
2023-01-21 20:52:45 <psykose> i couldn't really find anything else relevant upstream or whatnot
2023-01-21 20:53:03 <psykose> it looks like an abandoned xfce component compared to all the others
2023-01-21 20:53:11 <psykose> i would use a different screen locker personally
2023-01-21 20:56:12 <Saijin_Naib> It wasn't an issue until a week or so ago, though time is fuzzy a bit on that. Certainly not an issue for the entire past year, nor before the new year
2023-01-21 20:57:04 <psykose> maybe it's not the same issue
2023-01-21 20:57:10 <psykose> (with --debug you would see the same message)
2023-01-21 23:14:59 <mbx> minimal: clicksource=kvm-clock - is this still necessary if chronyc sources shows the PHC0 is followed properly? I think I got this.
2023-01-21 23:19:00 <minimal> mbx: clocksource is not directly related to NTP, on any Linux system there can be one or more clocksources
2023-01-21 23:19:34 <minimal> if you "cat /sys/devices/system/clocksource/clocksource0/available_clocksource" you'll see the list of ones available
2023-01-21 23:20:00 <minimal> typically for a KVM-based VM I'd use the "kvm-clock" clocksource
2023-01-21 23:58:01 <mbx> minimal: thanks. cat current_clocksource already returns kvm-clock, so I'm good to go. Is clocksource part of the kernel?
2023-01-22 00:29:36 <minimal> yeah, on any system if you do a "dmesg | grep clocksource" you can see the kernel deciding which one to use (it may pick one and then change during boot)
2023-01-22 00:30:11 <mbx> minimal: thanks
2023-01-22 00:30:20 <minimal> a physical machine might use "tsc" for instance
2023-01-22 00:30:41 <Habbie> my laptop (not alpine, just to be clear) goes tsc-early and then tsc, indeed
2023-01-22 00:30:48 <minimal> if you "cat" the current_clocksource file in the same /sys dir you'll see which one is currently in use
2023-01-22 00:31:55 <mbx> oh, now things are almost perfect here. I got my proxmox with an Alpine Linux KVM with Wayland+Sway and Firefox running in it. Previously Firefox was very slow with graphics until I increased the RAM, not there is plenty of RAM/storage/CPU power, the graphics are fast, but pages load so slow
2023-01-22 00:32:47 <mbx> I think it's the JavaScript engine having problems. My pfSense webConfigurator is taking long to load whenever JS executes
2023-01-22 00:40:41 <mbx> i continously read that wayland is not ready for kvm, so x11 is the way to go, bummer, but it is what it is
2023-01-22 08:34:49 <yjmbo> I've been trying to install mopidy, which in general seems to have been successful; this installs gstreamer, and that's where my biggest problem is. Although I've made sure that gst-plugins-base gst-plugins-good gst-plugins-ugly are all installed, and specifically /usr/lib/gstreamer-1.0/libgstsoup.so is present, gstreamer still insists that it doesn't have a URI handler for http or https. 
2023-01-22 08:36:07 <yjmbo> The sample command from gstreamer's docs is "gst-launch-1.0 playbin uri=https://www.freedesktop.org/software/gstreamer-sdk/data/media/sintel_trailer-480p.webm", and this just triggers an error, the core of which is "ERROR: from element /GstPlayBin:playbin0/GstURIDecodeBin:uridecodebin0: No URI handler implemented for "https"."
2023-01-22 08:37:01 <yjmbo> I'd welcome some advice on how to get this working :-)
2023-01-22 11:57:11 <adhawkins> yjftsjthsd: There is #gstreamer on this server. I've generally found them helpful.
2023-01-22 11:57:39 <adhawkins> It could be the way gstreamer is built for alpine I guess.
2023-01-22 13:26:20 <telmich> Who is running / populating https://hub.docker.com/u/alpine ?
2023-01-22 13:26:33 <telmich> It sounds as if it somebody from the alpine linux project, but the description of the user is somewhat empty
2023-01-22 13:28:50 <ikke> telmich: the source is https://github.com/docker-library/official-images/blob/master/library/alpine
2023-01-22 13:29:07 <ikke> So officially by docker itself, but triggered by ncopa making pull requests
2023-01-22 13:29:34 <telmich> I am not sure about that, ikke 
2023-01-22 13:29:38 <telmich> _ != alpine
2023-01-22 13:29:43 <ikke> ohh
2023-01-22 13:29:45 <telmich> I.e. above is the USER alpine
2023-01-22 13:29:45 <ikke> right
2023-01-22 13:29:48 <telmich> Not the library
2023-01-22 13:29:55 <telmich> That's why I am confused
2023-01-22 13:30:10 <ikke> Yeah, no clue who is behind that
2023-01-22 13:30:27 <telmich> I was actually checking out https://hub.docker.com/r/alpine/git/tags which seems to be very handy, but not knowing who/what is the alpine user, I'd rather stick to alpine:latest and do apk add git
2023-01-22 13:30:32 <telmich> :-)
2023-01-22 13:30:40 <ikke> yeah
2023-01-22 13:30:43 <ikke> We do have alpinelinux
2023-01-22 13:30:50 <ikke> https://hub.docker.com/u/alpinelinux
2023-01-22 13:30:54 <ikke> This is maintained by the alpine infra team
2023-01-22 13:31:57 <telmich> ok, alpinelinux/build-base has git, but seems to be a bit big for what I am looking for
2023-01-22 13:32:12 <ikke> yeah, if you just need git, build-base is not that great
2023-01-22 13:32:32 <telmich> My flow/idea for the background:
2023-01-22 13:32:40 <telmich> I am considering to use hugo and there is a nice hugo image
2023-01-22 13:32:49 <telmich> However that lacks git
2023-01-22 13:33:12 <telmich> So in k8s I was considering to use an initcontainer to git clone && git submodule init whatever is in the environment
2023-01-22 13:33:25 <telmich> And then pass that cloned content to the hugo container
2023-01-22 13:36:44 <telmich> Hmm, given that there is klakegg/hugo:alpine, I could in theory reuse that container, use apk add git in the initcontainer and then start it afterwards just regular
2023-01-22 17:09:05 <JanJ[m]> hmz
2023-01-22 17:59:01 <yjmbo> Thanks adhawkins, I might give them a try :-) 
2023-01-22 21:14:25 <KREYREN_oftc> howddya get GNOME on Teres-1 considering that i was told that setup-alpine is insane bcs reasons
2023-01-22 21:14:45 <KREYREN_oftc> and doing like setup-desktop gnome has conflicts
2023-01-22 21:15:27 <KREYREN_oftc> which is probably also insane bsc reasons O.o
2023-01-22 21:15:41 <psykose> what conflicts
2023-01-22 21:16:01 <KREYREN_oftc> something something eudev libudev-zero 
2023-01-22 21:16:26 <psykose> i just tried it and there are none
2023-01-22 21:16:37 <KREYREN_oftc> o.o
2023-01-22 21:17:10 <KREYREN_oftc> it is bcs i am installing it through http://ix.io/4lXp
2023-01-22 21:17:10 <KREYREN_oftc>  ?
2023-01-22 21:17:36 <psykose> probably
2023-01-22 21:17:44 <KREYREN_oftc> Lines 100~145 were followed
2023-01-22 21:17:51 <KREYREN_oftc> umm so like whaddya do?
2023-01-22 21:17:52 <KREYREN_oftc> O.o
2023-01-22 21:17:53 <psykose> i'm not reading that
2023-01-22 21:18:03 <KREYREN_oftc> it's 45 short lines!
2023-01-22 21:18:15 <psykose> all i did was run setup-desktop gnome and it worked
2023-01-22 21:19:18 <psykose> yeah you can't install libudev-zero and eudev at the same time
2023-01-22 21:19:21 <psykose> so just.. don't
2023-01-22 21:21:12 <KREYREN_oftc> ix.io/4lXq
2023-01-22 21:21:15 <KREYREN_oftc> output
2023-01-22 21:21:17 <KREYREN_oftc> hopefully
2023-01-22 21:21:24 <KREYREN_oftc> # setup-desktop gnome
2023-01-22 21:21:26 <KREYREN_oftc> psykose, ehh
2023-01-22 21:21:45 <KREYREN_oftc> i never used alpine beyond dockerfiles can you like dumb it down how to fix that
2023-01-22 21:21:53 <KREYREN_oftc> ~~by like a lot~~ @_@
2023-01-22 21:22:49 <psykose> apk del libudev-zero
2023-01-22 21:23:04 <KREYREN_oftc> done
2023-01-22 21:23:24 <KREYREN_oftc> what now sensei
2023-01-22 21:23:47 <KREYREN_oftc> do i like setup-desktop gnome
2023-01-22 21:24:08 <adhawkins> Hi all. Apologies if this is the wrong place to ask, but can't think of anywhere else.
2023-01-22 21:24:15 <KREYREN_oftc> O.o
2023-01-22 21:24:18 <KREYREN_oftc> 
ACTION invoked that

2023-01-22 21:24:30 <adhawkins> Trying to write a shell script where I use a variable to refer to the name of another variable.
2023-01-22 21:24:34 <adhawkins> Something like this: https://gist.github.com/adhawkins/312e7a10d415ff8e700c4422a494cb3c
2023-01-22 21:24:37 <KREYREN_oftc> seems to do gnome things, thanks! ^-^
2023-01-22 21:24:52 <adhawkins> However, I get './testscript.sh: line 12: syntax error: bad substitution"
2023-01-22 21:25:13 <adhawkins> Line 12 is the 'done' line, so I assume it refers to the previous line.
2023-01-22 21:25:25 <adhawkins> Anyone suggest how I can achieve this?
2023-01-22 21:25:32 <psykose> ${!varname} is not a thing
2023-01-22 21:25:48 <KREYREN_oftc> it's a thing in bash afaik, but he's using posix sh
2023-01-22 21:26:11 <adhawkins> Yes, it works in bash. Trying to do the same in ash (or sh on Alpine)
2023-01-22 21:26:42 <KREYREN_oftc> why? do you have like actual usecase or do you just expect us to give you an insane `eval` to show off? :p
2023-01-22 21:27:12 <adhawkins> I'm writing an entrypoint to a container, where the contents of a number of control files are stored in environment variables.
2023-01-22 21:27:21 <adhawkins> I want to populate those control files from the environment.
2023-01-22 21:28:05 <KREYREN_oftc> show up the assumed docker file and write to the header design intention which is: what you are doing, why and how do you expect to do it
2023-01-22 21:28:26 <adhawkins> All the variables are of the form 'NULLMAILER_REMOTE', 'NULLMAILER_DEFAULTDOMAIN' etc. It's be 'neater' if I could populate them inside a for loop iterating over 'remote defaultdomain ....' rather than referencing each variable and file individually
2023-01-22 21:31:00 <adhawkins> https://gist.github.com/adhawkins/3b448fb70bac84c693d0099b24282f17
2023-01-22 21:31:19 <adhawkins> I've put some comments in the bottom of 'ConfigureNullmailer' to show what I'd like to do
2023-01-22 21:31:46 <KREYREN_oftc> ehh so you are not doing that in a dockerfile?
2023-01-22 21:31:56 <adhawkins> No, it's the startup script for the container
2023-01-22 21:32:01 <KREYREN_oftc> what container
2023-01-22 21:32:36 <adhawkins> A container I'm writing to encapsulate borg and emborg. It needs to be able to send mail, so I'm using nullmailer. I want it to be relatively easy for them to configure nullmailer, so thought environment variables would be a good way to do it.
2023-01-22 21:33:58 <KREYREN_oftc> So the end-goal is to backup data right?
2023-01-22 21:34:09 <adhawkins> Yes, but I'm not sure that's relevant.
2023-01-22 21:34:47 <adhawkins> What I'm trying to achieve is an easy way to configure nullmailer. I could do that with lots of 'AddConfigLine' calls as I've done (there are 15 in total) but that looks messy.
2023-01-22 21:35:10 <KREYREN_oftc> checking
2023-01-22 21:35:30 <KREYREN_oftc> the shell file is kinda trash.. `local` is not a thing in POSIX sh check it against shellcheck.net later
2023-01-22 21:35:57 <adhawkins> If I can make my original script work, then I can easily adapt it to the full solution.
2023-01-22 21:36:17 <adhawkins> Focus on what I'm trying to achieve in the original gist, the rest is still in flux.
2023-01-22 21:40:55 <psykose> eval "echo \$$varname"
2023-01-22 21:41:25 <psykose> you can't do this without eval
2023-01-22 21:41:39 <KREYREN_oftc> adhawkins, try http://ix.io/4lXv
2023-01-22 21:41:45 <psykose> what you already wrote is ok though, no real point to add evals to it to save yourself a few lines of shell
2023-01-22 21:42:28 <KREYREN_oftc> note that eval IS INSANEEE and very easy to use for messing up a system with file="bad actor command" so be careful
2023-01-22 21:42:46 <psykose> > echo "$(eval echo
2023-01-22 21:43:07 <psykose> you don't get to insult people's shell scripts and then write that
2023-01-22 21:43:09 <KREYREN_oftc> psykose, ye attempt to avoid expansion of $NULLMAILER_
2023-01-22 21:43:21 <adhawkins> Thanks both, that works. I appreciate that eval isn't a good idea, which was why I was trying to find a way to do it without.
2023-01-22 21:43:32 <psykose> there's no way to get pointers in posix sh without eval
2023-01-22 21:43:36 <psykose> which is what this is
2023-01-22 21:43:36 <KREYREN_oftc> adhawkins, check it against the shellcheck.net
2023-01-22 21:43:41 <KREYREN_oftc> or you will have a really bad time later
2023-01-22 21:43:43 <psykose> so if you're sure you want it, then you have to use eval
2023-01-22 21:43:44 <KREYREN_oftc> and like fix the issues
2023-01-22 21:43:58 <KREYREN_oftc> +1 don't use eval
2023-01-22 21:44:04 <KREYREN_oftc> learn gnu guile
2023-01-22 21:44:06 <adhawkins> Is ${!xxx} 'safer' in bash?
2023-01-22 21:44:07 <KREYREN_oftc> 
ACTION runs to hide

2023-01-22 21:44:21 <KREYREN_oftc> adhawkins, ain't qualified to answer that
2023-01-22 21:45:04 <adhawkins> guile is a new one on me...
2023-01-22 21:45:08 <KREYREN_oftc> i would say bash scripting should always be avoided as if you can't do that in posix SH then trying to do that in bash is malpractice and wrong tool for the job
2023-01-22 21:45:33 <KREYREN_oftc> adhawkins, guile is the extensible language GNU uses in their software such as gimp, emacs-ish, guix etc..
2023-01-22 21:45:37 <adhawkins> I mean, I could write the whole thing very simply in Python I'm sure, but that needs python to be installed.
2023-01-22 21:46:02 <KREYREN_oftc> python's unmaintainable exponentially unmaintainable trash
2023-01-22 21:46:32 <adhawkins> I doubt that's a factor for a simple script like this.
2023-01-22 21:46:46 <psykose> can you go one sentence without being insulting of something or writing cringe
2023-01-22 21:47:01 <KREYREN_oftc> listen to the words of guy who has 14 years of experience in CS and had to rewrite 27 projects from python to other languages to avoid being a slave to technology each time the freaks make a new update
2023-01-22 21:47:04 <KREYREN_oftc> SAVE YOURSELFF
2023-01-22 21:47:19 <KREYREN_oftc> psykose, fine ;-;
2023-01-22 21:47:38 <adhawkins> 
ACTION has nearly 30 years experience. But lets not get into a willy-waving contest :)

2023-01-22 21:47:57 <KREYREN_oftc> 
ACTION stares

2023-01-22 21:48:17 <psykose> years of experience is an awesome cop out when you don't have an actual defence that what you're writing isn't wrong
2023-01-22 21:48:19 <psykose> now silence
2023-01-22 21:48:42 <KREYREN_oftc> so like i did the gnome thing and it installed it fine, but when i reboot it shows up CLI
2023-01-22 21:48:45 <psykose> adhawkins: i'm sure you know how to do it now :)
2023-01-22 21:48:51 <KREYREN_oftc> ehh i want it to like do GDM and let me login in gnome
2023-01-22 21:48:54 <psykose> i can't say if bash would be any safer, but what you have already is also fine
2023-01-22 21:49:01 <psykose> if you want to know anything else feel free to ask
2023-01-22 21:49:05 <adhawkins> Thanks psykose. Will do.
2023-01-22 21:49:27 <KREYREN_oftc> 
ACTION waives any responsibility and submitted the contribution UNDER NO WARRANTY AND LIABILITY permitted by law

2023-01-22 21:49:38 <adhawkins> :)
2023-01-22 21:50:06 <adhawkins> By the way, any 'simple' way to uppercase a string? I could just use tr, but if there's an 'ash' way then I'd prefer that.
2023-01-22 21:51:17 <psykose> it's indeed tr
2023-01-22 21:51:33 <psykose> | tr 'a-z' 'A-Z'
2023-01-22 21:51:47 <psykose> doesn't work on unicode of course
2023-01-22 21:51:49 <KREYREN_oftc> $ var="kreyren"; echo "${var^}"
2023-01-22 21:51:49 <KREYREN_oftc> Kreyren
2023-01-22 21:52:02 <KREYREN_oftc> not sure if that works in posix sh though
2023-01-22 21:52:27 <psykose> it does not
2023-01-22 21:53:36 <adhawkins> 	varname=$(echo "VAR_$var" | tr '[:lower:]' '[:upper:]')
2023-01-22 21:53:57 <adhawkins> Seems to do the trick. Don't care about unicode as these filenames are predefined for nullmailer
2023-01-22 21:55:49 <psykose> even better
2023-01-22 22:00:13 <KREYREN_oftc> adhawkins, $ var="kreyren"; case "${var:0:1}" in a) var="A${var:1}" ;; b) var="B${var:1}" ;; c) var="C${var:1}" ;; d) var="D${var:1}" ;; e) var="E${var:1}" ;; f) var="F${var:1}" ;; g) var="G${var:1}" ;; h) var="H${var:1}" ;; i) var="I${var:1}" ;; j) var="J${var:1}" ;; k) var="K${var:1}" ;; l) var="L${var:1}" ;; m) var="M${var:1}" ;; n) var="N${var:1}" ;; o) var="O${var:1}" ;; p) var="P${var:1}" ;; q) var="Q${var:1}" ;; r) 
2023-01-22 22:00:13 <KREYREN_oftc> Kreyren
2023-01-22 22:00:13 <KREYREN_oftc> var="R${var:1}" ;; s) var="S${var:1}" ;; t) var="T${var:1}" ;; u) var="U${var:1}" ;; v) var="V${var:1}" ;; w) var="W${var:1}" ;; x) var="X${var:1}" ;; y) var="Y${var:1}" ;; z) var="Z${var:1}" ;; esac; echo $var
2023-01-22 22:00:44 <adhawkins> I want to uppercase the whole string. And given that, I think I'll stick with 'tr' :)
2023-01-22 22:00:51 <KREYREN_oftc> 
ACTION nods

2023-01-22 22:01:21 <KREYREN_oftc> 
ACTION is tempted to make that infinite loop case statement, btu psykose wanted him to be less cringe

