2020-07-01 04:20:07 <lunaphyte> ah, bummer.  i only just noticed that the aws-cli package is  version 1
2020-07-01 06:42:05 <markand> hi there
2020-07-01 06:42:44 <markand> someone has an idea why I can't add my google account into GNOME? it opens the webpage to authenticate through google form, but after typing my address it simply says "sorry an error occured"
2020-07-01 06:54:40 <Cogitri> markand: Huh, that worked for me a few months ago
2020-07-01 06:54:55 <Cogitri> I'll check if when I'm Home
2020-07-01 07:29:34 <EF> So am in a bizarre situation. Pulling a disk from raid causes Alpine to go to the recovery prompt.
2020-07-01 07:48:49 <EF> Booted memory stick to partition new drive and rebuild array. Would installing syslinux be required?
2020-07-01 08:57:39 <EF> So WD red shr disks are slooooow
2020-07-01 11:35:10 <moosl> which is better w3m or elinks?
2020-07-01 12:06:42 <earnestly> elinks is closer to the definition of a browser, whereas w3m is more like a "viewer"
2020-07-01 12:06:42 <earnestly> Based on that you can pick whatever suits your requirements better
2020-07-01 12:55:19 <EF> Is there a link to the process fo requesting certain software is added to the repositories?
2020-07-01 12:56:24 <ikke> It's a matter of creating a ticket here: https://gitlab.alpinelinux.org/alpine/aports/-/issues
2020-07-01 12:59:30 <EF> Okay thanks. Will get this server configured and determine if there is a real need. sshl is one.
2020-07-01 13:06:41 <MY-R> EF, sshl or sslh?
2020-07-01 13:10:25 <yann-kaelig> Hi.
2020-07-01 13:21:52 <mckay-trabajo> guys: l
2020-07-01 13:21:53 <mckay-trabajo> ocalhost:~# cat > test << EOF
2020-07-01 13:21:53 <mckay-trabajo> EOF
2020-07-01 13:21:53 <mckay-trabajo> # $(cat /etc/alpine-release | cut -d'.' -f1,2)
2020-07-01 13:21:53 <mckay-trabajo> -ash: : not found
2020-07-01 13:21:53 <mckay-trabajo> $(cat /etc/alpine-release | cut -d'.' -f1,2)
2020-07-01 13:21:54 <mckay-trabajo> localhost:~# cat test
2020-07-01 13:21:54 <mckay-trabajo> #
2020-07-01 13:21:55 <mckay-trabajo> v3.8
2020-07-01 13:21:55 <mckay-trabajo> localhost:~#
2020-07-01 13:21:56 <mckay-trabajo> here why the first line give me error?
2020-07-01 13:25:20 <ikke> Do you happen to have carriage returns in there?
2020-07-01 13:33:57 <superduck> I'm trying to rebuild the boottime apk index included with the iso, but I get this error on bootup: "package mentioned in index not found"
2020-07-01 13:34:46 <superduck> I've generated a pub/priv key pair and stuck them in /etc/apk/keys, and I'm thinking that might not be the right spot.
2020-07-01 13:44:44 <id4rk> hello, i extrated the displaylink driver for ubuntu, and while installing manually it asks me a methotd on how it can be launched
2020-07-01 13:45:04 <id4rk> Please set an environment variable SYSTEMINITDAEMON to 'upstart', 'systemd' or 'runit' before running the installation script to force one of the options.
2020-07-01 13:45:29 <id4rk> i know alpine has not systemd, but what about upstart or runit?
2020-07-01 13:46:53 <superduck> It uses openrc, which is none of those. You can install runit independent of openrc, and then let runit manage your displaylink driver launching.
2020-07-01 13:48:18 <id4rk> superduck: brb
2020-07-01 13:54:52 <EF> MY-R: sslh, always get it backwards
2020-07-01 13:56:15 <id4rk> i installed runit
2020-07-01 13:56:33 <id4rk> how can i set this environment variable?
2020-07-01 13:56:58 <id4rk> export SYSTEMINITDAEMON=runit ?
2020-07-01 13:57:00 <MY-R> EF, package is in EDGE testing: https://pkgs.alpinelinux.org/packages?name=sslh*&branch=edge
2020-07-01 14:02:03 <EF> ok thanks. Am on 3.12 at the moment.
2020-07-01 14:10:44 <id4rk> how can I enable DKMS?
2020-07-01 14:12:58 <indy> hi all can apk download from password protected site?
2020-07-01 14:13:32 <indy> i tried /root/.netrc but not helped
2020-07-01 14:14:37 <Hello71> indy: did you try http://user@pass:whatever/
2020-07-01 14:14:51 <indy> no
2020-07-01 14:16:19 <indy> Hello71, another problem is that password contains '@' sign
2020-07-01 14:16:26 <indy> and it is https
2020-07-01 14:16:51 <Hello71> I think basic auth does not support username containing :
2020-07-01 14:16:56 <mckay-trabajo> ikke i resolved putting a echo in first line after the EOF of the cat line.. the problem was lack of bash in busibox fisrt minutes after install.. i cannot be installed .. thanks any way..
2020-07-01 14:17:00 <Hello71> probably some servers are more picky
2020-07-01 14:22:40 <earnestly> Hello71: (If you check Orange's blackhat talk about parsers you'll learn that implementations essentially ignored the RFC and proceeded to do the dumbest possible things they could)
2020-07-01 14:23:04 <Hello71> well I already know that URL parsing is done as shit as possible
2020-07-01 14:23:24 <Hello71> and I would be surprised if authentication was done better for some reason
2020-07-01 14:23:56 <earnestly> Hm, but my opinion has shifted from misinterpretation to the reckless and malicious
2020-07-01 14:25:23 <earnestly> E.g. a specification might say "the colon has to be last", to which implementations ignore and parse a colon however they want.  "A header must be terminated by CR LF", implementations do whatever they hell they like
2020-07-01 14:26:13 <earnestly> Even curl, as per that talk, can't even reliably parse a url whereby the bug report essentially told Orange to ensure you only give valid urls to curl
2020-07-01 14:30:03 <id4rk> Unsatisfied dependencies. Missing component: DKMS. <-- Any chance to fix this?
2020-07-01 14:49:42 <Cogitri> id4rk: Maybe a bit of context would help
2020-07-01 14:51:32 <Hello71> Cogitri: scroll up
2020-07-01 14:52:04 <Cogitri> Ah, oops
2020-07-01 15:02:40 <id4rk> I'm tying to make DisplayLink driver work.
2020-07-01 15:05:30 <telmich> I'm just trying to efi load alpine 3.12; however the screen stays black and the t490s then reboots; booting ubuntu, the grub menu is shown
2020-07-01 15:06:05 <telmich> I'm just reporting this, because it has happened for all devices I recently tried to install alpine to -> something is not good with the efi loader on many systems
2020-07-01 15:06:34 <eudaldgr> hi, everyone
2020-07-01 15:27:33 <Hello71> I think alpine uses grub for efi by default
2020-07-01 15:27:43 <Hello71> did you disable secure boot
2020-07-01 15:55:44 <moosl> I am gonna use musl as main install and see  what happens
2020-07-01 16:04:52 <moosl> what's a ultra ligth, non-mem hogger, torrent for musl?
2020-07-01 16:05:53 <fruchti> I'm running Alpine on a Raspberry Pi and am trying to find out how to persistently add a custom init script. lbu does not track /etc/init.d and I don't know how to convince it otherwise
2020-07-01 16:06:25 <fruchti> My goal here is to get console autologin. For that I'd like to have a copy (or symlink) of /etc/init.d/agetty as agetty-autologin.tty1
2020-07-01 16:07:14 <fruchti> I'm pretty much a noob here as Alpine is my first exposure to OpenRC and I've not been able to find an anwer on the wiki or elsewhere
2020-07-01 16:07:42 <moosl> fruchti: also look into /etc/inittab
2020-07-01 16:08:35 <ikke> moosl: what has that to do with it?
2020-07-01 16:09:14 <moosl> ikke: there are options for console login, plenty
2020-07-01 16:12:33 <fruchti> Oh, thanks, that's a lot easier actually!
2020-07-01 16:12:48 <ikke> moosl: I missed that part :)
2020-07-01 16:12:58 <ikke> You need agetty for autologin though
2020-07-01 16:13:02 <ikke> bb getty does not support it
2020-07-01 16:13:05 <moosl> ikke: I recalled that part from past battles
2020-07-01 16:14:08 <fruchti> ikke: Yeah, used agetty and had it working in no time
2020-07-01 16:15:07 <fruchti> While I'm here…  Is there a straight-forward way to defer network connections until after console login?
2020-07-01 16:15:38 <moosl> fruchti: rc-update del 'whatever netwoking service'
2020-07-01 16:15:59 <moosl> fruchti: look at 'rc-update show' or play 'rc-update'
2020-07-01 16:16:26 <fruchti> moosl: I guess moving them from 'boot' to 'default' won't make a difference? I'd still like to be able to SSH into the RPi
2020-07-01 16:16:49 <fruchti> I just won't need to usually and not having the boot time hit would be nice
2020-07-01 16:17:37 <moosl> adjust it to your own needs, if default is to much you could prolly do a simple script to trigger it late.
2020-07-01 16:48:12 <superduck> Anyone able to provide assistance with how to build an APKINDEX that doesn't break apk?
2020-07-01 16:51:06 <yann-kaelig> Well, I'm not sure about that, but for a subpackage I can not give a name like foo-bar() ? It must be _foo_bar and in subpackages list I have to made a transformation $pkgname-foo-ba:_foo_bar ?
2020-07-01 17:09:55 <superduck> Anyone know how to address "package mentioned in index not found" with a custom APK repository? I think it's a bad APKINDEX, but maybe not.
2020-07-01 17:17:46 <moosl> superduck: sorry I have no idea about custom repos in apk
2020-07-01 17:27:47 <afontain_> superduck: I just use abuild. It will add the package in ~/packages with a proper APKINDEX
2020-07-01 17:30:00 <superduck> Looks like I might need this extension:                 abuild-sign --installed
2020-07-01 17:46:27 <moosl> does windoze10 suffer mental retardation?
2020-07-01 17:47:21 <moosl> on an underpower machine musl is able to play 720p60 smoothly with 2GB ram, while windoze10x32 chokes and with 4GB of ram
2020-07-01 17:47:39 <mps> aiee, yet another chromebook converted to Alpine linux with kernel 5.7 \o/
2020-07-01 17:47:55 <mps> Acer R13
2020-07-01 17:49:14 <moosl> am about to do this on intel tablet as m$$10 gives really weird issues.
2020-07-01 17:52:11 <Hello71> mps: I heard that new chromebooks are not linux-friendly
2020-07-01 17:53:41 <mps> ehm, right, and even old ones are not friendly
2020-07-01 17:54:26 <mps> too much locked for making normal linux on them
2020-07-01 17:56:21 <mps> but I didn't found anything better with arm cpus to run linux and move it around, using them as notebooks about 5 years
2020-07-01 17:57:06 <moosl> mps: you can run linux on androids without a problem
2020-07-01 17:57:42 <moosl> plenty of excellent rooted devices out there
2020-07-01 17:58:07 <mps> moosl: this type of devices are not useful to me
2020-07-01 17:58:54 <mps> I want big screen, keyboard, touchpad, usb ports etc, like normal notebooks
2020-07-01 17:59:07 <moosl> oh I see.
2020-07-01 17:59:22 <moosl> but arm based
2020-07-01 18:01:01 <mps> my current notebook is less than 1Kg weight, have long time on battery, screen resolution 2400 x 1600
2020-07-01 18:01:42 <mps> with alpine on arm64, ofc
2020-07-01 18:03:01 <albr> unowe[d]: what notebook is that? sounds pretty decent
2020-07-01 18:04:34 <moosl> mps: what about battery life?
2020-07-01 18:04:43 <mps> samsung chromebook one plus, https://www.samsung.com/us/computing/chromebooks/12-14/xe513c24-k01us-xe513c24-k01us/
2020-07-01 18:05:46 <MY-R> moosl, and what did you use to play that video on windows and alpine?
2020-07-01 18:06:02 <mps> moosl: depends on work, compiling shorten time fast, but normal use keep it about 6-7 hours
2020-07-01 18:06:24 <moosl> mps: ha ha! arm macbook
2020-07-01 18:06:43 <moosl> MY-R: youtube 720p60 vids
2020-07-01 18:07:30 <MY-R> firefox? chromium? gpu acceleration on or off?
2020-07-01 18:08:08 <mps> MY-R: firefox
2020-07-01 18:08:25 <moosl> firefox
2020-07-01 18:08:30 <mps> no GPU drivers, panfrost for mali is still buggy
2020-07-01 18:09:03 <moosl> MY-R: I assume the wintel tablet turns it on by default
2020-07-01 18:09:27 <MY-R> not always
2020-07-01 18:11:42 <moosl> so turn it on in firefox prefernces?
2020-07-01 18:14:25 <moosl> i'll check, it doesn't matter because soon muls will nuke windoze
2020-07-01 18:59:24 <MY-R> mps, because of your performance edge kernel I had to install cpupower package! ;) to allow my cpu to drop frequencies down when idle :D
2020-07-01 19:01:21 <MY-R> but also learnt that intel_pstate doesnt support governors like "ondemand" :P
2020-07-01 19:02:56 <mps> MY-R: you are free to fill issue/bug about that
2020-07-01 19:05:05 <mps> moosl: btw, I forgot that my first arm based 'notebook' was asus transformer TF101 (iirc) with docking add on, long time ago
2020-07-01 19:05:37 <MY-R> it isnt a problem and not even sure if got some power usage impact because didnt notice any higher temperatures on cpu even if freqs were all time max
2020-07-01 19:05:45 <mps> I installed debian on it and used it about 4 years as notebook
2020-07-01 19:07:55 <mps> MY-R: yes, I wanted this kernel to be small and because that not much features are enabled, but I'm open to hear suggestions and ideas from anyone
2020-07-01 19:08:25 <mps> and, as you know, every alpine developer can make tweaks and push
2020-07-01 19:10:28 <yann-kaelig> When an abuild build success but the installation part failed, can I restart only the installation part withotu the need to rebuild fro mthe beginning ? I tried abuild package , but look like it's not the right command
2020-07-01 19:11:40 <mps> abuild rootpkg
2020-07-01 19:11:50 <mps> abuild index
2020-07-01 19:11:59 <mps> abuild -h
2020-07-01 19:12:10 <MY-R> mps, in my case I learnt something new but not sure how it work on for example ARM boards, arent temperatures there higher when idle and set default to "performance" ?
2020-07-01 19:12:37 <mps> yes
2020-07-01 19:12:48 <mps> but people like performance
2020-07-01 19:13:09 <yann-kaelig> thx  I know abuild -h, but sometime s Ican have some difficulties to understand things. Now, it's more clear for me
2020-07-01 19:13:14 <mps> and there are modules for those who want to change that
2020-07-01 19:15:22 <MY-R> yep
2020-07-01 19:17:04 <c705> mps: so i guess upsteam fixed grub. that was nice of them ;)
2020-07-01 19:17:36 <mps> nowadays people don't care about energy usage but abour performance and features
2020-07-01 19:17:40 <c705> also, i believe i already know the answer to this, but does apk do integrety checks on packages that i'm pulling in from apd add
2020-07-01 19:17:51 <ikke> yes
2020-07-01 19:18:14 <ikke> try removing the keys from /etc/apk/keys :)
2020-07-01 19:18:35 <c705> ikke: these keys are part of alpine-base or apk-tools I imagine?
2020-07-01 19:18:46 <ikke> alpine-keys is a separate package
2020-07-01 19:18:48 <ikke> but yes
2020-07-01 19:18:53 <c705> I see. cheers!
2020-07-01 19:19:58 <c705> Actually I lied. grub still broken (upstream I assume)
2020-07-01 19:26:56 <Hello71> mps: uh... think you should talk to the overwhelming majority of Linux system owners (cloud + Android)
2020-07-01 19:33:39 <yann-kaelig> I have a foo-dev and a foo-bar-dev which both are subpackages  from foo and some files  I want to foo-bar-dev are automaticaly installed in foo-dev. The foo-dev() is not explicitly written in the apkbuild. I have two ideas to move my files from foo-dev to foo-bar-dev. I'm asking first because the compilation is long. Can I inside a subpackage do a mv "pkgdir-dev"/a/b/c "subpackage"/a/b/ or should I have to write the foo-dev()
2020-07-01 19:33:40 <yann-kaelig> in the apkgbuild
2020-07-01 19:35:43 <ikke> if you place foo-bar-dev before foo-dev, then you can move it from $builddir to $pkgdir before they have been moved to foo-dev
2020-07-01 19:36:09 <ikke> in subpkgs
2020-07-01 19:36:35 <ikke> or from $pkgdir to $subpkgdir I mean
2020-07-01 19:37:18 <yann-kaelig> ha! ok so even of the foo-dev is not written in the apkbuild, there are always first in the foo package, but the move is done automatically  from foo to foo-dev
2020-07-01 19:37:37 <ikke> there is a default_dev function that does that
2020-07-01 19:39:05 <yann-kaelig> thx a lot ikke, I  will take a look at this function
2020-07-01 19:39:34 <ikke> there is also a helper function called amovre
2020-07-01 19:39:37 <ikke> amove
2020-07-01 19:39:54 <ikke> it can make it easier to move a file to another dir, including the parent directories
2020-07-01 19:40:24 <yann-kaelig> thx
2020-07-01 19:56:25 <monado> hi
2020-07-01 19:56:31 <monado> Oh it looks like albr hasn't died yet
2020-07-01 19:56:37 <ikke> heh
2020-07-01 19:56:50 <ikke> but not a lot of activity
2020-07-01 19:57:26 <monado> I see
2020-07-01 22:50:16 <bobbintb> Hello, I am having an issue and was hoping someone is available to help me.
2020-07-01 22:50:50 <bobbintb> I have VMWare Workstation 14. I downloaded both the virtual and standard ISOs for Alpine 3.12.0. The checksums match so I know the downloads are good. Whenever I try to install either version in a new VM, it never boots from the ISO. I don't have any issue booting other ISOs. Pressing ESC to specify boot device nor changing BIOS settings makes a
2020-07-01 22:50:50 <bobbintb> difference, and really, it should boot from CD by default anyway. I'm not sure what's the matter. Any ideas?
2020-07-01 23:01:27 <telmich> Aaaaand a success report:
2020-07-01 23:02:09 <telmich> I'm running alpine on the T490s. After changing the USB port (!) the loader worked in uefi mode
2020-07-01 23:24:45 <godot> howdy. I am curious what the status of grub2 support is with Alpine Linux, sufficient to support a completely encrypted /boot partition? The threads I've been reading give conflicting results, as it looks like grub 2.06 has not yet been integrated into Alpine yet?
2020-07-01 23:25:35 <joe9> anyone using alpine in bhyve (freebsd vm) to watch netflix?
2020-07-01 23:26:55 <Hello71> godot: grub doesn't really require distro support...
2020-07-01 23:27:07 <Hello71> joe9: didn't you just ask about this in #freebsd
2020-07-01 23:28:10 <godot> Well, pulling grub-current and doing the install would result in subsequent package updates nuking that config, no"
2020-07-01 23:28:12 <godot> ?
2020-07-01 23:31:12 <joe9> Hello71, on that forum to check if there was a way to do natively on FreeBSD. I cannot find any reference to widevine and FreeBSD. So, want to check if I can use alpine in a vm instead.
2020-07-02 05:16:33 <shrizza> Is there an updated definitive howto for persistent disk install for rpi3?
2020-07-02 05:17:20 <shrizza> I've tried the howto currently on the wiki a few times but something doesn't seem right.
2020-07-02 05:21:56 <albr> MrPotatoBob[d]: hello! anyone here use alpine as a desktop os, hows the experience?
2020-07-02 05:32:04 <maxice8> Me, it is pretty decent
2020-07-02 09:33:19 <EF> Are there any limitations to using syslinux with a seperate /boot and encrypted root please?
2020-07-02 10:00:46 <abenz> if I have only an apkovl from a usb that was setup nicely
2020-07-02 10:01:22 <abenz> now I wish to re-create said config, so since I have the apkovl, I setup alpine on a new usb (setup-bootable), and then remount as rw and copy the apkovl
2020-07-02 10:02:43 <abenz> I reboot, now my settings are back, but some kernel modules are not loaded and other things. So I think maybe its the missing cache, I setup apk cache and I see all my programs downloaded into the cache. Still not like the original USB though.. is there an obvious step I missed?
2020-07-02 10:06:43 <EF> To answer my own question, possibly but even the higher security option in the wiki works with a non-encrypted /boot (mine is on raid1 too).
2020-07-02 10:31:37 <OleTange> can you somehow manually verify apk packages for another architecture?
2020-07-02 10:32:19 <OleTange> when running "apk verify someaarch64package" on an x86_64 system i get an UNTRUSTED back
2020-07-02 10:32:40 <OleTange> x86_64 packages work fine
2020-07-02 10:37:11 <dngray> where is the best place to get the alpinelinux logo?
2020-07-02 10:37:38 <dngray> been browsing through https://gitlab.alpinelinux.org/alpine not sure which repo
2020-07-02 10:44:17 <dngray> i guess i could just edit https://alpinelinux.org/alpinelinux-logo.svg
2020-07-02 10:53:32 <godot> Hi again. I am trying to figure out what level of support Alpine Linux has for GRUB2 / grub 2.06, to facilitate a completely encrypted /boot partition? And in the event this is not a natively supported package, what issues there would be with pulling grub 2.06 manually during the install in terms of future package updates?
2020-07-02 10:56:17 <godot> As it looks like the most current grub integration is grub-2.04-r1.apk but from what I understand grub 2.06 is needed for encrypted /boot
2020-07-02 11:03:03 <MY-R> godot, 2.04 working just fine with fully encrypted /root , if using BIOS (MBR) then can have just one single partition with LUKS, for EUFI stuff of course need extra unencrypted partition with /boot/efi
2020-07-02 11:03:52 <MY-R> but installation process is more manual and done in chroot
2020-07-02 11:05:22 <MY-R> and if I remember correctly 2.04 don't have LUKS2 support like 2.06
2020-07-02 11:09:56 <OleTange> do  I need to import some keys to verify packages  for another architecture?
2020-07-02 11:11:27 <ikke> yes, each arch has its own keys
2020-07-02 11:11:51 <OleTange> are these the keys under /etc/apt/keys ?
2020-07-02 11:12:05 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/alpine-keys
2020-07-02 11:12:05 <OleTange> if so what package do i need to install for other archs?
2020-07-02 11:12:38 <ikke> OleTange: Each arch only has the keys relevant for that arch
2020-07-02 11:13:54 <OleTange> sure, but what if i need to cross verify from one arch for another arch?  I guess there is a key package for each arch, so I coulkd just extract those  to /etc/apk/keys and it should work, no?
2020-07-02 11:14:07 <ikke> yes
2020-07-02 11:14:19 <ikke> or get it from aports
2020-07-02 11:15:29 <abenz> ikke: does it make sense to copy over an apkovl (diskless) from one usb to another (after running setup-bootable on the second one) ?
2020-07-02 11:16:11 <ikke> Sure
2020-07-02 11:16:49 <abenz> because I've been trying to do that but the new usb always has problems loading kernel modules etc, as soon as I delete the apkovl it returns to normal. I extracted the tar.gz and I dont see any files that might cause this
2020-07-02 11:17:51 <abenz> problem is sometimes even the original usb fails to load some kernel modules, (after booting, if I say modprobe amdgpu it works).. dont know why it has this unpreditable behaviour, any idea?
2020-07-02 11:18:17 <abenz> tried multiple USBs btw, and even an sd card
2020-07-02 11:18:32 <abenz> to try rule out failure on the usb itself
2020-07-02 11:20:17 <godot> Thanks MY-R. Yes you are correct, only LUKS1 support is available with current grub-2.04, but /boot encryption is supported.
2020-07-02 11:20:48 <godot> MY-R: do you have any resources for chroot install?
2020-07-02 11:21:01 <godot> Or is that a part of the normal install process?
2020-07-02 11:21:15 <abenz> the bar that extracts the cache(?) ie right before kernel loads and openrc shows the messages about services starting, this bar sometimes zips very fast (here it fails to load kernel modules), other times it goes slower and here everything loads correctly
2020-07-02 11:24:27 <MY-R> godot, you can try with that but there is many steps which could just omit but it will give you some point of view https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop
2020-07-02 11:24:46 <godot> Thanks MY-R!
2020-07-02 12:00:42 <MY-R> godot, here you can use as a help that, is step by step and working fine: http://prettypasta.org/X0T33xs
2020-07-02 12:01:08 <godot> Thanks MY-R, I am going through the laptop link you sent now too.
2020-07-02 12:02:28 <MY-R> dunno how that laptop link will work for you, that second link is basic and simple stuff
2020-07-02 12:37:48 <godot> MY-R: have you ever done this setup with a fake raid (software) setup? It looks like the wiki discourages encryption or LVM for /boot if it's a RAID configuration.
2020-07-02 12:38:25 <godot> Your /boot partition should either not be on RAID, or else be on a RAID1 array, with no further layers of encryption or LVM. (Alpine's default bootloader extlinux can't handle either. Grub2 can handle /boot being on LVM.) The usual practice is to create a small (32--100 MB) partition for /boot. That can be a mirrored (RAID1) volume, however this is just for post-init access.
2020-07-02 12:42:56 <MY-R> godot, nope
2020-07-02 12:45:48 <MY-R> godot, but should work: https://wiki.archlinux.org/index.php/Dm-crypt/Encrypting_an_entire_system#LUKS_on_software_RAID
2020-07-02 12:46:33 <MY-R> I dont have any setup with qemu to check it for now
2020-07-02 12:53:18 <godot> Thanks MY-R!
2020-07-02 12:54:31 <EF> I am running /boot unencrypted on a raid 1 array and everying else on raid10 + luks + lvm using syslinux. Works fine. Grub was a PiTa, all on raid 10 was a PiTa...
2020-07-02 12:55:58 <TBB> if security is that much of a concern, get an USB stick that uses AES with a PIN code and use that as your /boot
2020-07-02 12:56:53 <godot> Yeah and Dell Tower fakeraid makes it worse. They market this as a full RAID controller but it presents two drives instead of RAID volume
2020-07-02 12:57:35 <EF> godot: just use mdadm or lvm as softraid.
2020-07-02 12:57:46 <godot> EF: which one is preferable?
2020-07-02 12:58:04 <godot> I am running LUKS1 with this, hopefully. So probably LVM?
2020-07-02 12:59:45 <EF> Depends on your usecase. mdadm with luks on top and lvm on that is simple. lvm raid with seperate luks containers is also simple but requires some extra work to overcome multiple keys
2020-07-02 13:00:55 <EF> Have a read on what order to use lvm, luks and raid. It can make a lot of difference apparently.
2020-07-02 13:03:23 <TBB> it shouldn't even require extra work; you'll get a device mapper entry you can then luks to your heart's content
2020-07-02 13:16:37 <godot> Thanks EF and TBB
2020-07-02 13:34:33 <Cogitri> So I just downgraded a rig from edge to 3.12 and I'm getting `Error code: NS_ERROR_NET_INADEQUATE_SECURITY` on all HTTPS websites now
2020-07-02 13:34:43 <Cogitri> Does someone else have the same erro?
2020-07-02 13:39:19 <Hello71> Cogitri: try creating a new profile
2020-07-02 13:41:28 <Cogitri> Seems like that didn't change anything about that
2020-07-02 13:42:00 <Cogitri> Sites like GitHub work but FF complains about weak encryption so I guess it can't use the right encryption algos for some reason?
2020-07-02 13:51:23 <Hello71> I heard in ff 78 there were some changes
2020-07-02 16:11:21 <webstrand> Does anyone have recommendations for putting sshd in my initramfs? I'd like to remotely luksOpen the root partition.
2020-07-02 16:12:09 <webstrand> From what I could tell mkinitfs handles decryption via a binary executable and doesn't have any hooks for alternative authentication mechanisms.
2020-07-02 16:33:50 <godot> Ok, so I've got a Dell Precision Tower which has a fakeraid type array; whenever I go to use mdadm I am getting the error that /dev/sda1 appears to be part of a raid array. It looks like dmraid is used for this type of configuration, as from what I understand fakeraid still uses software raid, but whenever a drive fails the replaced drive can be repopulated via BIOS without the need for software raid recovery. So does Alpine support any dmraid type
2020-07-02 16:33:50 <godot> configurations?
2020-07-02 16:43:51 <nacelle> godot: you can go into the bios and change it to AHCI mode for the sata ports
2020-07-02 16:44:06 <nacelle> much easier than trying to make fakeraid work
2020-07-02 16:48:50 <godot> And then use mdadm? But wouldn't that defeat the point of being able to reconstruct the RAID array via BIOS instead of OS-based recovery?
2020-07-02 16:51:05 <EF> godot: you should check if there is linux support for the hardware raid. See this which outlines the usual problems
2020-07-02 16:52:06 <EF> Okay, my copy and paste is broken
2020-07-02 16:52:07 <godot> it looks like there is a dmraid package in the Alpine package repo, I just haven't found any resources for how that would be implemented during install. How close is Arch linux to Alpine in terms of the install process?
2020-07-02 16:52:30 <EF> https://forums.freebsd.org/threads/dell-boss-vd-raid1.73878/
2020-07-02 16:53:39 <nacelle> godot: why would you want to reconstruct an array via BIOS?
2020-07-02 16:53:53 <nacelle> that takes the machine down, whereas doing OS based recovery has the machine up
2020-07-02 16:54:00 <godot> EF: yeah that's the problem. This controller does not present both drives as a single drive, either in RAID1 or striped RAID0. Both /dev/sda and /dev/sdb show up; but as long as one drive is still healthy, a replacement can be repopulated via device BIOS instead of mdadm-based recovery.
2020-07-02 16:54:27 <nacelle> also, you can recover on other machines with mdadm, with BIOS you're hosed into having to recover with another controller of the same kind (which can be harder to acquire than just another box with sata ports)
2020-07-02 16:54:46 <godot> nacelle: good point. This is actually just my workstation, so whatever is easiest to recover is the goal.
2020-07-02 16:54:55 <nacelle> you want to avoid fake raid
2020-07-02 16:55:09 <godot> ok
2020-07-02 16:55:16 <EF> Also read https://linux.die.net/man/8/dmraid
2020-07-02 16:55:57 <EF> Try to work out what chipset you have. There is a how-to dmraid on Arch's wiki
2020-07-02 16:56:24 <godot> ok thank you
2020-07-02 16:56:37 <inad922> hello
2020-07-02 16:57:01 <EF> My HP gen8 has a b120i which is not supported and the driver is propriatory (RedHat/Cent/Windows...)
2020-07-02 16:57:02 <inad922> Could someone tell me how to do proxy settings so that it's compatible with apk?
2020-07-02 16:57:26 <inad922> I tried to get this info from several github issue threads but no luck
2020-07-02 16:57:44 <inad922> I'm trying to build a docker image behind a proxy with basic authentication
2020-07-02 16:57:57 <EF> setup-proxy?
2020-07-02 16:58:15 <inad922> EF, I don't understand I'm sorry
2020-07-02 16:59:44 <inad922> What do you mean by setup-proxy?
2020-07-02 17:00:49 <inad922> ah ok so that's a program
2020-07-02 17:02:28 <inad922> which is not available in docker images of alpine
2020-07-02 17:02:47 <inad922> or at least not in 3.12
2020-07-02 17:03:44 <ikke> it's part of alpine-conf, which is not installed by default on docker images indeed
2020-07-02 17:04:06 <ikke> https://docs.docker.com/network/proxy/
2020-07-02 17:12:32 <inad922> Not sure if that has anything to do with it
2020-07-02 17:12:40 <inad922> But I already set that one to the correct values
2020-07-02 17:13:19 <inad922> I just find it puzzling that no one really seem to know how to use apk from behind a proxy. I mean even in the github issues people are just guessing.
2020-07-02 17:13:26 <ikke> Yeah, it's more about the docker client to use a proxy
2020-07-02 17:14:12 <ikke> setting http_proxy and https_proxy in the environment should be enough
2020-07-02 17:14:26 <inad922> but apparently it isn't
2020-07-02 17:15:15 <Hello71> did you write export
2020-07-02 17:15:39 <Hello71> and did you do docker run -e
2020-07-02 17:16:10 <inad922> docker run -it alpine:3.12 sh
2020-07-02 17:16:20 <Hello71> godot: the raid chipsets tend to be terrible
2020-07-02 17:16:24 <inad922> and exported the http_proxy and https_proxy variables
2020-07-02 17:16:29 <inad922> and apk update after
2020-07-02 17:16:42 <Hello71> did you try wget
2020-07-02 17:16:45 <ikke> https://gitlab.alpinelinux.org/alpine/apk-tools/-/blob/master/libfetch/http.c#L801
2020-07-02 17:16:51 <Hello71> wget inside or curl/wget outside
2020-07-02 17:17:05 <inad922> I get some "ERROR: http://dl-cdn...: Permission denied" error
2020-07-02 17:17:13 <ikke> https://gitlab.alpinelinux.org/alpine/apk-tools/-/blob/master/libfetch/http.c#L764
2020-07-02 17:17:14 <inad922> wget works
2020-07-02 17:17:15 <Hello71> also weren't you here earlier asking about a weird password
2020-07-02 17:17:18 <inad922> curl I don't have
2020-07-02 17:18:28 <inad922> hmm
2020-07-02 17:19:07 <inad922> When I try to hit http://dl-cdn.alpinelinux.org/alpine/v3.12/main
2020-07-02 17:19:12 <inad922> with wget that works
2020-07-02 17:19:14 <godot> Hello71: I'm still trying to figure out what support this chipset has. The dmraid docs seem to indicate that mdadm should be used if the BIOS can't reconstruct the RAID without extra software which makes sense.
2020-07-02 17:19:49 <inad922> But http://dl-cdn.alpinelinux.org/alpine/v3.12/main/x86_64/APKINDEX.tar.gz gives a 403
2020-07-02 17:20:16 <KatolaZ> inad922: are you sure your proxy is not filtering binary stuff?
2020-07-02 17:20:19 <KatolaZ> like tar.gz?
2020-07-02 17:21:13 <Hello71> "virus checking"
2020-07-02 17:21:33 <inad922> I don't know but I'm on RHEL7 behind the proxy and I just installed some newer packages with nix and that seems to work and it works from the nix cache which are binaries
2020-07-02 17:22:33 <inad922> Hah
2020-07-02 17:22:37 <inad922> You might be right
2020-07-02 17:22:43 <inad922> I mean I drop out of docker
2020-07-02 17:22:52 <KatolaZ> ...and it works...
2020-07-02 17:22:57 <inad922> And from RHEL7 from behind the proxy I try to wget the same url
2020-07-02 17:23:02 <inad922> and it doesn't work there either
2020-07-02 17:23:06 <KatolaZ> yep
2020-07-02 17:23:13 <inad922> ...
2020-07-02 17:23:15 <KatolaZ> so the issue is in the proxy conf, most probably
2020-07-02 17:23:30 <inad922> I don't have control over that
2020-07-02 17:23:43 <KatolaZ> 
ACTION shrugs

2020-07-02 17:24:21 <KatolaZ> inad922: your sysadmin is "protecting" you from doing something dangerous :)
2020-07-02 17:25:03 <inad922> KatolaZ, Sure.
2020-07-02 17:25:27 <inad922> Also I picked a random file from a github project release
2020-07-02 17:25:32 <inad922> again .tar.gz
2020-07-02 17:25:39 <inad922> and I can wget it
2020-07-02 17:25:48 <inad922> So that might not be the case
2020-07-02 17:34:48 <inad922> Ok, so if I change the link above to https it works. How can I make apk to download files via https rather than http
2020-07-02 17:38:52 <Hello71> /etc/apk/repositories
2020-07-02 18:05:24 <inad922> Yeah well
2020-07-02 18:05:29 <inad922> I figured that in the meantime
2020-07-02 18:05:42 <inad922> Not sure who is making the alpine images on docker hub
2020-07-02 18:06:09 <inad922> But if it's you guys could you change the links in /etc/apk/repositories from http to https?
2020-07-02 18:06:52 <inad922> I mean my proxy drops any request for a binary if it goes through http rather than https.
2020-07-02 18:07:08 <Hello71> our dl-cdn only recently got valid https certificates
2020-07-02 18:07:12 <inad922> It's not that I played with this for 2 hours but I guess other people have similar issues
2020-07-02 18:07:20 <inad922> Hello71, Ah I see.
2020-07-02 18:07:25 <Hello71> your proxy is extremely broken then
2020-07-02 18:07:42 <inad922> Nah, it's working fine
2020-07-02 18:07:49 <inad922> I think it's doing the right thing
2020-07-02 18:08:15 <Hello71> no.
2020-07-02 18:08:32 <inad922> Why?
2020-07-02 18:11:36 <Hello71> first off, blanket file type blocks are dumb. maybe there is some argument for blocking mismatched mime/content, but blanket blocks are dumb
2020-07-02 18:11:55 <Hello71> second, blocking only http and not https is a huge waste of time and resouces
2020-07-02 18:11:59 <Hello71> r
2020-07-03 03:32:01 <abenz> finally found out the culprit !
2020-07-03 03:32:25 <abenz> it was the symlink (/etc/apk/cache)
2020-07-03 03:34:08 <abenz> see my earlier ramdisk setups was on usb that had an extra partition (sdc1 vfat for alpine, sdc2 ext4 for storage).. this caused setup-bootable to refer to my partition with sdc1 (instead of the usual "usb").
2020-07-03 03:35:02 <abenz> on the new installation, I only had a single partition on the usb, so when I restored the apkovl (which was setup for sdc1), "usb" was broken on it. I updated lbu.conf and apk.conf, just that symlink
2020-07-03 04:42:38 <ikke> abenz: ok, nice
2020-07-03 04:43:47 <abenz> ikke: :)
2020-07-03 04:54:39 <mps> ikke in Dutch means 'I' or 'uncle'? trans isn't clear
2020-07-03 04:55:23 <ikke> 'I', it doesn't mean uncle
2020-07-03 04:56:20 <mps> google translate can be sometimes really wrong :)
2020-07-03 04:58:11 <mps> interesting thing is that my nick on some other chat system is also 'I' but in my language and written in script (alfabet) of second one lang :)
2020-07-03 05:00:37 <ikke> heh
2020-07-03 05:00:46 <ikke> cyrillic?
2020-07-03 05:10:02 <mps> yes, Russian, single letter 'Я'
2020-07-03 05:10:49 <joe9> is this a good resource to install alpine desktop?
2020-07-03 05:10:54 <joe9> https://wiki.alpinelinux.org/wiki/Desktop-notes
2020-07-03 05:11:54 <mps> joe9: it depends what 'desktop' means for you
2020-07-03 05:13:02 <mps> but, yes, it is fine guide for setup of basics
2020-07-03 05:14:48 <mps> yunfan: good morning. As of last night my working environment is intel (i386, amd64) free. only arm boxes
2020-07-03 05:14:49 <joe9> I am trying to set up a linux system for my kids to watch netflix.
2020-07-03 05:15:16 <joe9> Other than firefox and a point-and-click desktop, nothing much is needed.
2020-07-03 05:15:22 <mps> joe9: afaik, alpine doesn't have all needed for netflix
2020-07-03 05:15:36 <joe9> oh, shit.
2020-07-03 05:15:40 <joe9> thanks.
2020-07-03 05:15:57 <mps> something like closed source plugin is needed for netflix
2020-07-03 05:16:08 <mps> widewine?
2020-07-03 05:16:21 <mps> and it doesn't run on musl libc
2020-07-03 05:16:32 <joe9> ok, thanks for pointing it out.
2020-07-03 05:17:24 <mps> hmm, that could be good excuse to disallow children from bad content :)
2020-07-03 05:18:48 <mps> give books to children, paper and pencils or let them to play outside
2020-07-03 05:19:51 <ikke> widevine indeed
2020-07-03 05:21:32 <mps> ah nice ikke, my memory works though I never tried this and morning coffee not yet finished :)
2020-07-03 06:13:15 <id4rk> morning, ive downloaded the aarch64 image and flashed it on a sd card, the jetson nano had a v8 arm cpu, does not seem to boot
2020-07-03 06:37:27 <Phlogi> I’m unable to delete an outdated/old package: linux-hardened:
2020-07-03 06:37:49 <Phlogi> alpine:~# apk info | grep hard
2020-07-03 06:37:49 <Phlogi> linux-hardened
2020-07-03 06:37:51 <Phlogi> alpine:~# apk del linux-hardened
2020-07-03 06:37:52 <Phlogi> OK: 961 MiB in 158 packages
2020-07-03 06:37:53 <Phlogi> alpine:~# apk info | grep hard
2020-07-03 06:37:54 <Phlogi> linux-hardened
2020-07-03 07:53:59 <EF> pheoxy: it may be a dependancy issue. Have a read on "apk info" https://wiki.alpinelinux.org/wiki/Alpine_Linux_package_management
2020-07-03 08:50:59 <midasi> Hello, we opened several issues on gitlab.a.o months ago. We recently also attached .patch files. What can we do to get them priotized and included into 3.12.1?
2020-07-03 08:53:04 <midasi> most of them are only minor changes
2020-07-03 08:56:46 <ikke> midasi: creating merge requests for them will certainly help
2020-07-03 09:01:42 <midasi> ikke: we fixed the issues in your enterprise build environment. from there we cannot merge it to gitlab. that's the reason we created patches
2020-07-03 09:02:31 <Cogitri> How is making a MR different from posting a patch for you?
2020-07-03 09:03:06 <Cogitri> Reviewing patches in issues on Gitlab isn't really fun compared to MR
2020-07-03 09:06:14 <midasi> cogitri: To create a merge request we need a public fork of gitlab afaik...
2020-07-03 09:07:29 <Cogitri> You need to fork the repository, apply the patch on some branch and push that
2020-07-03 09:07:51 <midasi> exactly, but cannot do that in our enterprise environment. it doesn't allow to push
2020-07-03 09:08:20 <ikke> midasi: not over http(s)?
2020-07-03 09:08:34 <midasi> I assume I can do it as well manually in the gitlab UI
2020-07-03 09:08:41 <midasi> let me check that
2020-07-03 09:08:49 <clandmeter> push over https should work
2020-07-03 09:10:49 <midasi> well, there are some security restrictions, but we can try again...
2020-07-03 12:34:48 <telmich> What kind of mixer do you use for changing the volume? It seems alsamixer does not render correctly for me on alpine
2020-07-03 12:35:33 <godot> Goooooooooood morning #alpine-linux. After a night a fitful wrestling with that demon spawn from hell referred to as fakeraid on Intel RSTe, I have the following issue whenever I enable a RAID1 set with ondisk metadata detected by dmraid:
2020-07-03 12:35:44 <telmich> (I don't need pulse or anything, just want to tune the microphone)
2020-07-03 12:36:03 <mps> telmich: never had any issue with alsamixer on alpine
2020-07-03 12:36:20 <godot> RAID set "isw_ecijjbfbcg_Volume0" already active
2020-07-03 12:36:20 <godot> ERROR:  Unable to register a device mapper event handler for device "isw_ecijjbfbcg_Volume0"
2020-07-03 12:36:44 <godot> The relevant pastebin: https://pastebin.com/fUDpVRR3
2020-07-03 12:36:52 <godot> Any ideas?
2020-07-03 12:37:32 <telmich> mps: mine looks like this: https://www.nico.schottelius.org/temp/2020-07-03-143626_849x706_scrot.png
2020-07-03 12:37:56 <telmich> very consistent in every Alpine installation I have seen so far
2020-07-03 12:38:26 <mps> huh, looks ugly
2020-07-03 12:38:55 <telmich> :-)
2020-07-03 12:39:03 <telmich> I always thought that was a locale issue on alpine and took it as the default, but hearing you it should work is interesting
2020-07-03 12:39:14 <ReedWade_> Hello guys ! I need help.. I'm trying to boot with a separate usr partition but I doest not work. I don't have binaries before userspace. I can't find any usr feature on mkinitfs. How should I do ?
2020-07-03 12:39:35 <mps> no, for me it works and looks nice even over networks on servers
2020-07-03 12:39:46 <afontain_> nico: works for me
2020-07-03 12:40:43 <afontain_> ReedWade_: do you have any clue on where the process fails?
2020-07-03 12:41:05 <afontain_> like, could you read stdout from /init.sh?
2020-07-03 12:41:10 <telmich> Correction: it works in the console, it does not work in urxvt
2020-07-03 12:41:25 <ReedWade_> afontain_: by example I got "md5sum is missing, which suggests /usr is not mounted"
2020-07-03 12:42:00 <ReedWade_> I think usr should be mounted before all this
2020-07-03 12:42:10 <mps> telmich: and xterm? or other terminals
2020-07-03 12:42:41 <telmich> What terminal are you using? I don't even have xterm installed atm
2020-07-03 12:42:44 <afontain_> nico: could be related to the fact that urxvt is the only terminal I know to store his terminfo in /usr/share/terminfo/ in Alpine
2020-07-03 12:43:02 <telmich> It works in xterm
2020-07-03 12:43:28 <afontain_> you could try to create a symlink for debug purposes
2020-07-03 12:43:36 <mps> yes, urxvt want to be special
2020-07-03 12:43:43 <ReedWade_> afontain_: I think I'll just forfeit this separate usr partition
2020-07-03 12:44:04 <afontain_> I've never had the need for a separate usr partition
2020-07-03 12:44:47 <ReedWade_> I've done this but this seems not commun and not supported by all
2020-07-03 12:44:59 <afontain_> nico: for f in /usr/share/terminfo/r/*; do ln -s "$f" /etc/terminfo/r/"$(basename "$f")"; done
2020-07-03 12:45:13 <afontain_> maybe add a -v too
2020-07-03 12:45:39 <telmich> /etc? that's interesting
2020-07-03 12:47:22 <telmich> ok, just symlinked, but it dose not change anything
2020-07-03 12:47:42 <afontain_> well, so that's one thing ruled out I guess
2020-07-03 12:47:43 <telmich> stracing it it actually opens the terminfo file
2020-07-03 12:47:47 <afontain_> echo $TERM?
2020-07-03 12:47:51 <telmich> So it might be the right pacth
2020-07-03 12:48:09 <telmich> rxvt-unicode-256color
2020-07-03 12:48:17 <telmich> oh, is this one of the length issue?
2020-07-03 12:48:35 <afontain_> looks fine
2020-07-03 12:48:35 <mps> we install some rxvt terminfos in ncurses-terminfo-base
2020-07-03 12:49:05 <telmich> hmm. trying export TERM=rxvt-unicode does not fix it
2020-07-03 12:49:11 <afontain_> nico: did it actually symlink something btw?
2020-07-03 12:49:19 <telmich> Yes, I symlinked them
2020-07-03 12:49:48 <telmich> And alsamixer opened it, I can see it in the strace output
2020-07-03 12:50:51 <afontain_> 
ACTION uploaded an image: fractal-pasted-image (5KB) < https://matrix.org/_matrix/media/r0/download/gnugen.ch/KsltxjoazjKiOvixwbRbTvrY >

2020-07-03 12:50:53 <afontain_> wfm
2020-07-03 12:50:55 <afontain_> (with urxvt)
2020-07-03 12:51:18 <telmich> what?
2020-07-03 12:51:23 <telmich> wow
2020-07-03 12:52:04 <telmich> I setup the colours and remove the scrollbar in my .Xresources of urvt and I also setup a font
2020-07-03 12:52:08 <telmich> I'll check if that is the issue
2020-07-03 12:52:41 <telmich> removing font setting does not fix it
2020-07-03 12:53:15 <telmich> removing the scrollbar and co also does not fix it
2020-07-03 12:53:34 <telmich> and finally resetting the colours also does not fix it
2020-07-03 12:53:43 <afontain_> I also have rxvt-unicode-256color as TERM
2020-07-03 12:53:50 <telmich> It works for you even without symlinking?
2020-07-03 12:53:56 <afontain_> yes
2020-07-03 12:54:14 <afontain_> does TERM=linux alsamixer work better?
2020-07-03 12:54:45 <telmich> Does not change it
2020-07-03 12:55:08 <afontain_> seems like it may not be terminfo then
2020-07-03 12:55:16 <telmich> Just checked the strace, it correctly accesses l/linux terminfo
2020-07-03 12:55:23 <afontain_> which version of alpine btw?
2020-07-03 12:55:40 <afontain_> I've got edge
2020-07-03 12:56:08 <telmich> I'm on edge
2020-07-03 12:56:15 <telmich> But I've had this problem since at least 3.11
2020-07-03 12:56:20 <telmich> Probably even 3.09
2020-07-03 12:56:28 <afontain_> `apk upgrade -a`?
2020-07-03 12:57:03 <telmich> did not fix it
2020-07-03 12:57:06 <afontain_> also, `find / -name *.apk-new`
2020-07-03 12:57:19 <afontain_> (have fun)
2020-07-03 12:58:03 <mps> or install st
2020-07-03 12:58:06 <mps> :)
2020-07-03 12:58:08 <telmich> apk-new: some grub/chrony/motd/group/passwd/hostname changes
2020-07-03 12:59:30 <telmich> it also works in st
2020-07-03 12:59:42 <telmich> I'm now diffing the strace of alsamixer in xterm vs rxvt-unicode
2020-07-03 13:02:05 <telmich> No differences until it comes to the escape sequences
2020-07-03 13:03:33 <telmich> mps: can st resize the font dynamically? That's the only feature I use in rxvt-unicode frequently
2020-07-03 13:04:44 <mps> telmich: yes, but my locally built st
2020-07-03 13:04:47 <mps> :)
2020-07-03 13:05:01 <telmich> it actually...wow
2020-07-03 13:05:49 <telmich> damnit
2020-07-03 13:05:50 <telmich> that one is nice
2020-07-03 13:06:22 <mps> if you are ready to build it yourself I can post aports patch
2020-07-03 13:06:23 <telmich> ok... just need to fix its colours and then years of urxvt can be left behind
2020-07-03 13:06:40 <telmich> font resize works in the default install
2020-07-03 13:06:54 <telmich> c-s-pgup/pgdown
2020-07-03 13:07:22 <mps> yes, they added that in last releases
2020-07-03 13:08:08 <telmich> hmm, it does not honour St.background
2020-07-03 13:08:50 <mps> xresource patch is not integrated
2020-07-03 13:09:37 <telmich> Ah, I see, it comes as a patchset
2020-07-03 13:12:21 <telmich> is it possible to change the colours w/o xresources?
2020-07-03 13:12:51 <afontain_> I assume that would involve recompiling
2020-07-03 13:13:38 <telmich> quite a nice journey... I actually wanted to find out why my mic in firefox does not work :-)
2020-07-03 13:14:07 <afontain_> lol
2020-07-03 13:14:29 <telmich> AHAAAA
2020-07-03 13:14:30 <telmich> it works in chromium
2020-07-03 13:14:54 <afontain_> btw while we are talking about terminals, do someone know a good wayland-native terminal?
2020-07-03 13:16:44 <afontain_> I know there is is gnome-terminal, alacritty and termite
2020-07-03 13:16:58 <maxice8> I used alacritty with sway
2020-07-03 13:17:26 <afontain_> alacritty is quite exigent on the openGL support, so it doesn't work on pinephone
2020-07-03 13:17:37 <mps> telmich: st is best used if you build it yourself with patches you want/need
2020-07-03 13:18:38 <telmich> I think I am hitting https://bbs.archlinux.org/viewtopic.php?id=252378
2020-07-03 13:20:25 <telmich> Why do I need a terminal anyway? I've emacs!
2020-07-03 13:21:03 <mps> telmich: ah, then why do you need OS :)
2020-07-03 13:21:37 <telmich> You know it is possible to use init=/usr/bin/emacs ... ;-)
2020-07-03 13:21:45 <telmich> It actually works
2020-07-03 13:22:30 <mps> yes, and vim ;P
2020-07-03 13:22:55 <Hello71> init=X
2020-07-03 13:23:22 <telmich> https://www.nico.schottelius.org/temp/2020-07-03-152251_2559x1420_scrot.png
2020-07-03 13:23:58 <telmich> You know it always comes down to https://xkcd.com/378/
2020-07-03 13:43:16 <meehow> Can someone tell me how to umount /media/usb without root or sudo? I added `user` option in fstab and mounting is working now, but I can't unmount.
2020-07-03 13:48:00 <godot> question: from the Alpine install image running from USB, is there any way to reset to factory defaults, unload all packages etc without having to reboot
2020-07-03 13:49:12 <ikke> echo alpine-base >/etc/apk/world; apk fix
2020-07-03 13:49:31 <ikke> It does not reset the entire tempfs, just the installed packages
2020-07-03 13:51:32 <Hello71> yes u | update-conf
2020-07-03 13:51:58 <godot> ikke: thanks. Does that unload all packages from running memory as well? I am trying to quickly iterate through a build script without having to reboot
2020-07-03 13:52:10 <ikke> godot: no, it does not
2020-07-03 13:52:19 <godot> so reboot is the only option I take it
2020-07-03 14:11:48 <meehow> I added busybox-suid, but umount from user account still doesn't work (mount works). Any hints?
2020-07-03 14:12:24 <meehow> umount: can't unmount /media/usb: Operation not permitted
2020-07-03 15:18:46 <moosl> is it possible to re-compile entire installation?
2020-07-03 15:19:11 <moosl> with -march and -mtune?
2020-07-03 15:19:25 <Hello71> sure
2020-07-03 15:19:38 <Hello71> although at that point you might as well install gentoo
2020-07-03 15:20:15 <moosl> Hello71: well, gentoo's is a bit loosely put together, alpine feels more tightly fit
2020-07-03 15:20:52 <Hello71> uh
2020-07-03 15:21:15 <moosl> is there a guide or something, or an easy way?
2020-07-03 15:22:10 <moosl> I have the suspicion that the generic download is making musl underperform
2020-07-03 15:23:46 <Cogitri> I doubt it'll make a massive difference in anything but really, really CPU intensive things
2020-07-03 15:24:49 <Cogitri> See e.g. https://openbenchmarking.org/embed.php?i=1901086-SP-GCC9COMPI66&sha=22b4335&p=2
2020-07-03 15:25:49 <Cogitri> So I'd say selectively just recompiling a few things which take up a lot of CPU time for you should suffice
2020-07-03 15:28:24 <moosl> mm.. negligible margins
2020-07-03 15:28:40 <moosl> but x264 is the worst example to compile.
2020-07-03 15:29:27 <moosl> it's already heavily optimized in upstream.
2020-07-03 15:30:21 <afontain_> I wonder if there are such benchmarks with -Os
2020-07-03 15:31:44 <moosl> akp -o "gcc options" re-compile
2020-07-03 15:31:48 <moosl> *apk
2020-07-03 15:32:15 <moosl> that way we would clear any doubts
2020-07-03 15:33:06 <moosl> most importalty, firefox LTO + musl would rock!!
2020-07-03 15:33:14 <afontain_> wait, that works?
2020-07-03 15:33:20 <afontain_> are we gentoo now?
2020-07-03 15:36:26 <ikke> no
2020-07-03 15:39:00 <Cogitri> Well, you can set your own flags in /etc/abuild.conf and then recompile as usual with abuild
2020-07-03 16:05:40 <mps> telmich: I forgot that we already have st with xrdb patch in testing, apk info st-xrdb
2020-07-03 16:06:18 <telmich> Uhh, nice!
2020-07-03 16:06:29 <mps> I will upgrade it later today to 0.8.4
2020-07-03 16:13:50 <telmich> testing
2020-07-03 16:14:10 <telmich> OMG
2020-07-03 16:14:17 <telmich> it works instantly
2020-07-03 16:14:18 <telmich> nice
2020-07-03 16:14:20 <telmich> fresh
2020-07-03 16:14:21 <telmich> love it
2020-07-03 16:15:38 <telmich> It feels so good - https://www.nico.schottelius.org/temp/2020-07-03-181431_956x1041_scrot.png
2020-07-03 16:16:33 <afontain_> a terminal a gray background :o
2020-07-03 16:17:04 <telmich> Maybe some background: when I started using Linux with fvwm2, I really liked the background colour. Many Linux distros and terminals later I was actually looking for the "original" configuration and years later discovered that is lightyellow2. Since then all my terminals and emacs are configured that way
2020-07-03 16:17:59 <ikke> does st have link detection?
2020-07-03 16:18:32 <mps> ikke: url detection?
2020-07-03 16:20:15 <mps> telmich: just tested upgrade with new xresources patch, build flawlessly
2020-07-03 16:21:42 <afontain_> I did not notice that your #eeeed1 was slightly yellow btw
2020-07-03 16:21:45 <telmich> ha..
2020-07-03 16:22:07 <telmich> patching my cdist configuration for my notebooks to include st settings in .Xresources
2020-07-03 16:26:49 <mps> telmich: in about 15-30 minutes upgraded 0.8.4 st-xrdb will on mirrors
2020-07-03 16:36:06 <ikke> mps: That you can select urls and copy them / open in browser
2020-07-03 16:36:41 <mps> aha, thought so. there is a patch for this on st site
2020-07-03 16:37:00 <ikke> So you always end up having to build / host your own package of st
2020-07-03 16:37:31 <mps> yes, I have local build with patches I need and want to use
2020-07-03 16:37:48 <mps> even some my private for tweaking some things
2020-07-03 16:38:37 <mps> most important for me is change I made to use it on HHK (Happy HAcker Keyboard)
2020-07-03 16:39:16 <mps> no other term fits this like st
2020-07-03 16:56:29 <telmich> Nice - I'll certainly not run apk upgrade in a while loop until then...
2020-07-03 17:25:10 <joe9> telmich, you might like the plan9 colours. very similar to that background.
2020-07-03 17:56:41 <godot> Ok I give up. I have been trying to nail down this issue for hours without any luck. When populating efi using setup-disk -m sys /mnt/, I am getting the following error although both udev and eudev packages are installed:
2020-07-03 17:57:14 <godot> Looks like it's some type of a path issue, but strace doesn't show an error stat'ing udevadm
2020-07-03 17:57:26 <godot>  /var/cache/misc/device-mapper-udev-2.02.186-r1.post-install: line 4: udevadm: not found
2020-07-03 17:57:58 <godot> And I can't seem to catch this file in /var/cache/misc during the install if I suspend it, it must be written quickly then deleted.
2020-07-03 17:58:00 <godot> Any ideas?
2020-07-03 18:13:07 <Hello71> what happens when you run udevadm
2020-07-03 18:13:53 <godot> Hello71: it runs fine. It's on the path
2020-07-03 18:14:16 <godot> what is the difference between udev and eudev packages?
2020-07-03 18:14:33 <Hello71> https://git.alpinelinux.org/aports/tree/main/lvm2/device-mapper-udev.post-install
2020-07-03 18:14:40 <Hello71> well you shouldn't install both
2020-07-03 18:14:44 <godot> Ah
2020-07-03 18:14:49 <Hello71> but it shouldn't explode
2020-07-03 18:14:51 <godot> That might be it
2020-07-03 18:15:19 <godot> so which one is the better package to install?
2020-07-03 18:15:47 <Hello71> http://google.com/search?q=udev+vs+eudev
2020-07-03 18:16:03 <godot> Thanks Hello71
2020-07-03 18:33:57 <godot> Ok, another issue. When building mkinitfs.conf with base feature, that's not being found in /usr/local/bin/base /sbin/base etc. So what is the base feature? Is there a resource that describes all mkinitfs features?
2020-07-03 18:36:52 <ikke> godot: do you mean /etc/mkinitfs/features.d/ ?
2020-07-03 18:37:32 <godot> ikke: echo features="ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid" > /mnt/etc/mkinitfs/mkinitfs.conf
2020-07-03 18:37:52 <godot> mkinitfs -c /mnt/etc/mkinitfs/mkinitfs.conf -b /mnt/ $(ls /mnt/lib/modules/)
2020-07-03 18:38:20 <ikke> /etc/mkinitfs/features.d/base.files
2020-07-03 18:38:21 <godot> ==> initramfs: creating /mnt/boot/initramfs-lts
2020-07-03 18:38:22 <godot> cp: can't create '/tmp/mkinitfs.dgmHGm/lib/modules/5.4.43-1-lts/': No such file or directory
2020-07-03 18:38:22 <godot> cpio:   --no-auto-root  Do not automatically prefix input ELFs with ROOT: No such file or directory
2020-07-03 18:38:22 <godot> cp: can't create '/tmp/mkinitfs.dgmHGm/lib/modules/5.4.43-1-lts/': No such file or directory
2020-07-03 18:39:50 <godot> weird. So why is mkinitfs stat'ing /sbin/base as if it's a binary?
2020-07-03 18:41:41 <ikke> mkinitfs is a shell script, so you can inspect it
2020-07-03 18:42:52 <godot> ah ok, thanks ikke
2020-07-03 18:46:09 <Hello71> shell quoting doesn't work that way
2020-07-03 18:46:33 <Hello71> echo 'features="ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid"' > /mnt/etc/mkinitfs/mkinitfs.conf
2020-07-03 18:47:02 <Hello71> the way you've written it puts "features=ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid" in /mnt/etc/mkinitfs/mkinitfs.conf
2020-07-03 18:47:20 <Hello71> which means to run the command "base" with the variable "features=ata" and arguments "ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid"
2020-07-03 18:47:30 <godot> gah, my bad. should have been echo'ing into the file
2020-07-03 18:47:36 <godot> thanks Hello71
2020-07-03 18:48:44 <godot> echo """features="ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid"""" > /mnt/etc/mkinitfs/mkinitfs.conf
2020-07-03 18:49:08 <ikke> seems one " too many on the right side
2020-07-03 18:49:12 <Hello71> that also won't work but for slightly different reasons
2020-07-03 18:49:39 <Hello71> this isn't python
2020-07-03 18:49:42 <godot> haha
2020-07-03 18:49:51 <godot> I'm rusty on shell scripting
2020-07-03 18:50:19 <joe9> godot, learn plan9port's rc. It is simple and easy to remember.
2020-07-03 18:50:45 <godot> What package is that?
2020-07-03 18:50:48 <Hello71> I think only python uses """
2020-07-03 18:51:14 <Hello71> joe9: at that rate you might as well use xonsh or something
2020-07-03 18:52:43 <godot> echo features="\""ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid"\"" > /mnt/etc/mkinitfs/mkinitfs.conf
2020-07-03 18:54:46 <mps> echo 'features="ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey raid"'
2020-07-03 18:55:05 <mps> I hope it works in asj
2020-07-03 18:55:14 <mps> s/asj/ash/
2020-07-03 18:56:09 <ikke> echo features='"a b c"'
2020-07-03 18:56:18 <ikke> features="a b c"
2020-07-03 18:56:25 <mps> ikke: ash?
2020-07-03 18:56:38 <Hello71> POSIX sh
2020-07-03 18:56:43 <Hello71> well actually I think since bourne shell
2020-07-03 18:56:52 <mps> well, close then :)
2020-07-03 18:56:59 <ikke> mps: yes
2020-07-03 18:57:26 <mps> single quote, literal string
2020-07-03 18:57:38 <Hello71> /msg shbot b# echo a='"abc def"'
2020-07-03 18:58:57 <ikke>  /msg shbot b# echo a='"abc def"'
2020-07-03 19:00:49 <Hello71> uh
2020-07-03 19:01:52 <ikke> I accidentally copied a space :P
2020-07-03 19:03:01 <Hello71> also try /msg shbot triggers. it has some shells that are not that easy to install, like old ksh and bash
2020-07-03 21:17:27 <grharry> Hello ... just installed alpine edge + jack + audacity + ardour from edge testing ... I am seeing problems with memlock in limits.conf
2020-07-03 21:18:06 <grharry> changed it to @audio      hard      	memlock      	unlimited
2020-07-03 21:18:21 <grharry> but still same problem
2020-07-03 21:19:44 <grharry> anybody ??
2020-07-03 21:23:46 <Hello71> how young are you that you can't wait more than 1.3 minutes?
2020-07-03 21:24:27 <grharry> I wish !!