2023-01-22 22:02:08 <KREYREN_oftc> umm soo the gnome thing?
2023-01-22 22:02:09 <KREYREN_oftc> O.o
2023-01-22 22:02:40 <KREYREN_oftc> oh it fails bcs networking service fails and it has dependency on it
2023-01-22 22:02:42 <KREYREN_oftc> ehh
2023-01-22 22:04:24 <KREYREN_oftc> says it can't parse /etc/network/interfaces and that pathname doesn't exists 
2023-01-22 22:07:08 <KREYREN_oftc> oh i had to create that file
2023-01-22 22:07:09 <KREYREN_oftc> weird
2023-01-22 22:19:27 <adhawkins> psykose, KREYREN_oftc: Ended up with this: https://gist.github.com/adhawkins/c301390e434d158d0f9306b025957bd5
2023-01-22 22:19:36 <adhawkins> Passes shellcheck. Comments welcome!
2023-01-22 22:20:41 <KREYREN_oftc> adhawkins, declare `# shellcheck shell=sh # Written to comply with IEEE Std 1003.1-2017` below shebang
2023-01-22 22:20:54 <psykose> lol what
2023-01-22 22:21:02 <psykose> adhawkins: if you are happy with it then it's alright
2023-01-22 22:21:05 <KREYREN_oftc> replace shebang with `#!/usr/bin/env sh` if you want to avoid bites from nix users
2023-01-22 22:21:16 <psykose> bin/sh is valid on nix
2023-01-22 22:21:45 <KREYREN_oftc> lacks design intention at the header of the file so none really knows what are you doing, how and why -> less people willing to look at the code
2023-01-22 22:22:00 <psykose> what
2023-01-22 22:22:02 <KREYREN_oftc> bs that it passed shellcheck bcs it's using uppercase variables for things within function scope
2023-01-22 22:22:43 <KREYREN_oftc> psykose, /bin/sh is not valid, nixos just provides backwards compatibility which some users remove -> manage through environment instead of hardcoding binaries
2023-01-22 22:22:51 <psykose> it is valid
2023-01-22 22:22:54 <KREYREN_oftc> n
2023-01-22 22:23:12 <KREYREN_oftc> and imma guix user i have that intentionally removed to fix bad shebangs
2023-01-22 22:23:29 <psykose> bin/sh is literally specified in posix
2023-01-22 22:23:40 <adhawkins> KREYREN_oftc: Shellcheck outputs: 'No issues detected!'
2023-01-22 22:23:48 <KREYREN_oftc> psykose, not to my knowledge
2023-01-22 22:23:59 <psykose> adhawkins: you can ignore all of this terrible advice, the script is fine
2023-01-22 22:26:19 <KREYREN_oftc> weird that it passes on the variables.. the uppercase are reserved only for system-variables such as CURL, WGET, etc..
2023-01-22 22:26:34 <KREYREN_oftc> psykose, what terrible advice :C
2023-01-22 22:26:57 <psykose> there is no such thing as "system-variables" like "CURL"
2023-01-22 22:27:26 <KREYREN_oftc> psykose, https://pubs.opengroup.org/onlinepubs/9699919799/utilities/sh.html#tag_20_117_16 recommends /bin/sh or /usr/bin/sh.. fair
2023-01-22 22:28:22 <psykose> it actually says you cannot rely on /bin/sh or /usr/bin/sh and to require getconf PATH at installation time and to rewrite them
2023-01-22 22:28:46 <KREYREN_oftc> environment's more robust and doesn't cause issues for nix users though since it then recomments getconf for the reasons that usr/bin/env solves
2023-01-22 22:28:55 <psykose> in reality though /bin/sh works on every system
2023-01-22 22:29:05 <KREYREN_oftc> not on mine (%M)
2023-01-22 22:29:10 <psykose> that's a feature
2023-01-22 22:29:11 <KREYREN_oftc> *TM
2023-01-22 22:29:55 <KREYREN_oftc> psykose, so from the standard we can deduce that hardcoding binary paths is discouraged bcs some systems might return /usr/xpg4/bin/sh or alike thus environment-style management is preferred
2023-01-22 22:29:57 <KREYREN_oftc> :p
2023-01-22 22:30:44 <KREYREN_oftc> why is my gnome not working ;-;
2023-01-22 22:30:50 <psykose> yes, if you want to follow that then the correct shebang is not having one at al
2023-01-22 22:31:18 <KREYREN_oftc> > 	varname=$(echo "NULLMAILER_$file" | tr '[:lower:]' '[:upper:]')
2023-01-22 22:31:21 <KREYREN_oftc> how is this passing shebang
2023-01-22 22:31:24 <KREYREN_oftc> *shellcheck
2023-01-22 22:31:30 <KREYREN_oftc> should be double-quoted to prevent spliting
2023-01-22 22:31:33 <KREYREN_oftc> adhawkins, 
2023-01-22 22:31:36 <psykose> splitting of what
2023-01-22 22:31:48 <KREYREN_oftc> file="some thing"
2023-01-22 22:31:57 <psykose> it's a hardcoded list with no spaces
2023-01-22 22:31:58 <KREYREN_oftc> -> varname=some; thing
2023-01-22 22:32:10 <KREYREN_oftc> bad practice
2023-01-22 22:32:20 <psykose> it's literally a hardcoded list with no spaces
2023-01-22 22:32:31 <psykose> can you stop talking out of your ass
2023-01-22 22:32:56 <KREYREN_oftc> bad practice bcs you are supposed to assume that the list might at some point include item that has spaces in it
2023-01-22 22:33:16 <KREYREN_oftc> fix my gnome now? :(
2023-01-22 22:46:37 <KREYREN_oftc> hmm it starts into like black screen with a cursor line at the top left 
2023-01-22 22:47:26 <psykose> probably forgot to add yourself to video/input groups
2023-01-22 22:49:36 <KREYREN_oftc> what forgot it doesn't say that anywhere on https://wiki.alpinelinux.org/wiki/Gnome >_>
2023-01-22 22:51:28 <KREYREN_oftc> still doesn't seem to work
2023-01-22 22:51:29 <KREYREN_oftc> O.o
2023-01-22 22:51:35 <KREYREN_oftc> i was supposed to install gnome on root right
2023-01-22 22:51:42 <KREYREN_oftc> ye
2023-01-22 22:54:29 <KREYREN_oftc> xfce also doesn't work
2023-01-22 22:54:30 <KREYREN_oftc> @_@
2023-01-22 22:55:46 <valerius> switch to a tiling window manager to be elite
2023-01-22 22:55:57 <KREYREN_oftc> valerius, mehh
2023-01-22 22:56:23 <KREYREN_oftc> ooo lightdm works after reboot
2023-01-22 22:56:48 <KREYREN_oftc> and xfce4 launches from it :O
2023-01-23 07:52:12 <Forza> Good morning
2023-01-23 07:52:25 <Forza> I found the issue with the spontaneous reboots
2023-01-23 07:52:27 <Forza> https://dpaste.com/DM3LLSW7C
2023-01-23 07:52:55 <Forza> Seems a problem with iscsi and maybe ASM/ACPI?
2023-01-23 07:53:31 <psykose> just isci
2023-01-23 07:56:04 <Forza> Should I report this somewhere?
2023-01-23 07:56:31 <Forza> It's happeneing with 5.15 and 6.1.x kernels
2023-01-23 08:02:11 <psykose> somewhere, yet
2023-01-23 08:02:14 <psykose> yes*
2023-01-23 08:02:16 <psykose> where, i don't know
2023-01-23 08:03:53 <Forza> Seems a google group is dealing with open-iscsi parts in the kernel: https://groups.google.com/g/open-iscsi/about
2023-01-23 08:04:04 <Forza> But there ought to be a proper kernel mailing list? :D
2023-01-23 08:07:58 <psykose> it's probably that one
2023-01-23 08:12:44 <Forza> First useful thing I have used EFI for.. using efivar as pstore for kernel oupses
2023-01-23 08:12:46 <Forza> :)
2023-01-23 08:14:32 <psykose> how does one do that
2023-01-23 08:26:58 <Forza> psykose: how do you mean? Using pstore?
2023-01-23 08:27:03 <psykose> ye
2023-01-23 08:27:25 <Forza> efi_pstore module
2023-01-23 08:27:36 <Forza> check if pstore is mounted as /sys/fs/pstore
2023-01-23 08:29:07 <psykose> it is mounted
2023-01-23 08:29:20 <Forza> I should write some wiki abotu it =)
2023-01-23 08:29:25 <Forza> anything in it?
2023-01-23 08:33:42 <Forza>  -*-   Persistent store support
2023-01-23 08:35:16 <Forza>  http://dpaste.com/GZXT8NT59 relevant kernel config 
2023-01-23 08:43:11 <psykose> i don't have CONFIG_PSTORE_BLK but i do have the rest
2023-01-23 08:43:17 <psykose> but afaict that's just for a block device
2023-01-23 08:43:26 <psykose> i guess if i had an oops i would see something in it then
2023-01-23 08:43:34 <psykose> i haven't seen one in a real long time
2023-01-23 08:43:38 <psykose> thanks, should be fine then
2023-01-23 09:06:57 <ikke> Forza: how did you find out / confirm? 
2023-01-23 09:07:04 <ikke> Debug kernel? 
2023-01-23 09:20:52 <frag> what is the "Enable the built-in emacs(1) command line editor" in dash? thought it was ^e to jump to end of line etc
2023-01-23 09:27:23 <psykose> it is
2023-01-23 09:27:48 <psykose> the usual way anyone uses the shell i guess
2023-01-23 09:27:53 <psykose> -V / vi is a bit different
2023-01-23 09:28:18 <psykose> both are only available with --with-libedit, otherwise there is nothing
2023-01-23 09:31:52 <frag> psykose: when compiling? why is it not included in repo dash?
2023-01-23 09:31:58 <psykose> why would it be
2023-01-23 09:32:14 <psykose> nobody uses dash interactively
2023-01-23 09:32:19 <frag> also whats the default font for alpine (is there such a thing)?
2023-01-23 09:32:27 <psykose> there isn't
2023-01-23 09:32:35 <psykose> well
2023-01-23 09:32:45 <psykose> not sure if anyone tweaked the fontconfigs to set anything by defaulty
2023-01-23 09:32:55 <psykose> doesn't seem so
2023-01-23 09:32:57 <frag> if i uninstall the intentionally installed ones i get a bad font :p
2023-01-23 09:33:02 <frag> (in terminal)
2023-01-23 09:33:12 <psykose> because it resolves to some other font still installed
2023-01-23 09:33:25 <frag> mhmhm
2023-01-23 09:34:12 <frag> also link is compiled without the -g (graphical) option in repo?
2023-01-23 09:34:26 <psykose> link?
2023-01-23 09:34:34 <frag> links*
2023-01-23 09:34:39 <psykose> i think he was busy finding zelda
2023-01-23 09:34:51 <psykose> it is --without-x so probably not
2023-01-23 09:35:31 <frag> aha, otherwise it doesnt work in console phaps?
2023-01-23 09:36:28 <KREYREN_oftc> My GDM doesn§t work, http://ix.io/4lZt
2023-01-23 09:36:58 <KREYREN_oftc> teres-1 system (Allwinner 64 aarch64)
2023-01-23 09:38:10 <KREYREN_oftc> and if i do gnome-session it says that i can't open display
2023-01-23 09:38:19 <psykose> Failed to allocate onscreen framebuffer for /dev/dri/card0: Failed to create dumb drm buffer: Out of memory
2023-01-23 09:38:27 <psykose> out of vram i guess
2023-01-23 09:39:28 <KREYREN_oftc> O.o gnome works on ubuntu and debian though
2023-01-23 09:39:59 <psykose> frag: links -g works anyway though, you should've tested it
2023-01-23 09:42:12 <frag> psykose: i get "Graphics not enabled when compiling"
2023-01-23 09:42:51 <frag> links-2.28-r0
2023-01-23 09:43:01 <psykose> ah
2023-01-23 09:43:11 <psykose> it's because there's a links-graphics in testing/ with it
2023-01-23 09:43:18 <psykose> and it provides the same thing
2023-01-23 09:43:20 <psykose> epic
2023-01-23 09:43:48 <frag> oh cool
2023-01-23 09:48:20 <KREYREN_oftc> and the system has 80% free space so around 12GB
2023-01-23 09:48:31 <KREYREN_oftc> oh around 23GB
2023-01-23 10:05:34 <frag>  which fonts do you "need"?
2023-01-23 10:06:17 <psykose> the ones that display the characters you see
2023-01-23 10:06:36 <psykose> generally in both a monospace and non-monospace form
2023-01-23 10:07:08 <psykose> with perhaps both a serif and a sans variant if you care, and real italics/bolds
2023-01-23 10:20:09 <frag> i guess the dejavu font is kinda necessary/practical
2023-01-23 10:20:37 <frag> tried liberation, looks a bit weird in browser sometimes
2023-01-23 10:52:03 <aron> is there any way to know which kernel modules were loaded from the initramfs, and not from "normal" fs?
2023-01-23 10:54:03 <aron> my current problem: kernel stuck at the very beginning ("Starting kernel ...") with the initramfs generated by setup-alpine
2023-01-23 10:54:45 <aron> however, if I put my crafted initramfs with all modules included, it boots okay
2023-01-23 10:56:10 <aron> vmlinuz is not altered, only initramfs (and modloop is not mounted)
2023-01-23 11:09:31 <psykose> don't think there's tracing, but you could diff the modules
2023-01-23 11:09:38 <psykose> unless your "all" modules is.. literally all the modules
2023-01-23 11:10:39 <frag> mononoki looks kinda nice/minimalist ^^
2023-01-23 11:10:44 <psykose> you could boot with working initramfs, `lsmod`, and compare that set to the initramfs
2023-01-23 11:33:44 <aron> okay, thanks
2023-01-23 11:33:53 <aron> and yes, "all" modules is really all-all
2023-01-23 11:35:10 <aron> lsmod lists builtin modules as well, right?
2023-01-23 11:38:40 <psykose> not sure
2023-01-23 11:39:06 <psykose> ah
2023-01-23 11:39:09 <psykose> google saves ya
2023-01-23 11:39:10 <psykose> cat /lib/modules/$(uname -r)/modules.builtin
2023-01-23 11:39:13 <psykose> list of builtins
2023-01-23 11:40:39 <aron> hm, it seems builtins are not in lsmod
2023-01-23 11:41:38 <aron> yeah
2023-01-23 11:42:40 <aron> is there any preferred way to include more modules in initramfs?
2023-01-23 11:43:02 <aron> or I should make my own "feature" with my wanted modules listed in?
2023-01-23 11:43:09 <ikke> that would make sense
2023-01-23 11:54:08 <aron> bummer, it must be something else
2023-01-23 11:54:11 <aron> it still stuck
2023-01-23 11:54:22 <aron> (the new "feature" worked tho, just not enough)
2023-01-23 13:08:16 <KREYREN_oftc> My XFCE doesn't have window manager things around.. they are like decoration-less windows O.o
2023-01-23 13:24:23 <KREYREN_oftc> when i try to open the window manager settings in settings it tells me "these settings cannot work with your current window manager"
2023-01-23 13:24:26 <KREYREN_oftc> O_o
2023-01-23 13:24:50 <grawity> is the window manager running at all?
2023-01-23 13:24:57 <grawity> the xfwm4 process specifically
2023-01-23 13:25:35 <KREYREN_oftc> grawity, doesn't seem to be from `ps -el | grep xfwm4`
2023-01-23 13:25:46 <grawity> yeah, then run it
2023-01-23 13:25:54 <grawity> try starting it from a terminal in case it doesn't work
2023-01-23 13:26:34 <KREYREN_oftc> that just closed xfce4 panel and everything on the desktop
2023-01-23 13:26:36 <KREYREN_oftc> O.o
2023-01-23 13:26:48 <KREYREN_oftc> (installed through setup-desktop xfce and then launched from lightdm btw)
2023-01-23 13:27:18 <grawity> did it log you out of xfce entirely?
2023-01-23 13:27:48 <KREYREN_oftc> nah just i have the xfce4 background and desktop icons and i can summon terminal if i right click and press 'open terminal here'
2023-01-23 13:28:38 <grawity> with decorations this time?
2023-01-23 13:28:42 <KREYREN_oftc> nope
2023-01-23 13:28:55 <grawity> try to run the WM from a terminal window
2023-01-23 13:29:22 <KREYREN_oftc> tried the startx and that gets me a black screen with xfce4-panel and no cursor or othe rinput
2023-01-23 13:30:21 <KREYREN_oftc> errs are: no GPG agent found, no SSH authentification agent found, DRM_IOCTL_MODE_CREATE_DUMB failed: out of memory
2023-01-23 13:30:31 <KREYREN_oftc> xfsettingsd: No window manager registered on screen 0
2023-01-23 13:34:04 <KREYREN_oftc> https://gitlab.com/postmarketOS/pmaports/-/issues/1909
2023-01-23 13:34:08 <KREYREN_oftc> apparently it can be bcs of zram?
2023-01-23 13:34:09 <KREYREN_oftc> O.o
2023-01-23 13:35:11 <KREYREN_oftc> hmm says that it's only on edge i guess i can try downgrade
2023-01-23 13:35:16 <grawity> that DRM error really sounds like some kind of issue with GPU drivers (not sure whether it's more the kernel side or the Mesa side though)
2023-01-23 13:35:40 <psykose> it does just sound like some arm gpu memory thing with not enough memory reserved for it
2023-01-23 13:35:51 <psykose> i think you'd get this on e.g. the rpi4 with too small a gpu_mem for instance
2023-01-23 13:36:02 <psykose> idk how gpu memory works on other arm devices like here
2023-01-23 13:36:51 <psykose> aside from that could also be mesa itself, for which you can build older/newer versions yourself if you want
2023-01-23 13:38:02 <KREYREN_oftc> checking
2023-01-23 13:40:00 <KREYREN_oftc> umm how do i downgrade from edge t v3.17 ?
2023-01-23 13:40:21 <KREYREN_oftc> apk upgrade --update-cache --available?
2023-01-23 13:41:32 <ikke> Yes 
2023-01-23 13:42:25 <KREYREN_oftc> umm should /etc/apk/repositories include the 'releases' thing?
2023-01-23 13:42:52 <spazialecontinentale[m]> hi, i'm building a rust project that has openssl as dependencies and i keep getting segmentation fault.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JJWrZldDAgeznyHsCqvxqRGy>)
2023-01-23 13:44:37 <psykose> cargo -Vv
2023-01-23 13:46:32 <psykose> export RUSTFLAGS="-Ctarget-feature=-crt-static" if it's rustup
2023-01-23 13:49:54 <spazialecontinentale[m]> <psykose> "cargo -Vv" <- cargo 1.68.0-nightly (985d561f0 2023-01-20)... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/rCOdqhCupoOWwqNRNBvylcHO>)
2023-01-23 13:51:59 <KREYREN_oftc> hmm i get a black screen on latest-stable x_x
2023-01-23 13:52:14 <JanJ[m]> does anyone know, is there a diagramming app available on Alpine? like yEd?
2023-01-23 13:54:18 <JanJ[m]> hmz, maybe draw.io is what i need
2023-01-23 13:54:29 <spazialecontinentale[m]> <psykose> "export RUSTFLAGS="-Ctarget-..." <- seems to works thanks!
2023-01-23 13:54:42 <KREYREN_oftc> oh it loads now but same issue
2023-01-23 13:54:43 <KREYREN_oftc> hmm
2023-01-23 13:55:12 <lucidiot> JanJ[m]: there's dia in community
2023-01-23 13:57:26 <JanJ[m]> lucidiot: found it! thanks :)
2023-01-23 13:58:17 <KREYREN_oftc> How do i install a linux-sunxi-5.19.5-r0.apk from https://arvanta.net/mps/aports-local/aarch64/ without doing it manually like a caveman?
2023-01-23 13:58:22 <aron> about that kernel stuck during early boot: my module listing script had a flaw, lsmod list modules with "_" but actual names are with "-"; thus it boots perfectly now, yay
2023-01-23 13:58:31 <KREYREN_oftc> when i already have 6.0.8-0-sunxi kernel @_@
2023-01-23 13:59:30 <aron> on the other hand... if kernel missing mandatory modules in this case, shouldn't they are builtin?
2023-01-23 14:02:33 <KREYREN_oftc> `# apk add linux-sunxi=5.19.5-r0` seems to do stuff but complains about breaking the world
2023-01-23 14:08:15 <KREYREN_oftc> the internet is just bunch of confused docker users when i search for answers to that question @_@
2023-01-23 14:08:16 <KREYREN_oftc> *help*
2023-01-23 14:09:09 <aron> KREYREN_oftc: it should print a bit more stuff other than breaking world
2023-01-23 14:09:56 <aron> it maybe contain some info which is meaningless for you, but not for other ppl
2023-01-23 14:11:44 <KREYREN_oftc> aron, http://ix.io/4m0h
2023-01-23 14:14:27 <KREYREN_oftc> that's all it prints @_@
2023-01-23 14:14:54 <aron> and this is the result of # apk add linux-sunxi=5.19.5-r0
2023-01-23 14:15:10 <KREYREN_oftc> yes
2023-01-23 14:15:16 <KREYREN_oftc> i tried to add --force-overwrite too
2023-01-23 14:16:14 <KREYREN_oftc> do i like remove the old kernel or something first?
2023-01-23 14:16:14 <KREYREN_oftc> @_@
2023-01-23 14:17:25 <aron> (obviously, this is not something I'd like to try on my working system :P)
2023-01-23 14:17:54 <KREYREN_oftc> it's testing system to develop an alpine image for the thing so if it breaks i can just reflash the sdcard~
2023-01-23 14:18:09 <aron> removing the old kernel first seems to be a good start
2023-01-23 14:19:03 <aron> unless some pkg depends on it
2023-01-23 14:19:09 <KREYREN_oftc> same err
2023-01-23 14:19:28 <aron> "apk del linux-sunxi" yields the same error?
2023-01-23 14:19:32 <KREYREN_oftc> ye
2023-01-23 14:20:04 <KREYREN_oftc> it removed it successfully it ain't in the apk list -I thing
2023-01-23 14:21:10 <aron> do you have linux-sunxi-dev ?
2023-01-23 14:21:34 <KREYREN_oftc> nope
2023-01-23 14:22:30 <aron> check this:   grep linux-sunxi /etc/apk/world
2023-01-23 14:22:53 <KREYREN_oftc> aron, returns blank
2023-01-23 14:23:14 <aron> and it still does not want to install the 5.19 ?
2023-01-23 14:23:19 <KREYREN_oftc> ye
2023-01-23 14:23:23 <aron> odd
2023-01-23 14:23:28 <KREYREN_oftc> o.o
2023-01-23 14:23:53 <aron> is that 5.19.7 really available for apk?
2023-01-23 14:24:09 <aron> ->   apk search linux-sunxi
2023-01-23 14:24:34 <KREYREN_oftc> prints only 6.0.8-r0
2023-01-23 14:24:44 <aron> oh, that's a problem
2023-01-23 14:24:47 <KREYREN_oftc> o.o
2023-01-23 14:24:57 <aron> after  apk update   try search again
2023-01-23 14:25:19 <KREYREN_oftc> same shows only 6.0.8
2023-01-23 14:25:46 <aron> then your /etc/apk/repository lack the repo where 5.19.7 lives
2023-01-23 14:25:59 <KREYREN_oftc> https://arvanta.net/mps/aports-local/aarch64/ the repo has it though
2023-01-23 14:26:10 <qsx> is there any sensible documentation on the features of busybox ash?  it’s definitely more than POSIX sh, but not really bash either
2023-01-23 14:26:16 <KREYREN_oftc> http://ix.io/4m0J
2023-01-23 14:27:13 <qsx> shellcheck isn’t helpful, too, as it emits “yeah this might be a problem” warnings when using #!/bin/sh, and interpreting #!/bin/ash as dash which is known not to have some features that busybox ash does
2023-01-23 14:28:05 <KREYREN_oftc> qsx, sounds like a bug for shellcheck to handle that
2023-01-23 14:28:45 <aron> KREYREN_oftc: do you have the public key for the repo? i.e. apk update should whine if keys are missing
2023-01-23 14:28:56 <qsx> KREYREN_oftc: maybe.  but can’t blame them if i haven’t found any sensible documentation on busybox ash’s features as well
2023-01-23 14:29:04 <aron> s/whine/complain/
2023-01-23 14:29:04 <alpine-meetbot> aron meant to say: KREYREN_oftc: do you have the public key for the repo? i.e. apk update should complain if keys are missing
2023-01-23 14:29:20 <KREYREN_oftc> aron, it doesn't seem to whine and the pubkey was added during system build
2023-01-23 14:29:26 <qsx> (as a friend of mine often says, “there is always the _real_ documentation”, i.e., the source code, but oof.)
2023-01-23 14:29:33 <KREYREN_oftc> qsx, https://man.archlinux.org/man/busybox.1.en the only docs i knwo of
2023-01-23 14:30:30 <aron> KREYREN_oftc: can you post your /etc/apk/repository file?
2023-01-23 14:31:02 <KREYREN_oftc> aron, /etc/apk/repository doesn't exists, do you want /etc/apk/repositories?
2023-01-23 14:31:14 <aron> yeah, it way a typo, sorry
2023-01-23 14:31:18 <KREYREN_oftc> http://ix.io/4m0J
2023-01-23 14:31:48 <aron> it's blank :(
2023-01-23 14:32:11 <KREYREN_oftc> O.o
2023-01-23 14:32:55 <KREYREN_oftc> http://ix.io/4m0l
2023-01-23 14:33:07 <KREYREN_oftc> the '# Thing' is not part of the file.. just added bcs ix borked
2023-01-23 14:33:28 <aron> no prob, it's a comment anyway
2023-01-23 14:35:33 <aron> it seems the index file on that server is broken
2023-01-23 14:35:42 <aron> or... at least *incomplete*
2023-01-23 14:35:51 <KREYREN_oftc> oh
2023-01-23 14:36:27 <KREYREN_oftc> umm so i guess imma installing that manually @_@
2023-01-23 14:36:37 <aron> well
2023-01-23 14:36:40 <aron> partially manually
2023-01-23 14:36:50 <aron> wget the apk, and feed it to apk
2023-01-23 14:37:11 <KREYREN_oftc> mhm
2023-01-23 14:38:31 <aron> manually unpacking a package is usually a bad idea, but if you install a downloaded pkg with apk gives you a "rollback" feature, and you won't mess your system up _that_ much
2023-01-23 14:39:30 <aron> apk add path/to/linux-sunxi.apk
2023-01-23 14:39:49 <KREYREN_oftc> ye i meant to do that that way :D
2023-01-23 14:43:07 <aron> it's the lesser evil, I think, but there are other ppl, who might think otherwise
2023-01-23 14:50:13 <qsx> oh dear.  i just had a glance at the posix standard and the feature of bb ash/bash i’m using (string indexing, i.e., ${parameter:n[:m]}, for some numbers n and m) is indeed complicated in pure posix sh, as posix expr doesn’t even support `expr substr`
2023-01-23 14:51:11 <psykose> :)
2023-01-23 14:51:29 <qsx> (i’ve spent way too much time with the posix standard for my bachelor thesis and i stand my point that it isn’t a standard, it’s an accident report)
2023-01-23 14:51:53 <ikke> Heh
2023-01-23 14:51:58 <psykose> that's most standards
2023-01-23 14:52:11 <psykose> they describe a current state as a baseline more than anything
2023-01-23 14:52:39 <psykose> and then any potential addition gets rejected as "x already implemented this slightly different and it would be incompatible"
2023-01-23 14:52:42 <qsx> maybe, but posix is an exceptionally impressive example of that
2023-01-23 14:52:49 <psykose> reading the austin group requests is a trial in misery
2023-01-23 14:53:10 <qsx> “computers suck” “but you study computer science!” “exactly.”
2023-01-23 14:53:52 <psykose> heh
2023-01-23 14:54:08 <KREYREN_oftc> andd it doesn't boot anymore
2023-01-23 14:54:08 <KREYREN_oftc> xD
2023-01-23 15:04:44 <aron> KREYREN_oftc: with the sunxi 5.17?
2023-01-23 15:24:06 <KREYREN_oftc> aron, ehhh 5.19.10-r0
2023-01-23 15:24:43 <aron> yeah, right, that one
2023-01-23 15:25:35 <aron> your initramfs was regenerated?
2023-01-23 15:25:53 <KREYREN_oftc> ummm no idea
2023-01-23 15:26:00 <KREYREN_oftc> i use extlinux and jsut like populate it with thing
2023-01-23 15:26:07 <hodapp> as someone who also has a Comp Sci degree, yeah, computers do kinda suck but right now they're the fastest known way of doing computation
2023-01-23 15:26:47 <KREYREN_oftc> hodapp, you meant the most resource efficient right? Who gave you the degree to describe something in computer science as 'fastest' !
2023-01-23 15:26:56 <KREYREN_oftc> 
ACTION grins

2023-01-23 15:27:30 <aron> hodapp: s/kinda//
2023-01-23 15:27:30 <alpine-meetbot> aron thinks hodapp meant to say: as someone who also has a Comp Sci degree, yeah, computers do  suck but right now they're the fastest known way of doing computation
2023-01-23 15:27:51 <hodapp> KREYREN_oftc: fastest in terms of the fraction of the estimated time I have alive
2023-01-23 15:27:53 <hodapp> or something
2023-01-23 15:28:13 <KREYREN_oftc> 
ACTION stares and grins

2023-01-23 15:30:34 <aron> (ppl tend to develop even larger bloatwares, and last year's bloatware today is just a tiny component)
2023-01-23 15:30:43 <invoked> imagine the amount of human capital that has been spent theming desktop environments
2023-01-23 15:32:11 <aron> haha, yeah, they also redo GUIs all the time
2023-01-23 15:32:22 <psykose> time well spent
2023-01-23 15:32:44 <KREYREN_oftc> 5.18.10-r0 also doesn't work
2023-01-23 15:32:44 <KREYREN_oftc> hmmm
2023-01-23 15:33:07 <frag> hmm xmodmap works better with keykombinations? my terminal uses ^+ to increase font size, i have altgr+p=+  .. ctrl+altgr+p works when using xmodmap but not when using xkb, am i missing something?
2023-01-23 15:33:51 <aron> KREYREN_oftc: can you describe how it does not work?
2023-01-23 15:35:39 <KREYREN_oftc> aron, just black screen on boot
2023-01-23 15:36:03 <aron> do you have a serial output?
2023-01-23 15:36:40 <aron> it may emit something useful
2023-01-23 15:37:16 <KREYREN_oftc> aron, didn't try.. the sunxi wiki says to use the generic ARM image for alpine https://linux-sunxi.org/Bootable_OS_images#Alpine_Linux is there like a thing for it?
2023-01-23 15:37:23 <aron> also: did you edit your dtb files?
2023-01-23 15:37:38 <KREYREN_oftc> nah it spits out bunch of permission erros about changing ownership x>X
2023-01-23 15:43:16 <aron> on what files?
2023-01-23 15:44:35 <scorpion2185[m]1> I am trying swhkd does it use the same config as sxhkd?
2023-01-23 15:46:02 <scorpion2185[m]1> ERROR swhks] Server is already running!
2023-01-23 15:46:52 <scorpion2185[m]1> ERROR swhkd] Failed to send command to swhks through IPC.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/aSzVkuRYJkQcdiPGndVujNHU>)
2023-01-23 15:50:31 <KREYREN_oftc> aron, seems to be on all files in the /boot/dtsthing
2023-01-23 15:51:21 <minimal> aron: look at "dmesg" output, assuming the loaded module write anything there then it will be obvious which load from initramfs and which from full OS
2023-01-23 15:51:27 <aron> KREYREN_oftc: sounds reasonable, replacing kernel will overwrite all of them
2023-01-23 15:52:27 <KREYREN_oftc> 
ACTION broke his mounts again and has to reboot

2023-01-23 15:53:16 <minimal> aron: though you might need to add "log_buf_len=32768" to the cmdline to have "dmesg" store the full set of boot output
2023-01-23 15:54:04 <aron> minimal: thanks, but I have the module list now, I will build my own kernel, and those will be builtin
2023-01-23 15:55:55 <kreyren> 
ACTION back

2023-01-23 15:59:32 <kreyren> aron, i will wait for the guy who makes the kernel for the pcb he usually has lot of info O.o or like is there an alpine kernel i can use for Allwinner A64 ?
2023-01-23 16:04:03 <aron> for the record, I have Allwinner H616
2023-01-23 16:04:16 <kreyren> hm o.o
2023-01-23 16:04:35 <aron> but edge's linux-edge works with it
2023-01-23 16:04:49 <kreyren> i will try
2023-01-23 16:05:08 <kreyren> do you use any firmware on it?
2023-01-23 16:05:54 <aron> what do you mean?
2023-01-23 16:06:01 <kreyren> linux-firmware
2023-01-23 16:06:11 <kreyren> like the non-free things
2023-01-23 16:06:25 <aron> nothing is installed
2023-01-23 16:06:34 <kreyren> hm O.o
2023-01-23 16:08:51 <aron> at least, ls /lib/firmware/ is empty
2023-01-23 16:09:45 <aron> yeah, linux-firmware-none is the only installed pkg
2023-01-23 16:09:54 <kreyren> interesting
2023-01-23 16:10:38 <aron> I'm fine with that
2023-01-23 16:13:52 <kreyren> ye me too that's one of the reasons why i got teres-1 :D
2023-01-23 16:14:47 <witcher01> i've installed postgresql15 but running `postgres` as user `postgres` is telling me it doesn't know where the config file/data dir is. should this be happening?
2023-01-23 16:15:51 <aron> kreyren: what is that?
2023-01-23 16:16:36 <kreyren> aron, the system i am trying to get alpine to work https://www.olimex.com/Products/DIY-Laptop/
2023-01-23 16:16:45 <kreyren> free software and free hardware designs designed in KiCAD
2023-01-23 16:16:59 <aron> oh, an olimex board
2023-01-23 16:17:38 <aron> A64 is newer or older than H616?
2023-01-23 16:17:38 <kreyren> mhm
2023-01-23 16:17:45 <kreyren> ehh i think newer?
2023-01-23 16:17:48 <kreyren> 
ACTION goes to check

2023-01-23 16:17:54 <invoked> allwinner has been a bit sus
2023-01-23 16:18:24 <kreyren> invoked, hm?
2023-01-23 16:18:44 <invoked> look into allwinner's history.  i'm not alleging anything beyond what is already known.
2023-01-23 16:19:03 <kreyren> the backdoor in linux kernel and them violating GPLv3?
2023-01-23 16:19:49 <kreyren> aron, seems that yours is newer
2023-01-23 16:19:52 <invoked> among other things. but then everything is going fabless so you can never really know.
2023-01-23 16:19:54 <kreyren> 28 nm vs 40nm
2023-01-23 16:20:33 <kreyren> invoked, what other things O.o
2023-01-23 16:21:00 <kreyren> 
ACTION just uses A64 for prototypes and is now figuring out a new PCB for teres-1 likely NXP iMX8 TUSKLA

2023-01-23 16:21:06 <invoked> perform your due diligence, kreyren 
2023-01-23 16:21:20 <kreyren> invoked, i did and figured out only the linux backdoor and GPLv3 violation
2023-01-23 16:22:02 <kreyren> the backdoor not somethign that concerns me as they have documented root elevation for development, GPLv3 .. kinda ain't designed for hardware
2023-01-23 16:22:19 <aron> kreyren: then you have a chance your board will work with latest linux-lts, but let me check something
2023-01-23 16:22:33 <kreyren> o.o
2023-01-23 16:23:15 <aron> yup, with linux 5.14 you are good
2023-01-23 16:23:24 <kreyren> 5.14 and above?
2023-01-23 16:23:26 <aron> see: https://linux-sunxi.org/Linux_mainlining_effort
2023-01-23 16:23:29 <invoked> the nxp imx* are, from what everyone knows, objectively better on the open source end of things
2023-01-23 16:23:30 <aron> yeah
2023-01-23 16:23:43 <kreyren> https://github.com/armbian/build/search?q=teres seems to meet with the patches armbian is using too
2023-01-23 16:24:04 <kreyren> invoked, ye seems to be the best option in terms of open-source
2023-01-23 16:24:11 <aron> I always prefer mainline linux, not the patched ones
2023-01-23 16:24:16 <kreyren> me too
2023-01-23 16:24:19 <aron> (who knowns what's inside)
2023-01-23 16:24:23 <kreyren> well i prefer linux-libre
2023-01-23 16:24:36 <aron> don't get me wrong, mainline is not guarantee for anything either
2023-01-23 16:24:53 <aron> but I trust it more
2023-01-23 16:25:30 <kreyren> aron, seems that since 5.17  bcs of things for MBUS
2023-01-23 16:25:57 <aron> or that one :)
2023-01-23 16:26:38 <aron> if you need MBUS
2023-01-23 16:26:51 <kreyren> 
ACTION has no idea what MBUS is