2020-07-03 21:26:13 <grharry> I bet older than U btw !!!! ;)
2020-07-03 21:26:44 <albr> unowe[d]: after hearing that I have my doubts
2020-07-03 21:26:56 <albr> unowe[d]: > I am seeing problems with memlock in limits.conf
2020-07-03 21:26:57 <albr> unowe[d]: can you elaborate?
2020-07-03 21:27:19 <grharry> sure ...
2020-07-03 21:27:50 <grharry> xfce4 ... run qjackctl
2020-07-03 21:28:13 <grharry> as user foo in audio
2020-07-03 21:28:15 <grharry> group
2020-07-03 21:28:54 <grharry> jack complaints about not being able to lock 8xxxxxx amount of memory
2020-07-03 21:29:33 <grharry> same when running ardour!
2020-07-03 21:33:31 <grharry> "Cannot lock down 82280346 byte memory area (Out of memory)"
2020-07-03 21:33:42 <grharry> exact jack message
2020-07-03 21:34:48 <c705> Ok, but it tells you what the issue is
2020-07-03 21:34:52 <c705> Out of memory
2020-07-03 21:37:03 <Hello71> that's misleading, mlock returns ENOMEM when exceeding RLIMIT_MEMLOCK
2020-07-03 21:37:17 <Hello71> this is specified by POSIX
2020-07-03 21:37:43 <Hello71> grharry: did you log out and log in again after editing the file
2020-07-03 21:37:52 <grharry> 8G ram on board
2020-07-03 21:38:09 <grharry> yes did logout logged in ...
2020-07-03 21:38:11 <grharry> rebooted
2020-07-03 21:38:13 <grharry> and all
2020-07-03 21:38:14 <Hello71> also try changing soft limit as well as hard
2020-07-03 21:38:42 <grharry> did it for user foo
2020-07-03 21:38:51 <grharry> logged out
2020-07-03 21:38:54 <grharry> logged in
2020-07-03 21:39:01 <grharry> ulimit -l  = 64
2020-07-03 21:39:38 <grharry> Mem:              7           0           6           0           1           7
2020-07-03 21:39:42 <grharry> free -h
2020-07-03 21:39:46 <grharry> sorry ...
2020-07-03 21:39:50 <grharry> free -g
2020-07-03 21:40:05 <grharry> swap all free 16G
2020-07-03 21:44:20 <Hello71> I don't think alpine supports limits.conf unless you install linux-pam
2020-07-03 21:44:29 <Hello71> also please stop spamming the enter key
2020-07-03 21:45:26 <grharry> man ... give me a break !
2020-07-03 21:46:11 <grharry> linux-pam is installed
2020-07-03 21:47:40 <grharry> :-X
2020-07-03 21:48:07 <Hello71> s/install linux-pam/use pam/
2020-07-03 21:59:38 <grharry>  authpriv.err : PAM unable to dlopen(/lib/security/pam_elogind.so): Error loading shared library /lib/security/pam_elogind.so: No such file or directory
2020-07-03 22:05:39 <grharry> Nope ...
2020-07-03 22:07:02 <grharry> no luck ... I presume u have no idea ... so good night !!! and :-X
2020-07-03 22:46:37 <nsz> i did an apk upgrade and the xorg no longer works: graphics comes up but no input
2020-07-03 22:47:07 <nsz> i'm still investigating it seems no input device is detected, if somebody has an idea what it is let me know
2020-07-03 22:47:42 <Cogitri> Maybe you upgraded your kernel modules but not your vmlinuz (e.g. because you forgot to mount /boot)?
2020-07-03 22:47:50 <Cogitri> Then it can't load the modules for your input devices
2020-07-03 22:48:47 <Cogitri> That happened a few times for me already and played out exactly like what you described: graphics output works thanks to the vesa driver which is built into the kernel but input devices don't since they're built as modules
2020-07-03 22:49:25 <nsz> hm can be
2020-07-03 22:52:22 <nsz> thanks i think i need 'apk add linux-lts'
2020-07-03 22:56:11 <nsz> hm no that didnt fix it yet
2020-07-03 22:57:19 <mps> is you boot partition mounted under /boot
2020-07-03 22:57:47 <mps> and 'apk add -u linux-lts' to upgrade pkg
2020-07-03 22:58:54 <mps> and after kernel upgrade reboot is needed because old modules are removed
2020-07-03 23:00:23 <Hello71> well does the input work in tty
2020-07-03 23:00:27 <mps> also, udev and dbus should be runing
2020-07-03 23:00:37 <nsz> input works on tty
2020-07-03 23:00:46 <nsz> i'm typing on console now
2020-07-03 23:01:03 <Hello71> I think you need to install xf86-input-libinput now
2020-07-03 23:01:05 <nsz> i see now udev or evdev messages in the Xorg log and i'd expect some there
2020-07-03 23:01:15 <nsz> let me see
2020-07-03 23:01:18 <Hello71> I recall the keyboard driver was removed
2020-07-03 23:01:35 <Hello71> but not sure if that means base xorg doesn't support keyboard at all
2020-07-03 23:03:07 <mps> nowadays in most cases xf86-input-libinput is enough
2020-07-03 23:05:01 <nsz> it didnt fix it
2020-07-03 23:05:14 <Hello71> right but I don't think that comes with xorg-server
2020-07-03 23:05:28 <Hello71> oh, the replacement is evdev
2020-07-03 23:06:00 <nsz> i have xf86-input-evdev
2020-07-03 23:06:04 <Hello71> which is also packaged separately on alpine, but usually comes bundled with xorg-server
2020-07-03 23:06:10 <nsz> (and now *-libinput too)
2020-07-03 23:06:27 <Hello71> paste your xorg log
2020-07-03 23:07:18 <mps> xf86-input-evdev is not needed in most cases
2020-07-03 23:08:53 <mps> nsz: do you kernel modules loaded
2020-07-03 23:09:31 <Hello71> well I'm saying that evdev is the direct replacement of keyboard
2020-07-03 23:09:37 <Hello71> libinput is recommended but not built-in
2020-07-03 23:09:39 <nsz> e.g. i have evdev loaded
2020-07-03 23:09:49 <nsz> but i dont know what modules i have to look for
2020-07-03 23:09:59 <nsz> there are also device nodes in /dev/input
2020-07-03 23:10:30 <mps> Hello71: no, libinput is enough
2020-07-03 23:10:42 <Hello71> mps: seems like you are contradicting what I am not saying
2020-07-03 23:10:57 <mps> heh :)
2020-07-03 23:10:57 <Hello71> I am aware that both evdev and libinput can do all mouse/keyboard input
2020-07-03 23:11:06 <mps> ok
2020-07-03 23:11:56 <mps> nsz: if you don't have some special hardware it should work out-of-the-box
2020-07-03 23:12:02 <Hello71> but evdev is more "traditional"/"core", so it is more of a direct replacement than libinput, which is more "external" compared to fully built-in keyboard driver
2020-07-03 23:12:20 <mps> uhm
2020-07-03 23:13:11 <nsz> http://port70.net/~nsz/tmp/Xorg.0.log
2020-07-03 23:14:33 <Hello71> seems like evdev and libinput drivers are not being loaded?
2020-07-03 23:14:40 <Hello71> also fwiw modesetting usually works better than intel
2020-07-03 23:15:01 <Hello71> unless you really need fast xrender for some reason
2020-07-03 23:15:29 <mps> yes, X didn't tried input drivers
2020-07-03 23:17:02 <mps>    533.601] (II) The server relies on udev to provide the list of input devices.
2020-07-03 23:17:37 <mps> next msg I see first time ' If no devices become available, reconfigure udev or disable AutoAddDevices.'
2020-07-03 23:22:24 <mps> I mean, noticed it first time
2020-07-03 23:23:29 <mps> nsz: for some reason your X server didn't tried to load any input driver
2020-07-03 23:23:52 <nsz> yes that's how it seemed to me too
2020-07-03 23:24:27 <nsz> i straced it and it seems to stat all sorts of things in the /usr/lib/xorg/modules directories but dont load anything from under input/
2020-07-03 23:24:38 <Hello71> well it's based on udev reporting
2020-07-03 23:25:07 <nsz> ok so how do i debug udev?
2020-07-03 23:25:48 <Hello71> dunno
2020-07-03 23:26:32 <mps> I'm not sure is that could be issue but are in input group
2020-07-03 23:27:02 <mps> are you in input group*
2020-07-03 23:27:05 <nsz> no
2020-07-03 23:27:32 <nsz> how do i ask udev what devices it sees?
2020-07-03 23:28:18 <mps> udevadm monitor
2020-07-03 23:29:39 <nsz> what does it supposed to print?
2020-07-03 23:30:33 <mps> changes it detects, from kernel
2020-07-03 23:31:02 <nsz> but if nothing changes?
2020-07-03 23:31:16 <nsz> how does xorg uses it to load input drivers?
2020-07-03 23:31:42 <mps> example, KERNEL[151936.242161] change   /devices/platform/ff200000.spi/spi_master/spi2/spi2.0/cros-ec-dev.0.auto/cros-usbpd-charger.4.auto/power_supply/CROS_USBPD_CHARGER1 (power_supply0'
2020-07-03 23:32:14 <nsz> i suspect some udev related package may be missing or some related xorg config is missing
2020-07-03 23:32:15 <MY-R> mps asked good question, is your user in "input" group?
2020-07-03 23:32:22 <nsz> no
2020-07-03 23:32:26 <MY-R> so add it
2020-07-03 23:32:28 <mps> you can start 'udevadm monitor' and unplug/plug mouse
2020-07-03 23:32:36 <nsz> why?
2020-07-03 23:32:37 <MY-R> relogin and enjoy
2020-07-03 23:32:48 <nsz> and how do i debug this when it works?
2020-07-03 23:33:29 <nsz> i mean there is nothing on my system that runs as 'input'
2020-07-03 23:34:09 <MY-R> ls -l /dev/input/
2020-07-03 23:34:49 <mps> MY-R: you are faster, yes that
2020-07-03 23:35:09 <nsz> all root owned
2020-07-03 23:35:19 <nsz> and adding myself to input didnt fix it
2020-07-03 23:35:23 <mps> huh
2020-07-03 23:35:29 <nsz> (and i didnt see any eperm in strace)
2020-07-03 23:35:35 <mps> strange
2020-07-03 23:35:40 <MY-R> all inputs should be in "input" group by default
2020-07-03 23:35:49 <Hello71> try evtest
2020-07-03 23:36:17 <mps> here is my ls -l /dev/input/, https://tpaste.us/BjPb
2020-07-03 23:36:34 <Hello71> I think missing required input group would result in some xorg log message
2020-07-03 23:36:55 <Hello71> not skipping load entirely
2020-07-03 23:37:06 <MY-R> no
2020-07-03 23:37:40 <Hello71> once you find which device it is, paste udevadm info -n /dev/input/eventN
2020-07-03 23:38:05 <nsz> well /dev is devtmpfs so it's whatever the kernel puts there
2020-07-03 23:40:14 <mps> mdev.conf have input section where it changes groups
2020-07-03 23:40:38 <nsz> http://port70.net/~nsz/tmp/event1.info.txt
2020-07-03 23:40:59 <mps> for example, mouse[0-9]  root:input 0640 =input/
2020-07-03 23:41:53 <MY-R> thaaaat is why user have to be in groups like "input" and "video"
2020-07-03 23:41:57 <mps> sorry I can't help anymore now, I'm actually sleeping now, must go to bed. good night
2020-07-03 23:42:08 <nsz> ok thanks
2020-07-03 23:42:29 <nsz> i also have root:input lines there
2020-07-03 23:42:58 <nsz> maybe mdev.conf is not processed for me..
2020-07-03 23:44:53 <nsz> do i need mdevd?
2020-07-03 23:47:07 <mps> it is settled with /etc/init.d/mdev
2020-07-03 23:53:30 <Hello71> doesn't udev block mdev
2020-07-03 23:54:59 <Hello71> nsz: does evtest print the right keys if you select the keyboard?
2020-07-03 23:56:47 <nsz> yes
2020-07-03 23:57:01 <nsz> (when run as root)
2020-07-03 23:59:56 <nsz> should it work as normal user?
2020-07-04 00:00:38 <Hello71> well it should be able to read it if you put your user in input
2020-07-04 00:00:41 <Hello71> did you remember to log out
2020-07-04 00:01:34 <nsz> i dont have the input devices in 'input' group
2020-07-04 00:01:46 <nsz> so being in the input group does not help
2020-07-04 00:02:03 <nsz> i'm not sure what should change the group on the devices
2020-07-04 00:09:39 <Hello71> udev or mdev
2020-07-04 00:10:55 <Hello71> kinda strange, what interface is your keyboard on?
2020-07-04 00:10:57 <Hello71> laptop?
2020-07-04 00:13:14 <nsz> should 'libinput list-devices' command list something?
2020-07-04 00:14:12 <Hello71> probably
2020-07-04 00:15:48 <Hello71> isn't PNP0C0C the power button
2020-07-04 02:04:15 <nsz> i still dont know why this is happening but the fix was 'for i in /sys/dev/char/13:*;do udevadm test $i;done'
2020-07-04 02:05:07 <nsz> i.e. simply test the input char devices with udevadm and then udev does its magic and then everything works
2020-07-04 02:05:40 <nsz> but i still dont know why i have to trigger it manually or how to debug this at all
2020-07-04 02:39:18 <Hello71> I think you need to disable mdev
2020-07-04 02:40:01 <Hello71> enable one of mdev and udev and if you use mdev then probably uninstall udev
2020-07-04 07:23:09 <mps> Hello71: mdev is 'essential' for alpine, should be disabled or removed
2020-07-04 07:24:50 <mps> it have line 'provide dev' in init script, which says all
2020-07-04 07:25:28 <mps> and udev works fine with mdev, udev is started later
2020-07-04 07:34:10 <mps> mdev sets 'static' date in /dev on boot, and udev later set/change dynamically added/removed devices (mostly)
2020-07-04 07:34:38 <mps> so, they works fine in 'tandem'
2020-07-04 07:35:14 <mps> though some people run quite fine complete system without udev, even X
2020-07-04 08:20:03 <nacelle> X was around well before udev
2020-07-04 08:22:40 <mps> and well before most of current users :)
2020-07-04 08:23:13 <nacelle> hrmm, also true I guess, idk
2020-07-04 08:23:24 <nacelle> 
ACTION goes back under his smooth protective rock

2020-07-04 08:23:30 <KatolaZ> and we needed to edit XF86Config by hand...
2020-07-04 08:23:34 <KatolaZ> :P
2020-07-04 08:24:28 <mps> and that was not bad at all ;)
2020-07-04 08:24:38 <afontain_> soon there will be people to say they didn't even see X, short of XWayland
2020-07-04 08:25:00 <afontain_> I did edit the xorg.conf manually though
2020-07-04 08:25:40 <afontain_> s/even/ever, maybe?
2020-07-04 08:26:06 <mps> all my working life I read that C, Unix, Keyboards ... will soon be obsoleted :)
2020-07-04 08:26:44 <afontain_> meh, people are learning C
2020-07-04 08:28:40 <afontain_> you know, if it wasn't for Alpine, it wouldn't have seen some init that wasn't systemd
2020-07-04 08:28:45 <mps> there is saying, 'those who don't know/understand history are condemned to discover it, poorly'
2020-07-04 08:29:31 <afontain_> when I switched to Arch, it was already systemd and my phone was already systemd!
2020-07-04 08:29:37 <afontain_> (jolla phone)
2020-07-04 08:30:14 <KatolaZ> mps: https://katolaz.net/p/c-queen.jpg
2020-07-04 08:30:17 <KatolaZ> :P
2020-07-04 08:31:05 <mps> :)
2020-07-04 08:36:11 <TBB> shame Jolla's in the state it is
2020-07-04 08:36:51 <TBB> gotta appreciate still getting updates tho, no other phone (OS) maker does that
2020-07-04 10:03:21 <moosl> under what name is enlightenment available in apk?
2020-07-04 10:03:37 <moosl> apk search e17/enlight doesn't yield any results.
2020-07-04 10:08:50 <Cogitri> I don't think it's available in Alpine
2020-07-04 10:16:57 <moosl> also missing rox desktop manager
2020-07-04 10:17:18 <moosl> those are excellent candidates for musl
2020-07-04 10:17:20 <moosl> tbh
2020-07-04 10:20:15 <moosl> mind blowing!! --> https://www.youtube.com/watch?v=I1aoRQhqM1k
2020-07-04 11:12:03 <ThinkDao> ah this is the right channel - cool
2020-07-04 11:12:15 <ThinkDao> good morning all
2020-07-04 11:12:26 <ThinkDao> how to install / setup man pages in alpine-linux?
2020-07-04 11:13:18 <Cogitri> apk add mandoc and then apk add $package-doc
2020-07-04 11:14:09 <ThinkDao> oh... excellent ! thank You so much !
2020-07-04 11:16:12 <ikke> or apk add docs to automatically install them for all installed packages
2020-07-04 11:22:20 <ThinkDao> even better! thank you ikke
2020-07-04 11:23:29 <ThinkDao> oh.. even libraries.. yeah, not needed.
2020-07-04 11:44:34 <golb> What do I add to /etc/network/interfaces to get this command: "ip -6 route add <ipv6_gateway> dev br0"?
2020-07-04 11:45:13 <golb> Or do I have to use if-up to run iproute2 manually?
2020-07-04 11:45:33 <mps> 'up /sbin/ip address add dev ....'
2020-07-04 11:46:05 <mps> 'up /sbin/ip route add ....'
2020-07-04 11:46:15 <mps> or anything you need
2020-07-04 11:52:14 <golb> Ok thanks
2020-07-04 11:54:19 <golb> Are all the syntax from Debian's interfaces supported?
2020-07-04 11:56:56 <mps> golb: I think yes
2020-07-04 11:57:04 <mps> but not 100% sure
2020-07-04 12:05:58 <yunfan> mps: you dont use any x86 device?
2020-07-04 12:15:08 <mps> yunfan: well, I have one old asus tablet with phone (phonepad) which I use sometimes, and it is x86 with android
2020-07-04 12:15:49 <mps> and my customers/clients x86 servers
2020-07-04 12:16:41 <yunfan> mps: i got a atom based box recently
2020-07-04 12:16:57 <yunfan> and i had got a usb2serial adapter
2020-07-04 12:16:59 <mps> but I'm not owner of these servers
2020-07-04 12:19:08 <mps> I hope I will not have to by x86 anymore
2020-07-04 12:19:36 <mps> buy*
2020-07-04 12:32:08 <yunfan> i hope so, but the real life is not follow my wish :[
2020-07-04 12:46:47 <mps> real life is useless if people don't follow their wishes and dreams
2020-07-04 12:57:33 <earnestly> That's what buddhism tells us; the purpose of life is to satisify our desires
2020-07-04 12:59:15 <mps> hmm, my understanding of Buddhism is more aligned with Bodhisattvas moto, to save all conscious beings
2020-07-04 13:00:32 <mps> and to reach truth however it is hard to achieve that
2020-07-04 13:02:04 <mps> to satisfy our desires is westernization of Buddhism
2020-07-04 13:03:17 <mps> (some would tell 'plain lies')
2020-07-04 14:30:04 <afontain_> is there some way to allow rootless shutdown and reboot in Alpine? Short of adding /sbin/reboot and friends to sudoers?
2020-07-04 14:30:27 <ikke> elogind?
2020-07-04 14:30:44 <mps> acpid?
2020-07-04 14:31:33 <afontain_> 
ACTION sent a long message:  < https://matrix.org/_matrix/media/r0/download/matrix.org/eefvvtKfLOzKtDEeOEKuQFCK >

2020-07-04 14:31:45 <afontain_> ^ elogind seems installed
2020-07-04 14:32:46 <afontain_> sudo rc-service elogind start && reboot doesn't seem to work either
2020-07-04 14:33:04 <afontain_> acpid is started too
2020-07-04 14:33:16 <Cogitri> You shouldn't have to start elogind, it's started automatically during login
2020-07-04 14:33:18 <afontain_> I'll have a look to elogind config, then thanks for the pointer
2020-07-04 14:33:38 <Cogitri> That service file is just there in case you want to have elogind controlled by OpenRC
2020-07-04 14:33:49 <Cogitri> You probably have to install polkit-elogind
2020-07-04 14:34:04 <Cogitri> And the right polkit helper for your desktop (e.g. polkit-gnome), then it should just work
2020-07-04 14:34:47 <mps> huh, how much layers for simple tasks as shutdown/reboot :P
2020-07-04 14:37:36 <afontain_> well, I was really thinking of the /sbin/reboot command
2020-07-04 14:39:05 <Cogitri> mps: Well, it's two layers: elogind doing the shutdown and asking polkit whether that user really is permitted to do so
2020-07-04 14:39:37 <mps> 'echo poweroff >> /etc/acpi/PWRF/00000080'
2020-07-04 14:39:52 <afontain_> it's a bit weird to have gnome able to shutdown, the power button shutting down thanks to elogind, but /sbin/reboot not able to reboot
2020-07-04 14:40:45 <Cogitri> Ah well, GNOME uses elogind, whereas reboot doesn't then I guess
2020-07-04 14:41:04 <afontain_> worst case, `alias reboot="sudo /sbin/reboot"` is good enough, I guess
2020-07-04 14:41:05 <ikke> mps: doesn't that only handle the power button?
2020-07-04 14:41:26 <Cogitri> afontain_: try "loginctl reboot"
2020-07-04 14:41:58 <mps> ikke: yes, but there are other buttons which can be used, VOLUMEDOWN for example
2020-07-04 14:43:17 <mps> or, 'awful.key({                   }, "XF86PowerOff", function () os.execute('systemctl suspend -i') end),' in awesome wm.
2020-07-04 14:43:30 <mps> ikke: I think you understand this
2020-07-04 14:44:51 <afontain_> Cogitri: this is actually a good idea
2020-07-04 14:45:33 <ikke> mps: yes, I use it myself (on arch)
2020-07-04 14:45:49 <ikke> mps: but afontain_ was asking about the reboot command
2020-07-04 14:46:19 <afontain_> well, ideally, the reboot command would also work
2020-07-04 14:46:45 <afontain_> but that's already an improvement over raw polkit or sudoers
2020-07-04 14:46:52 <mps> I'm just giving some ideas, not final solution
2020-07-04 15:05:44 <moosl> ppsspp not in apk :-(
2020-07-04 15:07:21 <Cogitri> https://pkgs.alpinelinux.org/packages?name=ppsspp&branch=edge
2020-07-04 15:07:22 <Cogitri> ?
2020-07-04 15:07:54 <moosl> :)
2020-07-04 15:09:28 <moosl> is it possible to add ONE package of edge without having to convert the entire thing to edge?
2020-07-04 15:09:46 <moosl> mixed branches
2020-07-04 15:10:07 <afontain_> it is possible, and sane, if you build it locally
2020-07-04 15:10:24 <afontain_> otherwise, well, it will break someday
2020-07-04 15:13:20 <moosl> rofl! build logs on the site?
2020-07-04 15:13:23 <Cogitri> Yes, mixing branches is probably the one thing that causes the most breakage for users
2020-07-04 15:14:41 <afontain_> well, you don't really need to build it on the computer you'll use it on
2020-07-04 15:14:57 <afontain_> but you should build it yourself for this release
2020-07-04 17:29:22 <uplime>  /b 18
2020-07-04 17:29:26 <uplime> erm
2020-07-04 17:40:14 <mckaygerhard> GITEA wiki documented https://wiki.alpinelinux.org/wiki/Alpine_newbie_developer:_gitea
2020-07-04 20:36:03 <opal> ░░☰☰☰ did openssh-server recently get built without PAM support or is the 'UsePAM' option removed from openssh entirely now? ░░☰☰☰
2020-07-04 20:36:19 <opal> whoops
2020-07-04 20:36:36 <opal> but yeah
2020-07-04 20:37:38 <ikke> I don't see any changes in our package related to PAM
2020-07-04 20:37:54 <opal> odd
2020-07-04 20:38:10 <opal> i'll check openssh changelogs
2020-07-04 20:38:47 <opal> < afontain_> otherwise, well, it will break someday
2020-07-04 20:38:52 <opal> hasnt broken on me and i do that on a lot of systems
2020-07-04 20:39:00 <opal> and if something breaks its usually reversible
2020-07-04 20:39:02 <ikke> 6caa7e901fb0f63eefe2fde5be54687223709262
2020-07-04 20:39:08 <opal> ah ok thanks
2020-07-04 20:39:14 <ikke> that's 3 years ago though
2020-07-04 20:39:23 <opal> oh wild
2020-07-04 20:39:29 <opal> yeah i use pam for 2fa
2020-07-04 20:39:58 <opal> huh? i have that installed already
2020-07-04 20:43:21 <mps> openssh-server-pam pkg
2020-07-04 20:43:35 <opal> thats the one i have installed
2020-07-04 20:44:03 <opal> openssh-server-pam-8.1_p1-r0
2020-07-04 20:44:23 <mps> openssh-server-pam-8.3_p1-r0
2020-07-04 20:44:42 <mps> stable 3.12
2020-07-04 20:45:59 <opal> ah let me see
2020-07-04 20:49:03 <opal> mps yep now im on 8.3 and it works, thanks
2020-07-04 20:49:29 <mps> np :)
2020-07-04 20:53:58 <afontain_> opal : it mostly depends how much you rely on it, I guess
2020-07-04 20:54:17 <afontain_> I did pull python packages from edge in the past
2020-07-04 22:18:54 <amom> If I am claiming ownership of a package, do I need to bump the revision?
2020-07-04 22:20:03 <ikke> yes
2020-07-04 22:20:36 <ikke> that way, the package db gets updated
2020-07-04 22:20:43 <amom> ah, thank  you ikke
2020-07-04 22:50:39 <albr> c3r38r170[d]: Hi?
2020-07-04 22:59:00 <joe9> does sound over displayport work on alpine?
2020-07-04 22:59:22 <joe9> anyone with firefox on alpine? could you check if you can use netflix?
2020-07-04 22:59:33 <joe9> I asked this question yesterday and was told that it would not work.
2020-07-04 22:59:51 <joe9> I found that it needs ffmpeg. I am not sure if he had ffmpeg installed. Hence, want to check.
2020-07-04 23:00:09 <ikke> joe9: it requires widevine, which does not work with musl
2020-07-04 23:00:24 <joe9> that is what I thought.
2020-07-04 23:00:33 <joe9> but, it is being bundled with firefox and google-chrome now.
2020-07-04 23:00:43 <joe9> not separately, hence, wanted to check.
2020-07-04 23:02:00 <joe9> ikke, you know these things better. are you sure about that? there is no way to use a firefox binary executable from some other distro on alpine?
2020-07-04 23:02:12 <ikke> something like flatpack might work
2020-07-04 23:02:13 <joe9> or, google-chrome?
2020-07-04 23:02:22 <c3r38r170> Wait, how comes a website depends on a low level lib?
2020-07-04 23:02:41 <ikke> DRM
2020-07-04 23:04:46 <ikke> https://github.com/void-linux/void-packages/issues/6164#issuecomment-450681572 :)
2020-07-04 23:04:49 <joe9> ikke, have you tried firefox with ffmpeg installed?
2020-07-04 23:05:13 <ikke> no
2020-07-04 23:05:36 <joe9> is it possible to install ffmpeg on alpine?
2020-07-04 23:06:21 <c3r38r170> Oh wait, are you talking about the Netflix app, not the website?
2020-07-04 23:06:51 <ikke> https://pkgs.alpinelinux.org/packages?name=ffmpeg&branch=edge
2020-07-04 23:07:02 <ikke> c3r38r170: no, it's the website
2020-07-04 23:07:02 <joe9> no, I am talking about the website.
2020-07-04 23:07:08 <joe9> the app is obsolete.
2020-07-04 23:07:19 <c3r38r170> Then I can't believe it.
2020-07-04 23:07:25 <ikke> c3r38r170: widevine is google's DRM plugin
2020-07-04 23:07:40 <c3r38r170> I'll have to look up what DRM is, thanks.
2020-07-04 23:07:52 <joe9> ikke, is firefox built with musl libc on alpine?
2020-07-04 23:08:14 <ikke> joe9: the entire platform is musl
2020-07-04 23:08:39 <joe9> and, there is no google-chrome on alpine.
2020-07-04 23:08:56 <ikke> nope
2020-07-04 23:09:00 <joe9> so, that seals it. I guess. Thanks and Sorry for the bother.
2020-07-04 23:09:03 <ikke> because that's only built for glibc
2020-07-04 23:09:13 <c3r38r170> That's so weird.
2020-07-04 23:09:25 <c3r38r170> Wait, remember to try flatpak or something similar.
2020-07-04 23:09:35 <c3r38r170> If you want oc.
2020-07-04 23:10:24 <c3r38r170> Hey, I have a question. I am currently installing Alpine and I am at the NTP part, which one do you recommend me?
2020-07-04 23:10:48 <c3r38r170> busybox, openntp, chrony...? none...?
2020-07-04 23:11:02 <ikke> I tend to go with chrony
2020-07-04 23:11:14 <ikke> But whatever floats your boat :)
2020-07-04 23:11:18 <Hello71> if there was a recommended one, then it would be suggested as the default
2020-07-04 23:11:23 <Hello71> 
ACTION coughs

2020-07-04 23:11:28 <c3r38r170> lol
2020-07-04 23:11:36 <ikke> Hello71: I think chrony is the default suggested one :P
2020-07-04 23:11:37 <c3r38r170> I see, thanks!
2020-07-04 23:11:41 <c3r38r170> Yes hahhaa
2020-07-04 23:11:44 <Hello71> ikke: that's the joke
2020-07-04 23:11:52 <ikke> woosh
2020-07-04 23:11:58 <Hello71> interesting that ntpd isn't an option
2020-07-04 23:12:08 <c3r38r170> Isn't it busybox?
2020-07-04 23:12:12 <c3r38r170> "busybox"
2020-07-04 23:12:27 <Hello71> I mean real ntpd
2020-07-04 23:12:36 <c3r38r170> I see.
2020-07-04 23:12:45 <Hello71> I think busybox ntpd only does sntp? or is that systemd
2020-07-04 23:13:21 <c3r38r170> I guess I could choose none and get ntpd later, but... will it be a proble to not have any NTP for some minutes?
2020-07-04 23:13:45 <ikke> no
2020-07-04 23:13:45 <amom> anyone else configure logwatch on alpine?
2020-07-04 23:13:46 <Hello71> yes, your clock will start going backwards
2020-07-04 23:13:52 <ikke> lol
2020-07-04 23:14:04 <amom> I cannot seem to get sshd service to be monitored by logwatch
2020-07-04 23:14:05 <c3r38r170> lol
2020-07-04 23:14:19 <Hello71> amom: no, the package is just for fun, nobody uses it
2020-07-04 23:14:26 <Hello71> amom: well there's an actual question now
2020-07-04 23:14:40 <amom> Hello71: sorry. I type in multiple lines
2020-07-04 23:14:46 <amom> it's a really bad habit
2020-07-04 23:14:48 <Hello71> it would be more helpful if you provided some error message though
2020-07-04 23:14:49 <amom> shit, I did it again
2020-07-04 23:15:03 <amom> there's no error message
2020-07-04 23:15:16 <Hello71> did you start it and check /var/log/messages
2020-07-04 23:16:17 <amom> it just quietly ignores my sshd.conf in /etc/logwatch/conf/services
2020-07-04 23:17:08 <amom> I did test changing the parameters for zz-disk_space cause by default zz-disk_space uses df flags not in busybox
2020-07-04 23:17:12 <Hello71> c3r38r170: ntp would be pretty useless if it only worked while your computer was on and stopped working whenever it's off
2020-07-04 23:17:13 <amom> and I managed to fix that
2020-07-04 23:17:53 <amom> /var/log/messages has nothing.
2020-07-04 23:17:54 <Hello71> if you keep it off at least 16 hours a day then that's only 33% reduction in error, which is pretty useless
2020-07-04 23:18:38 <Hello71> dunno. I guess make sure you read the manual and if that still doesn't work try strace
2020-07-04 23:18:41 <Hello71> or debug options
2020-07-05 00:22:57 <legend> hey does anyone know how to limit the ram usage of domain 0 on the xen livecd?
2020-07-05 00:23:28 <legend> I'm trying to start a guest but it's complaining about being out of ram I think
2020-07-05 02:35:55 <golb> I tried to reinstall by using chroot. The issue is, rescue disk detected different device (starting with sdb), but when booting with Alpine before, it started with sda. This becomes a problem when installing boot loader. How can I change the device name so the bootloader thinks it is sda instead of sdb?
2020-07-05 02:47:28 <c705> golb: you can't rely on predictable disk names. you can use the uuids instead by using the symlinks in /dev/disk/ instead
2020-07-05 03:03:57 <golb> I use UUID for fstab. The issue is grub, for instance, detected it and decided to boot from hd1, not hd0.
2020-07-05 03:04:21 <golb> Grub uses UUID for detecting root partition though.
2020-07-05 03:07:31 <golb> Note that I have no access to console. So can't enter any option at boot prompt.
2020-07-05 04:43:36 <amom> Im trying to use rsnapshot to backup my alpine homeserver. Anytime I run it, I get: /usr/bin/rsync returned 255 while processing root@momi.ca:/etc/
2020-07-05 04:43:55 <amom> the computer running rsnapshot is running void.
2020-07-05 04:44:22 <amom> I can go into my root user and rsync /etc/nginx/nginx.conf  from my server but my rsnapshot cannot
2020-07-05 04:53:40 <amom> I think I fixed it. Idk how
2020-07-05 09:37:50 <cap3lla> hi all. I have a script that outputs to /var/log/foo.log and although I was in "rw" mode when I first touched this file and although executing "lbu commit -d" this file does not survive a reboot. After a reboot the file /var/log/foo.log is not there anymore. How can I solve this?any help appreciated
2020-07-05 11:03:03 <golb> Should /etc/fstab inside initramfs the same as /etc/fstab on the host?
2020-07-05 12:47:31 <Ashpool> What do I need to do in order to get the `w` command working on Alpine Linux (kernel linux-virt)? I installed it with the procps package, but it always shows an empty list, even though there're several users logged in
2020-07-05 12:48:52 <ikke> musl doesn't support it
2020-07-05 12:51:06 <Ashpool> ikke: ok, thanks
2020-07-05 12:52:26 <Hello71> you can install utempter and recompile all utmp/wtmp packages
2020-07-05 12:52:38 <Hello71> er, not utempter, the other one
2020-07-05 12:55:52 <mps> utmps, iirc
2020-07-05 12:58:02 <Ashpool> Right, I'll give it a try later, ty
2020-07-05 14:12:56 <golb_> Bad day today. Alpine didn't install on my bare metal server at OVH. Failed on VMWare Workstation either at initramfs.
2020-07-05 15:34:39 <afontain_> hey, if someone could have a look at https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/8689, it would be nice. I believe it is ready to be merged
2020-07-05 15:35:36 <afontain_> CI is failing, but I believe it is for unrelated questions
2020-07-05 15:50:04 <Cogitri> Seems like olm's test fail pretty hard, "afontain_" (https://matrix.to/#/@freenode_afontain_:matrix.exqa.de)
2020-07-05 15:50:07 <Cogitri> https://gitlab.alpinelinux.org/afontain/aports/-/jobs/159140#L522
2020-07-05 15:52:21 <afontain_> hmm, indeed
2020-07-05 17:20:38 <uwukorn> exit
2020-07-05 17:20:43 <uwukorn> oops new client i suck
2020-07-05 17:48:25 <kekmaster> has someone built cinnamon desktop on alpine?
2020-07-05 18:45:25 <dashavoo> Hi. Does alpinelinux have something like Debian's update-alternatives? I found an some old mailing list items discussing implementing it, but the trail ran dry.
2020-07-05 18:46:03 <ikke> dashavoo: no
2020-07-05 18:47:37 <dashavoo> OK thanks ikke. I was just checking whether there was a more "official" way before changing the iptables symlink to point to xtables-nft-multi instead of xtables-legacy-multi.
2020-07-05 18:55:19 <Hello71> better use /usr/local/sbin
2020-07-05 20:25:40 <adu> is there a apkbuild linter?
2020-07-05 20:25:50 <ikke> 0yes
2020-07-05 20:25:51 <maldridge> maxice8: 9767 has been rebased per your request.  Sorry it took so long, life's been a bit more exciting than I would have liked in my neck of the woods
2020-07-05 20:26:02 <ikke> adu: if you're on alpine, install atools
2020-07-05 20:26:07 <ikke> then there is apkbuild-lint
2020-07-05 20:58:22 <moosl> what's a good musl torrent client?
2020-07-05 21:00:09 <Cogitri> The same as on glibc ?
2020-07-05 21:00:41 <moosl> which one?
2020-07-05 21:01:22 <Cogitri> You can use Transmission, I guess
2020-07-05 21:04:06 <u0jQx9gPyrYg> i love aria2c
2020-07-05 21:09:26 <moosl> let's try
2020-07-05 21:09:47 <moosl> u0jQx9gPyrYg: jeez, what a weird nick, password generated?
2020-07-05 21:15:28 <u0jQx9gPyrYg> moosl: most of the time yes. sometimes i encode secret messages in my nick...
2020-07-05 21:16:45 <moosl> he he
2020-07-05 21:37:35 <mps> I thought it is pet name :)
2020-07-05 21:48:01 <chio> I am running into a problem with alpine upgrade. Current version is 3.8. While trying to upgrade anything I get ERROR: unsatisfiable constraints:
2020-07-05 21:48:02 <chio> required by: world[nextcloud-videoplayer]
2020-07-05 21:48:17 <chio> I could not find any documentation nor posts on how to resolve this issue.
2020-07-05 21:48:57 <chio> I should add, error occurs inside an LXC container.
2020-07-05 21:49:06 <chio> Host upgrade ran fine.
2020-07-05 21:56:34 <Cogitri> Well, that package has been removed since then, so you should go ahead and remove it from your world
2020-07-05 22:02:28 <chio> Cogitri remove how?
2020-07-05 22:03:29 <TBB> from /etc/apk/world, for example
2020-07-05 22:04:50 <chio> TBB, Cogitri, Thanks.
2020-07-05 22:57:03 <adu> What package has the headers for libcrypto1.1?
2020-07-06 00:58:59 <adu> What does "Has /home/... in rpath" mean?
2020-07-06 02:58:02 <adu> How do I add a description to a patch?
2020-07-06 02:58:05 <adu> "APKBUILD:Patch file ./xar.patch is missing a description"
2020-07-06 07:28:24 <yunfan> mps: the usb2serial works on x86 too :D
2020-07-06 07:31:59 <yunfan> the only problem is the postman must cause some of the components flexible, which will cause the kernel crash for ever, so i had bought another second hand board,hope it will well work
2020-07-06 07:38:26 <mps> yunfan: good, you are now hardware hacker
2020-07-06 07:38:29 <mps> :)
2020-07-06 07:39:23 <mps> bu careful with connecting proper pins and proper voltage levels
2020-07-06 07:39:35 <mps> s/bu/be/
2020-07-06 07:40:42 <yunfan> mps: nope, i am just a hardware monkey, i havnt even done a solder work
2020-07-06 07:41:01 <mps> these devices work on all architectures where drivers are compiled/included
2020-07-06 07:41:35 <mps> yunfan: everyone starts by making first steps
2020-07-06 07:42:30 <yunfan> but now i could make my own usb boot disk  :D , no need to put up with the iso image's default configuration
2020-07-06 07:43:29 <yunfan> the default config from grub dont provid console on serial port
2020-07-06 07:43:32 <mps> and you can hack your boards without attaching keyboards and monitors to them
2020-07-06 07:43:51 <yunfan> mps: yes, that;s the main purpose
2020-07-06 07:44:21 <yunfan> when my board arrived, i will try to build a 3 disk nas based on alpine
2020-07-06 07:44:34 <mps> yes, that's true, and I advocating long time to have serial console by default on all install media and tarballs
2020-07-06 07:44:39 <yunfan> its all garbage parts :]
2020-07-06 07:45:32 <yunfan> mps: a question, if the board dont have a com port, and i gave it a usb2ttl adapter, will the system use this as the console device?
2020-07-06 07:46:29 <mps> few days ago I took from my garage old linxsys WRT54G wifi router which is not used for about 5-6 years, attached serial console and tested it boots
2020-07-06 07:47:08 <mps> and I got old/new wifi AP :)
2020-07-06 07:47:49 <mps> yunfan: most (if not all) SBCs have serial console, but on some you have to solder pins on them
2020-07-06 07:48:25 <mps> you have to read documentation for any of them you work with
2020-07-06 07:49:34 <mps> where serial console is located, which pin is for what, what are voltage levels, speed (baud rate) of it etc
2020-07-06 07:50:58 <mps> openwrt.org have good docs for some of them, sunxi.org have for allwinners (sunxi) devices
2020-07-06 07:53:17 <mps> hmm, reading your question again, you connect usb part of usb2serial to your computer (workstation), not to SBC
2020-07-06 07:53:34 <yunfan> mps: well i mean x86 board
2020-07-06 07:53:50 <yunfan> one is e350 apu
2020-07-06 07:54:02 <mps> to SBC you connect these pins on adapter
2020-07-06 07:54:05 <yunfan> which had 4 sata 2 ports and 2 usb3 ports
2020-07-06 07:54:13 <yunfan> just enought for my disks
2020-07-06 07:54:26 <yunfan> SBC is easy on such works
2020-07-06 07:54:47 <yunfan> the only problem is the well tested cb2 boards is still too expensive
2020-07-06 07:55:54 <mps> I don't used x86 SBC boards for decades, so I don't know these new ones
2020-07-06 08:05:40 <mps> so, we know that alpine 3.12 works good on at least 10 arm devices and our infra team can add arm servers to this number
2020-07-06 08:23:26 <yunfan> mps: its just works, the setup scripts still need to be fixed
2020-07-06 08:25:14 <mps> nice to hear
2020-07-06 09:15:59 <moosl> can alpine/musl be localized?
2020-07-06 09:16:25 <moosl> if I install 'man pages' will it install many languages?
2020-07-06 09:16:34 <moosl> I am not sure how's localization sorted out in musl
2020-07-06 09:18:08 <moosl> also, is it to bloody to use edge?
2020-07-06 09:19:17 <Cogitri> I use edge on all of my systems, and it tends to work just fine
2020-07-06 09:19:52 <Cogitri> But be ready to deal with breakage in case it does happen
2020-07-06 09:20:57 <moosl> Cogitri: is it a hassle to convert from stable to edge? I prefer not re-installing
2020-07-06 09:21:45 <Cogitri> Just replace "v3.12" with "edge" and run "apk upgrade -a"
2020-07-06 09:23:59 <mps> before upgrade run 'apk update'
2020-07-06 09:24:05 <moosl> replace where?
2020-07-06 09:24:18 <moosl> in some file?
2020-07-06 09:24:29 <Cogitri> "/etc/apk/repositories"
2020-07-06 09:26:07 <mps> today 'good read' => wiki.alpinelinux.org ;)
2020-07-06 09:41:05 <moosl> mps: I am learning this new fangled apk/musl
2020-07-06 09:43:39 <moosl> I wonder how was it designed. The use is rather simple, no simple as under-feature, but simplicity in operation. Compaer to suse's yast, and apt-get which feel clunky as hell.
2020-07-06 09:44:01 <mps> moosl: it is not musl, by coincidence alpine is based on musl but it could be even on glibc
2020-07-06 09:44:43 <cgroot> Hi guys
2020-07-06 09:45:58 <Cogitri> 👋
2020-07-06 09:46:23 <cgroot> I've got a Talos 2 mainboard and want to use Alpine as root system but I have issues booting the Alpine image on it, can someone help me? I have a stack trace
2020-07-06 09:46:31 <cgroot> Seems like it crashes in some DMA operation
2020-07-06 09:46:50 <cgroot> https://0bin.net/paste/SjHSEyLvlwCGHxhc#1xYVY-iDlIf7g7YUkGAidkElNt5O0JaT/VkWdCEW9w1
2020-07-06 09:48:08 <mps> is that alpine kernel?
2020-07-06 09:49:32 <cgroot> Yes
2020-07-06 09:50:07 <cgroot> The current image from alpinelinux.org
2020-07-06 09:50:08 <mps> 'Linux version 5.5.0-openpower1 (root@raptor-build-public-staging-01) (gcc version 6.5.0 (Buildr' doesn't look like any alpine kernel
2020-07-06 09:51:01 <cgroot> Uhm... Ok, so the crash might happen before the Alpine kernel is loaded so the firmware kernel crashes?
2020-07-06 09:55:36 <moosl> Cogitri: firefox works on edge?
2020-07-06 09:55:50 <moosl> I am assuming edge uses musl 1.2
2020-07-06 09:55:57 <moosl> oh, and 32bit
2020-07-06 09:59:15 <cgroot> Ok, so I got it working. We have a SAS expander card in our setup and without the card it just boots. And now I can see that the crash happens before the kexec to Alpine... Thanks for your time!
2020-07-06 10:01:18 <Cogitri> moosl: Use pkgs.alpinelinux.og to see what version of musl edge uses (hint: not 1.2)
2020-07-06 10:01:25 <Cogitri> cgroot: Glad you got it solved :)
2020-07-06 10:01:44 <Cogitri> And yes, FF works, I use it as my main browser
2020-07-06 10:03:49 <mps> moosl: if you build musl 1.2 and install it should work. it works on my 4 machines, 2 armv7 and 2 aarch64
2020-07-06 10:04:07 <cgroot> Cogitri: thanks! Now I need to debug the PCI card...
2020-07-06 10:05:01 <moosl> zomg!! changing repo  is dead easy!! in apt-get and others you have to change files in a thousand places.
2020-07-06 10:06:08 <moosl> mps: does musl run better than bloatdroid on the arms?
2020-07-06 10:06:49 <mps> bloatdroid?
2020-07-06 10:07:58 <afontain_> android?
2020-07-06 10:08:04 <moosl> yeah
2020-07-06 10:08:28 <afontain_> well, there's a project that does just that and boots and thousands of arm devices
2020-07-06 10:08:36 <afontain_> (postmarketOS)
2020-07-06 10:08:45 <afontain_> so I'd say it works fine
2020-07-06 10:08:47 <mps> moosl: no, you can't install android apks but only alpine apks :)
2020-07-06 10:09:09 <mps> except when you install afontain_s anbox
2020-07-06 10:09:10 <moosl> I was trying to run youtube on firefox on android and the thing heated up some much black thick smoke  was comming out of the speakers.
2020-07-06 10:09:32 <mps> afontain_: Am I right?
2020-07-06 10:09:34 <amom> moosl: wow
2020-07-06 10:09:41 <amom> what phone is this?
2020-07-06 10:09:43 <afontain_> mps: It is possible
2020-07-06 10:09:54 <moosl> 4core 2GB
2020-07-06 10:09:58 <afontain_> mps: quite confusing by the way :P
2020-07-06 10:10:16 <moosl> the test was ONE app, firefox.
2020-07-06 10:10:19 <mps> afontain_: heh, why I expect this :)
2020-07-06 10:11:03 <moosl> afontain_: the smoke is a metaphor. It just heated up so much.
2020-07-06 10:11:32 <mps> moosl: yes, I use alpine on 32bit boxes as desktop for about two years
2020-07-06 10:11:32 <afontain_> "yeah, you install apk with `apk add ./something.apk`, but you should install this one with `adb install ./something.apk` after a `apk add cmd:adb`"
2020-07-06 10:12:07 <moosl> so no excuse, I am abusing alpine on an underpowered cpu and is blazing fast.
2020-07-06 10:12:49 <mps> actually most responsive GUI desktop is armv7 on old samsung peach-pi exynos 5800 board with emmc card
2020-07-06 10:13:10 <moosl> mps: which DE/WM?
2020-07-06 10:13:22 <mps> awesome wm
2020-07-06 10:13:33 <mps> but also tested with xfce
2020-07-06 10:13:49 <moosl> mps: how many cores? is it the big.little things?
2020-07-06 10:14:04 <mps> my son use xfce on armv7
2020-07-06 10:14:19 <mps> yes, it is big.little
2020-07-06 10:14:20 <moosl> mps: i'll be back, I have to reboot into edge,  later I'll ask you about gcc cflags for arm.
2020-07-06 10:14:39 <moosl> cflags for big.little are tricky as hell
2020-07-06 10:15:09 <moosl> I hate rebootings
2020-07-06 10:27:51 <nponaraHAa> I am back
2020-07-06 10:28:18 <nponaraHAa> mps: so about the arm big.little cflags? how do u use them?
2020-07-06 10:29:16 <mps> I use this only when build kernels for these machines
2020-07-06 10:29:44 <nponaraHAa> oh, no the apks?
2020-07-06 10:29:47 <mps> and don't activate it often
2020-07-06 10:30:33 <mps> are you sure it is needed for usespace
2020-07-06 10:30:39 <nponaraHAa> so edge booted just fine
2020-07-06 10:30:45 <nponaraHAa> I am impressed
2020-07-06 10:32:48 <nponaraHAa> rofl! free -m used: 375 buff/cache: 353
2020-07-06 11:50:58 <yann-kaelig> Hi
2020-07-06 11:51:59 <yann-kaelig> I have created a APKBUILD for libqb https://github.com/ClusterLabs/libqb, but failed for subpackage-dev with >>> ERROR: libqb-dev*: usr/lib/pkgconfig/libqb.pc: pkgconf version 2.0.0-yank is invalid
2020-07-06 11:52:04 <yann-kaelig> Any idea what the problem
2020-07-06 11:53:52 <yann-kaelig> here is the apkbuild http://dpaste.com/2QX4NHB
2020-07-06 11:55:42 <Cogitri> That it's 2.0.0-yank and not 2.0.0
2020-07-06 11:55:46 <mps> (hmm, could people use text/plain paste services)
2020-07-06 11:56:47 <Cogitri> It's not possible to tell whether 2.0.0-yank is a more/less recent version than 2.0.0-somethingElse, so the version is rightfully marked as invalid
2020-07-06 12:13:13 <yann-kaelig> got it, I took the wrong source archive
2020-07-06 13:32:41 <nponaraHAa> nice ranger available
2020-07-06 13:48:40 <gjabell> trying to run docker with overlay2 and it can't find the aufs kernel module
2020-07-06 13:48:48 <gjabell> does this need to be installed separately?
2020-07-06 13:49:22 <ikke> aufs is old and deprecated (never part of mainline linux)
2020-07-06 13:49:46 <gjabell> hmm, must be a red herring then
2020-07-06 13:56:11 <Hello71> "aufs is the preferred storage driver for Docker 18.06 and older, when running on Ubuntu 14.04 on kernel 3.13 which has no support for overlay2."
2020-07-06 13:57:41 <gjabell> yeah looks like the issue I'm running into is actually overlay2 not working on top of zfs
2020-07-06 13:58:02 <gjabell> which is unfortunate, because the software I need to run in docker doesn't seem to support the zfs backend :P
2020-07-06 14:01:11 <nponaraHAa> is there a reason why alpine doesn't have xfs option during install?
2020-07-06 14:01:28 <nponaraHAa> I need live resizing feature
2020-07-06 14:03:16 <Cogitri> Probably because no one implemented it
2020-07-06 14:03:28 <Cogitri> Also keep in mind that XFS can only grow
2020-07-06 14:03:55 <Cogitri> "gjabell" (https://matrix.to/#/@freenode_gjabell:matrix.exqa.de): huh, applications can opt to not support a storage backend?
2020-07-06 14:04:11 <nponaraHAa> I think you can also copy date only from xfs to xfs.
2020-07-06 14:04:47 <gjabell> well it's kubernetes in docker, think it's more an issue with kubernetes than them not supporting a backend
2020-07-06 14:05:30 <Cogitri> "nponaraHAa" (https://matrix.to/#/@freenode_nponaraHAa:matrix.exqa.de): That's only the case if you use xfs-send
2020-07-06 14:07:30 <mps> nponaraHAa: is your nick cyrillic try in latin
2020-07-06 14:09:12 <mps> пропагаНДа :)
2020-07-06 14:09:23 <ikke> mps: I had the exact same feeling
2020-07-06 14:09:38 <ikke> at least, that it was meant to resemble cyrilic
2020-07-06 14:09:47 <mps> propagaNDa
2020-07-06 14:10:22 <nponaraHAa> what makes you think so?
2020-07-06 14:10:27 <nponaraHAa> 
ACTION hides

2020-07-06 14:11:24 <mps> maybe we live in same or neighboring country
2020-07-06 14:12:00 <nponaraHAa> irc doesn't allow unicode in nicks
2020-07-06 14:12:11 <mps> I know
2020-07-06 14:12:18 <nponaraHAa> I wonder if other platforms such as discord allow full unicode use.
2020-07-06 14:12:50 <nponaraHAa> or matrix
2020-07-06 17:01:54 <lalo-salamanca> Guys, I want to install Alpine on a 16GB SD card
2020-07-06 17:02:15 <lalo-salamanca> should I make a 16GB bootable partition?
2020-07-06 17:02:49 <lalo-salamanca> I've heard bootable partitions are normally about 500mb - 1GB
2020-07-06 17:03:05 <c705> then why would you make a 16GB boot partition
2020-07-06 17:03:09 <c705> don;t be ridiculous
2020-07-06 17:03:11 <tehcloud> you can do whatever you please
2020-07-06 17:03:17 <tehcloud> as long as it works for you
2020-07-06 17:03:31 <Hello71> what is a "bootable partition"
2020-07-06 17:04:02 <tehcloud> I ran for years with a single data partition containing /boot and everything else
2020-07-06 17:04:05 <lalo-salamanca> So 500MB bootable FAT32 and rest ext4 would work right?
2020-07-06 17:04:14 <tehcloud> back in the days of MBR
2020-07-06 17:04:14 <lalo-salamanca> @tehcloud
2020-07-06 17:04:24 <tehcloud> that would work
2020-07-06 17:04:36 <tehcloud> there are lots of ways you can do this, try different methods out and see what's best for your use case
2020-07-06 17:05:00 <lalo-salamanca> Thanks. I want to set up a server
2020-07-06 17:05:25 <lalo-salamanca> Can alpine boot from usb on RPI3?
2020-07-06 17:06:12 <tehcloud> never tried it
2020-07-06 17:07:16 <lalo-salamanca> ok. thanks mate
2020-07-06 17:10:32 <nponaraHAa> w00t he left, it's just the kernel
2020-07-06 17:10:46 <nponaraHAa> uh?
2020-07-06 17:10:59 <nponaraHAa> congrats for what?
2020-07-06 17:11:32 <MY-R> w00t!?
2020-07-06 17:11:39 <ikke> it's an automatic response :P
2020-07-06 17:12:48 <nponaraHAa> ah yeah, I've see in other channels
2020-07-06 17:12:54 <nponaraHAa> it was a  fad a  few years ago
2020-07-06 17:13:49 <ikke> It's just a fun quirk, it normally does quite usefull things :)
2020-07-06 17:14:14 <ikke> eb7fa7a3a01c265db26584e6754695bd2cdb5302
2020-07-06 18:12:16 <lalo-salamanca> Guys, there is no py3-flask in apk search
2020-07-06 18:12:25 <lalo-salamanca> am I missing something?
2020-07-06 18:12:41 <lalo-salamanca> There's no pip
2020-07-06 18:13:10 <lalo-salamanca> There's no emacs !!!!!!!!!!!
2020-07-06 18:14:16 <nacelle> enable the community repo?
2020-07-06 18:14:16 <Hello71> yes there is !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!11
2020-07-06 18:14:59 <Hello71> (one exclamation mark at most is sufficient)
2020-07-06 18:18:23 <lalo-salamanca> Aah ok
2020-07-06 18:25:20 <mps> after so much answers 'enable community repo', maybe we should reconsider enabling it by default on installation
2020-07-06 18:31:44 <ikke> mps: next step: enable testing as well
2020-07-06 18:31:58 <mps> :)
2020-07-06 18:32:11 <ikke> And while we're at it, build testing for stable releases
2020-07-06 18:32:27 <mps> not much people ask for testing
2020-07-06 18:32:43 <ikke> they mix stable and edge to get testing packages
2020-07-06 18:33:17 <mps> I have split mind about testing
2020-07-06 18:33:20 <nacelle> maybe just make a suggestion of enabling the other repos when a package isnt found?
2020-07-06 18:34:04 <mps> testing could be enabled as pinned
2020-07-06 18:34:33 <mps> so, user must use 'apk add pkgname@testing'
2020-07-06 18:35:04 <mps> but that will fail on builders and CIs
2020-07-06 18:35:17 <ikke> Better something in the installer
2020-07-06 18:35:45 <mps> yes, better idea
2020-07-06 18:36:29 <Cogitri> Pinning is bound to break at some point too though
2020-07-06 18:36:45 <ikke> yeah, I don' think we should encourage testing for normal usage
2020-07-06 18:37:03 <Cogitri> I'd rather have apk print that mixing releases is unsupported and have a --allow-broken-mixed-releases flag for apk which you need to pass to mix releases or smth
2020-07-06 18:37:14 <mps> name doesn't sounds encouraging at the end
2020-07-06 18:37:42 <ikke> that's the goal
2020-07-06 18:37:51 <mps> testing name, I mean
2020-07-06 18:38:02 <ikke> The goal is to test the packages
2020-07-06 18:38:20 <mps> yes, but name says it is 'testing'
2020-07-06 21:53:41 <opal> anyone know off the top of their head the udhcpcd equivalent to dhcpcd.conf's 'nohook resolv.conf'?
2020-07-07 03:55:01 <opal> oh its just called udhcpd not udhcpcd
2020-07-07 03:57:53 <opal> guessing its handled with /usr/share/udhcpc/default.script ?
2020-07-07 03:59:14 <Hello71> dhcpcd is like udhcpc, udhcpd is like dhcpd
2020-07-07 03:59:18 <Hello71> yes it is confusing
2020-07-07 03:59:32 <Hello71> and yes
2020-07-07 03:59:51 <Hello71> you can also just install dhcpcd if you want
2020-07-07 04:00:27 <Hello71> or write your own script. not sure how to configure it though
2020-07-07 04:06:30 <maldridge> mps: a package like apt-suggestions would be nice
2020-07-07 04:06:52 <maldridge> 'foo' is not installed but can be installed with apk from the 'testing' repository
2020-07-07 04:15:47 <opal> Hello71 yeah if my change doesnt work i'll just switch to dhcpcd, but i think unsetting RESOLV_CONF should do the trick
2020-07-07 04:16:44 <opal> it's helpful to get comfortable with busybox's stuff when i find myself working on a router system or something where dhcpcd would be a stretch to install
2020-07-07 05:44:32 <Cogitri> "maldridge" (https://matrix.to/#/@freenode_maldridge:matrix.exqa.de):  I think the problem with having smth like apt-suggestions is that we'd need to fetch a list of packages and what repo they're in and that list would probably be about as big as the entire repo tarball is
2020-07-07 05:46:02 <maldridge> well yes, that is the problem with the original apt setup as well
2020-07-07 05:47:14 <maldridge> and why void provides this as a seperate package, but I think that apk could do it by doing the search in the cmd: namespace
2020-07-07 06:21:49 <mps> maldridge: Cogitri: as you know apk format don't have 'suggest/recomend' (or any tag) so such option/program would be hard to implement, maybe only with external DB
2020-07-07 06:23:21 <mps> but such external DB will always 'out-of-date', (who will maintain it)
2020-07-07 06:25:16 <mps> another option could be something which will read https://pkgs.alpinelinux.org/packages and filter according some search criteria
2020-07-07 06:25:31 <mps> like apk-file
2020-07-07 06:25:57 <Cogitri> pkgs.a.o doesn't offer an API though and it's not integrated into apk so next to no one will know about it and use it
2020-07-07 06:26:11 <maldridge> mps: pretty sure that all such systems like that are implemented with an external database
2020-07-07 06:26:19 <mps> parse html :)
2020-07-07 06:26:22 <maldridge> as there is a mapping of cmd to pkg
2020-07-07 06:26:46 <Cogitri> If we want people to not complain about missing packages which are in community we'll need apk to tell them what they need to do to get that pkg
2020-07-07 06:26:51 <maldridge> void's is regenerated on a crontab, I suspect something similar would be necessary to do for an apk pkg as well
2020-07-07 06:27:26 <Cogitri> Because most people just don't bother reading the docs anymore and just try things until it works
2020-07-07 06:27:55 <mps> actually, I'm for status quo
2020-07-07 06:28:41 <mps> (who cares for users who are lazy to read the docs)
2020-07-07 06:29:42 <mps> it is hard to have small and simple, and at the same big and easy to use
2020-07-07 06:29:52 <mps> same time*
2020-07-07 06:30:49 <EF> It would be lovely if pkgs.a.o accepted wildcards
2020-07-07 06:31:09 <mps> EF: it does
2020-07-07 06:31:25 <Cogitri> Even says so in the placeholder for the search :D
2020-07-07 06:32:40 <EF> It doesn't for me. Searching for "i3" in v3.12, everytihng else default, I get "No item found..."
2020-07-07 06:33:43 <Cogitri> Search for i3*
2020-07-07 06:33:47 <EF> Okay, ignore me! Mean partial searches. My error
2020-07-07 06:36:25 <Cogitri> Tbh that surprised me at first too
2020-07-07 06:46:51 <opal> < Cogitri> "maldridge" (https://matrix.to/#/@freenode_maldridge:matrix.exqa.de):
2020-07-07 06:46:55 <opal> did matrix-to-irc just get more busted
2020-07-07 06:50:51 <Cogitri> Huh, did it?
2020-07-07 06:51:06 <opal> seems like the nick mentions changed
2020-07-07 06:51:10 <Cogitri> Ohh, maybe that's RiotX being very mediocre at quoting
2020-07-07 06:51:14 <opal> oh ok
2020-07-07 06:51:23 <Cogitri> opal: Is it back to normal with this?
2020-07-07 06:51:27 <ikke> yes
2020-07-07 06:51:33 <Cogitri> Alrighty then
2020-07-07 06:51:35 <opal> yep
2020-07-07 06:51:43 <Cogitri> Thanks for the heads up :)
2020-07-07 06:51:57 <Cogitri> I better not use RiotX's built-in quoting then
2020-07-07 06:51:58 <opal> np
2020-07-07 06:52:01 <opal> heh
2020-07-07 06:52:10 <opal> seems like every client has a different way to quote
2020-07-07 06:52:18 <opal> or tab complete
2020-07-07 06:53:00 <opal> well, im getting alpine 3.12 armhf to try on my pi
2020-07-07 08:53:10 <opal> ay it works
2020-07-07 08:53:23 <ikke> nice
2020-07-07 09:03:28 <Piraty> <maldridge> and why void provides this as a seperate package, <...> --- which package?
2020-07-07 09:03:56 <maldridge> xlocate, which uses an external database for exactly the reasons that were described above
2020-07-07 09:04:12 <Piraty> ah yes
2020-07-07 09:04:20 <Piraty> daily indexed i think
2020-07-07 09:04:52 <Piraty> wasn't sure what you were talking about
2020-07-07 09:05:19 <ikke> I'm kind of opposed to automatically suggesting what package to install for a command unless it's opt-in. The delay it introduces (at least on ubuntu) is quite anoyg
2020-07-07 09:05:22 <ikke> anoying*
2020-07-07 09:05:47 <maldridge> hmm, it was always opt in on debian I thought
2020-07-07 09:06:00 <ikke> On Ubuntu it's enabled by default
2020-07-07 09:06:01 <maldridge> it printed some message about missing the suggestions package last time I tried
2020-07-07 09:06:49 <ikke> Well, that already means the mechanism is active
2020-07-07 09:07:24 <maldridge> or a default handler provided by a much slimmer package needs to be present
2020-07-07 09:12:00 <mps> yes, on debian and derivatives 'recomends' installed by default
2020-07-07 09:12:39 <mps> user have to disable this explicitly if don't want these pkgs installed
2020-07-07 09:15:02 <amom> I accidentally used make install for a package I installed via apk and had trouble with removing the mess make made. I ran apk audit --system to see packages that were broken. As long as there are no results when I run apk audit --system, i am ok right?