2023-01-23 16:27:02 <aron> I don't know if it has been backported to LTS (5.15)
2023-01-23 16:27:41 <aron> looks like some sort of smart metering thing
2023-01-23 16:28:00 <aron> probably uninteresting for your laptop project
2023-01-23 16:28:12 <kreyren> hm O.o
2023-01-23 16:34:31 <wakejagr> I'm following the wiki to build a custom ISO using mkimage.sh and I see how to add package(s) by creating a new profile, but I'm not finding a way to add a file to the custom ISO. Is there documentation somewhere that might help me figure out a method?
2023-01-23 16:35:38 <ikke> The expectation is that all files come from packages
2023-01-23 16:36:01 <wakejagr> ikke: i wondered if that might be the case. thank you for the confirmation
2023-01-23 17:00:24 <kreyren> hmm armbian is using custom kernel too over the linux-lts
2023-01-23 18:31:07 <aron> kreyren: one of the many reasons I don't like it ;)
2023-01-23 19:41:34 <scorpion2185[m]1> how can I install test man page?
2023-01-23 19:42:59 <ikke> apk add man-pages-posix
2023-01-23 19:43:05 <ikke> https://pkgs.alpinelinux.org/contents?file=test*&path=%2Fusr%2Fshare%2Fman*&name=&branch=edge&repo=main&arch=ppc64le
2023-01-23 19:43:23 <ikke> but note that most of the time, test is a shell built-in
2023-01-23 19:46:14 <scorpion2185[m]1> I can't use micro clipboard on river
2023-01-23 19:50:31 <scorpion2185[m]1> i have micro wayland app, i tried all micro values for clipboard
2023-01-23 19:50:42 <scorpion2185[m]1> it doesn't work outside micro
2023-01-23 21:00:31 <kreyren> aron, got a new kernel from the guy in olimex room and tried linux-lts both have the issue O.o
2023-01-23 21:00:54 <kreyren> this is on ehh latest-stable.. so probably something something mesa?
2023-01-23 21:15:31 <aron> kreyren: have you checked whether it is a devicetree issue?
2023-01-23 21:15:45 <kreyren> aron, devicetree in what sense?
2023-01-23 21:15:52 <aron> note that, replacing the kernel (or just even upgrade it) will overwrite dtbs
2023-01-23 21:16:09 <kreyren> ye that seems to have worked fine O.o
2023-01-23 21:16:32 <aron> dtb = devicetree blob
2023-01-23 21:16:53 <aron> do you have shell access?
2023-01-23 21:17:13 <kreyren> yep
2023-01-23 21:17:19 <aron> great
2023-01-23 21:17:50 <aron> then go to /sys/firmware/devicetree/base/soc
2023-01-23 21:18:15 <aron> and look around that related to video
2023-01-23 21:18:20 <aron> i.e. hdmi
2023-01-23 21:18:40 <aron> video-codec
2023-01-23 21:18:55 <aron> you what I'm saying
2023-01-23 21:19:26 <aron> these directories must have a "status" file containing "okay" to be enabled
2023-01-23 21:19:59 <aron> if some of them is not "okay", then linux will not enable hdmi (or whatever video) output
2023-01-23 21:20:30 <kreyren> umm i have black screen again after reboot @_@ i will recreate the image
2023-01-23 21:20:48 <aron> what I would do
2023-01-23 21:20:55 <aron> is copy the original dtb
2023-01-23 21:21:06 <aron> do the kernel replace
2023-01-23 21:21:24 <kreyren> aron, umm there ain't functional kernel on alpine though
2023-01-23 21:21:32 <aron> and copy back the dtb, overwriting th existing
2023-01-23 21:21:43 <aron> hm, ok
2023-01-23 21:21:55 <aron> but do you have a functional kernel from somewhere else?
2023-01-23 21:21:58 <kreyren> it only works on armbian and that seems to be doing some weird wizardry
2023-01-23 21:22:13 <kreyren> i asked armbian guy who worked on that to provide additional info 
2023-01-23 21:22:17 <aron> you can take the dtb from it as well
2023-01-23 21:22:36 <kreyren> o.o
2023-01-23 21:22:56 <aron> dtbs seem to compatible with a quite large spectrum of kernels
2023-01-23 21:23:08 <aron> they can break, but it worth a try
2023-01-23 21:23:28 <kreyren> i need it to work from alpine though as i want to make a CD to generate an img file for the system though
2023-01-23 21:24:09 <aron> whatever you want to achive, compiling a custom dtb is mandatory
2023-01-23 21:24:26 <aron> the stock ones are 99% is not what you want
2023-01-23 21:24:39 <kreyren> o_o
2023-01-23 21:25:29 <aron> take them as example of what your sbc is capable of
2023-01-23 21:25:35 <aron> most of things are disabled
2023-01-23 21:26:33 <aron> because those function can collide (i.e. multiple functionality won't work through the physical pin)
2023-01-23 21:26:48 <aron> **same physical pin
2023-01-23 21:28:04 <aron> as far as I can tell, your symptoms sounds like dtb/devicetree issue
2023-01-23 21:30:27 <kreyren> i guess
2023-01-23 21:30:29 <kreyren> trying
2023-01-23 21:33:43 <kreyren> aron18> these directories must have a "status" file containing "okay" to be enabled
2023-01-23 21:33:43 <kreyren> 18<aron18> if some of them is not "okay", then linux will not enable hdmi (or whatever video) output
2023-01-23 21:33:52 <kreyren> ehh the hdmi@1ee000/status has disabled
2023-01-23 21:34:07 <kreyren> hdmi-phy doesn't have status file
2023-01-23 21:34:21 <kreyren> gpu also doesn't
2023-01-23 21:34:39 <kreyren> video-codec doesn't
2023-01-23 21:35:06 <aron> yeah, then do not expect to see anything via hdmi ;)
2023-01-23 21:35:31 <kreyren> o.o
2023-01-23 21:37:20 <aron> afaik, there is 2 ways to fix this: (1) playing with dtb overlays, but I can't help you with that (2) decompile the dtb to dts, edit it, and recompile it to dtb -- I did the latter, might be not that elegant, but works
2023-01-23 21:38:20 <kreyren> umm i noted those, but prefer to wait for the guy in armbian to provide info and the guy from alpine who works on porting teres-1 to check those i guess O.o
2023-01-23 21:38:42 <aron> okay
2023-01-23 21:39:34 <aron> g2g now
2023-01-23 21:39:39 <aron> cya
2023-01-23 21:39:53 <kreyren> aron, oke thanks for helping me! <3
2023-01-24 01:50:57 <emonster> is it possible to edit kernel cmdline in disless mode?
2023-01-24 01:56:44 <emonster> in other words: how to do a diskless install on UEFI after booting the ISO??
2023-01-24 10:54:15 <WhyNotHugo> The font-twemoji package uses Mozilla's fork which basically only works on Firefox and Windows. Wouldn't it be best for the package to use the upstream source, which works everywhere (INCLUDING firefox)?
2023-01-24 10:54:17 <WhyNotHugo> https://github.com/twitter/twemoji
2023-01-24 10:54:33 <WhyNotHugo> Mozilla's fork says: Note that the resulting font will only be useful on systems that support layered color TrueType fonts; this includes Windows 8.1 and later, as well as Mozilla Firefox and other Gecko-based applications running on any platform.
2023-01-24 10:54:45 <WhyNotHugo> FWIW, the upstream/original works on GTK and also on Firefox.
2023-01-24 11:01:12 <psykose> that doesn't have a font
2023-01-24 11:02:50 <psykose> so it works on nothing, unless you mean something else
2023-01-24 11:05:15 <psykose> svg-in-ot also works on anything that supports svg-in-ot, which sounds fine to me
2023-01-24 11:05:18 <psykose> foot displays them fine
2023-01-24 13:10:00 <RoyalTiger[m]> I have a question: I tried to set up a Alpine SMB Server on a diskless installation (VM-Image) and therefore  I followed the "Setting up a samba-server"-Tutorial. Everytime I restart my machine and check pdbedit the SMB-Userlist is empty, I can add a smbuser without any problem and I'm able to reconnect from any other machine to my smb, but lbu status/commit shows no change. Any suggestions?
2023-01-24 13:13:42 <RoyalTiger[m]> * installation (VM-Image, separate disk for saving changes) and
2023-01-24 13:31:55 <markand> I have the feeling that libvirtd is overengineered piece of software
2023-01-24 13:32:25 <markand> countless of groups you need to add to your user, services to run and I still get error while trying to spawn a virtual machine
2023-01-24 13:32:34 <markand> I think I'll enable all /etc/init.d/*virt* 
2023-01-24 13:32:35 <markand> ¯\_(ツ)_/¯
2023-01-24 14:34:49 <WhyNotHugo> psykose: waybar does not render the one packaged in Alpine, but does render the one in AUR -> https://aur.archlinux.org/packages/ttf-twemoji
2023-01-24 14:35:09 <WhyNotHugo> The one is AUR is the upstream one.
2023-01-24 14:35:13 <WhyNotHugo> And the upstream one builds a ttf file.
2023-01-24 14:36:00 <WhyNotHugo> The odd part is that the fork is packaged as twemoji; so I'm not sure if I can submit an aport for the upstream (e.g.: non-fork) one.
2023-01-24 14:40:35 <WhyNotHugo> Oh, the Arch package just downloads the ttf built for the Fedora package.
2023-01-24 14:40:49 <WhyNotHugo> Which uses the upstream source, but uses the Makefile from NotoColorEmoji, lol.
2023-01-24 14:41:02 <WhyNotHugo> My main issue is that the Alpine one doesn't seem to work with waybar
2023-01-24 15:05:41 <pj[m]> sounds like a waybar skill issue
2023-01-24 16:49:23 <donoban> WhyNotHugo: your issue is with waybar and fonts rendering?
2023-01-24 16:49:57 <WhyNotHugo> donoban: Yes, emojis render blank.
2023-01-24 16:50:15 <donoban> uhM, in my case all letters are upper and I miss some symbols
2023-01-24 16:50:24 <donoban> it was broken in some upgrade months ago
2023-01-24 16:51:27 <WhyNotHugo> The README for mozilla's fork on the font says "the resulting font will only be useful on systems that support layered color TrueType fonts[...]Systems that do not support such color fonts will show blank glyphs if they try to use this font."
2023-01-24 16:51:31 <WhyNotHugo> So my guess is that this is the issue.
2023-01-24 16:54:08 <WhyNotHugo> Yeah, it also render blank on other GTK3 apps.
2023-01-24 16:54:31 <WhyNotHugo> Oh, wait, that's qt actually
2023-01-24 16:54:44 <WhyNotHugo> Yeah, both qt and gtk render emoji blank.
2023-01-24 17:07:29 <psykose> yes, only works for things with the svg-in-ot support
2023-01-24 17:08:50 <psykose> there aren't any other fonts though, and i'm not personally going to copy some cursed fedora build system that hammered twitter twemoji into a font file
2023-01-24 17:10:08 <WhyNotHugo> psykose: Yeah, that seems to be mostly Firefox and Windows.
2023-01-24 17:10:24 <WhyNotHugo> Lemme see if I can build twemoji in a way less ugly than Fedora
2023-01-24 17:10:47 <WhyNotHugo> Oh, foot works too, curious.
2023-01-24 17:11:50 <psykose> it's not necessarily ugly, i just don't want to unpack someones rpm file or create my own build system for it
2023-01-24 17:12:10 <psykose> but if you want to do some work, what would be nice is making that build system, then publishing a tag with just a font file somewhere
2023-01-24 17:12:13 <WhyNotHugo> Oh, but rpm spec files ARE ugly.
2023-01-24 17:12:17 <psykose> and then i'll gladly swap to it
2023-01-24 17:12:25 <WhyNotHugo> Yeah, that was my intent.
2023-01-24 17:13:02 <psykose> and for that you can even reuse the noto build most likely with the few patches like fedora does, no issues there. keeps the apkbuild curse-free
2023-01-24 17:13:17 <psykose> and yeah i know it's practically no support aside from foot, and godot master, and firefox (not even chromium), but
2023-01-24 17:13:17 <psykose> weh
2023-01-24 17:13:27 <psykose> i didn't find anything else at the time
2023-01-24 17:13:41 <psykose> to use as a font, that is
2023-01-24 17:13:47 <psykose> and conveniently i use foot, so it was better than nothin
2023-01-24 17:15:57 <psykose> .oO(svg-in-ot is also quite a cool format tbh, rather than hardcoded raster emojis, but since mozilla made it afaik obviously nobody is going to use it, sad)
2023-01-24 17:16:50 <grawity> on the other hand, svg-in-ot feels like ... you're putting a vector format inside another vector format?
2023-01-24 17:17:40 <grawity> kind of preferred MS's format of adding color layers
2023-01-24 17:46:31 <le_dude> hello
2023-01-24 18:31:17 <scorpion2185[m]1> I can't compile hyprland again
2023-01-24 18:31:29 <scorpion2185[m]1> meson _build
2023-01-24 18:31:29 <scorpion2185[m]1> protocols/meson.build:38:1: ERROR: File //usr/local/share/hyprland-protocols/protocols/hyprland-toplevel-export-v1.xml does not exist.
2023-01-24 18:31:29 <scorpion2185[m]1> ...
2023-01-24 18:48:29 <elly> that sounds like a bug in their build file?
2023-01-24 18:49:13 <psykose> it is
2023-01-24 18:51:18 <Ermine> 
ACTION opens issue "hyprland exists"

2023-01-24 18:51:59 <scorpion2185[m]1> for make install the dev stated: "missing opengl dep."
2023-01-24 18:52:44 <Ermine> do you have mesa installed?
2023-01-24 18:53:01 <scorpion2185[m]1> yes
2023-01-24 18:53:05 <scorpion2185[m]1> make error : /usr/lib/gcc/x86_64-alpine-linux-musl/12.2.1/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find -lOpenGL: No such file or directory collect2: error: ld returned 1 exit status ninja: subcommand failed .
2023-01-24 18:53:21 <scorpion2185[m]1> apk add mesa ?
2023-01-24 18:53:33 <scorpion2185[m]1> or mesa-gl?
2023-01-24 18:54:54 <Ermine> add everything to be sure
2023-01-24 18:55:41 <scorpion2185[m]1> is this for meson or make?
2023-01-24 18:55:43 <psykose> -lOpenGL doesn't exist
2023-01-24 18:56:09 <psykose> meson doesn't try to use it
2023-01-24 18:56:17 <psykose> it's just hardcoded in cmakee
2023-01-24 18:56:23 <scorpion2185[m]1> is there a way to install all mesa things ? like apk add mesa*
2023-01-24 18:57:08 <psykose> that works but you only need -dev
2023-01-24 18:57:30 <psykose> libOpenGL doesn't exist and you won't find it, just to save you some time
2023-01-24 18:57:31 <ikke> (-dev pulls in other mesa subpackages)
2023-01-24 18:57:43 <psykose> it's libglvnd's gl wrapper
2023-01-24 18:57:49 <scorpion2185[m]1> mesa-dev was already installed
2023-01-24 18:57:56 <psykose> then you have everything
2023-01-24 18:58:11 <psykose> i will repeat for the third time that that doesn't exist
2023-01-24 18:58:55 <scorpion2185[m]1> IOpenGL is cmake error
2023-01-24 18:59:25 <psykose> yes, as i said
2023-01-24 18:59:31 <psykose> i'm not sure if you're reading what i am saying
2023-01-24 18:59:55 <scorpion2185[m]1> yes
2023-01-24 19:07:32 <le_dude> I have a question regarding makeing a distro based on alpine linux
2023-01-24 19:07:40 <le_dude> I have seen this
2023-01-24 19:07:54 <le_dude> https://github.com/rancher/k3os
2023-01-24 19:08:11 <le_dude> Is this the way of least affort to make this?
2023-01-24 19:08:24 <le_dude> or the simplest?
2023-01-24 19:10:14 <ikke> Certainly the simplest to base it of an existing distribution, but that comes at the cost of flexibility
2023-01-24 19:10:34 <le_dude> yes it should be alpine linux
2023-01-24 19:10:49 <le_dude> I want to customize it for raspberry
2023-01-24 19:10:58 <le_dude> and do some configuration work
2023-01-24 19:11:14 <le_dude> so others do not have to jump trough the same hoops
2023-01-24 19:13:28 <ikke> maybe provide some apkovl with your customizations
2023-01-24 19:16:00 <le_dude> idk I want to make it its own distro
2023-01-24 19:16:09 <le_dude> I have to look in apkovl thx
2023-01-24 19:16:17 <le_dude> but Iam not sure if it is the answer
2023-01-24 19:16:21 <ikke> It's quite some effort to build your own distro
2023-01-24 19:16:29 <ikke> You need hardware to build the packages
2023-01-24 19:16:35 <ikke> You need to host the packages
2023-01-24 19:16:38 <le_dude> ok
2023-01-24 19:16:46 <le_dude> maybe I used the wrong word
2023-01-24 19:16:57 <le_dude> alpine linux version my own
2023-01-24 19:17:09 <le_dude> it uses the same infrastruktur
2023-01-24 19:17:14 <le_dude> *infrastructur
2023-01-24 19:17:18 <le_dude> etc
2023-01-24 19:17:25 <ikke> so the same existing packages
2023-01-24 19:17:31 <le_dude> yes
2023-01-24 19:17:45 <le_dude> it is more a configuration of alpine linux
2023-01-24 19:17:57 <le_dude> like alpine linux version x
2023-01-24 19:18:08 <ikke> The default rpi installation is run-from-ram
2023-01-24 19:18:08 <le_dude> is this a distro?
2023-01-24 19:18:12 <ikke> No
2023-01-24 19:18:20 <ikke> It would still be alpine linux
2023-01-24 19:18:22 <le_dude> ok
2023-01-24 19:18:29 <le_dude> so where should I start?
2023-01-24 19:18:46 <le_dude> documentation?
2023-01-24 19:18:48 <le_dude> etc
2023-01-24 19:18:54 <le_dude> I just want something to read
2023-01-24 19:19:13 <ikke> You need to decide how you want to apply your customizations
2023-01-24 19:19:24 <le_dude> Idk what are my options?
2023-01-24 19:19:30 <ikke> apkovl is one
2023-01-24 19:19:40 <ikke> custom image another
2023-01-24 19:19:57 <ikke> But a custom image already requires more infrastructure
2023-01-24 19:20:06 <le_dude> ?
2023-01-24 19:20:13 <le_dude> I can do this local can't I
2023-01-24 19:20:18 <ikke> sure
2023-01-24 19:20:23 <ikke> local is infra as well
2023-01-24 19:20:23 <le_dude> like doing the stuff in a virt machine
2023-01-24 19:20:31 <le_dude> and then make the image
2023-01-24 19:20:44 <ikke> You probably need to create packages that contain your customizations
2023-01-24 19:20:50 <le_dude> damm
2023-01-24 19:20:55 <ikke> Can be simple packages
2023-01-24 19:21:07 <ikke> that just place files somewhere / pull in packages
2023-01-24 19:21:16 <le_dude> Git?
2023-01-24 19:21:29 <le_dude> ok
2023-01-24 19:21:37 <le_dude> I understand
2023-01-24 19:21:39 <le_dude> thank you
2023-01-24 19:22:28 <elibrokeit> scorpion2185[m]1: yet again, people are discussing hyprland!!! Why is this piece of artisanal half-melted pottery so interesting.
2023-01-24 19:23:47 <le_dude> @elibrokeit well, it looks good
2023-01-24 19:24:18 <psykose> flashy cute stuff cool
2023-01-24 19:27:11 <elibrokeit> artisanal half-melted pottery can indeed look good on artistic merits, which is distinct from functioning good on technical merits -- and the whole "half-melted" thing sends a clear message on the latter topic.
2023-01-24 19:28:13 <elibrokeit> anyways, wow do people struggle to build it. So many channels.
2023-01-24 19:51:49 <abby> elibrokeit: those are artisanal half-melted build errors, tyvm
2023-01-24 19:57:16 <elly> basically a kintsugi build system right there
2023-01-24 20:03:27 <elibrokeit> :)
2023-01-24 20:08:46 <psykose> the build system is quite easy tbh
2023-01-24 20:09:28 <psykose> the average "makefile only" build system is much worse
2023-01-24 20:24:05 <elibrokeit> psykose: hyprland's build system is a Makefile that builds a bunch of .c files, then runs meson in a git submodule, then runs cmake, then runs make -C hyprctl that in turn runs:
2023-01-24 20:24:07 <elibrokeit> all:
2023-01-24 20:24:09 <elibrokeit>  g++ -std=c++23 ./main.cpp -o ./hyprctl
2023-01-24 20:24:20 <elibrokeit> sure, it has meson also :P
2023-01-24 20:24:56 <psykose> no, it builds fine with just meson
2023-01-24 20:25:10 <psykose> idc about all the random extras on top with varying quality
2023-01-24 20:25:41 <elibrokeit> all I'm saying is that it certainly does have an artisanal half-melted build system
2023-01-24 20:25:58 <elibrokeit> the meson stuff is, technically, an afterthought introduced later
2023-01-24 20:27:45 <elibrokeit> install guidelines say "CMake (recommended)" followed by additional instructions for Meson -- the cmake instructions are fantastic, it suggests to run `cd Hyprland && sudo make install`
2023-01-24 20:29:55 <psykose> sounds fine to me
2023-01-24 21:10:12 <bl4ckb0ne> just look at the aur pkgbuild
2023-01-24 21:11:02 <bl4ckb0ne> topkek https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=hyprland#n59
2023-01-24 21:12:09 <bl4ckb0ne> https://github.com/hyprwm/Hyprland/blob/e273717a27008e7bd6378bc8abe59b0dca3d9fef/Makefile#L200 
2023-01-24 21:12:10 <ikke> "that doesn't sacrifice on its looks."
2023-01-24 21:12:17 <ikke> But it does saccriface its build system
2023-01-24 21:12:22 <bl4ckb0ne> sacrifices on the packaging sanity
2023-01-24 21:12:23 <psykose> bl4ckb0ne: wow
2023-01-24 21:12:29 <psykose> i could not have written that any worse
2023-01-24 21:14:15 <bl4ckb0ne> there's about 4 different targets that buiild and install wlroots
2023-01-24 21:14:27 <bl4ckb0ne> with various flavours of horror
2023-01-24 21:14:43 <bl4ckb0ne> > cd subprojects/wlroots && meson ./build --prefix=/usr --buildtype=debug -Dwerror=false -Dexamples=false -Db_sanitize=address
2023-01-24 21:15:48 <Ermine> i'm concerned with hyprland getting traction
2023-01-24 21:16:07 <bl4ckb0ne> we've had ton of false wlroots issue because of it
2023-01-24 21:16:45 <bl4ckb0ne> not many relevant issues
2023-01-24 21:16:47 <psykose> what kind
2023-01-24 21:17:00 <abby> the hyprland devs got pissy at void because we refused to let a package that built its own unreleased version of wlr into the repos
2023-01-24 21:17:03 <bl4ckb0ne> psykose: nvidia users
2023-01-24 21:17:17 <bl4ckb0ne> https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3522
2023-01-24 21:17:30 <bl4ckb0ne> https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3549
2023-01-24 21:18:12 <bl4ckb0ne> got a few real ones but they were also reported from sway
2023-01-24 21:22:11 <WhyNotHugo> psykose: Well, I did it. I ported that horrible build script in a horrible language to be a horrible build script in a readable language -> https://paste.sr.ht/blob/08fd6795319e652ecf1969030f60b2a3470dd842
2023-01-24 21:22:49 <psykose> i did mean not as an apkbuild :p
2023-01-24 21:23:24 <Ermine> bl4ckb0ne: but those are not strictly hyprland-related, they are because of nvidia
2023-01-24 21:23:37 <bl4ckb0ne> they are reported by hyprland users
2023-01-24 21:24:51 <WhyNotHugo> psykose: Of course not. I merely wrote this to "translate" the recipe into something readable from which I can build on.
2023-01-24 21:24:59 <psykose> ah, right
2023-01-24 21:25:10 <psykose> then yeah, looks like what was in the spec but in shell directly
2023-01-24 21:25:13 <psykose> very cute
2023-01-24 21:25:13 <WhyNotHugo> Trying to replicate the steps from the spec file ina  shell fails.
2023-01-24 21:25:35 <WhyNotHugo> Yeah, that's pretty much it.
2023-01-24 21:25:51 <WhyNotHugo> I wonder if upstream twemoji would take a PR with an apache-licensed build system.
2023-01-24 21:27:47 <psykose> maybe
2023-01-24 21:27:52 <psykose> doesn't look very alive
2023-01-24 21:31:55 <WhyNotHugo> I wonder if anyone at the team working on it is still there or if they're all fired.
2023-01-24 21:32:38 <psykose> someone seems to have merge access i guess
2023-01-24 21:33:11 <WhyNotHugo> Yeah
2023-01-24 21:55:14 <elibrokeit> bl4ckb0ne: I imagine the hyprland author doesn't do the same thing and require specifying a CLI option to enable unsupported GPUs...
2023-01-24 21:55:22 <elibrokeit> *sane thing 
2023-01-24 21:56:18 <bl4ckb0ne> im glad I have the power to close/lock issues eh
2023-01-24 23:48:40 <WhyNotHugo> psykose: What kind of APKBUILD would be acceptable for twitter-twemoji? The build system is a mess, and while Fedora's approach is a pile of hacks, the build is so convoluted that I can't imagine anything simpler. It's piles of scripts on scripts on custom tooling just for building the font.
2023-01-24 23:49:00 <psykose> the one where the source= is https://you/you/font.ttf
2023-01-24 23:49:04 <WhyNotHugo> And honestly, I though of making a dedicated pipeline to build the TTF, but is downlaoding from that any different from downloading from fedoras output?
2023-01-24 23:49:44 <WhyNotHugo> Okay, I can do that. :P
2023-01-25 00:19:17 <pj[m]> 1. download rpm... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/QEDWejzHFOMuAKHDqmFksBXu>)
2023-01-25 00:28:42 <elibrokeit> theoretical purity concerns would be the only thing that springs to mind
2023-01-25 02:07:35 <WhyNotHugo> What's the difference between util-linux-login and shadow-login?
2023-01-25 02:16:00 <psykose> you mean aside from being separate projects with separate codebases?
2023-01-25 02:16:33 <psykose> probably just minor implementation differences
2023-01-25 02:22:06 <q66> one is terrible util-linux code and the other is terrible shadow code
2023-01-25 02:22:10 <q66> pick your evil
2023-01-25 02:22:59 <q66> shadow is what distros usually use 
2023-01-25 02:23:07 <psykose> what's terrible about it this time
2023-01-25 02:23:27 <q66> code is bad 
2023-01-25 02:25:20 <q66> but you want it anyway because you want pam support and busybox sucks 
2023-01-25 02:25:27 <q66> worse than either of those things 
2023-01-25 02:26:01 <q66> maybe I should write my own utilities 
2023-01-25 02:28:37 <psykose> what's bad about the code
2023-01-25 02:31:50 <q66> it's been months since I last dug around so i won't point out anything specific
2023-01-25 02:33:09 <q66> there have been several patches I've had to carry over time to fix an assortment of very basic bugs that shouldn't have been there 
2023-01-25 02:33:32 <q66> and then i went digging around in the codebase and got grossed out 
2023-01-25 02:39:43 <psykose> i can say that for shadow sure, not really the other though
2023-01-25 02:43:22 <q66> i haven't looked at the util-linux code for this but i've seen other util-linux code
2023-01-25 02:44:20 <psykose> util linux is util linux but at least i can say it's not coreutils and definitely not freebsd
2023-01-25 02:45:30 <q66> i guess util-linux is just a thoroughly mediocre collection of random junk where nobody thought about making something coherent 
2023-01-25 02:46:22 <elly> as are all util-whatever packages, honestly
2023-01-25 02:48:12 <q66> the pieces of it I replaced with freebsd ones have been a major improvement
2023-01-25 02:48:57 <q66> it's a small overlap but enough to unfuck the bootstrap path and make myself slightly happier 
2023-01-25 02:49:35 <q66> i should replace more of it over time
2023-01-25 09:26:02 <WhyNotHugo> q66: Honestly, I just want to raise the ulimit for all my sessions. I don't need pam suppot on login.
2023-01-25 09:27:55 <WhyNotHugo> Why the FreeBSD and not the OpenBSD one? OBSD usually has pretty good code standards.
2023-01-25 09:29:39 <spazialecontinentale[m]> hi, i'm trying to port a new package to alpine but i have some issue with openrc.
2023-01-25 09:29:40 <spazialecontinentale[m]> If run the program with my user everything works, but if I use rc-service mollysocket start it crash and --debug is pretty useless
2023-01-25 09:50:58 <sthlm> Hi, I am trying to install xorg with dwm but cant get the mouse and keybord to work. I installed xf86-input-evdev, xf86-input-libinput and xf86-input-synaptics. any ideas?
2023-01-25 09:52:33 <dwfreed> are you in the right group to be able to use the /dev nodes ?
2023-01-25 09:53:05 <dwfreed> usually this is input group, but I'm not sure what alpine uses
2023-01-25 09:53:15 <dwfreed> look at /dev/input/event*
2023-01-25 09:54:12 <sthlm> <usr> wheel audio video netdev
2023-01-25 09:55:53 <sthlm> should I add the user to INPUT group
2023-01-25 09:55:54 <sthlm> ?
2023-01-25 09:56:57 <dwfreed> well, look at what group /dev/input/event* has
2023-01-25 09:57:03 <dwfreed> and add yourself to that group
2023-01-25 09:57:51 <sthlm> after cat one of the file event0 all the text in terminal went bananas and looks crazy...
2023-01-25 09:58:10 <dwfreed> you're not supposed to cat them...
2023-01-25 09:58:20 <dwfreed> just look at the ls output to see what group owns them
2023-01-25 09:58:39 <sthlm> I have enevt files ending with 0-20
2023-01-25 09:58:54 <dwfreed> that's not uncommon
2023-01-25 09:59:07 <sthlm> how do I do that?
2023-01-25 09:59:17 <dwfreed> ls -l /dev/input/event*
2023-01-25 10:01:27 <sthlm> all of them have: crw-rw---- 1 root  input 13, 64 Jan 25 10:53 /dev/input/eventX
2023-01-25 10:02:32 <sthlm> btw, I also blacklisted i915 driver to even get into dwm. dont know if that helps..
2023-01-25 10:03:09 <dwfreed> irrelevant as far as input devices go
2023-01-25 10:03:15 <sthlm> ok
2023-01-25 10:03:35 <sthlm> but what about the group. is it root?
2023-01-25 10:04:06 <grawity> that shows 'root' as the owner user and 'input' as the group
2023-01-25 10:04:33 <sthlm> so I should add my user to the input group?
2023-01-25 10:05:15 <grawity> most likely yes
2023-01-25 10:05:31 <sthlm> I did not have this problem with void...
2023-01-25 10:05:43 <grawity> did void have something like seatd or elogind?
2023-01-25 10:06:24 <sthlm> not sure, but jumping to alpine make me realize how much comes as default in other distros
2023-01-25 10:06:38 <grawity> in some distros, Xorg gets access to /dev/input/eventX indirectly, by asking the logind/elogind/seatd service
2023-01-25 10:07:10 <sthlm> ok I see. and why does alpine not have this?
2023-01-25 10:07:24 <sthlm> is using group permissoons more safe?
2023-01-25 10:09:08 <sthlm> I just found this when trying to find info on adding user to group:
2023-01-25 10:09:10 <grawity> no, logind/seatd might be safer – but also more complex, with more moving parts involved
2023-01-25 10:09:10 <sthlm> "Only if elogind is not being used and running, then X users would need to be added to the video and input groups to be able to work with a graphical display: "
2023-01-25 10:09:44 <grawity> right
2023-01-25 10:11:04 <sthlm> I really like Alpine but I must say that the wiki could have a more noob safe structure
2023-01-25 10:12:01 <sthlm> like, the text I posted on what X users need to do should be in the "setup xorg" page.....
2023-01-25 10:12:33 <sthlm> I will try to add it there, but lets see first if this helps
2023-01-25 10:14:24 <sthlm> is this the way in Alpine:
2023-01-25 10:14:31 <sthlm> adduser <username> input
2023-01-25 10:14:32 <sthlm> ?
2023-01-25 10:16:09 <sthlm> IT WORKS
2023-01-25 10:16:13 <sthlm> 12h later
2023-01-25 10:16:16 <sthlm> thank you so much
2023-01-25 10:17:04 <WhyNotHugo> sthlm: Feel free to improve the wiki for the next person.
2023-01-25 10:17:39 <sthlm> great, I will
2023-01-25 10:18:40 <sthlm> next problem. I now tested to start mixxx in kiosk mode but it only renders on a part of the full screen. can I make it to use the full screen?
2023-01-25 11:23:08 <lnl> anyone else having mouse problems? my trackpoint just stopped being detected by sway
2023-01-25 12:32:54 <Forza> Hi. Apparently it indeed was a bug in the kernel that caused my reboots. I wonder if the fix will be back ported... https://marc.info/?l=target-devel&m=167463560822293&w=2
2023-01-25 12:34:19 <ikke> Forza: is there a way I could see if I'm affected? I have an ARM server that reboots (every couple of hours even)
2023-01-25 12:34:45 <ikke> But so far nothing in any log that confirms it (but I haven't tried remote syslog, though I did watch the serial console output)
2023-01-25 12:35:29 <Forza> ikke: the remote syslog didnt help me, I had to enable pstore to get the full trace. remote syslog only caught the first line
2023-01-25 12:35:38 <ikke> hmm, ok
2023-01-25 12:35:49 <ikke> Forza: do you use iscsi?
2023-01-25 12:36:01 <Forza> modprobe efi_pstore
2023-01-25 12:36:10 <Forza> yes, I use alpine as iscsi target
2023-01-25 12:36:28 <ikke> apparently efi_pstore is already loaded
2023-01-25 12:36:45 <Forza> check /sys/fs/pstore if it is mounted and has some entries
2023-01-25 12:36:47 <RoyalTiger[m]> <RoyalTiger[m]> "I have a question: I tried to..." <- I meanwhile found the easy solution for my problem: The Default-TDBSAM-File of SMB isn't in /etc/ but somewhere in /var/ (https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum/general-discussion/samba-smbpasswd-problem.html). After including that in lbu everything runs great. Sometimes it feels a little bit like the Diskless-Mode is not very often used or
2023-01-25 12:36:47 <RoyalTiger[m]> referred to, besides that this was mainly a SMB-Configuration-thing. 
2023-01-25 12:37:09 <ikke> Forza: empty
2023-01-25 12:37:35 <Forza> ikke: hmm
2023-01-25 12:37:35 <Forza>  http://dpaste.com/FC5U9TXQH
2023-01-25 12:38:20 <Forza> dmesg |grep -i pstore ?
2023-01-25 12:38:49 <ikke> nothign
2023-01-25 12:39:25 <ikke> But we do not use iscsi afaik
2023-01-25 12:39:34 <ikke> It's in anycase not a target
2023-01-25 12:40:34 <Forza> Then it is not the same issue
2023-01-25 12:40:39 <ikke> right
2023-01-25 12:40:56 <Forza> There is a use-after-free in the iscsi target code. It tries to free the seession after it is closed
2023-01-25 12:41:16 <Forza> but the pstore thing could help you if you have random reboots
2023-01-25 12:41:17 <ikke> But the debugging techniques can help
2023-01-25 12:41:19 <ikke> yeah
2023-01-25 12:41:35 <Forza> you can set it to efivar or any block device 
2023-01-25 12:41:42 <ikke> But not sure why there areno entriesi n there
2023-01-25 12:41:58 <ikke> I do see it mounted
2023-01-25 12:42:08 <ikke> pstore on /sys/fs/pstore type pstore
2023-01-25 12:42:29 <ikke> Forza: do you need kernel cmdline arguments?
2023-01-25 12:42:39 <Forza> The kernel .config I use is http://dpaste.com/AELTCJEGS
2023-01-25 12:42:45 <Forza> no cmdline arguments
2023-01-25 12:42:58 <ikke> ok
2023-01-25 12:43:23 <Forza> want the full .config?
2023-01-25 12:44:11 <ikke> CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE=y
2023-01-25 12:44:16 <ikke> ok, you have that as well
2023-01-25 12:44:34 <Forza> yea but it gets mounted anyways
2023-01-25 12:44:35 <Forza> :o
2023-01-25 12:44:50 <ikke> Forza: how do you set it to store it on efi?
2023-01-25 12:45:08 <Forza> I think the important lines are https://dpaste.com/AELTCJEGS#line-21 to 23
2023-01-25 12:45:56 <ikke> Ah ok, those are missing
2023-01-25 12:46:05 <ikke> CONFIG_PSTORE_CONSOLE is not set
2023-01-25 12:48:34 <Forza> Its kernel messages, user space console messages and function trace on oupses
2023-01-25 12:49:25 <ikke> Ok, I'll need to rebuild the kernel with those options enabled
2023-01-25 12:49:37 <ikke> thanks
2023-01-25 12:50:46 <Forza> :) good luck!
2023-01-25 12:51:34 <ikke> Appreciated
2023-01-25 12:51:53 <ikke> It's something that already happens for some time and as a result we haven't really been able to use that server
2023-01-25 12:52:20 <Forza> Such things are so annoying. I have been battling with this since I got this server a year ago
2023-01-25 12:52:41 <ikke> ouch
2023-01-25 12:53:53 <Forza> So annoying because it was long between each time, so it was hard to do a recording session over remote console etc
2023-01-25 12:54:06 <ikke> In our case it's not that long between reboots
2023-01-25 12:54:15 <ikke> average uptime is <24h
2023-01-25 12:57:23 <Forza> Check NMI or ECC logs
2023-01-25 12:57:33 <Forza> ECC/RAS (rasdaemon)
2023-01-25 12:59:12 <ikke> apparently only built for x86_64 on alpine
2023-01-25 12:59:21 <ikke> I'll check later, thanks for the pointers!
2023-01-25 13:03:21 <Forza> NP =)
2023-01-25 14:15:43 <JanJ[m]> hmz
2023-01-25 15:43:55 <blt[m]> hi, I tried to restart lxd where I got some containers running and got a "umount: /sys/fs/cgroup/systemd: target is busy." and now I am unable to start it again or stop it, any hints on how can I unlock my situation? I posted this to lxd channel, but it seems that this is some dependancies on cgroup/systemd in our init.d but I don't understand why, anyone of you has the same issue when trying to do a service lxd restart ?
2023-01-25 15:51:00 <blt[m]> if I comment the start_pre and stop_post it fixes it
2023-01-25 16:06:48 <scorpion2185[m]1> psykose can you still build hyprland head?
2023-01-25 16:07:33 <scorpion2185[m]1> meson _build      ?
2023-01-25 16:07:33 <scorpion2185[m]1> git clone --recursive https://github.com/hyprwm/Hyprland
2023-01-25 16:24:11 <kreyren> 
ACTION submitted https://gitlab.alpinelinux.org/alpine/aports/-/issues/14588