2020-07-07 09:15:21 <amom> specifically worried about stuff from the make command lingering on my system
2020-07-07 09:16:54 <mps> make uninstall?
2020-07-07 09:17:41 <amom> mps: there was part of the make file that had make uninstall unfortunately
2020-07-07 09:18:03 <mps> then 'by hand'
2020-07-07 09:19:56 <amom> lol ya, thats what I did
2020-07-07 09:26:07 <amom> oh well
2020-07-07 09:26:15 <amom> I am having issues with znc disconnecting.
2020-07-07 09:26:25 <amom> Does anyone else have that issue on musl?
2020-07-07 09:26:38 <amom> well, everyone is using musl here.... My bad
2020-07-07 09:37:46 <opal> lmao
2020-07-07 09:45:01 <nponaraHAa> is edge like a rolling release?
2020-07-07 09:45:11 <ikke> yes
2020-07-07 09:45:18 <nponaraHAa> excellent!
2020-07-07 09:45:42 <nponaraHAa> dumb mainstream distros, haven't discovered musl yet
2020-07-07 09:46:09 <maxice8> They have, they just prefer glibc
2020-07-07 09:52:26 <mps> debian have for more than 3 years musl
2020-07-07 10:13:32 <amom> It's not just musl. apk is a treat to use
2020-07-07 10:15:18 <mps> and !systemd
2020-07-07 10:15:45 <amom> ^ that too
2020-07-07 10:36:34 <nponaraHAa> mps: there's a debian musl iso?
2020-07-07 10:36:39 <nponaraHAa> I wish to try it.
2020-07-07 10:36:48 <nponaraHAa> I need like a stand alone apt-get
2020-07-07 10:45:05 <mps> no, musl lib and gcc deb pkg
2020-07-07 10:46:19 <mps> I used it to build go programs when I was on debian and worked with go
2020-07-07 14:02:12 <amazing1> i need bluetooth libbluetooth-dev, but apk add can't find those. how do i get them?
2020-07-07 14:02:56 <ikke> Is that part of bluez?
2020-07-07 14:03:30 <ikke> It seems so: https://pkgs.alpinelinux.org/contents?branch=edge&name=bluez-dev&arch=x86_64&repo=main
2020-07-07 14:03:33 <amazing1> i'm not sure. i need it for pip3 install pybluez to work
2020-07-07 14:03:34 <ikke> bluez-dev
2020-07-07 14:03:44 <ikke> that package contains libbluetooth.so
2020-07-07 14:03:53 <amazing1> i'll try.
2020-07-07 14:04:19 <amazing1> thanks!
2020-07-07 14:04:56 <ikke> yw
2020-07-07 16:13:16 <nponaraHAa> edge actually fixed some GL things. nice.
2020-07-07 16:14:09 <nponaraHAa> does alpine got xfs  support or I have to download something?
2020-07-07 16:19:04 <programmerq> my installed kernel package includes the xfs.ko module, but xfsprogs is a separate package
2020-07-07 16:21:03 <nponaraHAa> so mount should handle it without problems?
2020-07-07 16:31:01 <programmerq> is it giving you an error message?
2020-07-07 16:34:13 <nponaraHAa> never mind
2020-07-07 19:06:15 <nponaraHAa> 
ACTION resists the temptation to install texlive/latex

2020-07-07 19:07:15 <nponaraHAa> is there some metapackage?
2020-07-07 19:08:43 <nponaraHAa> I don't see the extra latex packages
2020-07-07 19:08:58 <nponaraHAa> what about math, science and the rest of stuff?
2020-07-07 19:16:50 <shiroyasha> nponaraHAa: What a coincidence, just two hours ago I installed latex.
2020-07-07 19:17:28 <shiroyasha> I think it was `texlive-full` the package name I used.
2020-07-07 19:17:44 <nponaraHAa> shiroyasha: is xetex reccomended even though it's not used? it complains if missing
2020-07-07 19:18:11 <nponaraHAa> shiroyasha: thanks, its there.
2020-07-07 19:22:31 <shiroyasha> Yeah I got some xetex/luatex errors with the `texlive` package, but the `texlive-full` fixed them. At least for my usecase.
2020-07-07 19:32:47 <nponaraHAa> :( mount needs explicit -t ??
2020-07-07 19:33:00 <mps> bb mount, yes
2020-07-07 19:33:33 <nponaraHAa> mps: wow, maddening. It should have some warning or something.
2020-07-07 19:34:09 <nponaraHAa> mps: a week of pulling my hair
2020-07-07 19:34:26 <mps> :)
2020-07-07 19:35:18 <nponaraHAa> I ended up typing random options
2020-07-07 19:35:45 <eydaimon> Cogitri:  I tried to clarify some points of confusion on your page on root ZFS with encryption but I don't have access :(
2020-07-07 19:58:22 <Cogitri> Huh, ikke ^ ?
2020-07-07 19:58:29 <jejej> where does one get iwlwifi firmware on alpine? i am trying to frantically find the package but i just cant find it
2020-07-07 19:58:50 <Cogitri> Not sure how permissions work in the wiki
2020-07-07 20:00:16 <ikke> Cogitri: What are you refring to?
2020-07-07 20:00:39 <ikke> wiki access?
2020-07-07 20:01:15 <ikke> registration is open, and you should be able to edit pages
2020-07-07 20:01:27 <ikke> but the first 5 hours you are not allowed to post links
2020-07-07 20:03:32 <Hello71> https://pkgs.alpinelinux.org/contents?file=iwlwifi*&path=&name=&branch=edge&repo=main&arch=x86_64
2020-07-07 20:04:25 <jejej> but i already have this package
2020-07-07 20:04:39 <jejej> yet it doesnt work
2020-07-07 20:07:41 <ikke> what network card do you have?
2020-07-07 20:08:09 <jejej> some intel dual band, but my networking menus are empty overall
2020-07-07 20:08:11 <jejej> not even wired shows up
2020-07-07 20:08:15 <jejej> not that i look at it
2020-07-07 20:08:48 <jejej> but i do have networkmanager setup with iwd
2020-07-07 20:41:47 <eydaimon> ikke: oh, I will wait a bit more and try then. I thought it said some smaller time.
2020-07-07 21:01:58 <eydaimon> Cogitri: trying to mount the rpool/ROOT/alpine after a reboot, but I'm getting permission denied as root. I did the zfs load-key -a step
2020-07-07 21:02:11 <eydaimon> Can you help ?
2020-07-07 21:06:17 <Cogitri> And it asked you to enter a password?
2020-07-07 21:11:05 <Cogitri> eydaimon: ^
2020-07-07 21:26:48 <opal> is diskless install default for raspberry pi builds just because of limited writes to sd card?
2020-07-07 21:27:50 <ikke> And because setup-disk doesn't know how to deal with these devices
2020-07-07 21:30:36 <opal> ah
2020-07-07 21:30:58 <opal> well it's a nice way to get me to learn the ins and outs of diskless installs
2020-07-07 21:31:03 <opal> think i'll stick with it
2020-07-07 21:32:50 <opal> hm i'm shopping for a router. openwrt immediately comes to mind when dealing with routers but i suppose alpine would work just as well or even better since i'm used to it?
2020-07-07 21:33:01 <opal> any known caveats compared to openwrt?
2020-07-07 21:33:13 <opal> other than i guess decreased arch support
2020-07-07 21:33:36 <nacelle> you should be just fine
2020-07-07 21:33:43 <ikke> no fancy web interface :)
2020-07-07 21:33:43 <nacelle> (with alpine)
2020-07-07 21:33:46 <opal> i'd probably be fine building against a different arch if the crossdev process is easy
2020-07-07 21:33:52 <opal> yeah i dont need a webui so thats fine
2020-07-07 21:38:00 <eydaimon> Cogitri: yes, it asked me and I got a few failures followed by a successful response
2020-07-07 21:38:11 <eydaimon> ( I had the password slightly wrong 😅 )
2020-07-07 21:42:41 <tsarna> I just replaced my old home router with a PC Engines apu3c4 running alpine
2020-07-07 21:42:42 <eydaimon-alpine> filesystem 'rpool/ROOT/alpine' can not be mounted: Permission denied
2020-07-07 21:43:27 <opal> oh yeah my friend said he heard good things about pc engines
2020-07-07 21:45:18 <nacelle> -sigh- I think they're a bit slow, but otherwise ok if you really need low power
2020-07-07 21:46:56 <tsarna> 4 cores but 1GHz... fast enough for me
2020-07-07 21:47:05 <bobby_> Hey there, I'm looking to rebuild the Alpine Xen ISO but with Xen 4.14-rc5.  I've found the Alpine image building script for VM targets, is there a similar script available for the Alpine Xen ISO?  Or an ISO building document?
2020-07-07 21:47:18 <tsarna> and way faster than the old Soekris
2020-07-07 21:56:15 <eydaimon> Cogitri: I just destroyed it and started over
2020-07-07 22:22:10 <eydaimon> I think there was an issue with the password. Does ZFS not like passwords ending with a space perhaps? 🤔
2020-07-07 22:41:01 <eydaimon> is using LUKS any easier ?
2020-07-07 22:41:26 <opal> i'd say luks is more apt for the purpose of encryption
2020-07-07 22:41:37 <opal> but thats just me and my stupid unix philosophy
2020-07-07 22:42:32 <eydaimon> but is it easier ?
2020-07-07 22:42:38 <opal> try it
2020-07-07 22:42:55 <opal> ive never used zfs
2020-07-07 22:57:36 <eydaimon> https://wiki.alpinelinux.org/wiki/LVM_on_LUKS ?
2020-07-07 23:05:41 <Hello71> opal: openwrt is designed as target-specific with buttload of patches and target-specific customizations (e.g. flash size)
2020-07-07 23:06:59 <Hello71> also for example vendor-specific firmware images
2020-07-07 23:07:47 <Hello71> hopefully we can get some new zen apus, but until then they're horribly slow
2020-07-07 23:10:34 <eydaimon> is fdisk scriptable ?
2020-07-07 23:12:15 <eydaimon> oh, looks like sfdisk is
2020-07-07 23:12:28 <MY-R> eydaimon, sure, and looks like this... echo -e "o\nn\np\n1\n\n\nt\n7\na\nw" | fdisk
2020-07-07 23:13:32 <MY-R> and at end device path | fdisk /dev/sdX
2020-07-07 23:15:36 <eydaimon> "sfdisk is a Scripted version of fdisk"
2020-07-07 23:15:48 <eydaimon> maybe easier :)
2020-07-07 23:16:30 <eydaimon> want to see if I can script Cogitri's ZFS encryption doc... doing it by hand is time consuming
2020-07-07 23:23:36 <MY-R> :)
2020-07-07 23:29:24 <Hello71> MY-R: printf '%s\n' o n p 1 '' t 7 a w
2020-07-07 23:29:35 <Hello71> maybe more ''
2020-07-07 23:30:56 <MY-R> :P
2020-07-08 01:54:55 <eydaimon> Following the instructions in https://wiki.alpinelinux.org/wiki/Alpine_Linux_in_a_chroot  but there is no /usr/share/syslinux/mbr.bin
2020-07-08 01:55:01 <eydaimon> so how do i make it bootable ?
2020-07-08 01:55:26 <eydaimon> also, could someone share the contents of their two first lines of /etc/apk/repositories ?
2020-07-08 01:55:53 <Hello71> this doesn't even work
2020-07-08 01:56:42 <Hello71> pretty sure you need to run extlinux
2020-07-08 02:03:02 <eydaimon> what ?
2020-07-08 02:03:16 <eydaimon> Hello71: can you still share your two first lines of the file ?
2020-07-08 02:03:20 <MY-R> eydaimon, be sure you have installed "syslinux" package if using it
2020-07-08 02:03:30 <eydaimon> MY-R: thank you
2020-07-08 02:04:07 <eydaimon> nm, I got the line I needed
2020-07-08 02:05:14 <eydaimon> or maybe not, is the line not http://dl-cdn.alpinelinux.org/alpine/v3.3/community/ ?
2020-07-08 02:05:40 <MY-R> 3.3?
2020-07-08 02:07:14 <eydaimon> oops
2020-07-08 02:07:16 <MY-R> http://dl-cdn.alpinelinux.org/alpine/v3.12/community/
2020-07-08 02:07:32 <eydaimon> thank you
2020-07-08 03:04:25 <eydaimon> Cogitri:  apk add linux-vanilla fails because there's no linux-vanilla package. Is there something else I should do instead ?
2020-07-08 03:10:06 <eydaimon> I'm following Cogitri's instructions and the /boot dir is empty. Would linux-vanilla install there ?
2020-07-08 03:10:53 <eydaimon> oh there are a few files there now. but I don't see any vmlinuz
2020-07-08 03:13:06 <MY-R> eydaimon, use "linux-lts" package
2020-07-08 03:15:21 <eydaimon> MY-R: cheers mate :)
2020-07-08 03:15:40 <MY-R> "linux-vanilla has been removed. Install linux-lts when upgrading." that is from changelog of v3.11
2020-07-08 03:16:37 <MY-R> you are welcome :)
2020-07-08 03:17:36 <eydaimon> going to reboot and cross my fingers :P If it works, I'll improve the docs :)
2020-07-08 03:47:18 <amom> Is there a mailing list for alpine linux security advisories?
2020-07-08 03:48:19 <opal> theres one for security advisories in general, dont think alpine operates anything specific for its packages though
2020-07-08 03:50:00 <opal> Hello71 yeah openwrt is maybe what i'll go with, kinda need a working router. i can play with alpine on a router im not actually using for my internet
2020-07-08 03:50:10 <opal> when i get one for that purpose which may be a while
2020-07-08 03:52:46 <amom> Hello71: I recommend openwrt too. I love luci is the best webgui for a router I have used. installing and using openwrt was plug and play all the way through
2020-07-08 03:54:12 <amom> opal: I cannot seem to find that mailing list. Do you have a link?
2020-07-08 03:55:28 <opal> nvd runs rss feeds, it isnt actually a mailing list
2020-07-08 03:55:54 <amom> oh rss is good too
2020-07-08 03:56:19 <opal> https://nvd.nist.gov/vuln/data-feeds
2020-07-08 04:23:58 <eydaimon> says missing operating system. I guess I will try more tomorrow :/
2020-07-08 04:33:22 <eydaimon> dd bs=440 count=1 conv=notrunc if=/mnt/usr/share/syslinux/mbr.bin of=/dev/sda
2020-07-08 04:33:26 <eydaimon> is this still valid ?
2020-07-08 04:34:34 <opal> sounds right
2020-07-08 04:35:56 <eydaimon> even for old bios ?
2020-07-08 04:36:31 <opal> "old bios"?
2020-07-08 04:36:58 <eydaimon> nevermind
2020-07-08 06:21:00 <mps> opal: depends how many ethernet ports you need for router an/or do you need wifi, but I'm using alpine on arm32 SBC about two years now
2020-07-08 06:21:23 <opal> well, friend said she would give me a standalone wifi AP
2020-07-08 06:21:31 <opal> so the router itself wont need that capability
2020-07-08 06:21:45 <opal> so far im using two eth ports but i'd like to have growing room
2020-07-08 06:22:03 <opal> its personal use so nothing crazy
2020-07-08 06:22:06 <opal> not like im rack mounting this
2020-07-08 06:22:29 <opal> also my internet plan is horse shit, nowhere near gbit
2020-07-08 06:22:44 <opal> (but if i move somewhere with gbit i'd like to be able to use that easily)
2020-07-08 06:23:03 <opal> (also LAN is a consideration)
2020-07-08 06:23:44 <mps> I'm switched from armbian to alpine 2 years ago on my old lamobo R1 (arm 32bit, 5 ports, allwinner SOC)
2020-07-08 06:24:11 <opal> mm
2020-07-08 06:24:29 <opal> yeah the tradeoff i'm most considering is "geared toward routers" vs "something i've used on desktop and server before"
2020-07-08 06:24:43 <opal> i feel i can learn openwrt fairly easily but do i benefit that much from doing so
2020-07-08 06:24:47 <mps> also, my internet is ADSL, and local net 100Mbit
2020-07-08 06:25:03 <opal> i love the alpine community and devel cycle, i can feel free to push to aports or spin my own repo
2020-07-08 06:25:07 <opal> like, i know it already
2020-07-08 06:25:18 <mps> actually I tried openwrt on this board but was not satisfied
2020-07-08 06:25:22 <opal> mm
2020-07-08 06:25:28 <opal> what missed the mark for you
2020-07-08 06:26:05 <mps> openwrt is to much trimmed down, I mean most software is bare minimum
2020-07-08 06:26:19 <mps> too much*
2020-07-08 06:26:28 <opal> ah
2020-07-08 06:27:17 <mps> for example, I need full postfix and dovecot functionalities, not bare minimum
2020-07-08 06:34:05 <opal> oh youre using software such as that
2020-07-08 06:34:19 <opal> i plan to have stuff like tor, i2pd, yggdrasil, wireguard
2020-07-08 06:34:52 <opal> stuff i can easily share through my lan and route transparently
2020-07-08 06:34:55 <opal> privoxy
2020-07-08 06:35:00 <opal> privoxy's light as hell so that isnt an issue
2020-07-08 06:35:09 <opal> dns
2020-07-08 09:14:52 <id4rk> alpine linux supports displaylink video docking stations?
2020-07-08 09:15:04 <id4rk> they are usb
2020-07-08 09:32:58 <albr> Puku puku[d]: hey folks, wondering if someone might be able to answer some of my questions regarding alpine linux.
2020-07-08 09:32:58 <albr> Puku puku[d]: 
2020-07-08 09:33:00 <albr> Puku puku[d]: I have a jailbroken kindle paperwhite 3 and am able to run alpine linux with DE on it. would it be possible to use run Anbox on this and then in turn be able to use 3rd party android ebook readers?
2020-07-08 09:38:49 <markand> someone has an idea for my wpa_supplicant problem described here? https://lists.alpinelinux.org/~alpine/users/%3C20200707131302.GA27249%40kiwi%3E
2020-07-08 10:12:15 <AnOddOne> hi
2020-07-08 10:14:52 <AnOddOne> so.. i booted custom alpine and i see that modloop hasn't been mounted, append line seems ok, cat proc/cmdline confirms, squashfs loaded, any ideas?
2020-07-08 10:38:48 <AnOddOne> anyone active here?
2020-07-08 10:39:35 <ikke> AnOddOne: there are people active, but at least I don't know what the issue is
2020-07-08 10:40:37 <AnOddOne> k, ty
2020-07-08 12:33:51 <kytart> hey, does anybody know if there's any info about how the rpi image is built? I'm trying to make my own version with newer kernel for rpi4 but I don't know how initramfs is built exactly in this case and also I don't know where to get modloop
2020-07-08 12:36:14 <Hello71> amom: openwrt does some weird and sketchy things, and their packaging system is meh
2020-07-08 12:36:32 <ikke> kytart: https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/scripts
2020-07-08 12:37:55 <kytart> @ikke: lol totally missed that, looks promising. Thanks, I'll take a look!
2020-07-08 12:49:40 <bovis> I'm trying to put the following lines in an APKBUILD, but the # is interpreted as a comment, which I don't want. How can I fix it?
2020-07-08 12:49:45 <bovis> https://bpa.st/QM3Q
2020-07-08 12:50:10 <ikke> quote it
2020-07-08 12:50:12 <ikke> "^#"
2020-07-08 12:59:41 <bovis> ikke: After that change, when running checksum, the script still fails. I can do "^# <<'EOF'" and it passes, but I don't know if that runs the same command.
2020-07-08 13:00:24 <ikke> no
2020-07-08 13:00:27 <ikke> it would not
2020-07-08 13:00:40 <afontain_> you should put the second quote after the #
2020-07-08 13:01:11 <afontain_> alternatively, I think you can write it « ^\# »
2020-07-08 13:01:38 <ikke> French quoting :)
2020-07-08 13:01:56 <bovis> afontain_: grep -v "^#" <<'EOF' does not work. Neither does escaping the pound
2020-07-08 13:02:00 <afontain_> indeed
2020-07-08 13:02:35 <afontain_> but both "^\#" and `^\#` could have been ambiguous
2020-07-08 13:02:56 <ikke> :)
2020-07-08 13:02:56 <afontain_> bovis: are you sure there isn't a problem elsewhere?
2020-07-08 13:03:17 <ikke> What error do you get?
2020-07-08 13:05:53 <bovis> afontain_: After removing that code from the script, it runs without an error.
2020-07-08 13:06:07 <bovis> ikke: line 125: syntax error: unexpected end of file (expecting ")")
2020-07-08 13:06:36 <afontain_> how about removing the quotes around 'EOF'?
2020-07-08 13:06:42 <bovis> That's the error without making any changes to what I presented originally. 125 is the final line of the file
2020-07-08 13:07:08 <afontain_> how did you end your heredoc?
2020-07-08 13:07:20 <afontain_> easiest is probably to link the source
2020-07-08 13:07:57 <bovis> afontain_: Removing those quotes doesn't help. Same error I just showed.
2020-07-08 13:08:16 <afontain_> well, add them back I guess
2020-07-08 13:08:20 <afontain_> could have been that
2020-07-08 13:08:33 <afontain_> I'm sure your error is several lines down
2020-07-08 13:09:01 <bovis> Here's the entire script: https://bpa.st/AP7Q
2020-07-08 13:09:37 <bovis> It's a direct copy of this, https://git.alpinelinux.org/aports/tree/community/wine/APKBUILD , but with the one change I've made we're discussing now
2020-07-08 13:09:57 <afontain_> `    EOF`
2020-07-08 13:10:02 <afontain_> ^ that's your error
2020-07-08 13:10:18 <afontain_> EOF must be the only thing on the line ever
2020-07-08 13:10:20 <afontain_> no spaces
2020-07-08 13:11:26 <bovis> afontain_: Yes the checksum runs without error now
2020-07-08 13:11:29 <bovis> Thank you
2020-07-08 13:20:34 <albr> Puku puku[d]: The `alpine-desktop` package doesn't exist anymore?
2020-07-08 13:21:48 <ikke> 59ed4dcbaef9dec859eb014a003cb7a4c89780b0
2020-07-08 13:21:53 <ikke> Indeed
2020-07-08 13:25:31 <albr> Puku puku[d]: what's the alternative? I don't see a setup script
2020-07-08 13:25:38 <ikke> setup-xorg
2020-07-08 13:25:49 <ikke> you need alpine-conf installed
2020-07-08 13:26:12 <albr> Puku puku[d]: install alpine-conf, then run setup-xorg-base again? (ive ran it once already)
2020-07-08 13:26:16 <ikke> ah ok
2020-07-08 13:26:18 <ikke> https://wiki.alpinelinux.org/wiki/Desktop-notes
2020-07-08 13:26:22 <ikke> then no, no need to run it again
2020-07-08 13:26:41 <albr> Puku puku[d]: Cheers, will follow along those notes
2020-07-08 13:26:41 <ikke> But you have to choose what desktop setup you want
2020-07-08 14:36:32 <wsinatra> Anyone here have any success using the apkovl=http://server/hostname.apkovl.tar.gz initramfs options?
2020-07-08 14:48:20 <albr> Puku puku[d]: afontain_: <basically install the package, install the image and start the service (anbox-container-manager) then start with anbox-launch> Would you able to give a little more info on this please? I'm up to the point where I've place the image in the correct location. What commands to run from that point?
2020-07-08 14:50:20 <afontain_> albr: arch?
2020-07-08 14:50:53 <afontain_> or rather puku puku, but I doubt pings will get through
2020-07-08 14:53:11 <albr> Puku puku[d]: arch? I'm trying to get it working on Alpine Linux
2020-07-08 14:53:24 <wsinatra> Puku puku: what isn't it doing precisely? also afontain means CPU type, arm, x86?
2020-07-08 14:54:15 <albr> Puku puku[d]: wsinatra: x86, I am testing on vmware player
2020-07-08 14:54:16 <afontain_> yeah. It's also the one that's given by the `arch` command
2020-07-08 14:54:29 <afontain_> 32bits?
2020-07-08 14:54:59 <albr> Puku puku[d]: afontain_: Oh right, I understand what you mean now. It's x86_64
2020-07-08 14:55:08 <afontain_> ah
2020-07-08 14:55:12 <afontain_> good
2020-07-08 14:55:23 <albr> Puku puku[d]: I've done apk add anbox. Then copied the android.img to the correct location.
2020-07-08 14:55:27 <albr> Puku puku[d]: Not sure what to do from that point.
2020-07-08 14:55:34 <wsinatra> did you start the service?
2020-07-08 14:55:39 <afontain_> do you know how openRC works basically?
2020-07-08 14:55:43 <wsinatra> rc-service start anbox-container-manager
2020-07-08 14:56:07 <afontain_> it's what "start the service (anbox-container-manager)" is about
2020-07-08 14:56:11 <wsinatra> if the manager isn't running you can't run the anbox container, if you want that to be persistent you need to also do rc-update add anbox-container-manager
2020-07-08 14:57:37 <albr> Puku puku[d]: Okay, I managed to start the service with `rc-service anbox-container-manager start` and have added it to be persistent.
2020-07-08 14:57:55 <wsinatra> oh yeah I always get that backwards, my bad
2020-07-08 14:58:26 <albr> Puku puku[d]: I now run anbox-launch?
2020-07-08 14:58:29 <wsinatra> run anbox-launch and it should work, assuming you have an android.img in /usr/share/anbox with the right perms
2020-07-08 14:59:00 <albr> Puku puku[d]: I did not set any perms, what perms do I set on the android.img?
2020-07-08 14:59:59 <wsinatra> I've got mine as 644 but also owned by my user
2020-07-08 15:00:12 <wsinatra> -rw-r--r--   1 wsinatra wsinatra 325902336 Jul  7 08:43 android.img
2020-07-08 15:01:11 <wsinatra> you'll probably also want this https://www.f-droid.org/ since the base container doesn't come with a whole lot installed in it, f-droid at least gives you a FOSS app store to work with
2020-07-08 15:01:20 <wsinatra> of course, you can just adb install whatever into the container once it's done
2020-07-08 15:02:12 <albr> Puku puku[d]: I think I've not set it up right, when I launch 'anbox-launch' it just hangs on the 'Starting...' screen.
2020-07-08 15:02:35 <wsinatra> It takes a bit to actually launch the container
2020-07-08 15:03:12 <wsinatra> that said, I'm not sure how well it'll play inside of a VM, I've only tested it on hardware up to this point
2020-07-08 15:03:46 <eydaimon> Cogitri: any idea why it would fail to  boot after following your instructions ?
2020-07-08 15:05:28 <Cogitri> Define "fail to boot" please :)
2020-07-08 15:06:03 <albr> Puku puku[d]: wsinatra: I'll let it sit for a while, but it looks pretty dead hehe
2020-07-08 15:07:12 <wsinatra> Puku puku: I actually pulled up an Alpine VM in qemu to test it there, and it went through in about a minute
2020-07-08 15:07:50 <wsinatra> that's on 2 vCUPs & 4G ram, but it's only using 1G for the LXC container + an XFCE4 desktop env
2020-07-08 15:08:24 <albr> Puku puku[d]: wsinatra: I see, I think I may have some other underlying issues, probably due to not having installed Alpine properly. The anbox window has now self closed too.
2020-07-08 15:08:26 <wsinatra> Not sure what kind of hardware you're trying to test it on though, which might have something to do with it.
2020-07-08 15:08:48 <eydaimon> Cogitri: one sec, I'll get you the exact wording
2020-07-08 15:08:53 <wsinatra> If you pull up a terminal and issue anbox-launch through there you'll get some level of error output at very least
2020-07-08 15:09:34 <albr> Puku puku[d]: Oh you are right: "Application manager service is not running yet"
2020-07-08 15:09:49 <eydaimon> Cogitri: Sorry I realize it was vague. It says: Missing Operating System, reboot and select proper boot device
2020-07-08 15:10:06 <wsinatra> rc-status will show you your running/enabled services
2020-07-08 15:10:36 <wsinatra> You might just need to restart the lxc manager, if you had your initramfs rebuilt recently (like after a large upgrade) you might also need to reboot
2020-07-08 15:10:51 <wsinatra> that happens pretty regularly with container/vm systems
2020-07-08 15:12:00 <albr> Puku puku[d]: wsinatra: "rc-service lxc-manager start" ?
2020-07-08 15:12:17 <wsinatra> should just be rc-service anbox-container-manager start
2020-07-08 15:12:26 <wsinatra> that needs to be done to let you launch the anbox container
2020-07-08 15:12:33 <wsinatra> since it's all kind of self contained
2020-07-08 15:12:37 <albr> Puku puku[d]: Yep, that one is running
2020-07-08 15:13:27 <wsinatra> If you run ip addr, do you have an anbox0 interface?
2020-07-08 15:13:56 <albr> Puku puku[d]: Yes, I do
2020-07-08 15:14:28 <wsinatra> Well at least part of it is definitely working then. restart the service and try to run anbox-launch from the terminal again
2020-07-08 15:14:41 <wsinatra> if it can bring up that interface it should in theory work just fine
2020-07-08 15:15:21 <albr> Puku puku[d]: "Application manager service is not running yet" I may not have set up my Alpine install correctly? Is this application manager referring to another component of the OS?
2020-07-08 15:17:19 <wsinatra> No I don't think that's the case, that error is specific to anbox
2020-07-08 15:17:24 <wsinatra> https://github.com/anbox/anbox/issues/1149
2020-07-08 15:17:44 <wsinatra> that said a lot of people see it, some suggest installing linux-headers, some suggest change cgroup settings
2020-07-08 15:19:54 <Cogitri> eydaimon: sounds like your EFI can't find something to boot from?
2020-07-08 15:20:12 <eydaimon> Cogitri: it's an old machine which uses BIOS
2020-07-08 15:29:02 <Cogitri> Ah, no clue then, haven't used BIOS for years now
2020-07-08 15:29:17 <Cogitri> But I guess it needs something to find your bootloader or smth
2020-07-08 15:32:15 <eydaimon> Cogitri: I was under the impression that BIOS can read MBR just fine. Also, the regular alpine install boots just fine
2020-07-08 15:32:29 <eydaimon> well, I know BIOS can read MBR just fine :)
2020-07-08 15:34:21 <Cogitri> Well, that msg means it doesn't even get into the bootloader, I guess
2020-07-08 15:40:00 <eydaimon> Cogitri: that's what it would seem. SO in addition to writing the MBR diwh  bs=440 count=1 conv=notrunc if=/mnt/usr/share/syslinux/mbr.bin of=/dev/sda is there something else? vmlinuz is installed on /boot at /dev/sda1. I'm not sure how to tell if the MBR is found or something else is missing
2020-07-08 15:43:52 <Cogitri> Sorry, I really have no clue about MBR or anything BIOS, but I don't think it should differ in the ZFS setup compared to normal Alpine
2020-07-08 15:44:07 <Cogitri> Maybe the /boot partition overwrote it or smth?
2020-07-08 15:45:45 <eydaimon> Not sure. it's the last step I performed :/
2020-07-08 15:46:54 <eydaimon> I'll try grub
2020-07-08 15:48:49 <Hello71_> eydaimon: for syslinux as I said you also need to run extlinux
2020-07-08 15:48:58 <Hello71_> also I think you need to set the boot flag on the partition
2020-07-08 15:49:29 <Hello71_> also technically BIOS doesn't read MBR at all. it just reads the first sector into memory and jumps to it
2020-07-08 15:49:45 <Hello71_> which I guess is the original sense of MBR, but not the modern sense
2020-07-08 15:55:35 <eydaimon> Hello71_:I missed that, sorry. I did run extlinux. and bootflag is set
2020-07-08 15:57:24 <PureTryOut[m]> Guess we'll need to support Flutter applications and libraries at some point in the future https://ubuntu.com/blog/canonical-enables-linux-desktop-app-support-with-flutter
2020-07-08 16:03:18 <maxice8> Oh dear
2020-07-08 16:05:47 <maxice8> Well I can postpone that until I personally need it
2020-07-08 16:07:04 <PureTryOut[m]> Haha exactly
2020-07-08 16:08:01 <Hello71_> as long as it's better than electron
2020-07-08 16:11:22 <afontain_> 
ACTION uploaded an image: fractal-pasted-image (18KB) < https://matrix.org/_matrix/media/r0/download/gnugen.ch/TVlpqFVDogQoxrGSORKOkAvY >

2020-07-08 16:11:43 <afontain_> the day you don't need anbox anymore
2020-07-08 16:11:44 <maxice8> Not very ambitious
2020-07-08 16:12:18 <afontain_> ^ wtf this is running on GTK3
2020-07-08 16:13:25 <afontain_> well, if the ecosystem is relying on snap…
2020-07-08 16:13:30 <eydaimon> hmm. If I run modprobe zfs in chroot I am met with this error: modprobe: FATAL: Module zfs not found in directory /lib/modules/5.4.0-26-generic  but the path is: /lib/modules/5.4.43-1-lts
2020-07-08 16:13:42 <eydaimon> is that to be expected ?
2020-07-08 16:14:00 <eydaimon> or am I crazy to load zfs in chroot?
2020-07-08 16:14:13 <ikke> yes, it doesn't make sense
2020-07-08 16:14:29 <ikke> the module version needs to match the kernel that is running
2020-07-08 16:14:45 <ikke> ie, uname -r
2020-07-08 16:14:59 <eydaimon> ok. I was trying to get grub going since the other way is failing, but no luck I guess
2020-07-08 16:20:33 <eydaimon> maybe this explains why I have issues:
2020-07-08 16:20:35 <eydaimon> Device     Boot  Start        End    Sectors   Size Id Type
2020-07-08 16:20:37 <eydaimon> /dev/sda1  *      2048     206847     204800   100M 83 Linux
2020-07-08 16:20:39 <eydaimon> /dev/sda2       206848 1250263039 1250056192 596.1G 83 Linux
2020-07-08 16:20:41 <eydaimon> sorry for spam
2020-07-08 16:20:51 <eydaimon> it wouldn't let me start on sector 1
2020-07-08 16:21:02 <eydaimon> could that be a problem ?
2020-07-08 16:21:16 <afontain_> isn't that expected?
2020-07-08 16:21:29 <afontain_> you have to write the partition table somewhere
2020-07-08 16:32:16 <eydaimon> oh nevermind I guess it is. the instructions even shows. it just wasn't letting me choose sector one with fdisk but the results shows 2048
2020-07-08 16:39:14 <Hello71_> what are you trying to do anyways
2020-07-08 16:42:46 <eydaimon> get alpine to boot ( https://wiki.alpinelinux.org/wiki/Alpine_Linux_with_root_on_ZFS_with_native_encryption )
2020-07-08 16:44:35 <eydaimon> since the steps there failed, I wanted to try grub. but grub complained about various things that I tried to address (and failed)
2020-07-08 16:44:42 <Hello71_> you said you were doing https://wiki.alpinelinux.org/wiki/Alpine_Linux_in_a_chroot
2020-07-08 16:45:19 <Hello71_> oh, this is a bad guide
2020-07-08 16:45:23 <Hello71_> well, both of these
2020-07-08 16:50:08 <Hello71_> oh, I see. June 1 was two weeks before 3.10 release
2020-07-08 16:51:24 <Hello71_> all you need to do now is use alpine 3.10+ (3.12 recommended obviously) iso and then skip setup-disk partitioning
2020-07-08 16:53:26 <eydaimon> Hello71: omg ❤️
2020-07-08 16:53:34 <eydaimon> I will try that. thank you :)
2020-07-08 16:53:46 <Hello71> maybe you can fix what you have now but it might be easier to restart
2020-07-08 16:54:14 <nponaraHAa> it got gimp, inkscape, krita
2020-07-08 16:54:26 <nponaraHAa> there's everything in here
2020-07-08 16:54:38 <eydaimon> Hello71: I'll give it a shot :)
2020-07-08 16:55:23 <Hello71> for example DISKOPTS='-m sys /mnt'
2020-07-08 16:55:27 <nponaraHAa> my barebones alpine is gaining weight :(
2020-07-08 16:55:33 <Hello71> see https://git.alpinelinux.org/alpine-conf/tree/setup-alpine.in and https://git.alpinelinux.org/alpine-conf/tree/setup-disk.in
2020-07-08 16:56:56 <maxice8> keepassxc is borked :D
2020-07-08 16:57:01 <c705> is anyone else having issues with libvirtd? I start the service but cannot connect with virsh -c. I want to blame upstream for this
2020-07-08 16:57:26 <c705> I'd share logs but I'm a bit busy now. Just wanted to take a poll and see if anyone else is having issues or if I'm doing something silly
2020-07-08 16:58:19 <wsinatra> Can confirm that I don't see any problems on the 3 edge systems I've got libvirt on
2020-07-08 17:01:16 <c705> wsinatra: interesting..you're just doing /etc/init.d/libvirt start? Did you ahve to do any mods after apk adding the packages?
2020-07-08 17:02:58 <wsinatra> Yeah haven't had to do anything weird, everything's status quo over here
2020-07-08 17:07:24 <c705> interesting. thanks for your observations
2020-07-08 17:07:49 <c705> if you wouldn't mind, would you share your package list? maybe i'm missing a dep or something
2020-07-08 17:09:30 <wsinatra> I used an ansible playbook to setup all three systems, but specifically I have this as for setting up libvirt
2020-07-08 17:10:12 <wsinatra> libvirt-daemon qemu-img qemu-system-i386 qemu-system-x86_64 qemu-system-arm qemu-system-aarch54 netcat-openbsd dbus polkit virt-manager
2020-07-08 17:10:18 <c705> I'm using ansible myself. Seems to be unstable 5.49 was fine, but with 5.50 now it's kaput
2020-07-08 17:10:25 <c705> cheers, let me take a look
2020-07-08 17:11:05 <eydaimon> Hello71: is that part of the setup then? The setup-disk partitioning ?
2020-07-08 17:11:14 <c705> I've got the same thing, but maybe i'm not starting dbus early enough. I haven;t really put all that much effort into looking at my issue yet..been a bit busy
2020-07-08 17:11:18 <Hello71> setup-alpine calls setup-disk
2020-07-08 17:11:43 <c705> wsinatra: thanks for your insight though. I'll look into this a bit later
2020-07-08 17:12:31 <eydaimon> Hello71: 👍
2020-07-08 17:12:34 <wsinatra> Hmm, I just checked the libvirtd version on mine, it's 6.5.0, looks like that's only one release behind, and 6.6.0 isn't out yet even.
2020-07-08 17:12:56 <c705> me too
2020-07-08 17:13:21 <c705> service starts OK, nothing funny in the logs, but both virsh and libvirt-manager can't connect to qemu
2020-07-08 17:14:37 <Hello71> rc-service libvirt restart
2020-07-08 17:14:43 <wsinatra> libvirtd*
2020-07-08 17:14:56 <c705> Hello71: no dice. If only it was that easy ;)
2020-07-08 17:15:09 <wsinatra> anything in any of the logs? Sounds like the way to go?
2020-07-08 17:15:15 <c705> I suspect virtlogd as well, but I am unsure
2020-07-08 17:15:17 <wsinatra> actually, is your user part of libvirt or kvm?
2020-07-08 17:15:20 <Hello71> doesn't libvirt use a local socket
2020-07-08 17:15:22 <wsinatra> can't remember which group it is
2020-07-08 17:15:23 <Hello71> not dbus
2020-07-08 17:15:27 <c705> user is in libvirt, not kvm,
2020-07-08 17:15:46 <wsinatra> wsinatra lp wheel audio kvm docker libvirt abuild <- that's a difference then, try adding to kvm
2020-07-08 17:16:05 <c705> Wilco - I'll add to kvm
2020-07-08 17:16:09 <wsinatra> I'm not sure that's actually necessary, but it's consistent across my machines
2020-07-08 17:16:32 <c705> I should see if root can connect actually
2020-07-08 17:16:35 <wsinatra> Hello71: I honestly don't remember. Pretty sure it does use a socket and the dbus is for something else
2020-07-08 17:16:55 <c705> for polkit maybe?
2020-07-08 17:16:56 <Hello71> find /var -type s
2020-07-08 17:18:17 <c705> will try much of this later on. cheers friends
2020-07-08 17:18:23 <c705> too bad i'm not in CET ):
2020-07-08 17:19:34 <wsinatra> Good luck! Happy to help if I can, I'm floating around here most days, just mention me if you see my nick
2020-07-08 17:27:10 <eydaimon> Hello71: it's asking me what type of install to do ... sys, data, lvm. I'm worrying about pickin gthe wrong option and erasing the partitions. Does the choice come later ? I don't recall
2020-07-08 17:27:36 <Hello71> setup-disk default creates a new partition table
2020-07-08 17:33:50 <eydaimon> Hello71: so how do I skip setup-disk ?
2020-07-08 17:33:58 <eydaimon> it didn't seem to be an option during the setup
2020-07-08 17:33:59 <Hello71> press ctrl-c
2020-07-08 17:34:05 <eydaimon> oh it's the last step ?
2020-07-08 17:34:18 <ikke> Yes
2020-07-08 17:34:18 <Hello71> https://git.alpinelinux.org/alpine-conf/tree/setup-alpine.in
2020-07-08 17:34:35 <Hello71> well if you use sys then it is
2020-07-08 17:35:15 <eydaimon> so how do I get it to do the rest of the setup  to get the ZFS enc to work ?
2020-07-08 17:35:22 <eydaimon> sorry if this is a stupid question
2020-07-08 18:38:54 <lalo-salamanca> Guys, there are so many packages for uwsgi
2020-07-08 18:39:21 <lalo-salamanca> I use http=0.0.0.0:2653
2020-07-08 18:39:56 <lalo-salamanca> in my configuration file and it complains of socket option is missing
2020-07-08 18:40:20 <lalo-salamanca> I installed uwsgi-http but error won't go away
2020-07-08 18:40:49 <lalo-salamanca> I think some plugin has to be installed but don't know which
2020-07-08 18:41:56 <lalo-salamanca> This configuration without a socket has worked for others
2020-07-08 19:00:18 <lalo-salamanca> .
2020-07-08 19:00:55 <lalo-salamanca> Was off for a while. Any solutions on uwsgi http socket problem
2020-07-08 19:03:04 <Cogitri> Huh, just works for me, I only installed uwsgi
2020-07-08 19:03:42 <Cogitri> I use it for searx
2020-07-08 19:12:04 <ikke> https://stackoverflow.com/questions/36474030/uwsgi-as-http-server-in-config-ini-option-is-called-http-socket-not-http/36474210#36474210
2020-07-08 19:12:10 <ikke> For when they come back
2020-07-08 20:17:19 <webstrand> Will busybox ever use PAM? Or will it always try and handle things internally
2020-07-08 20:19:10 <webstrand> I'm trying to figure out how to get /bin/su to use PAM so that I can enable 2-factor authentication
2020-07-08 20:20:07 <c705> webstrand: use sudo
2020-07-08 20:23:30 <webstrand> c705: I was actually trying to avoid installing sudo. But even if I did, the /bin/bbsuid binary could be used to bypass the 2nd factor, right?
2020-07-08 20:28:51 <c705> if that were the case, sudo would be useless. so I'm guessing no
2020-07-08 20:29:14 <Cogitri> Busybox could use pam, we just don't build it with pam to not pull in libpam into the base docker container
2020-07-08 20:29:35 <Cogitri> I guess maybe we could offer a busybox-pam edition
2020-07-08 20:30:06 <webstrand> I wouldn't mind switching to sudo, I really have nothing against it. But I'm not sure how I'd go about disabling `su` then?
2020-07-08 20:30:16 <ikke> disabling su?
2020-07-08 20:30:35 <webstrand> If I delete it, it'll just get recreated next time the busybox post install hook gets run.
2020-07-08 20:30:48 <ikke> why do you want to?
2020-07-08 20:31:01 <ikke> hmm
2020-07-08 20:31:06 <ikke> I see
2020-07-08 20:31:09 <webstrand> ikke: I want to force 2-factor authentication with PAM. If `su` doesn't use PAM to auth, then it can bypass 2fa
2020-07-08 20:31:30 <ikke> you can probably do chmod 600 on it
2020-07-08 20:31:42 <ikke> or remove setuid
2020-07-08 20:32:19 <ikke> oh no, for bb it's symlink to the general setuid binary
2020-07-08 20:32:22 <webstrand> that's the confusing part: /bin/bbsuid need the setuid for other commands, such as mount/unmount
2020-07-08 20:32:52 <ikke> yeah
2020-07-08 20:33:03 <ikke> so just removing su would be moot anyway
2020-07-08 20:33:12 <ikke> because bbsuid would still exist
2020-07-08 20:33:44 <c705> webstrand: https://github.com/gliderlabs/docker-alpine/issues/101
2020-07-08 20:33:51 <c705> is this what you're concerned about?
2020-07-08 20:34:04 <c705> all you need to do is lock root's password: passwd -l root
2020-07-08 20:36:40 <ikke> but you could still login as other users if you know their password
2020-07-08 20:39:32 <webstrand> c705: I think I see what you're getting at. Right now, I think I want root to have a valid password, though.
2020-07-08 20:39:52 <webstrand> I'll have to think about locking it, and using user passwords to escalate
2020-07-08 20:40:01 <c705> webstrand: imo, root should always be locked and sudo policies set for as little access as you require in your container
2020-07-08 20:40:09 <c705> then sudo up from the regualr user
2020-07-08 20:40:26 <c705> but do whatever you need to, be my guest
2020-07-08 20:40:42 <ikke> why do you assume docker containers?
2020-07-08 20:41:02 <webstrand> ah, that's one difference. I'm running alpine on bare metal
2020-07-08 20:41:23 <ikke> so root pw is not a bad idea in case you get locked out
2020-07-08 20:41:30 <c705> ikke: either way, my recommendation would be the same
2020-07-08 20:41:37 <ikke> ^^
2020-07-08 20:42:15 <c705> ikke: if you get locked out, init=/bin/sh or single user mode. it's sightly more of a pain, but a bit safer
2020-07-08 20:42:29 <ikke> does require downtime
2020-07-08 20:42:36 <ikke> which might or might not be an issue
2020-07-08 20:42:38 <c705> ikke: I suppose so
2020-07-08 20:43:26 <c705> idea of root password doesn;t sit well with me. especially if there is more than 1 admin
2020-07-08 20:43:36 <webstrand> not particularly an issue. Rebooting is already a pain due to an encrypted rootfs. I shouldn't be locking myself out, though.
2020-07-08 20:43:39 <c705> but obviously, you do whatever you need to do, right?
2020-07-08 20:47:01 <webstrand> another option would be to use a passwd hashing algorithm thats unsupported by busybox
2020-07-08 21:06:22 <opal> trying and failing to install ejabberd on armhf https://wowana.me/paste/m0EmN3.txt
2020-07-08 21:07:07 <ikke> opal: combining edge and stable is not supported
2020-07-08 21:07:13 <opal> i tried without edge/main and edge/community as well
2020-07-08 21:07:48 <ikke> and you have different repos pinne with the same name, not sure how that's supposed to work
2020-07-08 21:08:45 <ikke> I think each repository should have a unique name if you use pinnign
2020-07-08 21:09:13 <ikke> but as you see, apk does not try to find the dependencies in the pinned repos
2020-07-08 21:09:31 <ikke> and those dependencies might even conflict with already installed packages
2020-07-08 21:10:20 <ikke> opal: did you try on a complete edge system?
2020-07-08 21:16:56 <opal> no, i'd rather just roll my own repo entirely since i'm probably better off anyway. would like to get off openrc while im at it
2020-07-08 21:17:08 <ikke> nod
2020-07-08 21:17:14 <opal> >and you have different repos pinne with the same name,
2020-07-08 21:17:18 <opal> was just trying it to see if it worked
2020-07-08 21:17:31 <opal> >apk does not try to find the dependencies in the pinned repos
2020-07-08 21:17:31 <ikke> I don't think it would solve this issue
2020-07-08 21:17:32 <opal> ah
2020-07-08 21:17:46 <ikke> what would you use instead of openrc?
2020-07-08 21:17:49 <opal> s6
2020-07-08 21:17:52 <ikke> aha, ok
2020-07-08 21:17:54 <opal> no im not saying it would solve this issue
2020-07-08 21:17:59 <opal> just wanted to do it anyway
2020-07-08 21:18:32 <opal> might as well play with stuff while im not using my pi for anything serious, perfect opportunity
2020-07-08 21:18:46 <opal> ive used s6 for a user session so far, running it from xinitrc and it's worked fine for that purpose
2020-07-08 21:18:56 <opal> gonna see how s6-rc and s6 as init works for a system install
2020-07-08 21:19:25 <ikke> I can't recally why anymore, but I've heard that using s6 as a general init system doesn't work that well
2020-07-08 21:19:46 <opal> well, s6-rc is supposed to solve the service dependency issue
2020-07-08 21:19:53 <opal> other than that i cant think of possible pitfalls
2020-07-08 21:20:06 <ikke> Ah, right, that might have been the issue I've heard
2020-07-08 21:20:23 <opal> cgroups and other stuff can be managed externally
2020-07-08 21:20:30 <opal> so that shouldnt be an issue
2020-07-08 21:21:03 <opal> yeah, that along with static linking everything, perfect excuse to spin up a local repo :>
2020-07-08 21:22:24 <ikke> How do you manage security updates with statically compiled binaries?
2020-07-08 21:22:37 <ikke> sounds like a lot of work having to rebuild everythign constantly
2020-07-08 21:23:32 <maxice8> You don't, you pray
2020-07-08 21:25:06 <ikke> That's apears to me one of the biggest downsides of statically compiling everything
2020-07-08 21:25:30 <ikke> Even though it's every attractive due to not having to deal with dependencies
2020-07-08 21:30:03 <webstrand> opal: s6-rc looks pretty neat, do you know if there's an existing distro using it?
2020-07-08 21:37:45 <opal> not that i know of
2020-07-08 21:38:36 <opal> ikke well it's something a source-based distro would be better at, yeah
2020-07-08 21:43:00 <eydaimon> Hello71: did you see my stupid question ?
2020-07-08 21:49:13 <Hello71> didn't I already say to format yourself and then do setup-disk /mnt
2020-07-08 21:49:20 <Hello71> I guess not exactly
2020-07-08 21:55:59 <nshp> ikke: there was some analysis of that whole issue here: https://drewdevault.com/dynlib.html
2020-07-08 21:58:12 <ikke> Yes, I've read that
2020-07-08 22:09:09 <eydaimon> Hello71: thank you for clarifying :)
2020-07-08 22:15:24 <opal> nshp thanks for the link, havent seen it
2020-07-08 22:15:34 <opal> just seen uh, i think cat-v's writeup and maybe a few others
2020-07-08 22:15:47 <opal> both articles in defense for dynamic linking and against it
2020-07-08 22:17:20 <opal> how does alpine build process handle relinking against libraries whose ABIs change (qt comes to mind)
2020-07-08 22:17:53 <maxice8> Rebuild the package to link against the new version
2020-07-08 22:18:02 <opal> is this a manual process
2020-07-08 22:18:04 <ikke> yup, so a pkgrel bump for all dependencies
2020-07-08 22:18:05 <ikke> yes
2020-07-08 22:18:08 <opal> ouch
2020-07-08 22:18:18 <ikke> there are tools that list what depencies there are
2020-07-08 22:18:28 <ikke> ap (lua-aports) for example
2020-07-08 22:18:40 <maxice8> It is mostly automated
2020-07-08 22:19:10 <ikke> and apkgrel to update the pkgrel of packages
2020-07-08 22:19:19 <ikke> so it's quite easay to script
2020-07-08 22:19:22 <ikke> easy*
2020-07-08 22:21:18 <ikke> ap revdep qt5-qtbase-dev lists 372 packages
2020-07-08 22:21:52 <opal> i see
2020-07-08 22:23:09 <opal> would this scale for the purpose of static linking or am i better off with a package manager that's more aware of backtracking dependencies
2020-07-08 22:23:41 <opal> i know alpine is trying to add more -static packages so maybe this is something that could be improved
2020-07-08 22:23:45 <opal> unsure what your take is
2020-07-08 22:23:58 <maxice8> -static packages are mostly for static libraries some people use
2020-07-08 22:24:08 <maxice8> and they are added on demand and should be removed after some time
2020-07-08 22:24:11 <opal> ah
2020-07-08 22:24:19 <ikke> maxice8: why after some time?
2020-07-08 22:24:44 <maxice8> Because they are unwanted  and are only added after someone asks and people don't stick forever
2020-07-08 22:26:18 <ikke> opal: ncopa is planning to build a next-gen version of ap, so I think things in that regard should only improve
2020-07-08 22:26:23 <opal> i was under the impression that static binaries were a lasting effort in alpine as of late
2020-07-08 22:26:26 <ikke> but I think even ap already should get you pretty far
2020-07-08 22:26:35 <opal> ah ok
2020-07-08 22:26:46 <opal> ap being what here
2020-07-08 22:26:52 <opal> im conditioned to think of "activitypub" every time i hear it
2020-07-08 22:26:53 <maxice8> There are certain static stuff like busybox
2020-07-08 22:27:08 <ikke> opal: it's a tool provided by lua-aports
2020-07-08 22:27:14 <opal> o
2020-07-08 22:27:32 <ikke> it provides repo-wide data
2020-07-08 22:27:47 <opal> im gonna have to look more into this
2020-07-08 22:28:28 <ikke> the alpine builders rely heavily on ap
2020-07-08 22:33:30 <mps> maxice8: +1
2020-07-08 22:33:41 <ikke> mps: right on time
2020-07-08 22:33:51 <mps> :)
2020-07-08 22:33:58 <mps> you know
2020-07-08 22:35:21 <eydaimon> Hello71: just to be even more clear (sorry to bother you), would I still create /mnt as zfs and subdir /mnt/boot as ext4 ? Also, I can't load zfs ... modprobe fails because it's not in modules.dep
2020-07-08 22:35:36 <Hello71> did you apk add zfs
2020-07-08 22:35:41 <eydaimon> aye
2020-07-08 22:35:49 <Hello71> hm.
2020-07-08 22:35:52 <eydaimon> zpool.ko is in that file, but not zfs.ko :/
2020-07-08 22:35:57 <Hello71> try depmod -a
2020-07-08 22:36:40 <eydaimon> can't open modules.dep read-only filesystem
2020-07-08 22:37:28 <eydaimon> if I use the live CD, is setup-alpine available there ?
2020-07-08 22:37:32 <eydaimon> in the chroot ?
2020-07-08 22:39:15 <maxice8> I have no clue what you're '+1'-ing though mps
2020-07-08 22:39:34 <eydaimon> maxice8: take it and run with it :)
2020-07-08 22:39:48 <ikke> maxice8: removing -static :)
2020-07-08 22:39:59 <ikke> a pet-peeve from mps
2020-07-08 22:43:07 <maxice8> I think we want it for different reasons
2020-07-08 22:43:20 <ikke> yes
2020-07-08 22:47:28 <maxice8> A temporary alliance then someone backstabs the  other
2020-07-08 22:58:39 <Hello71> eydaimon: ... didn't I already say to use the alpine chroot
2020-07-08 23:13:50 <eydaimon> Hello71: not explitly and because I can't load ZFS, I am verifying
2020-07-08 23:13:56 <eydaimon> explicitly
2020-07-08 23:14:04 <Hello71> uh, s/chroot/iso/
2020-07-08 23:14:08 <Hello71> sorry
2020-07-08 23:14:20 <eydaimon> heh :)
2020-07-08 23:29:15 <alicemaz> hi all. couple questions, is there any way to track the status of an edge package making it into a release? looking forward to cloud-init and hoping it makes it in this year ehe. also, is it normal for *-dev packages of libraries to not include archive files for static linking?
2020-07-08 23:32:48 <eydaimon> Hello71: well, since I can't load ZFS, I'm a bit stuck doing it that way, no ?
2020-07-08 23:32:53 <Hello71> http://google.com/search?q=alpine+linux+release+cycle
2020-07-08 23:33:09 <Hello71> eydaimon: I forget how installing modules on the iso works
2020-07-08 23:34:25 <eydaimon> anyone here who would know ?
2020-07-08 23:35:24 <eydaimon> I searched on the wiki but I don't see anything obvious
2020-07-08 23:38:06 <c705> check ./modloop
2020-07-08 23:38:16 <c705> taht's where my modules land from the iso
2020-07-08 23:38:24 <eydaimon> c705: thank you
2020-07-08 23:39:04 <Hello71> oh, I see. you're supposed to download the extended iso
2020-07-08 23:39:10 <Hello71> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh#L32
2020-07-08 23:47:02 <alicemaz> oh I was confused about the distinction between edge/main and edge/testing. it's been in edge/testing for several years based on git logs
2020-07-08 23:51:20 <shiroyasha> Weird question about terminology. Is the "aports" term specific to Alpine's "aports tree", or is it some generic term? I'm making my own custom Alpine repositories for private packages (nothing from the official repos), so I'm not sure if it makes sense to call the Git repository "aports" too, or think of something else.
2020-07-09 00:10:04 <c705> I think aports is just short for alpine-ports
2020-07-09 00:15:21 <shiroyasha> Hmm yeah I'll just call it `ports` and move on.
2020-07-09 01:19:34 <singpolyma> Hello!  I am trying to build the unmaintained/gajim package manually (both to try it out on my pinephone, and maybe to get it back to maintained eventually?) but when I run abuild -r it complains that it can't find py3-nbxmpp, which is also in unmaintained.  Is there any way I can get abuild to search in the unmaintained dir for dependencies?
2020-07-09 01:21:20 <Hello71> abuild doesn't build recursively
2020-07-09 01:21:54 <singpolyma> Oh! I thought from the wiki that it did. It only installs recursively from things included in the main archive?
2020-07-09 01:26:40 <opal> this is tangential but if you dont need the administrative features gajim provides, try dino
2020-07-09 01:26:53 <opal> it isnt the best but neither is gajim, and dino works well enough
2020-07-09 01:27:45 <singpolyma> opal: dino is what I'm using on pinephone right now, yeah
2020-07-09 01:28:00 <singpolyma> but I want to try gajim. It's much more featureful and I'm more used to it
2020-07-09 01:28:29 <opal> ah
2020-07-09 01:49:37 <eydaimon> looks like zfs-lts (based on pkgs.alpinelinux.org) contains zfs.ko. Trying to install it using iso is giving me a lot of failures (failed to rename lib/something → lib/somethingelse
2020-07-09 01:54:55 <eydaimon> any idea how I'd go about resolving that? It doesn't give me the full path so I don't know what directories I could create to fix it
2020-07-09 04:18:19 <eydaimon> I want to switch to alpine. now :/
2020-07-09 04:18:22 <eydaimon> need to get over this hurdle
2020-07-09 04:48:21 <Marquel> eydaimon: you need to pastebin the full error message for someone to be able to actually help you.
2020-07-09 04:52:01 <opal> ikke i cant install ejabberd even when fully switching to edge
2020-07-09 04:52:07 <opal> same dependency errors
2020-07-09 04:53:49 <ikke> opal: right, I can reprpo
2020-07-09 04:53:56 <ikke> repro*
2020-07-09 04:54:05 <opal> ah
2020-07-09 04:54:16 <opal> the deps get removed at some point or what
2020-07-09 04:54:28 <ikke> I do recall a change in these dependencies
2020-07-09 04:54:41 <ikke> It might be that they are just now part of erlang itself
2020-07-09 04:54:45 <ikke> no longer subpackages
2020-07-09 04:55:23 <ikke> f65b878adaf5996b64bf8f6b11d04cbbceca865e
2020-07-09 04:57:57 <ikke> 53aebeddaa4ae6901b1c40652c1cdd21b57106cb
2020-07-09 04:58:00 <ikke> that's the actual change
2020-07-09 05:06:05 <ikke> without those dependencies it fails to build for me though
2020-07-09 05:06:09 <ikke> linking error
2020-07-09 05:06:26 <ikke> "cannot find -lerl_interface"
2020-07-09 07:11:48 <abenz> hey guys
2020-07-09 07:12:03 <abenz> I suspect my rare boot failures are due to my disks being mixed up
2020-07-09 07:12:12 <abenz> I'm using diskless mode btw
2020-07-09 07:12:46 <abenz> does it seem like a good idea to replace fstab entries, and modify grub.cfg to use UUIDs instead of the current block (sdc1) style naming?
2020-07-09 07:19:27 <afontain_> yes, it seems
2020-07-09 07:22:13 <afontain_> my debian-configured grub uses uuid btw
2020-07-09 07:23:59 <nacelle> by uuid has been the prerred way in most distros for a while now
2020-07-09 07:24:05 <nacelle> preferred
2020-07-09 07:24:19 <nacelle> (like, maybe 20 years or so)
2020-07-09 07:25:16 <abenz> boot failed, I modified grub.cfg (using UEFI), and fstab. Also updated /etc/apk/repos and relinked /etc/apk/cache to point to right place
2020-07-09 07:25:20 <abenz> what am I missing?
2020-07-09 07:25:40 <abenz> it says mounting /dev/sdc1 failed.. so sdc1 is stated somewhre I didnt modify
2020-07-09 07:26:24 <afontain_> what say that?
2020-07-09 07:36:09 <abenz> syslinux
2020-07-09 07:36:23 <abenz> I updated syslinux.cfg and added: root=UUID=xxx
2020-07-09 07:36:50 <abenz> but it fails again, this time with "mounting UUID=xxx on /sysroot failed: invalid argument"
2020-07-09 07:37:45 <ikke> I recall that bb mount does not support uuids
2020-07-09 07:38:53 <abenz> where can I change the current "sdc1" to "usb" ?
2020-07-09 07:39:23 <afontain_> maybe /dev/disk/by-uuid the way to go then
2020-07-09 07:40:53 <ikke> or install util-linux
2020-07-09 07:42:00 <ikke> abenz: iirc, you can specify alpine_dev to the kernel parameters
2020-07-09 07:42:12 <ikke> https://unix.stackexchange.com/questions/502610/alpine-persistent-disk-mounted-to-media-sda-even-though-fstab-has-media-usb
2020-07-09 08:06:56 <abenz_> ikke: specifying alpine_dev (in syslinux.cfg) made no difference..
2020-07-09 08:07:08 <abenz_> I still get error mounting /dev/sdc1 failed
2020-07-09 08:08:31 <abenz_> I put that at the end of append, retrying with alpine_dev being first
2020-07-09 08:09:49 <ikke> order should not matter
2020-07-09 08:10:29 <ikke> but if I look at https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L333, that's the only reference to it, so it might no longer be used
2020-07-09 08:11:00 <abenz_> I will try root=/dev/disk/by-uuid next
2020-07-09 08:11:03 <abenz_> is this the right format?
2020-07-09 08:12:20 <ikke> Unless you have udev, I don't think those nodes exist
2020-07-09 08:12:33 <ikke> "ls: /dev/disk: No such file or directory"
2020-07-09 08:12:53 <ikke> I would just install util-linux to get the full mount command
2020-07-09 08:13:01 <ikke> than you should be able to specify UUID in fstab
2020-07-09 08:14:05 <abenz_> I cant boot into my diskless usb now
2020-07-09 08:14:22 <abenz_> syslinux tries to find /dev/sdc1 which no longer exists
2020-07-09 08:14:39 <abenz_> I have to add disks until the usb becomes sdc1 to regain access
2020-07-09 08:15:17 <abenz_> this is why I'm trying to avoid this frequent change by using uuids..
2020-07-09 08:29:45 <abenz> I just ran grep -rn '.' -e '/dev/sdc1' on my usb and didnt get any hits
2020-07-09 08:30:00 <abenz> that means sdc1 is embedded into some binary/processed file
2020-07-09 08:30:18 <abenz> I guess I need to rerun syslinux to regenerate those files with the updated config
2020-07-09 08:30:26 <abenz> is this assumption correct?