2023-01-25 16:48:54 <frag> anyone tried PINEBOOK Pro? thoughts?
2023-01-25 16:50:26 <lopid> doesn't need all capitals
2023-01-25 16:50:57 <frag> i copied from https://pine64.com/product/14%E2%80%B3-pinebook-pro-linux-laptop-ansi-us-keyboard/ :>
2023-01-25 17:00:47 <elly> frag: I have; alpine does not boot on it by default, I had to use postmarketOS instead
2023-01-25 17:00:57 <elly> it's fine but not great, there are nicer linux laptops out there
2023-01-25 17:06:59 <frag> elly: oh, thats good to know! never even heard of postmarketOS
2023-01-25 17:07:11 <frag> elly: which laptops do you prefer btw?
2023-01-25 17:08:00 <elly> my favorite is my Framework
2023-01-25 17:10:11 <frag> a bit more pricy but interesting concept
2023-01-25 17:10:34 <frag> i like the alminium shell of the pine, the price, the lack of fan
2023-01-25 17:10:52 <frag> and minimalist power suply
2023-01-25 17:11:36 <frag> dunno if i will buy a fanned computer again
2023-01-25 17:12:48 <frag> elly: does the framework feels solid with all those modules?
2023-01-25 17:14:05 <frag> pine doesnt have alminium on the inside tho ..
2023-01-25 17:14:50 <elly> the framework feels okay to me physically
2023-01-25 17:14:56 <elly> it's pretty sturdy
2023-01-25 18:47:29 <frag> do you think disabling the fan will make most/all computers fanless (and ok)?
2023-01-25 18:49:56 <lopid> if you disable a computer's fan, its ability to compute will be compromised sooner or later
2023-01-25 18:53:22 <frag> lopid: you mean it will not work optimally or that it will be ruined?
2023-01-25 18:53:36 <lopid> yes
2023-01-25 18:54:18 <invoked> probably not ruined, but it will shut down if it gets too hot.
2023-01-25 18:57:13 <minimal> frag: generally speaking the fan(s) is there for a reason, and the reason is not (only) to make noise...
2023-01-25 18:58:15 <aron> you can replace a fan with a large enough heatsink tho
2023-01-25 18:58:52 <lopid> technically true
2023-01-25 19:00:28 <hodapp> technically true but still depends on airflow to move the heat out
2023-01-25 19:01:26 <aron> you can achive good enough airflow with good engineering :)
2023-01-25 19:01:40 <lopid> stick the other end of this theoretical heatsink in antarctica
2023-01-25 19:01:53 <hodapp> but what's an example of something that *can't* be achieved with good engineering?
2023-01-25 19:01:53 <aron> lol
2023-01-25 19:02:09 <aron> ok, you got me :P
2023-01-25 19:02:57 <hodapp> the only time I've tried to run something designed for a fan, without a fan, is when making some other changes - like underclocking the processor considerably
2023-01-25 19:03:49 <hodapp> and having the system run relatively open-air
2023-01-25 19:05:12 <aron> my system has oversized heatsink, so fan does almost nothing, makes the config super silent
2023-01-25 19:08:37 <frag> minimal: lol
2023-01-25 20:27:19 <frag> elly: why didnt alpine boot on pinebook, do you know?
2023-01-25 20:36:12 <elly> I don't know, it just did not and I was like "well that sucks"
2023-01-25 20:36:18 <elly> arm devices are sometimes like that
2023-01-25 20:39:11 <frag> mhmhm
2023-01-25 20:48:41 <psykose> probably missing some kernel options, and i'm not sure if all of the pinebook soc is upstreamed yet
2023-01-25 21:06:27 <Leolo> noob question : I have an archaic computer I want to set up as a router.
2023-01-25 21:06:36 <Leolo> becuase my current router is even more archaic
2023-01-25 21:07:16 <Leolo> anyway, I want to install the OS on a compact flash, using a CF to IDE adapter (that's how archaic it is)
2023-01-25 21:07:56 <Leolo> but the compact flash already has a apkovl on it, so when I boot from USB to do the install, it mount the partition as rdonly
2023-01-25 21:08:04 <Leolo> and loads the apkovl
2023-01-25 21:08:33 <Leolo> I tried ovl_dev=honk and it still mounted /dev/sda1
2023-01-25 21:08:44 <Leolo> (sdb = usb, sda = CF)
2023-01-25 21:10:03 <Leolo> oh wait
2023-01-25 21:10:23 <Leolo> I just did mout -o remount,rw and it worked.  that wasn't working before
2023-01-25 21:19:52 <Leolo> ha!  this is a 64MB CF of course I can't install to it :D
2023-01-25 21:49:41 <elly> you could probably put enough system on a 64MB CF card
2023-01-25 21:52:02 <psykose> certainly wouldn't be easy
2023-01-25 21:52:22 <psykose> need to build your own tiny kernels for that
2023-01-25 21:52:58 <elly> my kernel is 8.4MB, 14.8MB of initramfs... I bet you could get the system into < 64MB
2023-01-25 21:53:03 <elly> that's a stock alpine lts kernel
2023-01-25 21:53:55 <psykose> the initrd is much smaller than 14 when it doesn't have gpu firmware too :p
2023-01-25 21:54:03 <psykose> but it's still a lot of space after the modules in use and all
2023-01-25 22:21:07 <frag> alPINE
2023-01-26 11:44:26 <Apachez> elly: floppyfw were the days with 2.6 kernel + qos/tc + busybox + iptables etc and all fitted in a 1.44MB floppy :)
2023-01-26 13:12:49 <foobarbaz> Hello everyone! I have trouble with getting Xorg to work, will anyone help me please? Logs: https://0x0.st/oFT3.txt
2023-01-26 13:14:36 <ikke> foobarbaz: (EE) open /dev/fb0: Permission denied
2023-01-26 13:14:56 <ikke> foobarbaz: are you in the video group?
2023-01-26 13:16:32 <foobarbaz> The wiki said that it's required only if you don't have elogind installed, but i'll try that
2023-01-26 13:17:31 <ikke> Just installing elogind is not enough, it also needs to be running and be used
2023-01-26 13:17:41 <ikke> (though I have no experience with it)
2023-01-26 14:31:38 <elly> the ash sigwinch bug has now been around for long enough that it feels familiar
2023-01-26 14:32:49 <ikke> So we're getting in this territory? https://xkcd.com/1172/
2023-01-26 14:32:49 <alpine-meetbot> [xkcd] Workflow | Alt-text: There are probably children out there holding down spacebar to stay warm in the winter! YOUR UPDATE MURDERS CHILDREN.
2023-01-26 14:33:26 <elly> sorta, except that it has no useful functionality to me
2023-01-26 14:36:55 <invoked> installing linux/et al on a 2160p laptop is fun. had to use my phone's camera as a magnifying glass.
2023-01-26 14:37:25 <ikke> lol
2023-01-26 14:38:36 <elly> Apachez: I'm kind of nostalgic for it sometimes, tbh
2023-01-26 14:38:50 <elly> I see the 7MB kernel I have and I'm like "I remember when the whole install was that size"
2023-01-26 14:42:14 <minimal> I remember installing Slackware (in early 1990s) from approx 30 floppy disks...
2023-01-26 14:44:01 <hodapp> I ran ZipSlack back in the day on a craptop
2023-01-26 14:47:15 <invoked> a good design goal is to get your booted os to fit in l3 cache (or smart cache, or whatever it's called now). which if you don't change anything should get easier with time
2023-01-26 14:48:33 <alpine-meetbot> invoked meant to say: a good design constraint is to get your booted os to fit in l3 cache (or smart cache, or whatever it's called now). which if you don't change anything should get easier with time
2023-01-26 14:48:33 <invoked> s/goal/constraint/
2023-01-26 14:49:02 <elly> tired: run from RAM, wired: run from L2
2023-01-26 14:50:45 <lucidiot> isn't that what the james webb telescope does
2023-01-26 14:53:16 <invoked> i dunno what hardware they have on that, but i do remember hearing people say "javascript in space" so it can't be that small
2023-01-26 14:54:45 <lucidiot> yeah but it's at the L₂ lagrange point,
2023-01-26 14:59:55 <Apachez> elly: did you use floppyfw aswell? :)
2023-01-26 15:00:09 <Apachez> I found recently the tc howto I wrote like 20 years ago for floppyfw :D
2023-01-26 15:00:31 <Apachez> https://www.zelow.no/floppyfw/download/HOWTOS/tc-readme.txt
2023-01-26 15:00:36 <invoked> lucidiot: https://en.wikipedia.org/wiki/RAD750
2023-01-26 15:00:37 <alpine-meetbot> [wikipedia] RAD750 | "The RAD750 is a radiation-hardened single-board computer manufactured by BAE Systems Electronics, Intelligence & Support.  The successor of the RAD6000, the RAD750 is for use in high-radiation environments experienced on board satellites and spacecraft. The RAD750 was released in 2001, with the first units launched into space in 2005."
2023-01-26 15:01:13 <Apachez> costs an arm, a leg and your firstborn child?
2023-01-26 15:01:21 <elly> Apachez: not specifically, but I do remember tomsrtbt pretty well
2023-01-26 15:01:56 <elly> http://www.toms.net/rb/ good lord the website has a retro aesthetic
2023-01-26 15:02:00 <elly> there's a marquee
2023-01-26 15:02:39 <elly> download links to tucows' ftp server, an md5 hash so you can check the file isn't corrupt... it's like travelling in time
2023-01-26 15:03:28 <Apachez> :)
2023-01-26 15:03:44 <Apachez> only that gif "under construction" is missing ;)
2023-01-26 15:04:51 <lucidiot> the marquee has a <blink> in it too
2023-01-26 15:05:13 <Apachez> and magenta? ;)
2023-01-26 15:05:55 <lucidiot> invoked: from this reference in the article, it can have a 256KB to 1MB L2 cache https://web.archive.org/web/20090326020946/http://www.aero.org/conferences/mrqw/2002-papers/A_Burcin.pdf
2023-01-26 15:06:47 <scorpion2185[m]1> psykose did you use meson _build?
2023-01-26 15:06:52 <lucidiot> it's not said whether or not the JWST does have one since it seems optional. at the very least we know there's an L1 cache at lagrange L2
2023-01-26 15:06:55 <scorpion2185[m]1> after `git clone --recursive https://github.com/hyprwm/Hyprland` ?
2023-01-26 15:06:59 <Apachez> lucidiot: how is that radiation hardening achieved?
2023-01-26 15:09:49 <ikke> Apachez: https://en.wikipedia.org/wiki/Radiation_hardening
2023-01-26 15:09:49 <alpine-meetbot> [wikipedia] Radiation hardening | "Radiation hardening is the process of making electronic components and circuits resistant to damage or malfunction caused by high levels of ionizing radiation (particle radiation and high-energy electromagnetic radiation), especially for environments in outer space (especially beyond the low Earth orbit), around nuclear reactors and particle accelerators, or during nuclear accidents or nuclear […]"
2023-01-26 15:10:15 <lucidiot> i found some suggestions specific to this processor: https://space.stackexchange.com/a/50673
2023-01-26 15:12:41 <Apachez> doesnt tell HOW they are being made for withstand riddicilous amount of radiation without failing
2023-01-26 15:13:35 <Apachez> "they are being more rugged"
2023-01-26 15:13:38 <Apachez> sure but HOW
2023-01-26 15:13:44 <Apachez> adding 1 meter of lead?
2023-01-26 15:14:53 <minimal> Apachez: covering in a Deep Pile Rug? then they'd be more rugged ;-)
2023-01-26 15:14:53 <Apachez> some cool cold war technology https://www.space.com/soviet-venera-venus-missions-slideshow
2023-01-26 15:15:18 <elly> wider traces, higher separations, hardware error correction, component redundancy, shielding - some combination of those I guess?
2023-01-26 15:15:31 <lucidiot> there's a techniques section https://en.wikipedia.org/wiki/Radiation_hardening#Radiation-hardening_techniques
2023-01-26 15:15:32 <alpine-meetbot> [wikipedia] Radiation hardening - Radiation-hardening techniques | "Physical Hardened chips are often manufactured on insulating substrates instead of the usual semiconductor wafers. Silicon on insulator (SOI) and silicon on sapphire (SOS) are commonly used. While normal commercial-grade chips can withstand between 50 and 100 gray (5 and 10 krad), space-grade SOI and SOS chips can survive doses between 1000 and 3000 gray (100 […]"
2023-01-26 15:15:34 <nickersonm> Larger process nodes help with radiation hardening. Different substrates and other materials help as well.
2023-01-26 15:16:21 <Apachez> also fun how they did test those doses
2023-01-26 15:16:36 <elly> probably put them in a room with radiation sources and measured error rates
2023-01-26 15:16:39 <Apachez> https://www.youtube.com/watch?v=R0H-NWk_FDI
2023-01-26 15:16:49 <Apachez> and you thought your door was heavy? ;)
2023-01-26 15:17:51 <Apachez> 2.5 meter thick, 44 metric tons of weight
2023-01-26 15:17:56 <Apachez> now THATS a door =)
2023-01-26 15:18:31 <aron> lol
2023-01-26 15:19:30 <minimal> I used to work at a site with blast doors, don't think they were 2.5 metres thick though, more like 1-1.5m
2023-01-26 15:22:07 <invoked> i'm assuming they expect -- and build some tolerance for -- a little bit flipping. they can't plan for none.
2023-01-26 15:22:17 <invoked> there's probably a lot more to it.
2023-01-26 17:05:30 <Leolo> how portable are apkovl files?  specifically, I have a router that boots 2.4.5 from CD-R.
2023-01-26 17:06:23 <Leolo> if I was to burn and boot 3.17, would it use that apkovl "just fine" ?
2023-01-26 18:32:05 <frag> is you buy a laptop and store it for 5-10 years, will it be as good then as it is new?
2023-01-26 18:32:29 <blt[m]> am using linux-lts from edge as my laptop is pretty new, all the other package are from stable. from the latest linux 6.1.8.0 I got big issue with my mouse interupting to work, almost impossible to use it, is there a way to go back one version behind ? 6.1.7.0 ?
2023-01-26 18:33:07 <ikke> We don't keep older versions of packages atm, so you'd have to build it yourself
2023-01-26 18:33:51 <blt[m]> ☹️ that's the risk to play on edge so...
2023-01-26 18:37:56 <Leolo> frag - almost
2023-01-26 18:38:21 <Leolo> some components will age even if unused.  specifically capacitors.  
2023-01-26 18:39:14 <Leolo> also, how the laptop is stored will be important - specifcically condensation will bugger things up
2023-01-26 18:39:39 <frag> Leolo: __/\__
2023-01-26 18:40:58 <frag> ive had a laptop for 6 years ..... light use, as typewriter 1-2 times a week .. still works ..
2023-01-26 18:41:33 <frag> mostly keep the power plugged in .. isnt it weird its still alive??
2023-01-26 18:43:14 <Leolo> no
2023-01-26 18:43:29 <Leolo> I have computers from the 70s that still work
2023-01-26 18:43:39 <Leolo> well, still worked last time I plugged them in.
2023-01-26 18:45:37 <Leolo> there was a period in the 00's call the capacitor plague, where caps were dying after 5 years because of industrial espionage and counter espionage
2023-01-26 18:45:57 <frag> Leolo: wow, but i guess laptops are especially likely to not last long eh?
2023-01-26 18:46:21 <ikke> Leolo: how did espionage lead to short-living capacitors?
2023-01-26 18:47:48 <mercenary> ikke: $a stole the recipe for the dielectric from $b, not knowing that $b hadn't figured out the way to properly stabilize it yet.
2023-01-26 18:48:00 <psykose> elly: the sigwinch was fixed :p
2023-01-26 18:48:15 <ikke> mercenary: ah ok
2023-01-26 18:48:28 <ikke> stealing broken things
2023-01-26 18:48:34 <mercenary> So short-term it looked like it worked, and they started manufacturing based on that recipe. With disastrous results
2023-01-26 18:49:26 <ikke> classic
2023-01-26 19:03:47 <scorpion2185[m]1> psykose did you use meson _build?
2023-01-26 19:04:02 <scorpion2185[m]1> after `git clone --recursive https://github.com/hyprwm/Hyprland` ?
2023-01-26 19:04:33 <psykose> probably, idk
2023-01-26 19:04:41 <psykose> i'm not going to build half melted pottery again
2023-01-26 19:05:56 <scorpion2185[m]1> so you don't remember?
2023-01-26 19:09:34 <PureTryOut> so since I just had to reboot into a live CD of Kubuntu to get wifi to work again after I upgraded to a segfaulting iwd earlier, does apk keep a cache of previously installed packages somewhere you can downgrade too if needed?
2023-01-26 19:10:42 <lopid> /var/cache/apk
2023-01-26 19:11:50 <PureTryOut> I thought so but it didn't have an older copy of iwd in there
2023-01-26 19:12:03 <PureTryOut> In fact it has basically nothing on there, only kwin for some reason
2023-01-26 19:12:11 <PureTryOut> and an old version of that one as well
2023-01-26 19:12:13 <lopid> something cleaned it?
2023-01-26 19:12:26 <PureTryOut> I wouldn't know what
2023-01-26 19:18:14 <Leolo> mercenary - the way I heard it, A stole from B, but B knew and substituted a "wrong" recipe
2023-01-26 19:20:37 <mercenary> Leolo: there are a few versions of it, no way to know which one except to the people directly involved, but the net effect was A ended up with a recipe they shouldn't have, that turned into mush after a couple of years
2023-01-26 19:29:29 <psykose> PureTryOut: if /etc/apk/cache points to a cache then it keeps them around indefinitely
2023-01-26 19:30:44 <PureTryOut> Indefinitely, that's a bit too much 😅
2023-01-26 19:30:46 <psykose> but afaik downgrades won't work to those because they're not in an index, you'd have to copy the apk to somewhere and add it to an index or manually add it via path
2023-01-26 19:32:14 <elibrokeit> psykose: this is what you get for offering to build half-melted pottery even once.
2023-01-26 19:33:19 <psykose> no good deed, something something, punished unpunished, etc
2023-01-26 21:40:23 <elly> what happened to psykose?
2023-01-26 21:40:35 <psykose> i died
2023-01-26 21:40:39 <psykose> rip psykose 2022-2022
2023-01-26 21:40:41 <elly> oh: https://lists.alpinelinux.org/~alpine/devel/%3CCA%2BcSEmMAiEPvFo37DDKVLv6_mZez23NXopwseCBvRbEtouRwqw%40mail.gmail.com%3E
2023-01-26 21:40:53 <elly> well, congratulations :)
2023-01-26 21:41:52 <psykose> thanks mom
2023-01-26 21:42:05 <elly> very proud of you :)
2023-01-26 21:44:10 <elly> what can we expect from your new reign of terror?
2023-01-26 21:44:56 <psykose> i haven't a clue tbh :)
2023-01-26 21:45:33 <psykose> more bugs and fewer features i suppose
2023-01-26 21:45:55 <elly> oh no
2023-01-26 22:30:32 <Saijin_Naib> Damn, nice! Well-deserved! 
2023-01-27 00:09:19 <elly> hmmmm, busybox actually doesn't build under musl? perhaps I am holding it wrong
2023-01-27 00:09:40 <elly> init/init.c includes linux/vt.h which is not present
2023-01-27 00:10:07 <pj[m]> "busybox actually doesn't build.."
2023-01-27 00:15:23 <psykose> linux/vt is from linux-headers
2023-01-27 00:15:41 <psykose> same for anything under <linux/*
2023-01-27 00:19:49 <elly> now where on earth did linux-headers go for me then :P
2023-01-27 00:20:01 <elly> thanks psykose
2023-01-27 00:22:22 <elly> woo, they did fix the winch bug on busybox 1.37 branch
2023-01-27 00:22:23 <elly> soooon
2023-01-27 00:28:32 <psykose> it's also patched in edge another way
2023-01-27 00:30:16 <elly> ahh
2023-01-27 00:35:27 <EtherMan> Hi so I'm trying to get k3s to run, but when trying to deploy the CNI (calico), I run into errors that /sys and / are not shared. That's easily solved in any one bootup by simply running "mount --make-shared /" and for /sys... But how do I make that permanent across reboots so cluster doesn't require manual intervention to recover from a restart? :/
2023-01-27 00:35:58 <dwfreed> elly: was going to say, openwrt would have noticed too
2023-01-27 00:48:30 <minimal> ErterMan: perhaps edit /etc/init.d/sysfs and modify the sysfs_opts variable?
2023-01-27 00:48:38 <minimal> EtherMan: ^^^
2023-01-27 00:49:29 <minimal> as for "/" you could edit its entry in /etc/fstab
2023-01-27 00:49:57 <EtherMan> Wouldn't that just get replaced by updates? (for sysfs I mean, I assume fstab wouldn't be)
2023-01-27 00:54:18 <minimal> nope, I'd expect a updated /etc/init.d/sysfs file to be instead stored as /etc/init.d/sysfs.apk-dist (think that's the right extension)
2023-01-27 00:56:38 <EtherMan> Didn't work. Neither / or /sys actually become shared even if I set the mount option. I'd assume that has to do with https://man7.org/linux/man-pages/man8/mount.8.html "The command mount does not pass the mount options unbindable, runbindable, private, rprivate, slave, rslave, shared"... :/
2023-01-27 00:56:39 <minimal> oh, ".apk-new" is the correct extension
2023-01-27 01:00:52 <minimal> ok, write your own init.d script to change them and order it after init.d localmount and sysfs?
2023-01-27 01:09:33 <psykose> make-shared has to be set after mount indeed
2023-01-27 01:11:29 <dwfreed> or just make a start_post in conf.d/localmount and conf.d/sysfs
2023-01-27 01:14:20 <dwfreed> EtherMan: that comment is about fs mount helpers, which should not matter here
2023-01-27 01:14:25 <EtherMan> Well, did the basic openrc service method now and that works. Just seems a bit... Non standard :)
2023-01-27 01:16:26 <dwfreed> if you put /sys in fstab, the sysfs script will use whatever you put in there
2023-01-27 09:38:10 <adhawkins> I have a box currently running 'edge', but was considering switching it to a stable release. If I switch the 'repositories' file to point to the current release rather than edge, will it be Ok?
2023-01-27 09:38:12 <Piraty> psykose: congrats
2023-01-27 09:38:26 <adhawkins> I'm not expecting it to downgrade packages, just 'continue working' and upgrade any packages that become new.
2023-01-27 09:38:48 <adhawkins> Follow up question, will I get a notification that a new stable is available, or do I need to monitor that and change the repositories file when it's out?
2023-01-27 09:39:27 <Piraty> adhawkins: i never did this, but i assume the point of lowest breakage possible is whenever the next release is split from edge
2023-01-27 09:40:10 <Piraty> to do the switch (after upgrading everying right until very close to the split )
2023-01-27 09:40:57 <adhawkins> Ok, thanks Piraty. Is a release effectively just renaming edge to 'x.xx', then a new edge created from there?
2023-01-27 09:41:08 <adhawkins> Or is there a chance some packages might not make it from edge to the new release?
2023-01-27 09:42:54 <Piraty> adhawkins: take a look at branch points in aports repo to find out. and i assume (dunno) that the binary package repo is just copied over to match the new relase tag
2023-01-27 09:44:07 <adhawkins> Ok, thanks Piraty. What about the question about getting notified about new releaes?
2023-01-27 09:44:20 <adhawkins> Will apk 'know' that a new upstream is available, or do I need to monitor that and update the repos file?
2023-01-27 09:51:02 <Piraty> i assume there are RSS feeds to consume. not sure if alpine does [ANNOUNCE] on the ML, could monitor that if they do
2023-01-27 09:56:48 <Piraty> adhawkins: the last release is not too long ago, depending on the packages you may still be able to pull it off without too much pain. i would compare logs of both branches, filter for packages installed on the box and checking if core packages are affected that would make this too hard to accomplish
2023-01-27 10:00:06 <pj[m]> <adhawkins> "Follow up question, will I get a..." <- you can use `latest-stable` instead of specific version
2023-01-27 10:04:09 <pj[m]> <adhawkins> "Will apk 'know' that a new..." <- new versions are announced on ~alpine/announce@lists.alpinelinux.org / twitter / mastodon / alpinelinux.org website, APK doesn't have any way to know when there is new version or what even is a new version
2023-01-27 10:05:48 <Piraty> adhawkins: yes they do, just no [ANNOUNCE] https://lists.alpinelinux.org/~alpine/devel/%3C20221122163825.35366599%40ncopa-desktop.lan%3E
2023-01-27 10:06:24 <Piraty> announce of course
2023-01-27 10:19:19 <adhawkins> Thanks pj[m] and Piraty. Need to have a think. For some reason I thought releases were more often than every 6 months. What goes into the point releases? Is it just fixes for a package, or will new versions potentially end up in there too?
2023-01-27 10:19:41 <adhawkins> The box does my backups, so it's using borg and emborg, and would like to keep reasonably up to date with upstream versions.
2023-01-27 10:25:52 <WhyNotHugo> man pages for APKBUILDs are not published online, are they?
2023-01-27 10:29:31 <Piraty> they are in foo-doc usually
2023-01-27 10:30:40 <Piraty> WhyNotHugo: i do   podman run --rm -ti docker.io/library/alpine:latest sh -c 'apk add mandoc openssl-doc && man openssl'
2023-01-27 10:31:06 <Piraty> if i have no access to an alpine install
2023-01-27 10:31:07 <WhyNotHugo> I meant "online in HTML".
2023-01-27 10:31:13 <WhyNotHugo> E.g.: something that I can provide a link to.
2023-01-27 10:31:30 <Piraty> send this to a pastebin ;)
2023-01-27 10:34:54 <ikke> adhawkins: point releases are mostly relevant for installation media / docker images
2023-01-27 10:36:04 <ikke> If you follow a stable release, you will automatically get updates for that release, but we limit these mostly to security / bug fix updates. 
2023-01-27 10:36:52 <ikke> Rss for posts on the site: https://alpinelinux.org/atom.xml 
2023-01-27 10:38:44 <WhyNotHugo> How do I get init to reload /etc/inittab ?
2023-01-27 10:43:04 <grawity> telinit q
2023-01-27 10:55:30 <WhyNotHugo> greetd expects /run/greetd to exist, but /run is mounted tmpfs. What's a sane mechanism for creating this directory at startup?
2023-01-27 11:09:53 <donoban> it has checkpath -d -m750 -o greetd:greetd /run/greetd/ 
2023-01-27 11:10:01 <donoban> on initd
2023-01-27 11:14:56 <satmd> opentempfiles (port of a systemd equivalent) maybe, non-alpine suggestion, maybe still applies
2023-01-27 11:28:54 <WhyNotHugo> donoban: I'm running greetd via /etc/inittab.
2023-01-27 11:29:18 <WhyNotHugo> I'm not sure I fully graps how the openrc-service is supposed to work; it would collide with the getty instance definedin /etc/inittab
2023-01-27 11:29:32 <dwfreed> remove the getty defined in inittab then
2023-01-27 11:29:49 <WhyNotHugo> dwfreed: Yeah, that's what I did -- and replaced it with greetd
2023-01-27 11:30:03 <WhyNotHugo> Do you suggest deleting the line from inittab entirely and relying on the service itself instead?
2023-01-27 11:30:09 <dwfreed> yes
2023-01-27 11:30:40 <WhyNotHugo> Isn't there a chance of the service starting too early or something? I don't admit to fully understand this.
2023-01-27 11:30:47 <donoban> why don't just let openrc start it?
2023-01-27 11:31:06 <WhyNotHugo> donoban: I was mostly sticking to how things came out of the box --
2023-01-27 11:31:21 <WhyNotHugo> (e.g.: getty/login started via inttab)
2023-01-27 11:32:03 <WhyNotHugo> The openrc service would only start ONE greetd for the current tty. But tty2, tty3, etc wouldn't have a greetd instance.
2023-01-27 11:32:23 <dwfreed> you can instance openrc services pretty easily
2023-01-27 11:33:23 <WhyNotHugo> How's that?
2023-01-27 11:33:40 <dwfreed> ln -sv greetd /etc/init.d/greetd.tty1
2023-01-27 11:34:50 <dwfreed> then you can use either /etc/conf.d/greetd.tty1 or /etc/conf.d/greetd (and something like ${RC_SVCNAME#*.} to get what's after the .)
2023-01-27 11:35:35 <WhyNotHugo> Honestly, all that I want to do is replace getty+login with greetd. Shifting to running it via openrc seems to add a lot of complexity on top.
2023-01-27 11:35:57 <WhyNotHugo> There's no /etc/init.d/greetd.
2023-01-27 11:36:01 <dwfreed> you mean adds a lot of doing it the right way on top
2023-01-27 11:36:47 <WhyNotHugo> I don't follow.
2023-01-27 11:37:01 <WhyNotHugo> If this is "the right way", then why does alpine do it differently out-of-the-box with getty?
2023-01-27 11:37:24 <dwfreed> because getty has historically always been configured via inittab, because it has no requirements
2023-01-27 11:37:47 <dwfreed> also, the apkbuild for greetd installs the init.d script, so perhaps you deleted it?
2023-01-27 11:37:49 <dwfreed> https://github.com/alpinelinux/aports/blob/master/community/greetd/APKBUILD#L56
2023-01-27 11:38:02 <WhyNotHugo> Ah, I was missing greetd-openrc
2023-01-27 11:39:02 <donoban> altough your idea makes sense, I don't see too much support on it
2023-01-27 11:39:20 <WhyNotHugo> My idea?
2023-01-27 11:39:26 <donoban> yes
2023-01-27 11:39:41 <donoban> it seems more a replacament for gdm/sdm/ligthdm...
2023-01-27 11:40:10 <WhyNotHugo> Nope, it's a replacement to getty+login. It even looks the same!
2023-01-27 11:40:26 <WhyNotHugo> But with PAM (so closer to shadow-login)
2023-01-27 11:41:15 <donoban> Starting greetd
2023-01-27 11:41:18 <donoban> Enable greetd.service so greetd will be started at boot.
2023-01-27 11:41:25 <donoban> https://wiki.archlinux.org/title/Greetd there is no mention of inittab
2023-01-27 11:41:39 <WhyNotHugo> That's for systemd.
2023-01-27 11:41:51 <WhyNotHugo> systemd doesn't use inittab.
2023-01-27 11:44:25 <donoban> there is only a commit comment mentioning inittab on the greetd repository
2023-01-27 11:45:39 <donoban> anyway, whats your current problem? it fails due missing /run/greetd/?
2023-01-27 12:38:32 <WhyNotHugo> It shows an error on each run due to the missing $HOME...
2023-01-27 12:38:55 <WhyNotHugo> But actually, the user's home should not be in /run/ because it's often useful to have persistent files there.
2023-01-27 12:39:22 <WhyNotHugo> So probably moving it into /var/lib is best -- this is actually what the author suggested too.
2023-01-27 14:39:19 <vkrishn> userspace tools for fscrypt has not matured yet ? there is also fscryptctl
2023-01-27 15:28:13 <adhawkins> Thanks for the clarification re: releases ikke. I thought stable releases were more frequently than every 6 months. Probably best to stick with 'edge' for now. 
2023-01-27 15:28:50 <ikke> adhawkins: we rebuild from scratch for every stable release 
2023-01-27 15:30:46 <ikke> Note that edge may break occasionally
2023-01-27 15:31:31 <adhawkins> Appreciate that ikke. I don't update it every single day, and whenever I do I take a snapshot of the VM just in case. Should be safe.
2023-01-27 16:01:53 <elly> people who run edge in production are as brave as firefighters
2023-01-27 16:21:31 <salobulonet> 😊️
2023-01-27 16:28:30 <minimal> grawity: Alpine does not have "telinit"
2023-01-27 16:35:31 <minimal> vkrishn: I packages fscrypt a while ago. fscryptctl is a lot more limited so I didn't see the point in packaging it
2023-01-27 16:40:03 <invoked> been meaning to pick your brain minimal. on a lvm/luks/grub setup with cryptkey, cryptkey doesn't work. i don't really like using grub, but i've done this before and its always worked.
2023-01-27 16:40:19 <invoked> the only difference being that i'm using btrfs for root instead of what was normally ext4.
2023-01-27 16:40:48 <invoked> i don't know how to debug that.
2023-01-27 16:41:07 <minimal> invoked: what exactly do you mean that it doesn't work? I've used cryptkey before when using Grub's native LVM support to avoing the initramfs' init asking for the passphrase a 2nd time (after Grub asks)
2023-01-27 16:41:45 <invoked> exactly. it doesn't work. i get prompted
2023-01-27 16:42:03 <invoked> i've added the key/etc, i know how it's supposed to work
2023-01-27 16:42:05 <minimal> can you provide more info regarding the steps you took to set it up?
2023-01-27 16:43:24 <invoked> cryptkey is on the grub kernel line, crypto_keyfile.bin is in /, it's been added, etc. not much more to say
2023-01-27 16:43:43 <invoked> i'm just wondering if it's a btrfs thing.
2023-01-27 16:46:52 <minimal> invoked: I don't really use btrfs much. Let me spin up a VM with ext4/LVM/LUKS/Grub 1st and verify that it works as expected
2023-01-27 16:46:59 <minimal> then I can try btrfs
2023-01-27 16:48:34 <invoked> wait, don't do that. i was just wondering if you knew something off the top of your head
2023-01-27 16:48:43 <minimal> invoked: is this BIOS or UEFI booting? LUKS v1 or v2?
2023-01-27 16:48:46 <invoked> i don't really care enough to fix this right now, it's just annoying
2023-01-27 16:48:48 <invoked> uefi
2023-01-27 16:49:05 <minimal> invoked: I have the VM creation all scripted so it only takes a couple of minutes to test
2023-01-27 16:49:17 <invoked> roger that
2023-01-27 16:49:30 <minimal> LUKS v1 or v2?
2023-01-27 16:49:37 <invoked> 1
2023-01-27 17:03:05 <donoban> I'm using btrfs over LUKS and noticed some time ago that I can't use fstrim
2023-01-27 17:03:22 <donoban> fstrim: /: the discard operation is not supported
2023-01-27 17:03:35 <invoked> this is a new machine, let me try that.
2023-01-27 17:04:06 <minimal> invoked: ext4/LVM/LUKS/Grub boots as expected - only prompted by Grub for passphrase
2023-01-27 17:04:07 <donoban> I wonder if that its expected or I'm losing some free space, properly wearing or something like that
2023-01-27 17:04:50 <vkrishn> minimal: https://github.com/google/fscryptctl/tags is v1.0.0
2023-01-27 17:05:09 <vkrishn> fscrypt, v0.3.3
2023-01-27 17:05:30 <vkrishn> but yes I haven't looked at the code
2023-01-27 17:05:40 <minimal> vkrishn: as I said earlier fscryptctl has far less functionality than fscrypt
2023-01-27 17:06:01 <invoked> minimal: welp. thanks for checking that it's not a grub lvm issue. i don't know why it'd be a btrfs issue though.
2023-01-27 17:06:04 <vkrishn> yes, thanks
2023-01-27 17:06:30 <minimal> "fscryptctl does not handle key generation, key stretching, key wrapping, or PAM integration. Most users should use the fscrypt tool instead"
2023-01-27 17:07:31 <invoked> donoban: i forgot, it's nvme, nevermind
2023-01-27 17:08:17 <donoban> yes it's a nvme disk
2023-01-27 17:08:33 <invoked> yeah, you don't trim nvme devices
2023-01-27 17:08:57 <donoban> ahh, I didn't know
2023-01-27 17:09:18 <donoban> it's a SATA command?
2023-01-27 17:09:26 <invoked> so if you have discard in your fstab for your nvme device, remove it
2023-01-27 17:09:39 <donoban> I added, discard=async
2023-01-27 17:09:56 <donoban> but I didn't see anything related on dmesg
2023-01-27 17:10:05 <minimal> invoked: eh? NVME devices should be able to be trimmed
2023-01-27 17:10:45 <minimal> though I wouldn't do so via fstab discard option but rather running fstrim at intervals (i.e. weekly)
2023-01-27 17:10:56 <invoked> yeah rather than ircsplain this let me see if i can source something.
2023-01-27 17:15:48 <invoked> i can't find it, i've got people blowing up my phone, sorry (memoryhole my past comments and i'll try to come back to this.)
2023-01-27 17:17:30 <minimal> invoked: nvme-cli source code has multiple references to "trim"
2023-01-27 17:18:08 <f_> Hello! Does Alpine's linux kernel use the default arm64 defconfig for arm64=
2023-01-27 17:18:11 <f_> ?
2023-01-27 17:18:33 <minimal> though only in the WDC, OCP, Seagate, and Micron specific modules
2023-01-27 17:18:38 <invoked> minimal: it may really come down to specific hardware, which is why ... yeah
2023-01-27 17:18:54 <invoked> in other words, was was true yesterday is "mixed" now
2023-01-27 17:19:14 <invoked> but i want to be specific where i was speaking from, so i'll come back to it
2023-01-27 17:19:35 <ikke> f_: These are the settings we set for the aarch64 kernel (arm64): https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/linux-lts/lts.aarch64.config
2023-01-27 17:19:46 <ikke> not sure if and how much it deviates from defconfig though
2023-01-27 17:21:33 <minimal> donoban: do you have "discard" added in /etc/crypttab?
2023-01-27 17:21:57 <donoban> no, only on fstab, let me check anyway
2023-01-27 17:21:58 <minimal> also "lsblk -D" should give an indication if discard/trim is available
2023-01-27 17:22:34 <invoked> this type of thing happened in the early days of sata ssd and trim. some devices had their own gc and trim would just mess things up. eg sandforce
2023-01-27 17:22:39 <donoban> uhm, I don't have cryptab
2023-01-27 17:22:49 <donoban> I think its mounted via kernel params
2023-01-27 17:22:54 <minimal> donoban: as your btrfs is "layered" on top of LUKS then you need discard enabled for LUKS or discard/trim won't make it all the may to the underlying device
2023-01-27 17:23:28 <ikke> apparently there are security implications though
2023-01-27 17:23:49 <donoban> there isn't any mention to discard on blkid -d
2023-01-27 17:23:56 <minimal> donoban: then add "cryptdiscards=yes" to your cmdline
2023-01-27 17:24:23 <donoban> what would be exactly the benefit? prolong the lifetime for the disk?
2023-01-27 17:24:46 <invoked> performance mainly. chances are the disk will outlive you if it's made well
2023-01-27 17:24:49 <ikke> I think it helps wear leveling
2023-01-27 17:25:24 <donoban> ok, I will check the security implications and consider it
2023-01-27 17:25:28 <minimal> I also leave 10% of SSDs unallocated (unpartitioned) to help with wear-levelling
2023-01-27 17:26:04 <donoban> I read that some disk do that on the background too
2023-01-27 17:26:08 <donoban> not sure of my case
2023-01-27 17:26:27 <minimal> ikke: lts.aarch64.config defines *exactly* how it deviates from defconfig doesn't it? (ever since the kernel package switched from using full configuration settings)
2023-01-27 17:27:10 <ikke> minimal: oh, may be
2023-01-27 17:28:03 <minimal> donoban:  my earlier point is that there are 2 ways to use discard/trim - add it as an option in fstab for the filesystem and then trim is "continual", or run fstrim at intervals (i.e. via cron) instead. Many people recommend the fstrim method
2023-01-27 17:29:24 <donoban> I read that using discard could produce unexpected performance degradation when it gets activated, and doing it at fixed interval is more predictable
2023-01-27 17:29:50 <ikke> that's what I read as well
2023-01-27 17:30:04 <adhawkins> elly: I haven't had a single issue running edge. The collection of packages installed is minimal though. Basically a 'base' system plus emborg.
2023-01-27 17:30:05 <donoban> honestly on this laptop I don't care too much about some temporary performance degradation
2023-01-27 17:30:33 <ikke> adhawkins: be carefull around python upgrade times then :)
2023-01-27 17:30:56 <donoban> or enable cache for have a way to downgrade :D
2023-01-27 17:31:06 <adhawkins> Heh, fair point ikke. That's why I only update it maybe once a week, and then take a snapshot of it beforehand.
2023-01-27 17:50:57 <minimal> invoked: I hit an issue with my script and btrfs lol, haven't tested btrfs for a while so need to fix it.
2023-01-27 17:54:28 <minimal> invoked: anyway, just to check your setup, can you confirm the following: (a) that you created a keyfile, (b) that you added the keyfile to the LUKS device (cryptsetup luksAddKey), (c) that you put the keyfile somewhere on the filesystem and set its permissions to 400, (d) that you added an entry in /etc/crypttab for the LUKS partition which specified the keyfile
2023-01-27 17:58:03 <minimal> (e) that you added "cryptkey" to /etc/mkinitfs/features.conf and re-ran "mkinitfs" to regenerate the mkinitfs (so the keyfile is in initramfs file), so and a correction for (c) above - that you named the keyfile as "/crypto_keyfile.bin"
2023-01-27 18:08:42 <invoked> (a) yes (b) yes (c) yes, well 600 but fixed to 400 (d) no crypttab (e) cryptkey is already there but i will regenerate it, and retest as soon as i'm off this call.
2023-01-27 18:15:09 <invoked> ok. working now. ugh, it must have been the initramfs. but i added cryptkey to mkinitfs.conf (you wrote features.conf, heh) before i generated it the first time, pretty sure.
2023-01-27 18:15:32 <invoked> anyway, i don't know but obviously i did something wrong.
2023-01-27 18:15:53 <invoked> thanks for the assist, minimal
2023-01-27 18:15:55 <minimal> is working now, that's what matters ;-)
2023-01-27 18:16:10 <invoked> i'm really overjoyed that this is not some btrfs thing.
2023-01-27 18:18:00 <invoked> tomorrow: adding encfs to that (and screwing that up somehow)
2023-01-27 18:21:56 <minimal> invoked: why encfs rather than fscrypt?
2023-01-27 18:22:53 <minimal> especially as the last commit I see to encfs' github repo is May 2020...
2023-01-27 18:24:50 <invoked> misspoke, i meant encfs/fscrypt/et al generally
2023-01-27 18:24:58 <invoked> i haven't made decisions yet.
2023-01-27 20:17:06 <Leolo> so, turns out that the Extended ISO doesn't fit on a CD-R.  and it doesn't have Shorewall.  so I can't just boot it and have it work
2023-01-27 20:17:12 <Leolo> imma need to build my own ISO
2023-01-27 20:18:19 <Ermine> Leolo: try DVD disc. It should have enough space
2023-01-27 20:18:40 <Leolo> I have a lot of CD-R lying around.  I don't feel I should buy DVD-R
2023-01-27 20:18:49 <Leolo> also, my router has a CD-ROM drive in it, not DVD
2023-01-27 20:19:05 <Leolo> I remember booting my router from a write protected floppy :D
2023-01-27 20:22:08 <mor> 256kbit/s
2023-01-27 20:26:01 <Leolo> so my goal is to add screen, joe.  and get it onto a CD-R.  most of the large apks are kernel-firmware.  those aren't listed explicitly in aport.  I assume they are deps of kernel-lts 
2023-01-27 20:27:05 <donoban> they are automicatlly removed if you install some linux-firmware-xxx
2023-01-27 20:27:19 <donoban> you can get a try with linux-firmware-none
2023-01-27 20:30:13 <minimal> Leolo: wow another joe user! ;-)
2023-01-27 20:31:36 <Ermine> Leolo: there's also standard ISO, which is way smaller. It doesn't suit your needs I guess?
2023-01-27 20:39:40 <Leolo> minimal - I started using Wordstar in the 80s.  then qedit, the tse32, and now joe.  my fingers Just Know the key commands
2023-01-27 20:40:10 <Leolo> ermine - I will admit that I didn't try it.  I was hoping that Extended had what I wanted.
2023-01-27 20:41:08 <Leolo> I think this isn't a task I should start work on at 15h30 on a friday :D
2023-01-27 20:41:44 <minimal> Leolo: same here, too much time spent in Wordstar ;-)
2023-01-27 20:47:00 <minimal> Leolo: old-timers of the IT World unite! :-)
2023-01-27 21:58:01 <Guest28> Hello, suppose I want to automatically make a sysinstall image for pi3 / pi zw 2, but my main system is x86, do I need this and QEMU to do so? https://github.com/alpinelinux/alpine-chroot-install
2023-01-27 21:58:37 <ikke> Guest28: qemu-user-static + binfmt would probably work as well
2023-01-27 22:00:42 <Guest28> Interesting stuff, I'll check it out, thanks.
2023-01-27 22:01:51 <Habbie> that script sets those up if necessary, it appears
2023-01-27 22:02:48 <Guest28> Does binfmt require Docker?
2023-01-27 22:02:56 <ikke> no
2023-01-27 22:03:04 <ikke> binfmt is a kernel feature
2023-01-27 22:03:22 <Guest28> Oh then I was confusing it with this: https://github.com/tonistiigi/binfmt
2023-01-27 22:03:27 <ikke> (there is a docker image that sets it up, but it works perfectly fine without it)
2023-01-27 22:03:29 <ikke> right, that
2023-01-27 22:19:06 <tetrakist> Does alpine linux include a "printf" binary that accepts unicode codepoints like "\u2388"?
2023-01-27 22:23:33 <ikke> coreutils provides printf, but it seems to just print out the code verbatim
2023-01-27 22:23:56 <lopid> https://www.gnu.org/software/coreutils/manual/coreutils.html#printf-invocation suggests why
2023-01-27 22:25:01 <tetrakist> Thanks.
2023-01-27 22:25:18 <ikke> lopid: missing iconv?
2023-01-27 22:25:37 <lopid> it would seem so. that euro example works here
2023-01-27 22:26:03 <ikke> yes, for me as well
2023-01-27 22:26:09 <Leolo> ERROR: linux-firmware-other-20230117-r0: No space left on device
2023-01-27 22:26:38 <lopid> i don't appear to have iconv, though…
2023-01-27 22:26:45 <ikke> env printf '\u2388' works for me as well
2023-01-27 22:26:56 <ikke> but not command printf '\u2388'
2023-01-27 22:27:22 <lopid> ah
2023-01-27 22:27:53 <Mag242> hi
2023-01-27 22:28:17 <lopid> yes, it would appear the bash builtin supports it already
2023-01-27 22:28:36 <ikke> ash doesn't
2023-01-27 22:28:42 <ikke> but coreutils printf does
2023-01-27 22:29:15 <lopid> heh. /usr/bin/printf doesn't, here
2023-01-27 22:29:36 <ikke> does for me
2023-01-27 22:29:41 <ikke> after installing coreutils
2023-01-27 22:30:14 <lopid> oh, i don't have coreutils. i thought i had
2023-01-27 22:34:19 <tetrakist> lopid, what does printf "\342\201\204" output for you?
2023-01-27 22:35:01 <lopid> just a "/"
2023-01-27 22:35:09 <tetrakist> Ah, good.
2023-01-27 22:35:11 <tetrakist> Thanks.
2023-01-27 22:35:22 <Mag242> Got someone an irc-client working under iSH on iPad? If yes, which one?
2023-01-27 22:35:24 <lopid> do you now have a backdoor into my system?
2023-01-27 22:35:46 <tetrakist> No, now I know I can create files with "/" in the name.
2023-01-27 22:36:44 <ikke> sounds cursed
2023-01-27 22:37:14 <tetrakist> it is, but the alternative made me feel like the inside of a horse.
2023-01-27 22:54:27 <Quillith> If my repositories are set to https://[mirror]/alpine/edge/main and same but for community and testing, is it normal for my cat /etc/alpine-release to say 3.17.0? My partner is on latest-stable and his output is 3.17.1
2023-01-27 22:55:27 <ikke> Quillith: does /etc/alpine-release.apk-new exist?
2023-01-27 22:56:05 <Quillith> /bin/ash: /etc/alpine-release.apk-new: not found
2023-01-27 22:57:25 <ikke> Quillith: is your partner on a stable release?
2023-01-27 22:57:34 <ikke> apparently alpine-base on edge is still 3.17.0
2023-01-27 22:58:01 <ikke> For 3.17-stable, 3.17.1 was released
2023-01-27 22:58:11 <ikke> but that is not applied to edge
2023-01-27 22:58:13 <Quillith> his repositories say http://[mirror]/alpine/latest-stable/main
2023-01-27 22:58:17 <Quillith> ohhh okay, that makes sense
2023-01-27 22:58:31 <ikke> right, so it's expected
2023-01-27 22:58:45 <Quillith> Thank you for explaining
2023-01-27 23:42:57 <Leolo> following https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage I get that error
2023-01-27 23:49:07 <Leolo> ERROR: Failed to create boot/vmlinuz-lts: No space left on device
2023-01-27 23:49:07 <Leolo> ERROR: linux-lts-5.15.90-r0: No space left on device
2023-01-27 23:51:57 <minimal> perhaps there's no space left on the device? ;-)
2023-01-27 23:55:38 <bl4ckb0ne> i think your devicr has no space left
2023-01-28 02:05:01 <vkrishn> how to do something like aespipe in al ?
2023-01-28 02:14:52 <tetrakist> Leolo, one of the common reasons for this is that /boot partitions are often created too small. These days I wouldn't make one less than 500MB, with 1G being better.
2023-01-28 02:16:10 <tetrakist> vmlinuz + initrd can easily exceed 50MB, so if you have multiple kernels installed, then away goes the available space.
2023-01-28 02:18:22 <tetrakist> Uninstalling unused/unneeded kernels can help, though a large enough partition will avoid that altogether. Resizing the /boot partition is a possible solution, but as the /boot partition is generally first on the drive followed closely by other partitions, there may not be room to resize the partition directly. 
2023-01-28 02:19:54 <tetrakist> Finally, if you dig into the scripts used to create the initrds, sometimes they can be pointed to another directory like /tmp, instead of building the new image directly in /boot. This can allow the creation process to complete, but then requires you to manually move the created image to the /boot partition yourself, and to deal with any space issues.
2023-01-28 02:21:04 <tetrakist> This type of problem popped up on the postmarketOS installs on the pinephone, which had a boot partition of only 128MB, initially. If I recall correctly, that is.\
2023-01-28 02:28:35 <Leolo> "devicr" or "/boot" but I suspect it's because the VM I'm doing the build on doesn't have enough memory
2023-01-28 02:28:37 <Leolo> only 4GB
2023-01-28 02:30:10 <Leolo> make that 2GB and 4GB swap
2023-01-28 02:32:57 <Leolo> wait, virsh says the VM has 4GB memory, but free only shows 2GB.  this is a 32bit VM, to build 32bit ISO.  is there a hard limit to mem in a 32bit alpine install?
2023-01-28 02:33:28 <psykose> don't think s0
2023-01-28 02:33:30 <psykose> so*
2023-01-28 02:35:55 <Leolo> is at 3%, /boot is at 8%  so it's not the disk being full, but (I suspect) initrd build or other
2023-01-28 02:43:23 <Leolo> aha, /tmp is only 1GB.  
2023-01-28 02:43:28 <Leolo> 
ACTION sets up a ~/tmp

2023-01-28 02:48:40 <Leolo> yeah, that fixed it
2023-01-28 02:52:07 <Leolo> but now it dies with Could not open file or uri for loading private key from
2023-01-28 03:44:05 <Leolo> problem is that PACKAGER_PRIVKEY isn't set anywhere
2023-01-28 03:44:44 <psykose> you do need one
2023-01-28 03:44:57 <Leolo> it wasn't in the instructions :)
2023-01-28 03:45:21 <psykose> SUDO=doas abuild-keygen -ai sets it up, assuming a user with configured doas
2023-01-28 03:48:16 <Leolo> doas: Operation not permitted
2023-01-28 03:56:20 <Leolo> SUDO=sudo abuild-keygen did the trick
2023-01-28 03:57:35 <Leolo> thank fuck these programs are shell scripts
2023-01-28 04:44:31 <Leolo> ok, now I'm even more confused; it's not creating ISOs, it's creating .tar.gz files
2023-01-28 04:48:36 <psykose> it's the default for arm images or minirootfs's
2023-01-28 05:02:48 <Leolo> gotcha.  I was using a small image so my tests go faster
2023-01-28 08:53:02 <Saijin_Naib> unrar was dropped after Alpine v3.14?
2023-01-28 08:56:52 <psykose> ye
2023-01-28 08:57:15 <Saijin_Naib> lepoop
2023-01-28 08:57:31 <Saijin_Naib> Our 7zip implementation is built without RAR support it seems
2023-01-28 09:01:49 <Saijin_Naib> Thankfully WINE exists. Windows 7zip to the rescue
2023-01-28 09:02:30 <psykose> bsdtar has some support
2023-01-28 09:02:39 <psykose> 7zip rar support is the same unrar code
2023-01-28 09:03:24 <Saijin_Naib> Does it make sense to file an Issue to have 7zip built with that feature enabled?
2023-01-28 09:03:35 <Saijin_Naib> I'm not sure many would know that bsdtar has that support, haha
2023-01-28 09:08:33 <psykose> no, because it's the same as unrar
2023-01-28 09:08:40 <psykose> like it's literally the same code
2023-01-28 09:08:45 <psykose> just copied into 7zip
2023-01-28 11:57:22 <mitsukage> hi everyone
2023-01-28 12:41:57 <Apachez> greetings earthling
2023-01-28 14:12:30 <minimal> tetrakist: Alpine's mkinitfs utility creates a directory inside /tmp/ for assembling the new initramfs unless you specify the "-t <directory>" option. Also no a Sys-mode install /tmp is typically a tmpfs-based directory rather than part of rootfs
2023-01-28 14:13:24 <tetrakist> minimal, excellent.
2023-01-28 14:14:21 <minimal> tetrakist: also Alpine only keeps 1 version of a kernel package (unless you install linux-lts, linux-virt, and/or linux-edge which are each distinct packages) so you won't have more than 1 kernel installed and taking space in /boot (except briefly when a kernel package is updated)
2023-01-28 14:16:13 <tetrakist> minimal, even more excellenter.
2023-01-28 14:17:11 <tetrakist> I am definitely reminded of the old adage that the best way to get a question answered is to post the wrong answer. ;)
2023-01-28 14:17:45 <tetrakist> I think my info was a year or two out of date. :)
2023-01-28 14:20:14 <minimal> tetrakist: so the minimal (pun intended) size of /boot depends on (a) whether UEFI is in use and if /boot is also the ESP partition and so storing EFI files, (b) the size of the kernel file (but NOT modules are they are in /lib/modules), the size of any ucode files, and the size of the initramfs (which depends on firmware packages installed, modules selected, compression used, etc). I'm not sure exactly how "apk upgrade" works and so how 
2023-01-28 14:20:14 <minimal> the any replacement kernel files end up in /boot so I assume /boot also needs to have sufficient freespace for another (temporary) set of those kernel files
2023-01-28 14:21:13 <minimal> s/size of kernel file/size of kernel, config, and System.map files/
2023-01-28 14:21:31 <minimal> oh and also size of bootloader files as well
2023-01-28 14:23:42 <mitsukage> i wanted to install Alpine Linux with awesome window manager and failed
2023-01-28 14:24:03 <minimal> awesome! ;-)
2023-01-28 14:25:04 <mitsukage> i chosen lightdm as login manager
2023-01-28 14:25:04 <mitsukage> lightdm didn't showed
2023-01-28 14:25:04 <mitsukage> and after reboot
2023-01-28 14:25:20 <mitsukage> i saw - blinking
2023-01-28 14:25:40 <mitsukage> and black screen
2023-01-28 14:25:55 <mitsukage> how i fix that ?
2023-01-28 14:26:25 <minimal> channel Groundhog Day? ;-) discussion about RAR again today, this was previously discussed on 16th
2023-01-28 14:27:25 <mitsukage> is that xorg problem or a missing package ?
2023-01-28 17:11:04 <Saijin_Naib> <minimal> "channel Groundhog Day? ;-) discu..." <- Didn't see it then 🤣
2023-01-28 17:11:55 <Saijin_Naib> First time since switching FROM Windows that I got Geospatial data in RAR and was surprised that I couldn't open it with 7zip library and XArchiver
2023-01-28 18:16:30 <invoked> crap bug report: noticed /sbin/setup_interfaces wouldn't find (or populate a list correctly) with an ssid i could otherwise find and connect to using 'iw scan' normally
2023-01-28 18:17:00 <invoked> i peeked at the script briefly but can't figure out where it goes wrong.
2023-01-28 18:18:33 <invoked> also it's a 4-character ssid, all lowercase letters, nothing strange.
2023-01-28 18:46:07 <ikke> invoked: what authentication method does it use?
2023-01-28 18:57:33 <invoked> ikke: http://ix.io/4mse
2023-01-28 19:00:19 <ikke> So WPA1 PSK?
2023-01-28 19:00:41 <invoked> if what iw is reporting is correct.  it's a wifi6 router.
2023-01-28 19:02:43 <invoked> it's a white-labeled nokia 5g internet router
2023-01-28 19:02:51 <invoked> isp-controlled
2023-01-28 19:09:49 <ikke> invoked: appears the "SSID List" in the output breaks it
2023-01-28 19:11:12 <invoked> good catch
2023-01-28 19:11:32 <invoked> i figured you needed the whole thing.
2023-01-28 19:12:27 <ikke> invoked: can you try https://tpaste.us/0lNx
2023-01-28 19:13:11 <invoked> sure. i need about 10min though, cooking
2023-01-28 19:13:14 <ikke> np
2023-01-28 19:16:15 <invoked> ikke: fixed
2023-01-28 19:16:19 <ikke> cool
2023-01-28 19:16:46 <ikke> I'll make an MR for that
2023-01-28 19:17:04 <invoked> sorry
2023-01-28 19:18:06 <ikke> Why?
2023-01-28 19:18:55 <invoked> normally i would try to do a proper bug report, etc
2023-01-28 19:19:08 <ikke> Oh, don't worry
2023-01-28 19:22:45 <minimal> invoked: BTW I fixed the script and confirmed btrfs+Grub+LVM+LUKS using keyfile ;-)
2023-01-28 19:24:41 <invoked> nice
2023-01-28 19:28:24 <invoked> my next adventure with that laptop is getting all the keys to work, and then power management. power management in linux, one of my least favorite things
2023-01-28 19:30:05 <scorpion2185[m]1> can you hibernate?
2023-01-28 19:30:41 <invoked> in theory :)
2023-01-28 19:30:44 <scorpion2185[m]1> what does it mean when you launch an  ELF 64-bit LSB pie executable, and it says file not found? but it's there
2023-01-28 19:31:38 <ikke> scorpion2185[m]1: compiled against glibc?
2023-01-28 19:31:41 <ikke> Wrong arach?
2023-01-28 19:32:26 <minimal> scorpion2185[m]: why do you frequently ask a question and then immediately follow it with another completely unrelated question?
2023-01-28 19:32:39 <scorpion2185[m]1> i downloaded the bin so glibc thing I guess
2023-01-28 19:33:44 <scorpion2185[m]1> right arch
2023-01-28 19:35:09 <scorpion2185[m]1> https://github.com/hyprwm/Hyprland/releases/tag/v0.21.0beta
2023-01-28 19:35:56 <scorpion2185[m]1> ikke so "file not found" means glibc?
2023-01-28 19:36:12 <ikke> It means it cannot find the dynamic linker that is set in the binary
2023-01-28 19:36:53 <scorpion2185[m]1> interpreter /lib64/ld-linux-x86-64.so.2
2023-01-28 19:37:08 <ikke> yes
2023-01-28 19:37:12 <ikke> that does not exist on alpine
2023-01-28 19:37:36 <ikke> as that's glibc
2023-01-28 19:37:52 <scorpion2185[m]1> can I put a link to the alpine version?
2023-01-28 19:38:12 <ikke> no, it does not work like that
2023-01-28 19:38:38 <scorpion2185[m]1> I can't build it
2023-01-28 19:38:40 <ikke> scorpion2185[m]1: try `ldd Hyprland` and you'll see that it is missing a lot more
2023-01-28 19:39:23 <scorpion2185[m]1> ldd Hyprland
2023-01-28 19:39:24 <scorpion2185[m]1> /lib/ld-musl-x86_64.so.1: cannot load Hyprland: No such file or directory
2023-01-28 19:39:38 <invoked> what do you mean you can't build it
2023-01-28 19:40:07 <scorpion2185[m]1> invoked protocols/meson.build:38:1: ERROR: File //usr/local/share/hyprland-protocols/protocols/hyprland-toplevel-export-v1.xml does not exist.
2023-01-28 19:40:23 <ikke> invoked: the build system is "half-melted pottery" according to someone here
2023-01-28 19:40:43 <invoked> that's a metaphor, i guess
2023-01-28 19:40:48 <ikke> yes
2023-01-28 19:41:45 <invoked> scorpion2185[m]1: sounds like you probably need this first.  https://github.com/hyprwm/hyprland-protocols
2023-01-28 19:41:51 <invoked> that's as much as i'm going to help you though
2023-01-28 19:42:31 <scorpion2185[m]1> thanks
2023-01-28 19:44:35 <scorpion2185[m]1> uhm better ask there are no instruction
2023-01-28 19:44:48 <scorpion2185[m]1> they don't say that you need it anywhere
2023-01-28 19:47:49 <invoked> maybe it's not written for people who don't know how these programs are put together
2023-01-28 19:48:28 <ikke> Or it's 'open source' but we are not making it easy for you to build yourself
2023-01-28 19:50:09 <psykose> it's a subproject of hyprland in a submodule you probably forgot to update
2023-01-28 20:11:34 <scorpion2185[m]1> the wiki doesn't mention submodules
2023-01-28 20:12:38 <scorpion2185[m]1> git clone --recursive https://github.com/hyp...
2023-01-28 20:13:32 <invoked> if you know what you're doing, they don't need to mention it. (hint: .gitmodules)
2023-01-28 20:14:27 <scorpion2185[m]1> git submodule update --init?
2023-01-28 20:17:50 <scorpion2185[m]1> A wiki mentions all the steps
2023-01-28 21:35:42 <vkrishn> codeberg is up for me, i guess it was upgrade to forgejo
2023-01-28 21:37:14 <ikke> vkrishn: context?
2023-01-28 21:38:51 <vkrishn> it was in-accessible from my ip couple of days back , i am guessing region/country wise(due to upgrade)
2023-01-28 21:40:38 <vkrishn> if someone is free, pls pkgs aespipe, if something similar is not available !
2023-01-28 21:49:02 <psykose> not against packaging this necessarily, but it just looks more outdated than using the regular crypto utils
2023-01-28 21:49:26 <psykose> openssl can stream data from i/o with a cipher and password too
2023-01-28 21:49:38 <psykose> same for e.g. the gnutls utils pretty sure, etc
2023-01-28 21:49:54 <psykose> obv not as ergonomic but, for a crypto utility i'd probably trust that more than this sourceforge page :p
2023-01-28 21:50:33 <psykose> https://stackoverflow.com/questions/10106771/encrypt-a-string-using-openssl-command-line small example
2023-01-28 21:55:41 <vkrishn> ok, that should work, trying to figure out a to-from compat method for string encoded in web(javascript aes)
2023-01-28 21:58:59 <psykose> ah, that's easily decoded pretty sure
2023-01-28 21:59:31 <psykose> if you know the cipher and the encoding into text (e.g. base64) then it's very simple with the above
2023-01-28 22:03:18 <vkrishn> i did that couple of years back, but with all args iv,hmackey, salt...
2023-01-28 22:04:01 <vkrishn> even made a nice common login app for all my web applications
2023-01-28 22:04:33 <vkrishn> think , i should upload that soonish
2023-01-28 22:18:46 <vkrishn> trying to modifying my https://gist.github.com/insteps/db43bdf0fcbca1ac57afc6277332ffee script 
2023-01-28 23:37:20 <boombim[m]> Are software from edge branch tested by someone?
2023-01-28 23:37:51 <Leolo> you 
2023-01-28 23:37:53 <Leolo> :D
2023-01-28 23:38:06 <boombim[m]> Really?
2023-01-28 23:43:34 <elly> well, I tried booting 3.17 on my 2014 macbook pro, and it gets as far as "Booting `Linux lts'"
2023-01-28 23:45:04 <elly> I wonder if it's actually hung or it's just VERY slow
2023-01-28 23:47:55 <elly> aw yeah it is just slow! score
2023-01-28 23:47:56 <elly> it booted
2023-01-28 23:48:56 <invoked> could be a device holding it up.
2023-01-29 00:06:09 <vkrishn> psykose: thanks, i think sjcl(browser) is quite updatee, will try doing thing around it, should work !
2023-01-29 00:06:22 <vkrishn> updated^
2023-01-29 00:41:11 <elly> rats, this macbook's wifi doesn't work in the installer and it is so old that it doesn't have usb-c
2023-01-29 00:43:47 <psykose> more likely that ya think, something something
2023-01-29 00:47:50 <invoked> probably a broadcom softmac
2023-01-29 00:47:55 <invoked> b43 or something
2023-01-29 00:48:13 <invoked> then you need the tool to extract the firmware for your particular build
2023-01-29 00:49:03 <invoked> https://wiki.archlinux.org/title/Broadcom_wireless this might help if that being the case
2023-01-29 01:06:07 <Quillith> Hello, I think there may be a small bug in whatever script makes the /etc/apk/repositories because I've done a few new installs over the last few days and I've found that there's a double set of the same repos, so like two paragraphs of the same thing. I don't think it changes or hurts anything but thought I'd mention it
2023-01-29 01:08:05 <minimal> Quillith: how did you do the installs? setup-alpine? in which mode - Sys mode (run from disk), Data mode, or run-from-ram mode?
2023-01-29 01:08:07 <dwfreed> invoked: if it's a bcm4360 like mine, it is not supported by anything other than the broadcom proprietary driver
2023-01-29 01:08:22 <Quillith> setup-alpine in cryptsys
2023-01-29 01:09:12 <dwfreed> elly: check the device ID, if it's 4360 or 43a0, wifi will only work with the wl driver
2023-01-29 01:10:11 <dwfreed> thunderbolt does work, but not hotplug as far as I know, so you'd have to boot the device with any thunderbolt devices you want to use
2023-01-29 01:10:29 <Quillith> This newest install is aarch64 on a raspberry pi if it matters
2023-01-29 01:10:33 <dwfreed> (eg, I have a thunderbolt ethernet adapter I use sometimes)
2023-01-29 01:12:15 <invoked> dwfreed: true. there is no choice but nonfree on some of them, which i had to do until i could get a travel router and use that instead
2023-01-29 01:35:10 <minimal> Quillith: setup-apkrepos appends to the /etc/apk/repositories file so if it was called multiple times (e.g. you ran setup-alpine more than once) then yes you would get multiple sets of entries
2023-01-29 01:41:29 <elibrokeit> ikke: it's not half-melted pottery, it is ***artisanal*** half-melted pottery :P
2023-01-29 01:50:35 <Quillith> minimal: thank you for explaining, my power did cut out partway through, that's probably why. I appreciate your time, thanks
2023-01-29 04:33:09 <nic> hello new to alpine, migrating from debian. stuck on install on old hp laptop, xorg wont start i think it isnt finding my intel driver or device?
2023-01-29 05:27:11 <Saijin_Naib> Welcome, nic! Does your laptop happen to have the awful NVIDIA Optimus platform? That was a struggle for me when getting a friend setup in Alpine 
2023-01-29 05:27:34 <Saijin_Naib> It'd be a Windows 7-era machine
2023-01-29 05:28:27 <Saijin_Naib> Also, did you happen to run the setup-xorg script to ensure all required packages were installed for you? 
2023-01-29 07:17:43 <ikke> Saijin_Naib: fyi, nic left
2023-01-29 10:26:56 <quinq> Hello
2023-01-29 10:27:40 <quinq> I upgraded my Apline install, and postgresql version has been upgraded do 15 (was 14 before)
2023-01-29 10:28:04 <quinq> Now the DB contents are empty, but I suppose data is still available on the 14 directory
2023-01-29 10:28:17 <quinq> Question is how to move data from 14 to 15?
2023-01-29 10:28:39 <quinq> I need to somehow dump data from 14, but not sure how
2023-01-29 10:38:45 <hjaekel> quinq You should have dumped the database before upgrading.
2023-01-29 10:40:46 <quinq> apk didn't warn me, it just upgraded directly
2023-01-29 10:41:20 <quinq> I would have for sure otherwise
2023-01-29 11:02:31 <Ermine> don't you need to switch version explicitly with pg_versions? There's also an utility to migrate to newer version
2023-01-29 11:03:07 <quinq> No, it's been done automatically
2023-01-29 11:03:22 <quinq> I managed to reinstall manually the older version
2023-01-29 11:03:40 <quinq> But the pgsql migration utility didn't work
2023-01-29 11:04:11 <quinq> I finally managed to dump the data from old db and import it back in the new one though
2023-01-29 11:42:54 <ikke> Note that you can still install v14 explicitly 
2023-01-29 11:43:36 <quinq> 12:03:22  quinq$ I managed to reinstall manually the older version
2023-01-29 11:43:39 <quinq> ^-- yes
2023-01-29 13:04:54 <tetrakist> Hi. Is there any reason one wouldn't be able to install pulseaudio-bluez?
2023-01-29 13:05:25 <tetrakist> When I run "apk install pulseaudio-bluez" it outputs:
2023-01-29 13:05:39 <tetrakist> After this operation, 0 B of additional disk space will be used.
2023-01-29 13:05:39 <tetrakist> OK: 1301 MiB in 598 packages
2023-01-29 13:05:45 <tetrakist> And nothing else.
2023-01-29 13:05:59 <ikke> it means it's already installed
2023-01-29 13:06:33 <tetrakist> # apk list --installed | grep bluez
2023-01-29 13:06:33 <tetrakist> bluez-5.66-r2 aarch64 {bluez} (GPL-2.0-or-later AND BSD-2-Clause AND MIT) [installed]
2023-01-29 13:06:33 <tetrakist> bluez-openrc-5.66-r2 aarch64 {bluez} (GPL-2.0-or-later AND BSD-2-Clause AND MIT) [installed]
2023-01-29 13:07:10 <tetrakist> doesn't show up. And the files it is supposed to contain are not present in /usr/lib/pulseaudio/modules/
2023-01-29 13:08:01 <ikke> What's in your /etc/apk/repositories
2023-01-29 13:09:49 <tetrakist> postmarketos/master, alpine/edge/main, alpine/edge/community, and alpine/edge/testing
2023-01-29 13:10:22 <ikke> ok, so pmos as well
2023-01-29 13:12:10 <tetrakist> correct.
2023-01-29 13:13:58 <ikke> what does `apk policy pulseaudio-bluez` return?
2023-01-29 13:14:26 <tetrakist> pulseaudio-bluez policy:
2023-01-29 13:14:26 <tetrakist>   16.1-r7:
2023-01-29 13:14:26 <tetrakist>     http://dl-cdn.alpinelinux.org/alpine/edge/community
2023-01-29 13:14:26 <tetrakist>   9999_git20220528-r0:
2023-01-29 13:14:26 <tetrakist>     http://mirror.postmarketos.org/postmarketos/master
2023-01-29 13:14:40 <tetrakist> >.> 
2023-01-29 13:14:51 <ikke> so it will install the pmos version
2023-01-29 13:15:21 <tetrakist> Well, I wish it would. But nothing happens.
2023-01-29 13:16:29 <tetrakist> oh wait.
2023-01-29 13:17:16 <tetrakist> I have pipewire-pulse installed... not the actual pulseaudio package. the pmOS package has "install if: pulseaudio=9999_git20220528-r0"
2023-01-29 13:17:23 <tetrakist> would that prevent installation?
2023-01-29 13:17:30 <ikke> No, that should not prevent it
2023-01-29 13:17:32 <ikke> the install_if part
2023-01-29 13:17:34 <tetrakist> weird.
2023-01-29 13:17:51 <tetrakist> Ok, well, I just strace'd the apk add, so gonna go grepping through that.
2023-01-29 13:21:14 <ikke> hmm, pulseaudio-bluez does seem to depend on pipewire-pulse
2023-01-29 13:23:03 <ikke> tetrakist: pipewire-pulse provides pulseaudio-bluez
2023-01-29 13:23:11 <ikke> so for what apk cares about, it's alreaady installed
2023-01-29 13:23:15 <tetrakist> oh.
2023-01-29 13:25:46 <tetrakist> ikke, where do I find that pipewire-pulse provides pulseuadio-bluez? The package details from pkgs.alpinelinux.org doesn't show anything about that, that I can see.
2023-01-29 13:26:00 <ikke> apk info -P pipewire-pulse
2023-01-29 13:27:20 <tetrakist> ah, thanks.
2023-01-29 13:27:45 <tetrakist> It would be nice if package details in pkgs.alpinelinux.org showed stuff like that.
2023-01-29 13:27:56 <tetrakist> heh.
2023-01-29 13:28:24 <tetrakist> ugh. there's probably some *other* arcane reason why this headset won't connect, lol.
2023-01-29 13:30:28 <tetrakist> Hmm.
2023-01-29 13:33:56 <tetrakist> well, installed pipewire-spa-bluez and now it connects. Just no audio. heh.
2023-01-29 13:36:53 <tetrakist> Ah. Reboot fixed that.
2023-01-29 13:37:59 <tetrakist> thanks for your help ikke. :)
2023-01-29 13:41:04 <ikke> np
2023-01-29 15:30:52 <WhyNotHugo> Polkit is not usable without (e)logind, right?
2023-01-29 15:31:16 <WhyNotHugo> Actually, wrong question; my real question is: can I use fprintd without polkit?
2023-01-29 15:41:37 <elly> I am trying to put a package (gnupg) onto an alpine system with no networking - is there a way to get apk on another system to download and stash the gnupg apk + apks for its dependencies, so I can copy them to removable media?
2023-01-29 15:44:07 <WhyNotHugo> elly: Maybe https://wiki.alpinelinux.org/wiki/Local_APK_cache#Enabling_Local_Cache_with_current_releases ?
2023-01-29 15:44:27 <WhyNotHugo> A cache will help since it should fetch any dependencies too.
2023-01-29 15:44:39 <ikke> elly: combined with apk cache download <package..>
2023-01-29 15:44:49 <elly> aha, `apk cache download` is the wisdom
2023-01-29 15:44:50 <elly> let's see
2023-01-29 15:45:23 <elly> hmmm, I expected this to work:
2023-01-29 15:45:29 <elly> ~ $ apk cache --cache-dir /tmp/packages download gpg
2023-01-29 15:45:29 <elly> ERROR: Package cache is not enabled.
2023-01-29 15:47:34 <ikke> create the dir first
2023-01-29 15:48:01 <elly> unfortunately that yields:
2023-01-29 15:48:02 <elly> WARNING: opening from cache http://dl-cdn.alpinelinux.org/alpine/v3.17/main: No such file or directory
2023-01-29 15:48:11 <ikke> yeah, I get the same
2023-01-29 15:48:14 <elly> and then no actual downloading
2023-01-29 15:48:41 <elly> I think it's expecting that if a cachedir exists it has a certain structure already maybe
2023-01-29 15:49:29 <elly> if I run it without --cache-dir it works fine, but it puts them in /var/cache/apks where they are intermingled with all my other apks, so not ideal
2023-01-29 15:49:33 <ikke> it works if you run update first
2023-01-29 15:49:53 <elly> as root, you mean?
2023-01-29 15:50:14 <elly> `apk --cache-dir /tmp/packages update` complains that it can't open the apk database
2023-01-29 15:50:16 <ikke> the cache-dir is also what contains the cached repository indexes
2023-01-29 15:50:38 <ikke> elly: maybe easier to create a symlink at /etc/apk/cache
2023-01-29 15:50:52 <elly> I already have one of those pointing to my actual apk cache
2023-01-29 15:51:02 <elly> ... actually, I can just empty my actual apk cache, then use `apk cache download` normally
2023-01-29 15:51:07 <elly> instead of fiddling with --cache-dir
2023-01-29 15:53:06 <WhyNotHugo> Keeping a cache would be useful to rebuild a system from scratch now that I think of it. I should keep that enabled.
2023-01-29 15:57:21 <novorol[m]> I'm having trouble running NetworkManager with iwd backend, running nmcli device wifi list returns the list of SSIDs, but BSSIDs are certainly incorrect, like 00:01:02:00:01:00 and
2023-01-29 15:57:34 <elly> hmm, now that I have the set of apks it is still not obvious how to get them into the offline system's cache
2023-01-29 15:57:58 <ikke> elly: You do need the index as well afaik
2023-01-29 15:58:05 <ikke> rsync the cachedir
2023-01-29 15:58:08 <elly> the APKINDEX files? I also copied those over, yeah
2023-01-29 15:58:25 <ikke> point /etc/apk/cache to it and you should be able to install those packages
2023-01-29 15:58:33 <elly> however, `apk update` on the offline system still just says "60 packages available", which is how many the base installed system has
2023-01-29 15:58:49 <ikke> did you try apk add gpg?
2023-01-29 15:58:55 <elly> I did - one sec
2023-01-29 15:59:52 <elly> so, `apk update` says: "OK: 60 distinct packages available"; `apk add gnupg` says: "no such package"
2023-01-29 16:00:08 <ikke> and you did setup /etc/apk/cache?
2023-01-29 16:00:23 <novorol[m]> also trying to connect with nmtui returns 'Could not activate connection: request succeeded with /org/freedesktop/NetworkManager /ActiveConnection/5 but object is in an unsuitable state
2023-01-29 16:00:30 <elly> I have /etc/apk/cache as a symlink to /var/cache/apk
2023-01-29 16:01:48 <elly> and indeed I see all of the gnupg apks and stuff in /var/cache/apk, along with five different APKINDEXes
2023-01-29 16:01:57 <elly> (from the five repos I have configured on the online system)
2023-01-29 16:02:12 <ikke> what'
2023-01-29 16:02:19 <ikke> what's in your /etc/apk/repositories?
2023-01-29 16:02:25 <elly> on the offline machine, it's empty
2023-01-29 16:02:33 <ikke> I guess that's the issue
2023-01-29 16:02:33 <elly> or well, a commented-out #/media/usb/apks
2023-01-29 16:02:39 <elly> on the online machine I have those 5 repos
2023-01-29 16:02:54 <elly> I wonder if I need to add them on the offline machine, even though it can't fetch from them, for it to use those APKINDEXes
2023-01-29 16:03:02 <ikke> yeah
2023-01-29 16:04:02 <ikke> yeah
2023-01-29 16:04:05 <ikke> lol
2023-01-29 16:04:18 <ikke> It works if I empty /etc/resolv.conf in a container and add the usual repos in /etc/apk/repositories
2023-01-29 16:04:27 <ikke> apk update will fail, but apk search returns packages
2023-01-29 16:04:43 <ikke> and I can install gpg
2023-01-29 16:04:47 <elly> yup, now I see "39594 distinct packages available"
2023-01-29 16:04:55 <elly> humorous
2023-01-29 16:05:07 <ikke> otherwise it would not consider those indexes
2023-01-29 16:05:11 <elly> yeah
2023-01-29 16:05:32 <elly> I wonder if perhaps apk should warn if it sees an APKINDEX in the cache that doesn't correspond to a live repo
2023-01-29 16:05:38 <elly> maybe that's too niche
2023-01-29 16:05:43 <elly> yay, gnupg success, thanks :)
2023-01-29 16:08:38 <elly> so the general formula for copying packages to an offline device is this: https://bpa.st/WTXW2
2023-01-29 16:09:29 <ikke> is the last update required?
2023-01-29 16:09:47 <elly> on the offline machine? I'm uncertain
2023-01-29 16:09:54 <elly> it may not be
2023-01-29 16:11:42 <elly> I still don't quite understand the behavior of --cache-dir
2023-01-29 16:12:02 <elly> I would naively expect that if I run `apk` with --cache-dir set to a nonexistent directory, it would act as though it had never been run on the system before basically
2023-01-29 16:13:50 <ikke> elly: I think it does, but apk cache does not seem to automatically update the indexes
2023-01-29 16:14:04 <ikke> If I run apk add --cache-dir /tmp/packages gpg, it works
2023-01-29 16:14:13 <ikke> but you obviously need to do that as root
2023-01-29 16:14:15 <elly> but `apk cache download` doesn't?
2023-01-29 16:14:17 <ikke> correct
2023-01-29 16:14:22 <elly> that's the part that is odd
2023-01-29 16:14:22 <ikke> as it can be done as none-root
2023-01-29 16:14:47 <elly> `apk cache --cache-dir /tmp/packages download gpg` seems like it should just produce a new apk cache in /tmp/packages with gpg and all its deps
2023-01-29 16:16:18 <ikke> I have to run update as root before that
2023-01-29 16:17:02 <elly> hmm
2023-01-29 16:17:37 <ikke> update (as root) downloads the indexes, then you can fetch any package as non-root
2023-01-29 16:17:43 <elly> aha!
2023-01-29 16:17:49 <elly> the update also needs to be run with --cache-dir
2023-01-29 16:17:51 <elly> obvious in retrospect
2023-01-29 16:17:55 <ikke> yeah, that's what I meant
2023-01-29 16:17:59 <ikke> sorry, I assumed that was clear
2023-01-29 16:18:08 <elly> no worries
2023-01-29 16:18:15 <elly> I'm surprised the `apk update` needs to be run as root though
2023-01-29 16:18:20 <elly> since it's not doing anything privileged really
2023-01-29 16:19:09 <elly> hmm, it's trying to open /lib/apk/db/lock
2023-01-29 16:19:42 <ikke> to prevent concurrent conflicting operations
2023-01-29 16:19:56 <elly> right, but I'm surprised the lock file is not a function of the cache dir
2023-01-29 16:20:04 <elly> since you can operate on multiple cache dirs concurrently safely (right?)
2023-01-29 16:20:16 <elly> I guess you still can't actually add a package concurrently
2023-01-29 16:20:17 <elly> hum
2023-01-29 16:20:23 <elly> oh well, I got it to work :)
2023-01-29 16:20:29 <ikke> It's primarily locking the main db
2023-01-29 16:20:35 <ikke> which there is only one instance of
2023-01-29 16:22:25 <elly> ya that makes sense
2023-01-29 16:22:41 <elly> and I guess a cache update of the main cache must be locked
2023-01-29 16:28:52 <WhyNotHugo> `powertop` fails with `modprobe cpufreq_stats failed`. Is this driver not availale on linux-edge?
2023-01-29 16:28:58 <WhyNotHugo> I don't see it on some other package either.
2023-01-29 16:32:27 <minimal> # CONFIG_CPU_FREQ_STAT is not set
2023-01-29 16:32:50 <minimal> from the linux-edge kernel config for x86_64
2023-01-29 16:33:01 <WhyNotHugo> My issue right now is that my desktop uses 50W when idle, and I want to kinda figure out why it's so hungry.
2023-01-29 16:33:08 <WhyNotHugo> Any good guidelines to inspect this?
2023-01-29 16:34:15 <minimal> it is enabled (and built in) on linux-lts
2023-01-29 16:34:52 <WhyNotHugo> I had linux-lts earlier this month but had lots of issues with gpu (my GPU is relatively new).
2023-01-29 16:35:13 <minimal> so then request it be built for linux-edge?
2023-01-29 16:35:55 <WhyNotHugo> Oh, yeah, that should be reasonable :P
2023-01-29 16:36:06 <WhyNotHugo> The simple answers often elude me.
2023-01-29 16:41:01 <Waitingfortherain[m]> Hey I was wondering if someone could guide me on how to setup alpine with btrfs instead of the normal ext during the setup alpine process
2023-01-29 16:41:12 <Waitingfortherain[m]> I would really appreciate it
2023-01-29 16:42:10 <Waitingfortherain[m]> And also since im new to alpine I was also wondering if I was on version 3.x and alpine upgrades to 4.x would upgrading be as simple as a few commands or I would need to do a complete migration
2023-01-29 16:42:26 <WhyNotHugo> Waitingfortherain[m]: Format the partition, mount it into `/mnt`, and then run `setup-disk /mnt`.
2023-01-29 16:42:51 <Waitingfortherain[m]> ah great thank you very much
2023-01-29 16:42:58 <Waitingfortherain[m]> is there a link of this on the official website?
2023-01-29 16:43:05 <WhyNotHugo> You'll need to install btrfs-progs on the installation media.
2023-01-29 16:43:16 <Waitingfortherain[m]> right
2023-01-29 16:43:39 <WhyNotHugo> If you want to use subvolumes make sure you set them up BEFORE you do this mounting;setup-disk.
2023-01-29 16:43:58 <Waitingfortherain[m]> yeah of course
2023-01-29 16:44:10 <WhyNotHugo> I don't think there's a full guide. I've a lot of notes around here from my recent setup and keep postponing publishing them in a tidy shape.
2023-01-29 16:44:27 <minimal> run "ROOTFS=btrfs setup-disk" ?
2023-01-29 16:44:33 <Waitingfortherain[m]> haha I am making notes on it as well
2023-01-29 16:44:50 <minimal> look at "setup-disk -h" output, this is documented
2023-01-29 16:44:56 <Waitingfortherain[m]> oh thanks
2023-01-29 16:45:08 <Waitingfortherain[m]> so after I do setup-disk
2023-01-29 16:45:10 <Waitingfortherain[m]> I can do setup-alpine?
2023-01-29 16:46:36 <minimal> I'd think it would more a case of running setup-alpine (so that mirror, keyboard, etc are configured) and then at the point it asks about disk configuration you quite it and run setup-disk manually
2023-01-29 16:46:46 <minimal> s/quite/quit/
2023-01-29 16:46:46 <alpine-meetbot> minimal meant to say: I'd think it would more a case of running setup-alpine (so that mirror, keyboard, etc are configured) and then at the point it asks about disk configuration you quit it and run setup-disk manually
2023-01-29 16:47:16 <Waitingfortherain[m]> oh right
2023-01-29 16:47:57 <Waitingfortherain[m]> do i ever need to go back to setup-alpine after setup-disks to do the remaining stuf
2023-01-29 16:47:57 <Waitingfortherain[m]> s/stuf/stuff/
2023-01-29 16:47:57 <alpine-meetbot> Waitingfortherain[m] meant to say: do i ever need to go back to setup-alpine after setup-disks to do the remaining stuff
2023-01-29 16:47:58 <Waitingfortherain[m]> if there is anything
2023-01-29 16:50:03 <minimal> setup-disk is one of the last things that setup-alpine runs (apart from setup-lbu and setup-apkcache which are not relevant if you're doing a Sys-mode install)
2023-01-29 16:50:35 <Waitingfortherain[m]> ah thanks minimal I will definately do this
2023-01-29 17:11:04 <zeke> I am trying to get bcache working on alpine, and have loaded everything how I usually would on another linux distro
2023-01-29 17:11:09 <zeke> however there are no `/dev/bcache{0,1,2,3}` devices present
2023-01-29 17:15:26 <minimal> zeke: using eudev or mdev?
2023-01-29 17:17:36 <minimal> I assume you are using mdev and I don't see any mdev rules to create /dev/bcache* devices
2023-01-29 17:17:50 <minimal> try installing and enabling eudev?
2023-01-29 17:18:44 <zeke> yeah I was using mdev
2023-01-29 17:18:49 <zeke> ill try that thanks
2023-01-29 17:32:00 <zeke> works as intended
2023-01-29 17:33:10 <newalpuser> hello
2023-01-29 17:43:54 <Quillith> I want to check and make sure I'm understanding this right. AppImages aren't compatible with Alpine because they use glibc instead of musl. So if there was an AppImage that didn't have a flatpak version, I would just install gcompat?
2023-01-29 17:54:40 <invoked> the glibc shim *might* work
2023-01-29 17:54:46 <invoked> i guess it depends...
2023-01-29 17:55:36 <invoked> i haven't tried using appimages in quite some time, so i don't know what's true/false anymore
2023-01-29 17:56:12 <invoked> but i wouldn't count on appimages working as some kind of solution
2023-01-29 17:56:21 <invoked> even if you do get a glibc shim working today
2023-01-29 17:56:47 <invoked> seems like a brittle setup
2023-01-29 18:59:27 <Quillith_> Okay, sounds like it's not a very reliable way to do things. I'll wait for the app to be released as a flatpak, seems my best bet. Thanks 
2023-01-29 19:30:44 <elibrokeit> Quillith_: alternatively, package it properly as a distro package instead of an embedded container
2023-01-29 19:53:46 <Guest28> Hello, anyone here successfully installed Alpine on Pi ZW2? I tried following the official docs here: https://wiki.alpinelinux.org/wiki/Raspberry_Pi#Preparation - But main difference is that I specified was +1024M during fdisk and after mounting the partition: https://paste.debian.net/plain/1268892 - When the Pi starts up it does light up the green LED, but I can't get anything to show up on the UART reader (screen in my case)
2023-01-29 19:54:22 <Guest28> While that does work with the official raspios. Am I doing something wrong or are the docs maybe wrong?
2023-01-29 20:02:39 <Quillith_> @Guest28, I installed Alpine on Pi4 with aarch64 architecture. I used the wiki page you linked as well as the one it connected to, https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device#Manually_copying_Alpine_files, for putting the tarball onto the pi. I mostly used this wiki page, https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi, since I was looking for a classic 
2023-01-29 20:02:45 <Quillith_> installation instead of diskless. It seemed to have worked, my pi runs and boots ok
2023-01-29 20:08:11 <Quillith> elibrokeit: I thought about that, and I looked at the Creating an Alpine Package page, but I can barely keep my own computers running; I can't imagine being responsible for a distro package
2023-01-29 20:08:43 <Guest28> @Quillith: Wait, am I listed in IRC as Guest28 ?
2023-01-29 20:08:48 <ikke> Guest28: yes
2023-01-29 20:08:59 <Guest28> Ugh this stupid Matrix bridge again.
2023-01-29 20:11:25 <elibrokeit> Quillith: it's usually fairly simple, you should try it out :)
2023-01-29 20:11:50 <Quillith> well I guess inexperienced hasn't stopped me yet, so why not try
2023-01-29 20:15:16 <ForeverNoob[m]> @Quillith: Are you connecting via UART too?
2023-01-29 20:17:50 <Quillith> ForeverNoob[m]: I made a text file usercfg.txt with 1 line saying enable_uart=1 when I was setting up
2023-01-29 20:18:26 <Quillith> I don't know if it was necessary or what it does, but the page said it avoids a bug, and the pi hasn't broke yet
2023-01-29 20:21:43 <ForeverNoob[m]> Yeah that setting makes it so one can connect to the Pi via UART (a low level serial connection, pretty handy). I've added that line to `config.txt` but I'll now try adding it to `usercfg.txt` and see what happens.
2023-01-29 20:22:11 <Quillith> Okay, thanks. Good luck!
2023-01-29 20:22:33 <elly> making packages is not that hard :) https://elly.town/d/blog/2022-10-31-packaging-for-alpine.txt
2023-01-29 20:24:51 <Quillith> elly: Thanks for the link! I'll give it a shot. It's probably simple, but I have no background in computer science or developing. I'll start with your blogpost, re-read the wiki page, and give it a go
2023-01-29 21:31:33 <spazialecontinentale[m]> hi, where to add env variable for alpine?
2023-01-29 21:31:45 <spazialecontinentale[m]>  * hi, where do i add env variable in alpine?
2023-01-29 21:32:14 <spazialecontinentale[m]> i need it for an APKBUILD script
2023-01-29 21:33:16 <pj[m]> inside APKBUILD
2023-01-29 21:35:01 <spazialecontinentale[m]> pj[m]: well, i'm porting a software that as a weird path for config file. I put it on /etc/ so i need a global env var
2023-01-29 21:35:20 <spazialecontinentale[m]> for openrc i have no issue
2023-01-29 21:35:22 <pj[m]> why would you need it
2023-01-29 21:35:43 <pj[m]> where and when
2023-01-29 21:36:00 <spazialecontinentale[m]> my initd... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/ffveBQoSosisrYEgsNokeGPm>)
2023-01-29 21:37:05 <tomk44> greetings to all
2023-01-29 21:37:17 <spazialecontinentale[m]> with this solution i need to add MOLLY_CONF=/etc/mollysocket before every mollysocket command
2023-01-29 21:39:03 <tomk44> DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD
2023-01-29 21:39:51 <tomk44> sorry folks! my keyboar stuck
2023-01-29 21:40:56 <ForeverNoob[m]> At least you didn't forget the D
2023-01-29 21:41:32 <elibrokeit> spazialecontinentale[m]: I assume you can do something like stick that in /etc/profile.d/ but perhaps alternatively, this software should re-evaluate its life choices
2023-01-29 21:41:47 <elibrokeit> this is not very nice to impose upon the system as a whole
2023-01-29 21:42:14 <spazialecontinentale[m]> elibrokeit: right
2023-01-29 21:42:24 <pj[m]> I was about to say that this program is terrible
2023-01-29 21:42:24 <spazialecontinentale[m]> i will try to propose it 
2023-01-29 21:42:36 <pj[m]> and you would end up better by patching it
2023-01-29 21:42:48 <pj[m]> to use etc/mollysocket.conf or alike
2023-01-29 21:43:49 <pj[m]> well, imo, you should patch it to read from etc/filename if you want to add it to alpine repo
2023-01-29 21:44:33 <elibrokeit> seems like this upstream is, fairly typically, only interested in running the software as a docker container :D
2023-01-29 21:45:11 <elibrokeit> or well, kubernetes
2023-01-29 21:48:20 <tomk44> Could anybody help me with Alpine Linux xrdp session configuration? I am running alpine Linux on Windows 11 WSL 2. I could not figure out remote desktop connection. I googled and followed available instructions. I could not connect to the remote session although I can ping the Alpine WSL 2 instance. I am using KDE Plasma as a desktop manager on alpine Linux 3.17. I can seamlessly connect to wsl 2 remote session with Ubuntu or fedor
2023-01-29 21:48:20 <tomk44> a without any issues.
2023-01-29 21:49:15 <pj[m]> I don't think xrdp ever worked on Alpine
2023-01-29 21:52:47 <latex> Does anyone know why nginx (either the program or the RC service) changes the permissions of user directories?
2023-01-29 21:53:03 <latex> And also why openrc says a process crashed while it's actually still running?
2023-01-29 21:53:40 <latex> Now I have a cron job called fuck-off-nginx that resets the owner of user directories every minute
2023-01-29 21:53:48 <latex> This is ridiculous
2023-01-29 21:54:25 <tomk44> I like Alpine Linux so much on WSL 2. It is very fast and small. KDE desktop runs well with some issus using Vcx server. Based on my experience XRDP sessions run very well on Fedora and Ubuntu on WSL 2. Wish I could make it work on Alpine Linux.
2023-01-29 21:55:23 <latex> location /uploads/ {
2023-01-29 21:55:25 <latex>         auth_basic "Anonymous uploads";
2023-01-29 21:55:27 <latex>         auth_basic_user_file /home/anonymous/.htpasswd;
2023-01-29 21:55:29 <latex>         root /home/anonymous;
2023-01-29 21:55:31 <latex>         autoindex on;
2023-01-29 21:55:33 <latex> }
2023-01-29 21:55:53 <latex> Why does this change the owner of /home/anonymous to nginx every 5 minutes?
2023-01-29 22:00:23 <ikke> latex: The only path the nginx initd file modifies the permissions for is the path the pidfile is located, and only when the service is started
2023-01-29 22:01:02 <ikke> And I've never heard of nginx itself changing permissions
2023-01-29 22:01:37 <latex> Well on my system the /home/anonymous/uploads directory mysteriously changes to nginx every 5 minutes
2023-01-29 22:01:41 <latex> or so
2023-01-29 22:01:48 <ikke> did you change the pidfile path?
2023-01-29 22:01:59 <latex> Why would it be there?
2023-01-29 22:02:06 <spazialecontinentale[m]> <elibrokeit> "spaziale.continentale: I..." <- The programmes says that the user running the bin must have write access to the directory, that is why it is set in a user config dir by default. Commonly, linux distributions write a wrapper in this case. 
2023-01-29 22:02:06 <spazialecontinentale[m]> Can this be done throught APKBUILD?
2023-01-29 22:02:07 <latex> I don't think I changed it from the defaults
2023-01-29 22:02:38 <pj[m]> > <@spaziale.continentale:matrix.org> The programmes says that the user running the bin must have write access to the directory, that is why it is set in a user config dir by default. Commonly, linux distributions write a wrapper in this case.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/EPPGcfLSrmCNItAjXwPvxUJG>)
2023-01-29 22:03:00 <spazialecontinentale[m]> pj[m]: ok thanks
2023-01-29 22:04:21 <elibrokeit> hold on, who runs this program
2023-01-29 22:06:24 <pj[m]> mollysocket looks like user service
2023-01-29 22:12:18 <hallogallo> Hi, startx wont work after having dbus started and added the user to the input group. any ideas?
2023-01-29 22:32:32 <spazialecontinentale[m]> <elibrokeit> "hold on, who runs this program" <- a user called mollysocket
2023-01-29 22:33:55 <spazialecontinentale[m]> but normal user need to config it through command  
2023-01-29 22:37:46 <spazialecontinentale[m]>  * but normal user need to config it through command
2023-01-29 22:37:46 <spazialecontinentale[m]> https://gitlab.com/Marc0x/mollysocket-alpine-build ( nothing professional, just use it as backup)
2023-01-30 03:09:19 <SF> Hi, is this an ok place to report a problem with alsa in alpine 3.17.1?
2023-01-30 03:12:17 <pj[m]> yes and no
2023-01-30 03:12:30 <pj[m]> it should be (preferably) reported on gitlab.a.o
2023-01-30 03:12:57 <pj[m]> or gitlab.alpinelinux.org/alpine/aports when irc bot is dead
2023-01-30 03:13:13 <SF> I have no clue how to deal with gitlab
2023-01-30 03:19:54 <SF> When exiting alsamixer, settings are not saved. I'd had added alsa to rc-update and the service was running. Workaround was to touch /var/lib/alsa/asound.state and lbu include it, and then to rc-sevice alsa stop which then triggered state save. 
2023-01-30 03:39:00 <u0_a273> hello
2023-01-30 07:23:46 <bulle_> hey. I just installed alpine with a sys install. followed the instructions for installing dwm with xorg but startx wont work. I also put the user in the group "input", but no success. can you help?
2023-01-30 07:29:59 <psykose> could you pastebin the xorg.log it outputs
2023-01-30 07:42:14 <bulle_> [   933.437]  X.Org X Server 1.21.1.6 X Protocol Version 11, Revision 0 [   933.439] Current Operating System: Linux emcmp06 6.1.8-0-lts #1-Alpine SMP PREEMPT_DYNAMIC Tue, 24 Jan 2023 10:09:31 +0000 x86_64 [   933.439] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-lts root=UUID=4af62754-c9dd-470b-a57a-d1324a2ee9c6 ro modules=sd-mod,usb-storage,ext4 quiet rootfstype=ext4 [   933.441]   [   933.442] Current version of pixman: 0.42.2 
2023-01-30 07:43:24 <psykose> that is missing like the entire log
2023-01-30 07:43:28 <psykose> this is also not a pastebin
2023-01-30 07:43:38 <psykose> http://ix.io/ , etc
2023-01-30 07:45:38 <bulle_> i fetched: ~/.local/share/xorg/Xorg.0.log
2023-01-30 07:46:34 <bulle_> never used a log file tbh. can you tell me what file you need and how to get it?
2023-01-30 07:47:25 <psykose> the startx output usually tells you where it is
2023-01-30 07:48:10 <bulle_> right, and it told me /home/USERNAME/.local/share/xorg/Xorg.0.log
2023-01-30 07:48:58 <bulle_> when I past the text here it renders as one long text unfortunatly
2023-01-30 07:49:54 <bulle_> can I send a link to the text file in any way?
2023-01-30 07:55:12 <bulle_> I copy the log file to my PC now, how do I upload it to ix.io?
2023-01-30 07:55:53 <lopid> go to http://ix.io and find out
2023-01-30 07:56:03 <bulle_> https://logpaste.com/FUICfAAs
2023-01-30 07:56:16 <bulle_> is this ok?
2023-01-30 07:58:08 <psykose> ah the ole "exited signal 6" with no more reasoning
2023-01-30 07:58:14 <psykose> no idea then
2023-01-30 07:58:32 <bulle_> is that a common unfixable error?
2023-01-30 07:59:39 <lopid> (EE) Failed to load module "fbdev" (module does not exist, 0)
2023-01-30 08:01:15 <bulle_> lopid so I install fbdev???
2023-01-30 08:02:12 <lopid> xf86-video-fbdev perhaps
2023-01-30 08:02:34 <lopid> vesa, too. isn't there a meta package for xorg?
2023-01-30 08:03:16 <bulle_> yes, setup-xorg-base. and that is installed
2023-01-30 08:08:14 <psykose> that's not a real error
2023-01-30 08:08:30 <bulle_> I installed both, still dont work. here is a new log:
2023-01-30 08:08:30 <bulle_> https://logpaste.com/nizq2Pav
2023-01-30 08:08:39 <psykose> it just tries to load every driver regardless
2023-01-30 08:10:48 <bulle_> dont understAND WHAT YOU MEAN..
2023-01-30 08:10:55 <bulle_> oops sorry caps
2023-01-30 08:11:06 <bulle_> is this something? :
2023-01-30 08:11:07 <bulle_> [  2705.152] xf86EnableIO: failed to enable I/O ports 0000-03ff (Operation not permitted)
2023-01-30 08:12:04 <bulle_> and this:
2023-01-30 08:12:05 <bulle_> [  2705.153] (WW) VGA arbiter: cannot open kernel arbiter, no multi-card support
2023-01-30 08:13:12 <psykose> afaik no
2023-01-30 08:13:33 <psykose> unless you were also not in the video group or something i have no idea why it doesn't work
2023-01-30 08:15:26 <bulle_> ok. is dueal graphics card generally a problem?
2023-01-30 08:17:12 <bulle_> *dual
2023-01-30 08:19:22 <psykose> never had dual graphics to test
2023-01-30 08:19:56 <Ermine> which cards do you have?
2023-01-30 08:23:31 <bulle_> Radeon HD 8730M and integrated Intel
2023-01-30 08:23:52 <bulle_> I installed many drivers, is that a problem?
2023-01-30 08:24:31 <bulle_> xf86-video-radeon, xf86-video-ati and xf86-video-amdgpu
2023-01-30 08:25:25 <bulle_> I also installed xf86-video-intel ofc
2023-01-30 08:28:43 <Ermine> I guess it picks the right drivers
2023-01-30 08:29:18 <Ermine> did you re-login after adding yourselves to audio/video/input groups?
2023-01-30 09:11:04 <qaqland[m]> Is there any package for https://github.com/gollum/gollum?
2023-01-30 09:13:52 <psykose> gem install --user-install gollum
2023-01-30 09:40:39 <vkrishn> is it a good idea to do  "dmesg -c" or "dmesg -C" on low ram devices ?
2023-01-30 09:44:34 <ikke> It's a ring buffer, so the amount of memory is should be constant 
2023-01-30 09:47:24 <vkrishn> thanks
2023-01-30 12:48:58 <Hello71> the previous xorg problem was  [  2705.197] (EE) Caught signal 6 (Aborted). Server aborting
2023-01-30 12:50:35 <qaqland[m]> <psykose> "gem install --user-install..." <- why some software's install need system package management, some do not
2023-01-30 12:51:51 <psykose> that's a good philosophical question i don't have the answer to
2023-01-30 13:27:48 <qaqland[m]> :-)
2023-01-30 13:40:25 <qsx> i think i don’t quite see the wood for all the trees.  if i read mkinitfs.trigger correctly, a) it (and not linux-*) contains the mechanism to create the initramfs when installing linux, and b) if mkinitfs.conf contains 'disable_trigger=1' or so it doesn’t.  cf. https://git.alpinelinux.org/aports/tree/main/mkinitfs/mkinitfs.trigger?h=3.17-stable
2023-01-30 13:40:39 <qsx> is this right, and where is this documented?
2023-01-30 13:41:33 <qsx> (i’m trying to disable initfs generation on my rpi, as the kernel contains all modules to boot and reliably finds the rootfs when using root=PARTUUID=…)
2023-01-30 13:41:51 <ForeverNoob[m]> Findings: Specifying start_x=1 in combination with Alpine Linux borks the UART (and the boot process as well I think) - https://www.raspberrypi.com/documentation/computers/config_txt.html#start_x-start_debug
2023-01-30 13:42:08 <ForeverNoob[m]> Happens on Pi ZW 2 - Is this expected though?
2023-01-30 13:45:23 <psykose> qsx: yes, mkinitfs has the hook, not the kernel
2023-01-30 13:45:27 <psykose> there's multiple initfs generators
2023-01-30 13:52:51 <ForeverNoob[m]> That feel when the setup-alpine script is not able to find your hidden network, guest network or your phone hotspot :(
2023-01-30 14:00:20 <psykose> there's always the manual way
2023-01-30 14:12:06 <ForeverNoob[m]> I suppose (reluctantly)
2023-01-30 14:15:16 <satmd> 
ACTION adds: hidden networks are not much hidden (until a client connects), but will make the client to broadcast requests for this network whereever it is - thus making it very easy to track. Hidden SSIDs are bad.