2020-07-09 08:32:11 <nacelle> usually one does update-grub
2020-07-09 08:32:52 <ikke> Not for diskless setups
2020-07-09 08:35:37 <abenz> upon this failed boot, I tried exploring the available commands
2020-07-09 08:35:47 <abenz> sdc1 is infact the right partition
2020-07-09 08:36:00 <abenz> and when I attempt to mount it to /sysroot, I get: invalid arguments
2020-07-09 08:36:17 <abenz> so the failure is not caused by wrong device name
2020-07-09 08:37:39 <dominykas> hey everyone, just looking to confirm - are any of these organizations in any way official? https://hub.docker.com/u/alpine, https://github.com/alpine-docker?
2020-07-09 08:42:05 <ikke> dominykas: no
2020-07-09 08:42:20 <ikke> https://github.com/alpinelinux/
2020-07-09 08:42:29 <dominykas> thank you
2020-07-09 08:42:36 <ikke> and https://hub.docker.com/u/alpinelinux
2020-07-09 08:43:56 <abenz> once I removed root=UUID=xxx from grub.cfg it worked
2020-07-09 08:44:46 <abenz> my fstab has entries for UUIDs though
2020-07-09 08:45:10 <abenz> I will restart a few times and if it doesn't fail, the solution was simply to update fstab without touching grub.cfg or syslinux.cfg
2020-07-09 08:48:15 <abenz> 3 reboots and it works, I think its fixed. Thanks for the help :)
2020-07-09 08:49:50 <ikke> so what was the final solution?
2020-07-09 08:50:12 <abenz> remove /dev/sdXX from fstab, replace with UUIDs
2020-07-09 08:50:20 <ikke> ok
2020-07-09 08:50:30 <maxice8> nice
2020-07-09 08:51:55 <abenz> I have to mention, if you use setup-bootable on a usb drive with more than one partition (say one for alpine and one for storage) it will default to using /dev/sdXx. Only if its a single partiton stick, will it use /media/usb..etc
2020-07-09 08:52:53 <abenz> mine has an ext4 partition for storage, so I had to update /etc/apk/repo, and /etc/lbu/lbu.conf, and /etc/apk/cache ---> /media/usb/cache
2020-07-09 08:53:03 <abenz> and that is the full solution in detail :)
2020-07-09 09:40:05 <id4rk> solutions in detail are great solutions, like long cat is long
2020-07-09 09:54:06 <lalo-salamanca> Guys, in FreeBSD, I used to add ddclient_enable=YES at /etc/rc.conf
2020-07-09 09:54:25 <lalo-salamanca> I use inadyn at Alpine
2020-07-09 09:54:46 <lalo-salamanca> How can I do this in Alpine?
2020-07-09 09:56:34 <ikke> Do what exactly?
2020-07-09 09:57:07 <lalo-salamanca> Start inadyn daemon in background at boot
2020-07-09 09:57:27 <ikke> rc-update add inadyn
2020-07-09 09:58:08 <lalo-salamanca> I need to reboot then right?
2020-07-09 09:58:14 <lalo-salamanca> Thanks
2020-07-09 09:58:30 <ikke> if you just want to start it
2020-07-09 09:58:38 <ikke> you can run rc-service inadyn start
2020-07-09 09:58:48 <ikke> (asuming the service name is called inadyn)
2020-07-09 09:59:29 <lalo-salamanca> Yes, it is.
2020-07-09 09:59:33 <lalo-salamanca> Thanks man
2020-07-09 10:56:28 <lalo-salamanca> There are so many uwsgi plugins guys. I'm confused
2020-07-09 10:56:47 <lalo-salamanca> My site redirects http to https but it doesn't load
2020-07-09 10:57:39 <lalo-salamanca> I have used https=0.0.0.0:2554,example.com.cer,example.com.key
2020-07-09 10:57:54 <lalo-salamanca> configuration for https
2020-07-09 10:58:38 <ikke> and does curl --resolve 127.0.0.1:2553:example.com  https://localhost:2554 work?
2020-07-09 10:58:45 <lalo-salamanca> Do I need additional plugin for https, I have installed uwsgi-sslrouter
2020-07-09 10:58:46 <ikke> sorry
2020-07-09 10:58:51 <ikke> 2554 in the first case as well
2020-07-09 10:59:18 <ikke> lalo-salamanca: I would suggest putting something on front of it that does the tls part
2020-07-09 11:00:24 <lalo-salamanca> ? I use just uwsgi. I don't want to use nginx
2020-07-09 11:01:32 <ikke> lalo-salamanca: did you try the command I suggested?
2020-07-09 11:02:52 <lalo-salamanca> I don't understand that command. I'm reading curl help
2020-07-09 11:05:03 <lalo-salamanca> No I don't think this is the problem. I'll look into this with curl
2020-07-09 14:19:14 <eydaimon> omg, the extended ISO. re-reading  helps sometimes.
2020-07-09 14:21:03 <eydaimon> Marquel: there was context form above. but Hello71 had already given me the solution :/
2020-07-09 14:28:21 <dustfinger> If I fetch a package to a local directory and run sha512sums package-name.tar.gz, then that should give me the same 512sums that I see in the APKBUILD file of the commit referenced by the packages PGKINFO. Is that correct?
2020-07-09 14:29:13 <ikke> no
2020-07-09 14:29:46 <eydaimon> are there plans to put ZFS in the standard ISO ?
2020-07-09 14:29:47 <dustfinger> ikke: OKay, that is good. So how does it really work? How can I manually verify a package with what I have installed?
2020-07-09 14:30:19 <ikke> dustfinger: the hashes in the APKBUILD verify the upstream sources
2020-07-09 14:31:21 <dustfinger> Okay, thank you. I will check that as well. How about the package itself. Is there anyway of verifying the downloaded package? Or perhaps the individual binaries in the downloaded package?
2020-07-09 14:36:47 <webstrand> Cogitri: busybox also provides /bin/login which needs to be replaced to be PAM-safe. It might be a good idea to have a busybox-pam, just for safety's sake.
2020-07-09 14:37:14 <ikke> dustfinger: it's broken now, but once fixed, you can use abuild-gzsplit to properly extract the package
2020-07-09 14:37:22 <ikke> then you can verify it's signature manually
2020-07-09 14:37:43 <ikke> or you can use apk verify on the package
2020-07-09 14:57:39 <mps> webstrand: apk add linux-pam
2020-07-09 14:58:16 <Cogitri> mps: That doesn't cause busybox to be built against PAM, which is the problem here
2020-07-09 14:58:40 <mps> we don't need pam (any) in base installations
2020-07-09 14:58:50 <ikke> mps: that's not what's being asked
2020-07-09 14:59:03 <mps> ikke: aha. sorry then
2020-07-09 14:59:05 <Cogitri> webstrand: Can you maybe open an issue for that on our Gitlab? I don't have time to look into it right now (and make sure it doesn't break login :)
2020-07-09 14:59:20 <webstrand> Cogitri: sure
2020-07-09 14:59:25 <mps> what was question, I don't see original
2020-07-09 14:59:34 <ikke> mps: it started yesterday alreayd
2020-07-09 14:59:45 <webstrand> mps: sorry, it was sometime yesterday. I was just following up with my discoveries
2020-07-09 14:59:54 <mps> ah, have to look backlog, sorry again
2020-07-09 15:00:05 <ikke> mps: the basic idea is to require 2fa
2020-07-09 15:00:09 <webstrand> I'm trying to add 2fa, but have run into trouble with su and login
2020-07-09 15:00:10 <ikke> for things like su
2020-07-09 15:00:37 <mps> hmm, understand now
2020-07-09 15:00:43 <dustfinger> ikke: yah, I just looked at the source for apk verify. Thank you foro your time and for letting me know about abuild-gzslit, I will take a look at that also.
2020-07-09 15:01:02 <mps> so, su doesn't work without pam?
2020-07-09 15:01:34 <webstrand> the issue is, it uses it's own auth and bypasses pam. So whatever 2fa I configure doesn't get applied
2020-07-09 15:02:20 <mps> su and login from busybox intended to work without pam
2020-07-09 15:03:14 <Cogitri> mps: Yes, but you need PAM for it to work with your PAM config..
2020-07-09 15:03:29 <webstrand> mps: but if you've enabled PAM, they represent a vulnerability, since they ignore PAM config
2020-07-09 15:03:30 <Cogitri> And we don't enable PAM to not increase the size of the Docker baseimage
2020-07-09 15:03:43 <mps> then why linux-pam wouldn't be installed
2020-07-09 15:04:14 <webstrand> Cogitri: Do you think this is out-of-scope for this distro? I could switch to gentoo
2020-07-09 15:05:13 <mps> and I think we should not 'enable' pam in base installs
2020-07-09 15:07:02 <mps> webstrand: you can fill bug/issue report as Cogitri told, so more people can be involved in discussion
2020-07-09 15:08:43 <webstrand> If I included a patch to build busybox-pam, would you find that useful?
2020-07-09 15:10:01 <mps> busybox already have to much patches, and maintainer expressed it is not good idea to add more
2020-07-09 15:10:23 <mps> but, sure you can post patch for discussion
2020-07-09 15:10:25 <webstrand> sorry, I mean a patch to the APKBUILD, not an additional busybox patch
2020-07-09 15:10:51 <mps> ah, config option? then it could be easier, maybe
2020-07-09 15:11:31 <webstrand> yeah, busybox has support for PAM, it just needs to be built with CONFIG_PAM=y
2020-07-09 15:11:44 <Cogitri> Yes, it's one config switch we have to change and it wouldn't end up in the base install, I guess we'd need a separate package
2020-07-09 15:12:30 <mps> maybe for busybox-extra
2020-07-09 15:12:45 <webstrand> you wouldn't want to build it as a subpackage?
2020-07-09 15:14:25 <Cogitri> Wouldn't the subpackage replace files from the mainpackage?
2020-07-09 15:14:26 <mps> bug/issue report would be best to discuss this futher
2020-07-09 15:14:47 <mps> not all devs are here
2020-07-09 15:14:49 <webstrand> sure. trying to hunt down my gitlab account
2020-07-09 15:15:42 <mps> and that will be relatively important change so broader discussion will be needed
2020-07-09 15:16:33 <mps> though I can express my position right now, i dislike pam
2020-07-09 15:17:28 <Cogitri> Well, you don't have to install the busybox-pam package then, I guess
2020-07-09 15:19:13 <mps> Cogitri: :)
2020-07-09 15:19:48 <mps> I dislike whole idea called 'pam'
2020-07-09 15:20:01 <webstrand> apparently, gitlab deleted my account >.<
2020-07-09 15:20:10 <ikke> Wut?
2020-07-09 15:20:41 <Cogitri> mps: Well, PAM is pretty neat, how would you manage 2fa, kerberos, settings things during login in a generic way otherwise?
2020-07-09 15:20:54 <mps> pam is bad for security, imo
2020-07-09 15:21:02 <dustfinger> ikke: I noticed that the .SIGN.RSA.*.pub is not an RSA file. Is that normal. It is binary and full of unprintable bytes
2020-07-09 15:21:45 <dustfinger> s/is not an RSA file/is not an RSA public key/
2020-07-09 15:21:52 <mps> I alredy elaborated my position about pam few times, so I'd rather not repeat again
2020-07-09 15:22:46 <ikke> dustfinger: why can't it be? It's jusr not encoded in printable characters
2020-07-09 15:23:25 <ikke> Most likely DER encoded
2020-07-09 15:23:30 <ikke> Not PEM
2020-07-09 15:23:39 <dustfinger> ikke: okay, I never thought about that. Thanks again!
2020-07-09 15:23:44 <mps> PEM
2020-07-09 15:24:05 <ikke> mps?
2020-07-09 15:24:16 <mps> PEM encoded
2020-07-09 15:24:26 <ikke>  https://www.cryptosys.net/pki/rsakeyformats.html
2020-07-09 15:24:45 <dustfinger> Thanks!
2020-07-09 15:25:01 <ikke> PEM, encoded would be base64 encoded
2020-07-09 15:25:41 <ikke> But for something that is rarely read by humans DER encoding makes more sense I guess
2020-07-09 15:25:45 <mps> hmm, DER then
2020-07-09 15:29:34 <mps> ikke: yes, it is DER
2020-07-09 15:32:20 <nponaraHAa> has anybody tried alpine in  m$$10.dll WSL thingy?
2020-07-09 15:32:48 <ikke> only very briefly
2020-07-09 15:37:21 <nponaraHAa> did it work?
2020-07-09 15:38:01 <ikke> yes
2020-07-09 15:38:25 <nponaraHAa> ikke: I wonder if it would be a bettor choice than cygwin
2020-07-09 15:58:47 <webstrand> ikke: I just phished myself... I thought I was logging into gitlab.com, not a privately hosted instance.
2020-07-09 16:10:46 <Hello71> password manager
2020-07-09 16:16:41 <webstrand> Hello71: oh, I have one. I was confused enough that I defeated its protection and manually copied my password out.
2020-07-09 18:25:23 <nponaraHAa> has anybody attempted to do this? run firefox in fbdev
2020-07-09 18:25:53 <nponaraHAa> the thing just devours ram
2020-07-09 18:27:34 <distortedsignal> Hey all, I'm trying to build a Docker container for the UPX utility (https://github.com/upx/upx) and I'm trying to build it with Alpine Linux. Building with Alpine version 3.10 works, but when I try to upgrade to 3.11, I get a compile error with the following error message: /usr/lib/gcc/x86_64-alpine-linux-musl/9.2.0/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find -lz \n collect2: error: l
2020-07-09 18:27:35 <distortedsignal> ery unfamiliar with building large C/C++ projects, so you might have to dumb it _waaaaaaaaaay_ down for me.
2020-07-09 18:27:35 <distortedsignal> d returned 1 exit status \n make: *** [Makefile:87: upx.out] Error 1 \n. If you want to try building the container on your local system, the dockerfile is here: https://raw.githubusercontent.com/distortedsignal/Dockerfiles/add-upx/upx/Dockerfile. I'm curious what changed between version 3.10 and version 3.11 that causes this to break this build. I am very unfamiliar with the upx codebase, and I'm v
2020-07-09 18:28:12 <distortedsignal> Looks like my error message got hit by the linebreak demon, so I'll add them back in.
2020-07-09 18:28:26 <nponaraHAa> distortedsignal: compile what where?
2020-07-09 18:28:41 <ikke> distortedsignal: are you missing zlib-dev?
2020-07-09 18:29:47 <distortedsignal> I don't think so - lines 13-17 of the linked dockerfile reduce to "apk add --no-cache build-base git ucl-dev zlib-dev"
2020-07-09 18:30:38 <opal> < nponaraHA>…the thing just devours ram
2020-07-09 18:30:42 <opal> idk sounds like firefox
2020-07-09 18:31:19 <ikke> distortedsignal: you know we have upx in the repositories?
2020-07-09 18:31:24 <ikke> https://pkgs.alpinelinux.org/packages?name=upx&branch=v3.11
2020-07-09 18:31:36 <nponaraHAa> opal: I read somewhere on firefox site it already reccomends 8GB minimum of system ram.
2020-07-09 18:32:18 <nponaraHAa> opal: alpine have been selectively OOM killing firefox tabs. Which quite surprising.
2020-07-09 18:32:47 <nponaraHAa> somehow doesn't kill firefox entirely but searches bloated tab and kills them.
2020-07-09 18:32:56 <nponaraHAa> I didn't know that was possible.
2020-07-09 18:33:00 <opal> check dmesg
2020-07-09 18:33:06 <opal> its likely the oomkiller doing that
2020-07-09 18:33:11 <opal> and not just one tab but a tab group
2020-07-09 18:33:15 <nponaraHAa> opal: I did, that's what I am saying.
2020-07-09 18:33:21 <distortedsignal> @nponaraHAa - I'm trying to compile the upx utility inside of a container that is the alpine linux image. I feel like the dockerfile does a pretty good job of explaining what I'm trying to do.
2020-07-09 18:33:24 <nponaraHAa> opal: exactly tab groups
2020-07-09 18:33:39 <opal> its how firefox's content processes work, they each control a set of tabs
2020-07-09 18:33:50 <opal> this has been a thing ever since firefox rolled out e10s
2020-07-09 18:34:09 <ikke> distortedsignal: any reason you are trying to compile it yourself? (just curious)
2020-07-09 18:34:13 <nponaraHAa> interesting behaviour
2020-07-09 18:34:43 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/upx/APKBUILD
2020-07-09 18:34:43 <opal> so, e10s works properly on musl now?
2020-07-09 18:34:46 <nponaraHAa> ikke: I was about to ask the same at risk of looking dumb
2020-07-09 18:34:48 <opal> thats cool i guess
2020-07-09 18:34:55 <mps> firefox on alpine runs even with 2GB of ram
2020-07-09 18:35:08 <opal> i have firefox capped to a 2gb memory cgroup
2020-07-09 18:35:19 <opal> it "runs" but heavy tabs will cause issues
2020-07-09 18:35:30 <opal> and, well, the web is full of unavoidable heavy tabs
2020-07-09 18:35:33 <opal> god i hate the web
2020-07-09 18:35:36 <nponaraHAa> but really it isn't all firefox fault, but insane js and flashy whatnots on webpages.
2020-07-09 18:35:39 <distortedsignal> ikke: It's an interesting exercise. I want to run this on a company-owned build server, and I think the only way I'm going to be able to sneak this on is if I compile it/export it myself.
2020-07-09 18:35:47 <ikke> ok
2020-07-09 18:35:49 <nponaraHAa> one example level1tech forums.
2020-07-09 18:35:54 <opal> firefox could more-aggressively garbage collect but it isn't
2020-07-09 18:36:02 <ikke> Look at that APKBUILD, perhaps it has some clues
2020-07-09 18:36:11 <nponaraHAa> reddit needs hundreds of MBs
2020-07-09 18:37:29 <distortedsignal> I'll check it out - thanks ikke.
2020-07-09 18:56:07 <bovis> I'm receiving a "prepare_subpackages failed" error with an APKBUILD. Where should I begin looking to resolve the issue? Here's the build file: http://sprunge.us/wJn6mj
2020-07-09 18:56:48 <ikke> bovis: what is the exact error?
2020-07-09 19:02:45 <bovis> ikke: Here are the last several lines of the build output http://sprunge.us/mnLAct
2020-07-09 19:03:07 <bovis> I'm seeing "prepare_subpackages" about 40 lines from the bottom
2020-07-09 19:10:35 <ikke> the source is missing
2020-07-09 19:13:28 <Cogitri> «««ERROR: wine-libs*: Found textrels:»»»
2020-07-09 19:19:49 <bovis> ikke: I have the source of the local tar.gz. The package builds fine. This is the only source that's listed in previous APKBUILDs for wine.
2020-07-09 19:20:02 <bovis> Cogitri: I don't know how to resolve that issue.
2020-07-09 19:21:06 <Cogitri> Ideally ask upstream to fix it
2020-07-09 19:21:13 <ikke> bovis: I tried to build it from the APKBUILD, but that obviously fails
2020-07-09 19:21:20 <Cogitri> In the meantime you can add "options="textrels""
2020-07-09 19:21:43 <ikke> https://wiki.gentoo.org/wiki/Hardened/Textrels_Guide
2020-07-09 19:21:54 <bovis> ikke: I downloaded the xz file from winehq, packaged it locally, and built it locally
2020-07-09 19:28:09 <ikke> https://flameeyes.blog/2016/01/16/textrels-text-relocations-and-their-impact-on-hardening-techniques/
2020-07-09 19:53:59 <Winst> how do I setup my card before setup
2020-07-09 19:54:10 <ikke> what card>
2020-07-09 19:56:47 <Winst> wifi
2020-07-09 19:58:44 <ikke> Is your wificard recognized?
2020-07-09 20:00:24 <Winst> no it doesn't come up in iwconfig
2020-07-09 20:04:00 <ikke> and neither does ip link?
2020-07-09 20:05:07 <Winst> nope just lo
2020-07-09 20:05:14 <ikke> did you use the standard or extended image?
2020-07-09 20:05:26 <ikke> So no ethernet port either?
2020-07-09 20:08:07 <Winst> standard yep and no ethernet either
2020-07-09 20:08:26 <ikke> Maybe you should try the extended image
2020-07-09 20:08:33 <ikke> it has more things by default
2020-07-09 20:16:41 <bovis> ikke Cogitri: My build worked with "options=textrels" added. Here's an APKBUILD linked to the Wine source I used here locally, if you'd like to try it for any reason: http://sprunge.us/Xov0Lc
2020-07-09 20:17:25 <bovis> Unfortunately I'm also seeing a problem now where, when I run "winecfg" for instance, or any other wine command, nothing happens. It's like hitting enter on a blank line in the console.
2020-07-09 20:17:39 <bovis> If that's more a winehq matter, I'll ask about it there.
2020-07-09 20:25:46 <Hello71> echo $?
2020-07-09 20:28:51 <bovis> 0
2020-07-09 20:44:30 <opal> how do i set up my credit card
2020-07-09 20:58:33 <c705> opal: talk to your bank
2020-07-09 20:59:03 <c705> new cards come with a bloody sticker on them instructing you on what to do
2020-07-09 21:49:03 <Thalia> Hey Guys, anybody ever experienced the Problem udhcpc no leasebug when installing?
2020-07-09 22:18:43 <AndyAndyBoBandy> Is virtualenv for py2 packaged in 3.12? Or something that provides pip for py2?
2020-07-09 22:28:11 <opal> pretty sure most/all py2 stuff is being pruned out of repos
2020-07-09 22:28:42 <maxice8> py-pip isn't available?
2020-07-09 22:43:05 <AndyAndyBoBandy> maxice8: doesn't seem so. python2 is, but not supporting stuff beyond setuptools. But using py2-setuptools' easy_install to get virtualenv yields a broken virtualenv tool.
2020-07-09 22:44:45 <Hello71> python2 -m ensurepip
2020-07-09 22:48:27 <AndyAndyBoBandy> then using that pip to install virtualenv will work better, you think?
2020-07-09 22:58:04 <Hello71> yes
2020-07-09 22:58:15 <Hello71> assuming ensurepip actually works on musl
2020-07-09 22:59:21 <c705> virtualenv is kind of old now. if you're using python3 (you should unless you're maintaining legacy), then python3 -mvenv <path> is the way to go
2020-07-09 23:02:21 <AndyAndyBoBandy> excellent, thanks, I think it's working. It's not so much that I am/want to use py2, but I'm generating containers to demo some tooling, and that tooling has some py2 support, and I want the containers to have proper instructions.
2020-07-09 23:03:00 <c705> understood
2020-07-10 06:40:47 <lalo-salamanca> Ok. So I tried to debug my problem.
2020-07-10 06:41:20 <lalo-salamanca> When I try export FLASK_APP=myapp.py
2020-07-10 06:41:34 <lalo-salamanca> then flask run
2020-07-10 06:41:53 <lalo-salamanca> Server runs on 127.0.0.1:5000.
2020-07-10 06:42:16 <lalo-salamanca> curl 127.0.0.1:5000 gives my index page
2020-07-10 06:42:28 <lalo-salamanca> But same thing with uwsgi doesn't
2020-07-10 06:43:27 <lalo-salamanca> There are so many packages on uwsgi here. Please tell me which to install for core server
2020-07-10 06:44:41 <lalo-salamanca> https://stackoverflow.com/questions/62814165
2020-07-10 12:16:27 <albr> MrPotatoBob[d]: Anyone know how I can add a user with the usal permessions?
2020-07-10 12:16:31 <albr> MrPotatoBob[d]: Usual*
2020-07-10 12:18:11 <albr> unowe[d]: https://wiki.alpinelinux.org/wiki/Setting_up_a_new_user
2020-07-10 12:18:12 <albr> unowe[d]: you can also see the groups here you can add to the user
2020-07-10 12:20:06 <albr> MrPotatoBob[d]: Thanka
2020-07-10 12:20:09 <albr> MrPotatoBob[d]: Thanks
2020-07-10 13:50:23 <telmich> Is there any font package in alpine that supports emoji? I checked for the ones seen on https://wiki.archlinux.org/index.php/fonts#Emoji_and_symbols but apk search did not find any
2020-07-10 13:51:12 <ikke> telmich: https://pkgs.alpinelinux.org/package/edge/community/armv7/font-noto-emoji ?
2020-07-10 13:56:11 <telmich> err.. I was blind, thanks, @ikke
2020-07-10 14:00:11 <webstrand> for subpackages, is there any way to set subpackage specific variables, such as replaces or provides?
2020-07-10 14:02:58 <Cogitri> Yes, just define them in the subpackage's function
2020-07-10 14:03:16 <webstrand> oh, neat. thanks
2020-07-10 14:18:10 <yunfan> mps: i just installed alpine to my sata disk for the pc
2020-07-10 14:18:26 <yunfan> i think the setup script should be adjust for it use the whole disk currently
2020-07-10 14:18:51 <yunfan> i knew this before, so i had choose a second disk, but for others , its quite incovenient
2020-07-10 14:35:57 <webstrand> Hmm. I created a package `busybox-pam` which has provides="busybox-1.32.0"
2020-07-10 14:36:01 <webstrand> It replaces base busybox perfectly, but running `apk add busybox` doesn't let me switch to the non-pam version.
2020-07-10 14:36:46 <ikke> I think you need to remove busybox-pam again
2020-07-10 14:36:47 <webstrand> is this fixable? Or would base busybox need a different package name to be selectable?
2020-07-10 14:36:54 <ikke> does that work?
2020-07-10 14:37:01 <ikke> apk del busybox-pam?
2020-07-10 14:37:16 <ikke> (only do it on a trashable system :))
2020-07-10 14:37:42 <webstrand> oops :)
2020-07-10 14:38:03 <webstrand> now to figure out how to recover
2020-07-10 14:38:21 <ikke> apk add busybox?
2020-07-10 14:38:42 <afontain_> lol
2020-07-10 14:39:02 <Cogitri> Fwiw you probably want "provides=busybox-$pkgver=r$pkgrel"
2020-07-10 14:39:05 <webstrand> that results in an error. Looks like tar still works
2020-07-10 14:42:03 <Hello71> if you still have ash open can't you just do cp /proc/self/exe /bin/busybox
2020-07-10 14:42:23 <Hello71> busybox --install
2020-07-10 14:42:23 <webstrand> yep. I also had a copy in the pkg dir
2020-07-10 14:42:47 <webstrand> Is there a way to print the "1 error" from apk? It's still failing to install busybox
2020-07-10 14:58:39 <webstrand> ah, I just needed to reinstall the version of busybox in the repos.
2020-07-10 14:59:08 <ikke> apk fix
2020-07-10 14:59:57 <webstrand> apk fix was producing this error: (1/1) [APK unavailable, skipped] Reinstalling busybox (1.32.0-r0)
2020-07-10 15:00:23 <webstrand> I'm not sure why. busybox was already installed. downgrading to the repo version cleared the error, though.
2020-07-10 15:02:11 <wago> anyone use Widevine plugin for Firefox? I'm getting segfault on /usr/lib/firefox/libxul.so. i'm running Alpine Linux edge, linux 5.7.8 and gnome 3.36.4
2020-07-10 15:05:09 <webstrand> It looks like, when I install busybox-pam, a virtual busybox package gets installed along with it? Which is why `apk add busybox` doesn't work.
2020-07-10 15:05:37 <webstrand> But why can I purge busybox-pam, the only provide for the virtual busybox package?
2020-07-10 15:05:48 <webstrand> it should produce conflicts
2020-07-10 15:06:22 <Cogitri> wago: I don't think widevine works on musl, or does it?
2020-07-10 15:09:32 <wago> Cogitri: not sure? I'm new to Alpine. It seems like it does not work, but I thought i would check here.
2020-07-10 15:09:59 <howudodat> quick question.  everytime I reboot my alpine server it re-writes /etc/network/interfaces.  I have commented out one of my bond interfaces to do some testing and at reboot, it's uncommented again.  ideas?
2020-07-10 15:10:37 <Cogitri> wago: To be honest, I haven't tested it since I just didn't expect it to work from the start, so maybe I'm wrong about it
2020-07-10 15:10:57 <Cogitri> But FF without the widevine plugin certainly works fine on GNOME for me
2020-07-10 15:13:54 <wago> that's true, i haven't seen any other issue with FF yet. I have a DAC hooked up over USB to the Alpine laptop and was using it for Tidal music (before I switched to Alpine).
2020-07-10 15:16:05 <Cogitri> Huh, what a coincidence, I started using Tidal yesterday :)
2020-07-10 15:16:28 <Cogitri> I use it via MellowPlayer's flatpak and Widevine works great in that
2020-07-10 15:17:28 <wago> on Alpine?
2020-07-10 15:20:57 <wago> I've been using Tidal since 2015. I rationalize the subscription as it being "my" constantly updated music collection. but if they vanish in the middle of the night, there goes the collection  :) anyway. i will check mellowplayer out.
2020-07-10 15:21:21 <Cogitri> Yes
2020-07-10 15:21:58 <Cogitri> See https://mellowplayer.readthedocs.io/en/stable/users/install.html#flatpak and the other widevine support section in there
2020-07-10 15:23:56 <wago> hmm. i see that. thanks for the pointer, i will check it out.
2020-07-10 15:32:53 <ikke> howudodat: disk install or disk less?
2020-07-10 15:43:21 <wago> it's interesting that widevine works with mellowplayer. MellowPlayer is looking for kde, and i only have gnome installed on that laptop. I will maybe try it later when i have more time. it looks like libwidevinecdm.so references strtoll_ and strtoull_l which are in glibc but not in musl. if i had the source it would be an easy fix i think, but it's a closed source DRM thing from google. :)
2020-07-10 15:43:34 <mps> yunfan: yes, but alpine is 'simple, small' and setup-alpine is intended for small systems, and to be simple
2020-07-10 15:44:04 <mps> yunfan: so, simple things are easy and complicated things are possible
2020-07-10 15:44:43 <howudodat> ikke: diskinstall
2020-07-10 15:45:15 <ikke> howudodat: strange, I cannot think of anything out-of-the-box that would overwrite /etc/network/interfaces
2020-07-10 15:45:35 <howudodat> https://pastebin.com/YdHyVFey
2020-07-10 15:45:52 <ikke> modloop
2020-07-10 15:45:56 <ikke> so that indicated run-from-ram
2020-07-10 15:47:30 <howudodat> yeah, this is interesting (it's not my server, helping someone with an issue).   I dont see a mountpoint for /
2020-07-10 15:47:58 <howudodat> I'm wondering if it's booting from usb?
2020-07-10 15:48:05 <ikke> yes, that's most likely
2020-07-10 15:48:32 <ikke> /media/usb is present
2020-07-10 15:48:53 <howudodat> yes it is present, and tmpfs is mounted on /
2020-07-10 15:49:10 <howudodat> https://pastebin.com/cuhBWdhj
2020-07-10 15:49:10 <ikke> so yes, it's run-from-ram / live usb
2020-07-10 15:49:28 <ikke> so that explains why your change is not persisted
2020-07-10 15:50:32 <Cogitri> wago: I feel like it's fine to use MellowPlayer in GNOME too, it integrates well with the desktop with MPRIS
2020-07-10 15:50:44 <Cogitri> I use it all the time for playing music
2020-07-10 15:51:46 <howudodat> ikke, the usb has a couple of apk's then /apks /boot /cache ldlinux.c32 and ldlinux.sys  I dont spot anything that looks like a rootfs or expandable image
2020-07-10 15:52:00 <ikke> is there an apkovl file?
2020-07-10 15:54:43 <howudodat> ikke: no
2020-07-10 15:54:49 <howudodat> ls -l
2020-07-10 15:54:56 <howudodat> (hah, oops sorry)
2020-07-10 15:55:10 <ikke> ok
2020-07-10 15:56:01 <howudodat> ikke: I might have found one
2020-07-10 15:56:10 <howudodat> how do I know if this is getting extracted?
2020-07-10 15:56:10 <ikke> This looks like hybrid setup
2020-07-10 15:56:17 <ikke> try lbu status
2020-07-10 15:56:35 <ikke> /var is mounted on disk, rest is in ram
2020-07-10 15:56:55 <howudodat> U etc/network/interfaces
2020-07-10 15:57:00 <ikke> yes
2020-07-10 15:57:13 <ikke> so when you make changes, you need to run lbu ci (lbu commit) to commit them
2020-07-10 15:58:24 <ikke> lbu commit updates the apkovl file
2020-07-10 15:58:29 <ikke> so the next time you boot, it gets applied again
2020-07-10 16:00:28 <yunfan> mps: i hope it could use only 1 partitions :D
2020-07-10 16:01:39 <howudodat> ikke: cool, I am rebooting now
2020-07-10 16:03:38 <mps> yunfan: boot partition is always good to have
2020-07-10 16:04:14 <eydaimon> Wow, it's so much easier to setup alpine with ZFS encrypted on root now. I'm going to do  it a second time and update the documentation.