2023-01-30 14:47:27 <ForeverNoob[m]> @satmd: Isn't this dependent on wifi stack though? For example GrapheneOS doesn't broadcast the SSID IIRC. Also, if one disconnects using nmcli for example, would the wifi stack still broadcast it?
2023-01-30 14:55:12 <invoked> if anything, the sort of people you don't want to take interest in your network will only be more interested if you make it hidden. and they will find it easily.
2023-01-30 15:04:40 <satmd> ForeverNoob[m]: you cannot use a hidden ssid as client if you do not send probes as a client. if you're sending probes, you're trackable.
2023-01-30 15:05:21 <satmd> either the AP says "here I am" or the client will need to poll for its presence
2023-01-30 15:05:43 <satmd> and the latter will happen independent of location and make it trackable
2023-01-30 15:06:29 <satmd> an AP not saying "here I am" will make it hard for regular users, but will still be visible if you have time to observe the air traffic
2023-01-30 15:06:44 <satmd> also: hidden ssids cause more battery drain for regular probes too
2023-01-30 15:06:49 <ForeverNoob[m]> Will the latter happen if I explicitly state to connect with said hidden AP using nmcli con up SSID ?
2023-01-30 15:06:56 <satmd> yes
2023-01-30 15:07:16 <satmd> well, most likely
2023-01-30 15:07:27 <ForeverNoob[m]> Ok, so if I don't explicitly state to connect to it, then probes are not sent?
2023-01-30 15:07:32 <satmd> yeah
2023-01-30 15:07:50 <satmd> well... point is "hiding the ssid" does not gain you anything positive, only potentially negative
2023-01-30 15:08:08 <satmd> you don't win anything
2023-01-30 15:09:14 <ForeverNoob[m]> To me it's equivalent of setting SSH to a non-standard port. Will it deter hackers? No, but it will lessen the noise of failed attempts.
2023-01-30 15:09:35 <ForeverNoob[m]> *sshd
2023-01-30 15:27:59 <psykose> only if you track noise
2023-01-30 15:28:19 <psykose> i've never seen someone track home wifi login attempts tbh
2023-01-30 15:32:54 <elly> imo if you're using pubkey-only auth for sshd (which you should) then tracking failed logins is pretty uninteresting
2023-01-30 15:35:26 <psykose> everyone goes through that phase where you first get looking at some logs and realise how much spam there is
2023-01-30 15:35:43 <psykose> 99% of email is spam? 99.99% of everything else is, too
2023-01-30 15:36:18 <psykose> 
ACTION glances at alpine wiki http logs

2023-01-30 15:36:22 <psykose> "GET /w/index.php?title=Buy_Winter_Shoes_Brands_As_Lowest_As_50__Off_Online_In_Dubai&action=edit HTTP/1.0"
2023-01-30 15:36:23 <psykose> profound
2023-01-30 15:36:28 <elly> lol
2023-01-30 15:37:31 <elly> # grep 'Disconnected from authenticating user' auth.log | wc -l ; grep 'Accepted publickey' auth.log | wc -l
2023-01-30 15:37:34 <elly> 5308
2023-01-30 15:37:37 <elly> 3
2023-01-30 15:38:04 <elly> according to dc, 99.94% of login attempts on my digitalocean vm are indeed spam
2023-01-30 15:38:16 <elly> or rather people trying to guess root's password or whatnot
2023-01-30 15:39:05 <elly> oh, I forgot to count the over 1100 invalid user attempts as well
2023-01-30 15:39:10 <elly> so rather north of 99.97%
2023-01-30 15:39:21 <valerius> years back I brought up a DO instance and it immediately got exploited and a bitcoin miner installed
2023-01-30 15:39:23 <elly> and this log file starts *yesterday afternoon*
2023-01-30 15:39:28 <valerius> didn't do my updates fast enough :p
2023-01-30 15:39:28 <psykose> mine is all "Unable to negotiate" instead
2023-01-30 15:39:45 <elly> valerius: rip
2023-01-30 15:40:28 <elly> so in conclusion, dudes are hella trying to log in, and pubkey auth is good
2023-01-30 15:42:17 <elly> the most popular users they're trying are 'root', 'admin', 'user', 'ubuntu', and 'debian'
2023-01-30 15:45:00 <invoked> 'pi' is a popular one too
2023-01-30 15:45:32 <elly> I'm actually surprised that alpine is not supported by digitalocean directly
2023-01-30 15:45:51 <psykose> small indie distro
2023-01-30 15:46:08 <lucidiot> i have a blog that returns HTTP 418 for all paths starting in /wp-
2023-01-30 15:46:13 <elly> "you probably haven't heard of it" psykose
2023-01-30 15:46:23 <psykose> who has tbh
2023-01-30 15:46:29 <lucidiot> what's alpine
2023-01-30 15:46:47 <invoked> https://www.redbubble.com/shop/alpine+linux
2023-01-30 15:49:09 <elly> my "REDUCE ATTACK SURFACE" laptop sticker is prompting a lot of questions already answered by my sticker
2023-01-30 15:49:16 <psykose> how the hell do you pass CXXFLAGS to like an AC_CHECK_HEADER
2023-01-30 15:51:25 <elly> don't you set CXXFLAGS for the whole configure run and it uses them when checking headers?
2023-01-30 15:52:17 <psykose> one would think so
2023-01-30 15:52:20 <psykose> hmm
2023-01-30 15:52:29 <invoked> elly: the alpine linux dog mat, for canine alpine linux enthusiasts
2023-01-30 15:54:47 <minimal> elly: I'm guessing in the past it was due to lack of cloud-init in Alpine releases prior to 3.13
2023-01-30 15:54:54 <elly> what is cloud-init?
2023-01-30 15:55:12 <Ermine> elly: at least DO supports custom images with cloud-init
2023-01-30 15:55:31 <minimal> elly: tail g
2023-01-30 15:55:46 <minimal> oops, ignore that lol
2023-01-30 15:55:48 <minimal> https://cloudinit.readthedocs.io/en/latest/
2023-01-30 15:55:53 <psykose> the configure.ac already does CXXFLAGS= and even sets them to like CXXFLAGS=$CXXFLAGS $CXXFLAGS_CXX11 and the _CXX11 one is set in 20 places and yet.. in the end they are somehow without the -std
2023-01-30 15:55:54 <psykose> classic
2023-01-30 15:56:13 <psykose> there's even a flag to "set to latest -std" and even that does not do anything
2023-01-30 15:56:14 <elly> yeah, DO supports custom images but then you have to use some janky third-party config for it, no?
2023-01-30 15:56:17 <minimal> elly: is it what most cloud providers use for configuration instances upon 1st boot
2023-01-30 15:56:31 <elly> TIL
2023-01-30 15:57:55 <Ermine> minimal: iirc DO uses their own do-agent in images they provide
2023-01-30 15:58:01 <minimal> elly: so when you create a DO droplet, AWS EC2 instance etc you can provide user-data (which is typically YAML) to define what users/groups to create, the SSH pubkey(s) to permit access, NTP settings, etc and cloud-init acts upon this (typically on 1st boot)
2023-01-30 15:58:58 <psykose> tis what powers the fancy webui
2023-01-30 15:59:34 <minimal> Ermine: that's not currently packaged for Alpine is it?
2023-01-30 15:59:43 <elly> that makes sense but is not something I have much inclination to mess with :P
2023-01-30 16:00:27 <invoked> it always struck me as odd that cloud-init started as an ubuntu thing.
2023-01-30 16:00:44 <psykose> every villain has their origin story
2023-01-30 16:00:44 <Ermine> minimal: I guess no
2023-01-30 16:00:57 <minimal> invoked: well somebody had to start it
2023-01-30 16:01:01 <Ermine> psykose: oic
2023-01-30 16:02:36 <minimal> Ermine: also the DO Agent files in github only seem to go back to May 2021 so its a somewhat recently think (or recently opensourced at least)
2023-01-30 16:18:43 <minimal> invoked: cloud-init appears to have started out as ec2-init written for Ubuntu
2023-01-30 16:30:29 <invoked> interesting. i can't recall the details anymore. ~2012-2013 was a noisy time. k8s, openstack, all that stuff was landing because entities like ibm saw big money in it.
2023-01-30 16:31:43 <minimal> invoked: there are copyrights in files going back to 2008. The oldest version I can see for download is 0.5.14 released in August 2010
2023-01-30 16:32:11 <f_> Hello. I have been blocked from fetching stuff on dl-cdn.alpinelinux.org for some weird reason (idk why it happened or how), as such I cannot work with `pmbootstrap` (pmOS). Who manages Alpine infrastructure?
2023-01-30 16:32:34 <ikke> f_: I do
2023-01-30 16:32:48 <f_> Sure.
2023-01-30 16:33:10 <f_> For some reason I cannot access http://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/APKINDEX.tar.gz
2023-01-30 16:33:11 <invoked> minimal: probably. i got approached by amazon early (like "pre-beta"), like they flew 3 people out to pitch me aws in 2005.
2023-01-30 16:33:35 <ikke> f_: what happens if you try?
2023-01-30 16:33:57 <f_> Access denied.
2023-01-30 16:34:07 <f_> Works fine when fetching from Tor.
2023-01-30 16:34:24 <elly> O_o
2023-01-30 16:34:28 <f_> o_O
2023-01-30 16:34:29 <ikke> dl-cdn is backed by fastly
2023-01-30 16:34:39 <f_> Oh¿
2023-01-30 16:34:43 <invoked> maybe your isp's dns is funky
2023-01-30 16:34:55 <f_> invoked: No it worked yesterday.
2023-01-30 16:35:03 <ikke> f_: what does dl-cdn.alpinelinux.org resolve to for you?
2023-01-30 16:35:05 <f_> Oh well maybe.
2023-01-30 16:35:08 <f_> Thanks!
2023-01-30 16:35:15 <f_> 
ACTION fixes their router

2023-01-30 16:35:34 <invoked> this bug report sucks
2023-01-30 16:59:43 <Apachez> maybe your isp is funky today?
2023-01-30 17:03:07 <latex> Here in The Netherlands we have a cancer ISP called Ziggo that has CG-NAT on IPv6
2023-01-30 17:03:23 <latex> Even KPN is cheaper than Ziggo nowadays
2023-01-30 17:03:44 <grattykraat900[m]> hi, im trying to run pw-jack catia on alpine linux, but its not showing everything
2023-01-30 17:03:44 <grattykraat900[m]> it doesnt show the pulseaudio stack or any applications playing or recording audio
2023-01-30 17:03:47 <grattykraat900[m]> 
ACTION uploaded an image: (47KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/uEJIMWoKjNMtjXIreKrhXyQY/image.png >

2023-01-30 17:12:17 <psykose> works for me so you might be missing something else
2023-01-30 17:16:03 <frag> has the raspberry pi 4b gotten slower after upgrade? notice specially in mpv/movies (could be something else i guess tho)
2023-01-30 17:16:04 <grattykraat900[m]> any ideas?
2023-01-30 17:23:25 <grattykraat900[m]> it works now, i think i needed to reboot after installing pipewire-pulse
2023-01-30 17:23:50 <psykose> frag: from what to what
2023-01-30 17:24:52 <psykose> mpv 0.35.0 specifically (and maaybe .1) have a chance to be slower on some hardware without --vd-lavc-dr=no
2023-01-30 17:24:54 <psykose> you can try that
2023-01-30 17:32:38 <frag> psykose: ahh, seems to be much better :))
2023-01-30 17:32:55 <psykose> is it .0 or .1
2023-01-30 17:34:19 <frag> psykose: i have mpv 0.35.0
2023-01-30 17:34:30 <psykose> upgrade first then try without that arg at all
2023-01-30 17:35:54 <frag> psykose: ah yes, .1 seems to work well without that option too
2023-01-30 17:36:16 <psykose> :)
2023-01-30 17:37:23 <psykose> it should now be =auto by default with auto meaning only when dr buffers are host cached
2023-01-30 17:49:48 <Guest2906> looking for a little help with awall any resources beside alpine wiki out there?
2023-01-30 17:50:52 <elly> this very IRC channel
2023-01-30 18:29:55 <ikke> The readme as well
2023-01-30 20:42:21 <WhyNotHugo> What writes /var/log/messages ?
2023-01-30 20:43:56 <psykose> syslog
2023-01-30 20:47:12 <minimal> or rsyslog, or syslog-ng, or syslog-really-the-ng, or syslog-the-final-generation ;-)
2023-01-30 20:51:35 <Ermine> minimal: syslog-the-final-generation (1)
2023-01-30 20:53:42 <ForeverNoob[m]> Given this setup on my sdcard which I use in the PI ZW 2 - https://paste.debian.net/1269023/ - setup-alpine is still telling me that no disk is found and whether I want to use /media/mmcblk0p1 - Is that normal? Did I forget something?
2023-01-30 20:54:21 <Forza> ikke: how did it go with the pstore thing on your server? 
2023-01-30 20:54:59 <Forza> Seems (knock on wood) that the iscsi patch I got solved my issues
2023-01-30 20:55:11 <ikke> Forza: nice, I haven't looked at it myself to be honest
2023-01-30 20:55:44 <ikke> one thing I'm wondering about that even when I tail dmesg output from a serial console, all I see is a sudden reboot
2023-01-30 20:55:52 <Forza> :) was just curious. Hope you find a solution. 
2023-01-30 20:55:53 <ikke> so not sure if pstore is going to add anything
2023-01-30 20:56:35 <Forza> That's how I had too. It rebooted too quick to get the trace on the sol or console 
2023-01-30 20:56:40 <ikke> ok
2023-01-30 20:57:35 <ikke> Forza: did you just manually enabled those extra modules, or did you use kernel config menu?
2023-01-30 20:57:53 <Forza> 
ACTION back in the day it used to be Linux that was stable and Windows blue screened... 

2023-01-30 20:57:55 <ikke> (I haven't extensively done a lot with kernels yet)
2023-01-30 20:58:16 <Forza> I used menuconfig
2023-01-30 20:58:21 <ikke> ok
2023-01-30 20:58:46 <Forza> But it ought to work by editing the .config 
2023-01-30 21:00:35 <ikke> hmm, your paste expired apparently
2023-01-30 21:00:58 <Forza> I'll have to check again tomorrow
2023-01-30 21:01:04 <Forza> Remind me :) 
2023-01-30 21:01:28 <Forza> I'm half asleep in bed already
2023-01-30 21:03:42 <ikke> heh
2023-01-30 21:03:45 <ikke> nite
2023-01-30 21:13:05 <Borley> I can't find any documentation on whether Alpine repositories distinguish between free and nonfree software
2023-01-30 21:15:01 <ikke> Borley: in general all packages in the repositories should be open source / floss, but we do have an exception for linux-firmware
2023-01-30 21:24:36 <Borley> Even better. Thanks.
2023-01-30 22:09:31 <minimal> ForeverNoob[m]: I thought you were using run-from-ram? Doesn't the "No disks found" message come from setup-disk (when disk mode is Sys or Data rather than run-from-ram) which setup-alpine calls
2023-01-30 22:24:32 <ForeverNoob[m]> @minimal: I wasn't quite sure. I just copied the contents of the image tar file to that first FAT32 partition (along with a few config modifications) and booted from it. I also partition + formatted the 2nd partition with EXT4 (in fdisk I specified it as a primary partition instead of an extended one because otherwise it didn't work for some reason)
2023-01-30 22:26:47 <minimal> ForeverNoob[m]: I'm confused - you copied files to the partition, so in other words you manually installed stuff, so why are you also running setup-alpine/setup-disk then?
2023-01-30 22:29:19 <ForeverNoob[m]> It's my understanding that for manual install, you first have to copy the image contents and then once booted, run alpine-setup to install it to either of those modes: https://wiki.alpinelinux.org/wiki/Raspberry_Pi#Installation
2023-01-30 22:33:57 <ForeverNoob[m]> (Please do correct me if I'm wrong tho)
2023-01-30 22:50:40 <ForeverNoob[m]> Ah, apparently it's a bit more involved: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12353#note_164623
2023-01-30 23:22:16 <minimal> ForeverNoob[m]: if you write all the files to SD card and boot it then why do you need to install it (for run-from-ram) as it is already installed on the SD card?
2023-01-30 23:24:01 <minimal> that Wiki article is more about *configuring* the Alpine that is already on the SD card, rather than installing it to the SD card
2023-01-30 23:25:38 <minimal> ForeverNoob[m]: that comment/note you referred to is about installing in Sys mode, not run-from-ram mode, that's a completely different scenario
2023-01-30 23:26:02 <ForeverNoob[m]> Yeah but I do intend to install it in sys mode :)
2023-01-30 23:26:48 <minimal> ForeverNoob[m]: I asked you earlier regarding run-from-ram or Sys mode and you didn't indicate Sys mode
2023-01-30 23:26:55 <ForeverNoob[m]> I want to have some more storage space for packages and definitely not want to run it on a FAT32 FS :p
2023-01-30 23:27:20 <minimal> e.g. "<minimal> ForeverNoob[m]: I thought you were using run-from-ram? Doesn't the "No disks found" message come from setup-disk (when disk mode is Sys or Data rather than run-from-ram) which setup-alpine calls"
2023-01-30 23:27:26 <ForeverNoob[m]> I thought you asked whether I knew what mode I was booting it from, but I wasn't sure. But my intention is to run it in sys mdoe.
2023-01-30 23:32:04 <minimal> so then install it in Sys mode via setup-alpine...which will overwrite the SD card
2023-01-30 23:41:50 <ForeverNoob[m]> Yeah so that's the thing... despite having 2 (properly) formatted partitions - https://paste.debian.net/1269023/ - `setup-alpine` is apparently only seeing 1 (that first FAT32 partition). And so apparently installing in sys mode is a bit more involved than just running `setup-alpine`, but now that I've seen that GitLab issue I can try and follow the steps there.
2023-01-30 23:42:34 <minimal> setup-alpine/setup-disk will set up the partitions itself
2023-01-30 23:44:40 <minimal> the only "more involved" part of that gitlab comment is pointing out that you need to answer "yes" to the "No disks available. Try boot media? (y/n)" because it can't use/repartition the SD card unless any filesystems on the SD card are unmounted first - which is what it will do if you answer "yes" to that prompt
2023-01-31 00:28:02 <ForeverNoob[m]> @minimal: Hmm, so I did... and apparently that borked my UART lol: https://paste.debian.net/plain/1269042/
2023-01-31 00:28:32 <ForeverNoob[m]> So I'm guessing it overwrote that first FAT32 partition. Or eh... perhaps a part of it :p
2023-01-31 00:32:30 <minimal> ForeverNoob[m]: yes I did say it would repartition the SD card. as in delete everything on it
2023-01-31 00:34:04 <minimal> ForeverNoob[m]: note in the output you shared: "WARNING: Erase the above disk(s) and continue? (y/n) [n] y" where you typed "y"? ;-)
2023-01-31 00:34:32 <minimal> specifically the use of the word "erase"
2023-01-31 00:35:09 <ForeverNoob[m]> And maybe also specifically the disk instead of partition :p
2023-01-31 00:36:07 <ForeverNoob[m]> But that's ok, luckily I wrote a script so that everything gets configured like I intend, maybe I can salvage a part of it to only do the 1st partition.
2023-01-31 00:36:13 <minimal> well my use of the work repartition earlier covered this....
2023-01-31 00:36:28 <alpine-meetbot> minimal meant to say: well my use of the word repartition earlier covered this....
2023-01-31 00:36:28 <minimal> s/work/word/
2023-01-31 00:38:52 <ForeverNoob[m]> Yeah luckily this is my messaround stage so not much of value was lost anyway.
2023-01-31 01:15:32 <ForeverNoob[m]> Ah yep it kinda worked. I modified my script a bit to reflect that: https://paste.debian.net/1269045/ - But now it's booting into emergency mode lol - https://paste.debian.net/1269046/
2023-01-31 01:16:18 <ForeverNoob[m]> Probably messed something up.
2023-01-31 01:16:48 <ForeverNoob[m]> Oh well, a new day tomorrow. Good night everyone!
2023-01-31 01:55:29 <minimal> ForeverNoob[m]: what "image file" are you copying?
2023-01-31 01:57:47 <minimal> ForeverNoob[m]: your script does not look correct for either a run-from-ram or a Sys mode layout - which is it supposed to setup?
2023-01-31 09:50:07 <ForeverNoob[m]> @minimal: This time I didn't use an image as argument. Initially when I prepared the sdcard for first boot, the argument was: `sh -x prepare_alpine.sh --image alpine-rpi-3.17.1-aarch64.tar --device /dev/mmcblk0 --model pi3` - That would essentially follow this part of the docs: https://wiki.alpinelinux.org/wiki/Raspberry_Pi#Preparation
2023-01-31 09:50:30 <ForeverNoob[m]> (along with some modification to config files that made stuff like UART serial possible)
2023-01-31 09:53:43 <ForeverNoob[m]> Then, after I did the `setup-alpine` thing, which then formatted the entire sdcard, I wanted to have those config modifications again, so since I didn't want to run the entire script again, I basically split the script into 2 functions. One partitions & formats the sdcard, and the other just copies / modifies config files. I used the latter function like this: `sh -x prepare_alpine.sh --device /dev/mmcblk0 --model pi3 --justcopy`
2023-01-31 10:51:46 <ForeverNoob[m]> I'm trying to setup a sys install for now, but using the methods I've just described I'm ending up booting into emergency mode for some reason.
2023-01-31 11:28:23 <wart_> hi folks. i have a patch to nerd-fonts-stable (version bump). I did a git clone on the aports, modified it, etc., but I'll be honest I'm not sure what the next step is to make a push request
2023-01-31 11:28:37 <wart_> (I installed it local and tested most of it and it is fine, probalby :))
2023-01-31 11:29:23 <wart_> do i git format-patch and e-mail it or git send-email or something? thanks in advance, and thanks for allthe work :)
2023-01-31 11:32:36 <Newbyte> wart_: there used to be a mailing list for this, but as I understand it's no longer used
2023-01-31 11:32:43 <Newbyte> you have to use GitLab
2023-01-31 11:33:29 <Newbyte> on GitLab you can fork aports and then make a merge request to the canonical one
2023-01-31 11:33:48 <wart_> thanks! i get it
2023-01-31 11:34:03 <Newbyte> great :)
2023-01-31 11:34:27 <wart_> been a while since i did the gitlab stuff :P
2023-01-31 11:46:35 <wart_> do i have to fork the whole aports or can i just fork one little bit :)
2023-01-31 11:49:52 <Habbie> you can fork the whole, but then you can do a shallow clone to your system, does that help?
2023-01-31 11:50:10 <Habbie> **
2023-01-31 11:50:12 <Habbie> you fork the whole, but then you can do a shallow clone to your system, does that help?
2023-01-31 11:50:57 <ikke> git clone --depth=1 <url>
2023-01-31 12:04:51 <ForeverNoob[m]> @minimal: Ultimately, I'm trying to setup a home security cam / webcam. Sys install is an in-between goal which may or may not suit me in the long run. I'm trying to mess around and install stuff manually so that I can know when/if stuff breaks and perhaps even improve the documentation regarding that. The scripting I do is mostly for personal use so that I can document while automating as I go.
2023-01-31 12:06:04 <ForeverNoob[m]> I think in the end I'll be using something to make a ready-made image using one of those cloud init tools, but I'll have to see how / if it suits my needs.
2023-01-31 12:27:15 <hhhhhhhhhh> hi, is there an easy way to either downgrade a single package to the last version i had installed, or make apk upgrade not just immediately kill itself like this:
2023-01-31 12:27:29 <hhhhhhhhhh> . /var/cache/apk # apk upgrade ERROR: unable to select packages:   so:libtiff.so.6 (no such package):     required by: imagemagick-7.1.0.60-r0[so:libtiff.so.6]
2023-01-31 12:28:18 <ikke> hhhhhhhhhh: what's in your /etc/apk/repositories? 
2023-01-31 12:28:51 <hhhhhhhhhh> edge/community latest-stable/main latest-stable/community edge/testing
2023-01-31 12:29:17 <hhhhhhhhhh> ok i removed the edge ones
2023-01-31 12:29:20 <hhhhhhhhhh> and now i can upgrade
2023-01-31 12:29:47 <ikke> :-)
2023-01-31 12:32:00 <hhhhhhhhhh> ty
2023-01-31 12:50:03 <markand> sometimes at shutdown mount-ro fails saying that / is already "used" (if I can read fast enough, I should take a picture)
2023-01-31 12:50:09 <markand> ending in a unproperly unmount of /
2023-01-31 12:50:31 <markand> 
ACTION hates more and more openrc

2023-01-31 12:51:34 <markand> could be nice if we can start shipping runit service files as -runit packages for those who prefer this
2023-01-31 12:53:35 <dwfreed> runit doesn't fix an in use filesystem
2023-01-31 12:55:17 <grawity> is it even possible to umount / normally?
2023-01-31 12:55:27 <grawity> oh mount-ro, nvm
2023-01-31 12:57:56 <markand> dwfreed, sure, but I (myself) think that runit is order of magnitude better than openrc anyway
2023-01-31 13:23:33 <ecraven> hello ;) a quick question, after `apk add curl which`, running `which curl` says it doesn't find anything... why is that? /usr/bin/curl is definitely there?
2023-01-31 13:24:37 <ecraven> it looks like which can't find *any* program... `which which` doesn't work either, it runs, but doesn't find anything.. the paths it looks in seem fine
2023-01-31 13:25:21 <ecraven> (running inside podman run -ti --rm alpine:lastest)
2023-01-31 13:32:04 <dwfreed> ecraven: command -v which
2023-01-31 13:34:00 <ecraven> just tried with 3.16, same: http://ix.io/4mHb
2023-01-31 13:36:00 <ForeverNoob[m]> What about type curl ?
2023-01-31 13:37:08 <ecraven> > curl is /usr/bin/curl
2023-01-31 13:37:11 <ecraven> but I want which to work :-/
2023-01-31 13:40:03 <dwfreed> is /usr/bin/curl executable
2023-01-31 13:40:14 <dwfreed> does it actually run
2023-01-31 13:40:21 <ecraven> yes and yes
2023-01-31 13:40:46 <ecraven> it's not a symlink to busybox, it's an actual executable ;)
2023-01-31 13:40:52 <dwfreed> (which actually tests executable flag, type does not)
2023-01-31 13:40:53 <ecraven> -rwxr-xr-x    1 root     root      241.9K Dec 21 07:34 curl
2023-01-31 13:41:24 <ecraven> hm.. maybe some problem with podman? all of this is inside rootless podman
2023-01-31 13:54:15 <ecraven> very strange, a colleague (running just about the same setup as me) cannot reproduce this. so it's somehow related to my system. sorry about the noise, I'll try to find out the rest myself
2023-01-31 14:04:30 <spazialecontinentale[m]> hi how can I run doas with some environment variable?
2023-01-31 14:04:31 <spazialecontinentale[m]> doas: MOLLY_CONF=/etc/mollysocket/default-config.toml: command not found
2023-01-31 14:04:31 <spazialecontinentale[m]> doas -u mollysocket MOLLY_CONF=/etc/mollysocket/default-config.toml mollysocket server
2023-01-31 14:21:38 <lopid> sh -c '…' ?
2023-01-31 14:26:13 <spazialecontinentale[m]> lopid: i'm writing a wrapper for an app:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/mgnlMPbgLtkzIgGatSSQoKfS>)
2023-01-31 14:26:35 <spazialecontinentale[m]> s/MOLLY_CONF/MOLLY\_CONF/, s/kn/know/
2023-01-31 14:27:27 <elly> spazialecontinentale[m]: your doas.conf needs either keepenv or setenv in it for that command
2023-01-31 14:27:37 <elly> doas doesn't propagate env vars by default
2023-01-31 14:27:42 <elly> see doas.conf(5) keepenv section
2023-01-31 14:28:22 <elly> setenv { MOLLY_CONF } or somesuch may work
2023-01-31 14:45:32 <spazialecontinentale[m]> <elly> "setenv { MOLLY_CONF } or..." <- this approach need the edit of the config file
2023-01-31 14:46:56 <spazialecontinentale[m]> * config file. I'm writing a APKBUILD that i want to push
2023-01-31 14:47:39 <spazialecontinentale[m]>  * this approach needs the edit of the config file. I'm writing a APKBUILD that i want to push
2023-01-31 14:50:05 <elly> oh
2023-01-31 14:50:28 <elly> setuid wrapper script I guess?
2023-01-31 14:51:35 <spazialecontinentale[m]> elly: can you explain?
2023-01-31 14:58:08 <elly> like, a script that doas is supposed to run that sets the needed env variable and then invokes mollysocket, or something
2023-01-31 14:58:21 <elly> and having that script itself be runnable by doas instead
2023-01-31 15:01:47 <spazialecontinentale[m]> elly: well this is already a wrapper script.... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/IGJGDhQRvNTZaNLmeVsUizxl>)
2023-01-31 20:22:29 <Waitingfortherain[m]> When can we see an updated version of luarocks and lua-unbound to arrive on alpine linux
2023-01-31 20:22:40 <Waitingfortherain[m]> since prosody depends on that for a lot of its modules it would be great to see it
2023-01-31 20:23:33 <ikke> Waitingfortherain[m]: !27474
2023-01-31 20:23:39 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/27474
2023-01-31 20:39:09 <Waitingfortherain[m]> hopefully soon!
2023-01-31 21:59:25 <tomk> Did anyone sucessfully connect to alpine linix via XRDP from Windows 10 or 11? The instructions found in the Wiki doesnot provide any successful solutions.
2023-01-31 23:14:43 <elly> well, wifi doesn't seem to work on my old macbook with the 3.17 extended installer
2023-01-31 23:14:49 <elly> but on the upside, it's air gapped for added security!
2023-01-31 23:15:52 <psykose> it is probably the custom b43 firmware case :p
2023-01-31 23:16:16 <elly> hehe, it is a very old macbook so it is possible
2023-01-31 23:16:19 <elly> I decided to just leave it :)