2020-07-10 16:04:21 <eydaimon> Hello71: ☝️  thank you so much
2020-07-10 16:04:23 <yunfan> but the script dont provide the options, looks like i had to read the code of it
2020-07-10 16:04:30 <mps> export SWAP_SIZE=0 will skip create swap partition
2020-07-10 16:05:39 <mps> yunfan: and actually it is possible to install all on one partition, it is explained somewhere on wiki
2020-07-10 16:05:48 <yunfan> mps: also i had installed X and i3wm too, and got firefox run perfectly
2020-07-10 16:05:55 <yunfan> it only lack of my chinese input method
2020-07-10 16:06:05 <yunfan> but for non-chinese, i think its perfect
2020-07-10 16:06:20 <yunfan> mps: aha, will check it
2020-07-10 16:06:46 <mps> nothing is perfect, especially 'crap' called computers/software
2020-07-10 16:07:43 <mps> yunfan: this https://wiki.alpinelinux.org/wiki/Setting_up_disks_manually
2020-07-10 16:07:52 <yunfan> mps: thx
2020-07-10 16:08:57 <mps> it is not out-of-the-box working guide but by reading this you will know what are steps
2020-07-10 16:09:06 <yunfan> mps: these partitions job is easy to me, i think the most headache issue is install for uefi support
2020-07-10 16:09:43 <mps> yunfan: yes, uefi is idiotic invention
2020-07-10 16:10:11 <eydaimon> there's no chsh command for alpine I notice. How are users changing shell ?
2020-07-10 16:10:12 <yunfan> mps: ok . i could use it at my own risk, its another harddisk i founded in my parent's garbage box :D
2020-07-10 16:10:36 <yunfan> eydaimon: editing /etc/passwd ?
2020-07-10 16:10:39 <mps> eydaimon: apk add shadow
2020-07-10 16:11:08 <mps> but better is that what yunfan proposed :)
2020-07-10 16:12:20 <eydaimon> Yes, but I asked how users change shell. Only root can edit /etc/passwd :)
2020-07-10 16:12:26 <eydaimon> I tried to be specific 😅
2020-07-10 16:12:42 <eydaimon> I fail
2020-07-10 16:12:49 <mps> yunfan: re Chinese input is the real issue on alpine, I agree
2020-07-10 16:13:14 <mps> eydaimon: shadow pkg contains chsh
2020-07-10 16:13:15 <eydaimon> mps: thank you for shadow :)
2020-07-10 16:14:35 <mps> one can search for commands and in which pkg they are by 'apk search cmd:chsh' for example
2020-07-10 16:14:50 <yunfan> mps: i am trying to solve that from other side
2020-07-10 16:15:45 <yunfan> mps: there were firefox extentions to change textarea tag into  a vim/emacs invoking which i think could be study to change to a js input method
2020-07-10 16:16:01 <mps> yunfan: sorry I don't have any idea how input for Chinese and similar scripts/langs works at all
2020-07-10 16:16:24 <yunfan> mps: its not a problem to me, because the display works :D
2020-07-10 16:17:07 <yunfan> i wrote all my code comments in english since the broken support of input method from a decade
2020-07-10 16:17:54 <howudodat> ikke: that worked thanks, I have added that to my documenation
2020-07-10 16:19:04 <mps> yunfan: I've seen some effort to add Chinese input to alpine but didn't touched it or even read it
2020-07-10 16:19:46 <mps> I dislike to read software which I don't understand (if I understand anything :) )
2020-07-10 17:07:15 <eydaimon> "The researchers also found that none of those passengers in the two buses who wore face masks were infected.
2020-07-10 17:07:17 <eydaimon> They said it vindicated the decision to ask people to wear a face mask in public."
2020-07-10 17:07:19 <eydaimon> oops
2020-07-10 17:07:40 <eydaimon> sorry 😅
2020-07-10 17:07:52 <eydaimon> I meant to paste this:  df: /sys/kernel/debug/tracing: Permission denied
2020-07-10 17:08:02 <eydaimon> why do I get this when running df ?
2020-07-10 17:23:09 <ikke> eydaimon: what environment?
2020-07-10 17:24:04 <eydaimon> ikke: regular user: https://i.postimg.cc/KYNqtjQ0/Screen-Shot-2020-07-10-at-10-06-24.png
2020-07-10 17:24:17 <eydaimon> ikke: after regular boot
2020-07-10 17:25:06 <eydaimon> drwx------   29 root     root             0 Jul 10 09:02 /sys/kernel/debug/
2020-07-10 17:25:28 <eydaimon> ikke: install via extended iso
2020-07-10 17:25:45 <ikke> dr-xr-xr-x    2 root     root             0 Jul  4 09:41 /sys/kernel/debug
2020-07-10 17:25:45 <eydaimon> fresh install. just did it today
2020-07-10 17:26:15 <eydaimon> using this image: alpine-extended-3.12.0-x86_64.iso
2020-07-10 17:26:31 <eydaimon> oh I think I know why
2020-07-10 17:26:41 <eydaimon> I'll reinstall and see if it happens again
2020-07-10 18:13:45 <Hello71> that's normal
2020-07-10 18:14:00 <Hello71> er
2020-07-10 18:47:01 <amazing1> hi, im running alpine within docker and i'm trying to play some sound. root can play sounds fine. regular user cannot.
2020-07-10 18:47:09 <amazing1> i added it to the audio group
2020-07-10 18:47:50 <amazing1> getting error: Can't open default sound device! and error: Found no driver out of [alsa,oss] working with device <default>.
2020-07-10 18:50:06 <amazing1> if anyone has an idea..
2020-07-10 19:18:33 <Hello71> did you pass through /dev/snd
2020-07-10 19:18:41 <Hello71> also dmix won't work
2020-07-10 19:18:45 <Hello71> nor pulse
2020-07-10 19:28:22 <amazing1> yeah, root inside the container can play sound.
2020-07-10 19:30:20 <amazing1> i guess ill just run stuff as root there -_-
2020-07-10 19:30:24 <amazing1> thanks anyway
2020-07-10 19:31:02 <c705> bad idea, but do what you will
2020-07-10 19:32:18 <amazing1> it's in a container that's not exposed to the internet so yeah i think its fine
2020-07-10 19:32:59 <Hello71> yes, uaccess won't work in a container unless you run udev
2020-07-10 19:32:59 <amazing1> i mean its better as a regular user sure, but considering i cant get it to work with that...
2020-07-10 19:33:01 <c705> if you say so
2020-07-10 19:33:17 <amazing1> Hello71 i mounted /dev/snd under devices
2020-07-10 19:33:24 <c705> so put some effort into getting it to work instead of giving up and exposing a secrity hole
2020-07-10 19:33:37 <amazing1> c705 how is this a security hole
2020-07-10 19:34:01 <Hello71> fwiw it's pretty easy to get out of a container as root
2020-07-10 19:34:08 <Hello71> otoh it's not that hard to get out as non-root either
2020-07-10 19:34:19 <c705> Hello71: trivial, even
2020-07-10 19:34:25 <c705> root in the container is root out of the container
2020-07-10 19:34:28 <amazing1> if you have an idea how to get it to work, i'd love to hear you.
2020-07-10 19:35:00 <c705> https://docs.docker.com/engine/security/security/
2020-07-10 19:44:11 <amazing1> there's no need to convince me it's better off running as a normal user.
2020-07-10 19:45:47 <Hello71> did you try just chown -R user:user /dev/snd
2020-07-10 19:47:04 <amazing1> that'll prevent other apps from accessing it.
2020-07-10 19:53:52 <amazing1> with help of #docker i solved it. it a docker related issue not alpine.
2020-07-10 19:53:59 <amazing1> sorry for bother you and thanks for the help
2020-07-10 19:54:08 <amazing1> to*
2020-07-10 19:57:27 <Hello71> yes
2020-07-10 19:57:38 <Hello71> as I said, uaccess requires udev
2020-07-10 21:34:03 <awilfox> my alpine VM seems to have broken itself.  booting up I get:
2020-07-10 21:34:27 <awilfox> mount: mount /dev/sda2 on /sysroot failed: No such file or directory
2020-07-10 21:34:54 <awilfox> it drops me to initramfs emergency shell, which shows /dev/sda2 and /sysroot both exist
2020-07-10 21:35:07 <ikke> and if you run mount -a there?
2020-07-10 21:35:23 <awilfox> nothing happens.  no error, but /sysroot is not mounted
2020-07-10 21:35:37 <ikke> and mount /dev/sda2 /sysroot?
2020-07-10 21:35:43 <awilfox> and running # mount /dev/sda2 /sysroot    still yields No such file or directory
2020-07-10 21:36:24 <awilfox> https://i.imgur.com/7BPfptz.png
2020-07-10 21:37:08 <Hello71> dmesg
2020-07-10 21:38:32 <Hello71> mount errors get dumped in dmesg, because... ugh
2020-07-10 21:39:00 <Cogitri> Maybe you don't have the required helper binary for your FS installed
2020-07-10 21:39:43 <awilfox> dmesg is clear
2020-07-10 21:40:35 <awilfox> ugh... ext4fs isn't built in to the kernel? :/  it doesn't look like the module is in the initramfs.
2020-07-10 21:41:06 <Cogitri> Hm, mkinitfs has the ext4 feature enabled by default IIRC
2020-07-10 21:47:27 <Hello71> I wish apk had at least basic config file determination (don't keep apk-new if the new file is identical to the old base)
2020-07-10 21:47:31 <mps> that is why I set ext4 and f2fs =y in linux-edge for armv7 and aarch64
2020-07-10 21:48:50 <mps> thinking to do same for x86_64
2020-07-10 21:49:35 <awilfox> please do.  idk how this even happened, it was running fine, then I upgraded 3.11.2 to 3.11.6, rebooted, and boom
2020-07-10 21:50:07 <awilfox> it wasn't even 3.11->3.12, which I was afraid of doing for this reason...
2020-07-10 21:50:28 <awilfox> (I was bitten by some LVM bug when upgrading a cluster from, iirc, 3.5 to 3.6)
2020-07-10 21:50:50 <mps> awilfox: linux-edge is in testing and is not used by install media or script
2020-07-10 21:51:00 <awilfox> oh.
2020-07-10 21:51:19 <mps> linux-lts is in main and it is used
2020-07-10 21:51:49 <awilfox> I think I will just boot a distro kernel with -kernel option in qemu, to try and recover this VM
2020-07-10 21:51:53 <mps> though I agree with youj=
2020-07-10 21:51:57 <mps> uhm
2020-07-10 21:52:00 <mps> with you
2020-07-10 21:52:29 <mps> most used FS drivers should be =y in kernel config
2020-07-10 21:53:28 <awilfox> haha, that is a weird prompt:  Alpine Linux 3.11.6 Kernel 5.4.5-mc0-adelie-easy on an x86_64 (/dev/tty1)
2020-07-10 21:53:49 <mps> :)
2020-07-10 21:54:04 <awilfox> looks like that did get me in though.  now I should be able to fix up mkinitfs config.
2020-07-10 21:58:09 <awilfox> that worked-ish... I had to manually modprobe ext4 in the initramfs shell, but after that it worked fine
2020-07-10 21:59:35 <awilfox> going to try upgrade to 3.12 and see if that helps
2020-07-10 22:00:45 <mps> if you ext4 in /etc/mkinitfs/mkinitfs.conf it should work
2020-07-10 22:00:55 <mps> fetures, I mean
2020-07-10 22:01:07 <awilfox> I did.
2020-07-10 22:01:17 <mps> but I know you know this better than me :)
2020-07-10 22:01:35 <awilfox> it did not help :/  still drops to emergency shell and requires me to modprobe ext4
2020-07-10 22:01:39 <awilfox> on 3.12 now
2020-07-10 22:01:51 <mps> uhm
2020-07-10 22:02:22 <mps> is the linux-lts upgraded during upgrade
2020-07-10 22:02:36 <awilfox> yes.  and then mkinitfs trigger ran and generated new initramfs
2020-07-10 22:02:54 <mps> hm
2020-07-10 22:03:10 <mps> strange
2020-07-10 22:03:24 <awilfox> features="ata base ext4 keymap kms virtio"
2020-07-10 22:03:36 <mps> didn't had these issues in few last upgrade cycles
2020-07-10 22:05:01 <awilfox> I tried rerunning `mkinitfs` myself and now /dev/sda isn't even found \o/
2020-07-10 22:05:04 <mps> stupid qestion, but did the bootloader use upgraded kernel and initramfs
2020-07-10 22:05:42 <awilfox> it shows in dmesg as   scsi 0:0:0:0 Direct Access    ATA   QEMU HARDDISK
2020-07-10 22:05:45 <awilfox> but nothing about 'sda'
2020-07-10 22:06:22 <mps> maybe scsi need to be added to features
2020-07-10 22:06:28 <awilfox> yeah.  I looked, and there's only one (vmlinuz-lts and initramfs-lts)
2020-07-10 22:07:22 <mps> I 'ditched' my last x86_64 box where I had qemu's so can't check now
2020-07-10 22:07:48 <awilfox> mhm, I'm more "at home" on power hardware myself, but I need this x86_64 thinger for work
2020-07-10 22:09:02 <mps> ah, on one of my linode VM (qemu) i have this 'features="ata base ide scsi usb virtio ext4"'
2020-07-10 22:10:48 <mps> (now wonder why usb and and ide are there)
2020-07-10 22:13:42 <mps> cat /sys/block/sda/device/model => QEMU HARDDISK
2020-07-10 22:13:47 <awilfox> I didn't put usb but I put ide and scsi in and now it's back to normal
2020-07-10 22:13:51 <awilfox> that is, I have to modprobe ext4
2020-07-10 22:13:54 <awilfox> but everything else works
2020-07-10 22:15:35 <mps> that should be investigated
2020-07-10 22:16:08 <mps> out of curiosity, why do you don't use linux-virt kernel for VM
2020-07-10 22:18:38 <awilfox> looking at git log, wrong assumption: I thought only 'linux-lts' package would track 5.4 branch
2020-07-10 22:18:50 <awilfox> I do not trust linux branches that are not LTS any more, after so many data losses
2020-07-10 22:19:41 <mps> linux-virt is also 5.4 LTS
2020-07-10 22:20:02 <awilfox> yes I see that
2020-07-10 22:21:53 <mps> about data loss, ime there is no much differences in kernels I use
2020-07-10 22:45:44 <bovis> In a 32-bit chroot on this 64-bit laptop, running "wineserver" segfaults with this: https://bpa.st/OQKQ
2020-07-10 22:46:02 <bovis> Running the same version of wine 5.0.1 in the 64-bit natively, it runs fine.
2020-07-10 22:46:10 <bovis> Am I doing something wrong in my chroot?
2020-07-10 22:48:10 <Hello71> awilfox: I think you need to specify modules=ext4 on kernel command line
2020-07-10 22:48:22 <Hello71> or load-modules or something
2020-07-10 22:49:10 <Hello71> hm, not sure why the kernel doesn't do it. try checking sysctl kernel.modprobe
2020-07-10 22:52:06 <mps> Hello71: yes, I forgot this, 'modules=sd-mod,ext4 rootfstype=ext4'
2020-07-10 22:52:29 <Hello71> but also kernel should auto-probe the modules
2020-07-10 22:52:41 <Hello71> even without udev
2020-07-10 22:53:04 <mps> not sure
2020-07-10 22:55:03 <mps> last time I tested this without adding explicitly needed module in APEND  for boot loader kernel didn't loaded them
2020-07-10 23:05:40 <Hello71> maybe modprobe is not in /sbin in alpine initramfs
2020-07-11 03:05:55 <LACampbell> do people use something other than sudo on alpine technically? I always noticed it was a bit weird to set up a user account, what with sbin not being in the path etc
2020-07-11 03:33:35 <eydaimon> OK, here are the new instructions for encrypted ZFS on root with alpine. Anyone Want to check it out before I update the old wiki?    https://gist.github.com/danielb2/597c175145d6b1ee39207668807d2e3d
2020-07-11 03:33:40 <eydaimon> Cogitri: ☝️
2020-07-11 09:27:06 <hansili> hi, i'm running 3.12.0 diskless.  when i do  'rm /etc/runlevels/boot/hwclock ; lbu commit"  then the overlay archive does not contain that symlink as expected.  but when i reboot afterwards, that file shows up again.  why is that?
2020-07-11 09:35:25 <Cogitri> eydaimon: I think it'd be best to include the "Installing Alpine Linux" section should still contain that you need to enable openRC services, add features to mkinitfs.conf, set rootfstype and need to install the zfs package
2020-07-11 09:45:42 <ikke> hansili: it gets restored by apk
2020-07-11 09:46:07 <ikke> hansili: I'm not sure if lbu can explicitly commit a deletion of a file
2020-07-11 09:46:25 <ikke> most likely not
2020-07-11 09:47:01 <ikke> hmm
2020-07-11 09:47:22 <ikke> on the other hand, /etc/runlevels/boot/hwclock is not managed by apk
2020-07-11 09:52:25 <hansili> ikke: yes, and for /etc/runlevels/default it works just fine.
2020-07-11 10:29:58 <yunfan> ikke: why alpine extend version dont provide wireless-tools package default?
2020-07-11 10:30:12 <ikke> yunfan: I don't know
2020-07-11 10:32:51 <yunfan> the download page said it's Most common used packages included. Suitable for routers and servers.
2020-07-11 10:33:05 <yunfan> but for routers, i thinks wireless tools is a basic needs
2020-07-11 10:59:34 <hansili> ikke: which apk does include /etc/runlevels/boot/hwclock?  (i couldn't find it.)
2020-07-11 11:00:21 <mps> hansili: /etc/runlevels/boot/hwclock symlink target is owned by openrc-0.42.1-r10
2020-07-11 11:00:36 <mps>  apk info -W /etc/runlevels/boot/hwclock
2020-07-11 11:00:38 <ikke> the target, but not the symlink itself
2020-07-11 11:01:37 <hansili> mps and ikke: thanks
2020-07-11 11:01:45 <mps> ok, apk info -W /etc/init.d/hwclock
2020-07-11 12:05:54 <nponaraHAa> does muls benefit from LTO?
2020-07-11 12:17:02 <Cogitri> LTO isn't dependant on the libc
2020-07-11 12:47:43 <eydaimon> Cogitri: it boots fine without all of those. Also, I do add the zfs package here: https://gist.github.com/danielb2/597c175145d6b1ee39207668807d2e3d#add-needed-packages
2020-07-11 12:47:58 <eydaimon> Cogitri: since it boots fine, do I need the mkinitfs etc ?
2020-07-11 13:06:46 <eydaimon> what firewall do you ya'll use ?
2020-07-11 13:09:40 <albr_> unowe[d]: ufw gets the job done well]
2020-07-11 13:11:03 <ikke> awall
2020-07-11 13:11:47 <Cogitri> <eydaimon "Cogitri: it boots fine without a"> It boots fine, but zfs won't mount any volumes but your root volume without the OpenRC services
2020-07-11 13:12:02 <Cogitri> He next kernel upgrade your system won't boot without the zfs features
2020-07-11 13:12:19 <Cogitri> I use ufw as well, it's super easy to setup
2020-07-11 13:44:11 <eydaimon> ufw ?
2020-07-11 13:45:14 <eydaimon> Cogitri: you mean these things, right? rc-update add zfs-import sysinit rc-update add zfs-mount sysinit Edit the /etc/mkinitfs/mkinitfs.conf
2020-07-11 13:46:38 <Cogitri> Yup
2020-07-11 13:46:56 <Cogitri> https://pkgs.alpinelinux.org/package/edge/community/x86/ufw
2020-07-11 13:48:58 <eydaimon> Cool, thanks for pointing that out :)
2020-07-11 13:49:27 <eydaimon> thank you on suggesting ufw
2020-07-11 13:51:19 <Hello71_> firewall doesn't really improve security in most applications
2020-07-11 13:53:01 <Cogitri> Can be useful to avoid accidentally exposing services to the internet tho
2020-07-11 13:53:05 <eydaimon> not sure what you mean, but my NAS will be accessible to the outside. I only want to be able to SSH in and no one to be able to access my services from outside the network. What else woudl I use the firewall for ?
2020-07-11 13:55:29 <eydaimon> Cogitri: nice, ufw has the same syntax as ipfw (a FreeBSD firewall I use)
2020-07-11 13:55:42 <eydaimon> it's a winner :)
2020-07-11 13:58:27 <Hello71_> better to use bind addresses
2020-07-11 13:58:40 <eydaimon> Hello71_: tell me more :)
2020-07-11 14:03:58 <Cogitri> You could make services only listen on certain addresses, e.g. 127.0.0.1:port instead of letting it listen on 0.0.0.0 (so any address)
2020-07-11 14:04:51 <Cogitri> Or add two interfaces to your NAS and then have one go into your private net with all services listening on that and one interface that goes to your NAT and only sshd listens on that interface
2020-07-11 14:05:16 <Cogitri> But IMHO a firewall still doesn't hurt, if only to prevent you accidentally misconfiguring a service and as such leaving it exposed
2020-07-11 14:05:17 <eydaimon> Ah. I guess I could do that, but I don't know that all services allow that configuration. wildcarding is allowed so I can use internal network, right ?
2020-07-11 14:07:44 <Cogitri> In my case I only have sshd listening and nginx (which is a reverse proxy for everything else), so it wasn't hard to set that up
2020-07-11 14:09:45 <eydaimon> I have at least 6 I don't want to be exposed
2020-07-11 14:14:17 <adu> So I have a fork of aports, how do I add packages, do I add them to testing, commit, then make a merge request?
2020-07-11 14:16:32 <maxice8> yes
2020-07-11 14:38:00 <mps> Hello71_: but firewall is a 'good selling product' :)
2020-07-11 14:38:06 <Hello71_> what?
2020-07-11 14:38:48 <mps> you can get a lot of money by selling firewall
2020-07-11 14:45:58 <mps> jokes aside, firewall can be very useful, and it is
2020-07-11 14:47:58 <mps> ping (icmp echo) rate limiting, ssh access only from particular addres/networks/ports, auto block ssl/password crackers etc
2020-07-11 14:52:34 <u0jQx9gPyrYg> and also blocking various unexpected ports being open to the internet that you don't know about....
2020-07-11 14:53:03 <u0jQx9gPyrYg> it's called defense-in-depth
2020-07-11 14:59:19 <mps> yes, control outgoing traffic also, especially if there devices in local network which 'call home'
2020-07-11 15:11:16 <eydaimon> Cogitri: I updated the gist
2020-07-11 15:19:16 <eydaimon> how do I get info about a package with apk ?
2020-07-11 15:19:22 <eydaimon> like what's the home url etc
2020-07-11 15:19:50 <eydaimon> dero
2020-07-11 15:20:02 <eydaimon> apk info... I made a typo and typed zfs info 😂
2020-07-11 15:20:03 <u0jQx9gPyrYg> apk info $pkg
2020-07-11 16:36:28 <eydaimon> visual mode doesn't seem to work in neovim. vim shows it just fine. What could cause that ?
2020-07-11 16:50:32 <c705> seems like its neovim
2020-07-11 17:13:35 <eydaimon> but I have neovim running on macos and freebsd with no such issues
2020-07-11 17:13:47 <eydaimon> so maybe linux only problem ?
2020-07-11 17:59:01 <eydaimon> really liking alpine :)
2020-07-11 17:59:26 <eydaimon> Cogitri: you OK with me updating your doc as it stands with that update ?
2020-07-11 18:02:16 <Cogitri> Sure, looks good. Thanks for working on it
2020-07-11 18:20:28 <bovis> Using the chroot script on the alpine website (https://wiki.alpinelinux.org/wiki/Chroot#Performing_the_chroot), I'm unable to connect to dbus. What else should the script do to get it running? The error, for example, is here: https://bpa.st/V5FQ
2020-07-11 18:27:26 <Hello71_> connect to dbus... where?
2020-07-11 18:32:36 <bovis> Hello71_: I don't know what you're asking. Can you be more specific?
2020-07-11 18:45:38 <tartar> i need some help fixing ImportError: Error loading shared library ld-linux-armhf.so.3: No such file or directory. i installed libc6-compat.
2020-07-11 18:47:54 <eydaimon> Cogitri: I forgot I didn't follow the instructions. turns out that /etc/mkinitfs/mkinitfs.conf  already had zfs in it :) one less step to do. and sure, you're welcome
2020-07-11 19:04:27 <joe9> is there a telnet command on alpine?
2020-07-11 19:04:43 <joe9> I cannot find anything in the default for which telnet or netcat.
2020-07-11 19:05:08 <joe9> I just want to check if I can connect to a port on the same machine. Something simple.
2020-07-11 19:06:33 <ikke> openbsd-netcat
2020-07-11 19:21:13 <mps> joe9: busybox-extra have it, also latest screen
2020-07-11 19:23:59 <joe9> mps, ikke, thanks.
2020-07-11 19:47:07 <joe9> the busybox-extra telnet, can it read from stdin and send the commands to the telnett'ed port?
2020-07-11 19:47:24 <joe9> I am trying to figure out if I can use that telnet in a pipe
2020-07-11 19:48:03 <joe9> https://unix.stackexchange.com/questions/86556/testing-remote-tcp-port-using-telnet-by-running-a-one-line-command
2020-07-11 19:48:14 <mps> it is barebone telnet client, afaik. never tried anything except basic usage
2020-07-11 19:49:09 <mps> for this some of netcat version looks more apropriate
2020-07-11 19:49:12 <joe9> mps, I am trying to connect to a tcp server running on alpine. Something basic to test connectivity. The first command that the tcp server exists is this string
2020-07-11 19:49:23 <joe9> API\0\0\0v100..1
2020-07-11 19:49:41 <joe9> And, I cannot figure out how to send that with the telnet client on alpine.
2020-07-11 19:49:57 <joe9> mps, ok, thanks. will check netcat
2020-07-11 19:50:17 <mps> openbsd-netcat can be easily 'scripted'
2020-07-11 19:50:46 <joe9> ok, will try that. Thanks.
2020-07-11 20:42:03 <joe9> I need to send some NUL characters from a shell script, read -c /dev/zero. read on alpine does not have the -c. Just curious if there is any other tool that can do something similar.
2020-07-11 20:43:14 <joe9> got it, echo -e
2020-07-11 20:44:37 <mps> dd if=/dev/zero bs=1 count=1
2020-07-11 20:44:51 <joe9> mps, that works too. cool. Thanks.
2020-07-11 21:02:15 <helby> hey ladies, gentlemen and others
2020-07-11 21:03:00 <helby> tried to use efibootmgr while chrooted to my broken shit, but was not able to use it, got error 'no efivar in this filesystem' or something like that
2020-07-11 21:03:28 <helby> what could cause that? wrong mounting some parts? or is it normal behaviour in chroot
2020-07-11 21:04:04 <helby> I mounted just root and efi, did not touch any sys or other things
2020-07-11 21:04:53 <helby> I already reinstalled, whole issue was with grub obviously bc. that crap is useless
2020-07-11 21:06:02 <helby> also found one interesting things ... default install somehow put 'acct' accounting package in world and install it :) not sure in what step it come there
2020-07-11 21:32:54 <TBB> helby, if I remember correctly (and it's been a while since I've done UEFI stuff on Alpine) there's a specific erivarfs mount that needs to exist to gain access to efi variables; it can be dangerous to your system so it's often mounted read-only
2020-07-11 21:34:25 <helby> TBB: thought something like that, thanks
2020-07-11 21:35:44 <TBB> (why I wrote erivarfs instead of efivarfs, I have no idea, must be I should be asleep already)
2020-07-11 21:36:17 <helby> I did not notice ;)
2020-07-11 21:36:40 <TBB> ah, you read around typos without noticing, I do the same thing :)
2020-07-11 21:36:55 <helby> that 'chrooting' then reinstalling took a lot of energy from me :d
2020-07-11 21:37:19 <helby> I tried to save some work, but then deciced just backup home and install crap again :)
2020-07-11 21:38:01 <helby> now thinking to get rid of grub,  but ...
2020-07-11 21:38:38 <helby> first need to change /boot/efi nonsense to just /boot but I am afraid fstab will be broken then
2020-07-11 21:38:58 <helby> also not sure what will put vmlinuz and other crap inside of /boot
2020-07-11 21:39:24 <helby> mkinitfs will do?
2020-07-11 21:42:46 <helby> for fuck sake if wiki is not dump
2020-07-11 21:44:36 <TBB> usually /boot yes, and whatever does the early boot into /boot/efi, that's pretty much how the spec does it. but like I said I haven't been doing much of UEFI+Alpine lately, it's been one and a half years since the last time
2020-07-11 21:46:09 <helby> what else if not uefi
2020-07-11 21:46:14 <helby> it's 2020 :)
2020-07-11 21:47:03 <TBB> well, my work laptop is uefi but it runs Ubuntu for work reasons, and my other system, also an Ubuntu, is UEFI without Secure Boot so I don't even know exactly how it boots
2020-07-11 21:49:11 <helby> TBB probbly like my with shitty grub
2020-07-11 21:49:21 <helby> I turned off secure boot to, as it was useless
2020-07-11 21:49:38 <helby> don't tell me about ubuntu, have one mail server with that crap, it's ...
2020-07-11 22:11:39 <TBB> I'm not sure about this one, but basically you could just use the kernel as the boot binary and provide it with parameters via efi boot params
2020-07-11 22:12:24 <TBB> and if you're feeling really courageous, you can combine the kernel and the initramfs into one. but that probably goes beyond what mkinitfs can handle
2020-07-11 22:18:57 <helby> TBB yes, I will try, maybe tommorow, I used just direct boot for years with any distro
2020-07-11 22:19:15 <helby> TBB: see fucking bootloader as something not necessary at all
2020-07-11 23:57:36 <joe9> is there any place I can find more details on what these options do? od [-aBbcDdeFfHhIiLlOovXx] [-t TYPE ] [ FILE ]
2020-07-11 23:58:35 <joe9> got it.
2020-07-12 00:36:18 <albr_> R&[d]: When installing apks within a chroot, how would I get them to install in the main environment instead of the temp sandbox?
2020-07-12 01:06:18 <bovis> albr_: Sorry, I don't have an answer to your question. But I'm curious about how you enter the chroot. Do you use a script? I'm just trying to diagnose issues with my own chroot, and a comparison could be helpful.
2020-07-12 01:16:31 <albr_> R&[d]: Yes I used a slightly modified script from https://github.com/alpinelinux/alpine-chroot-install/blob/master/alpine-chroot-install
2020-07-12 01:19:39 <bovis> albr_: I can't get my chroot to recognize dbus. Do you do anything in particular in your script to make dbus work?
2020-07-12 01:27:01 <albr_> R&[d]: I don't have any idea what dbus refers to on a Linux platform, but I read your earlier question. What does 'dbus start' do?
2020-07-12 01:44:54 <bovis> albr_: "dbus start" returns 'not found', but running "dbus-launch" gives me this successful output: https://bpa.st/4GBA
2020-07-12 01:45:39 <bovis> Running an application to access dbus *after* running dbus-launch, though, doesn't help. I still see an error that it fails to connect to a dbus session.
2020-07-12 05:00:42 <amom> I get a segmentation fault when running git log
2020-07-12 05:00:48 <amom> Im running aarch64.
2020-07-12 05:01:07 <amom> I have tried compiling git using aports. It passed the tests fine.
2020-07-12 05:01:22 <amom> I then ran `doas apk fix -d --repository /home/anjan/packages/main/ git`
2020-07-12 05:01:29 <amom> still having the segfault
2020-07-12 05:01:45 <amom> I also have the segfault when running other git commands (git status, etc.)
2020-07-12 05:01:54 <amom> any help?
2020-07-12 05:10:14 <adu> interesting
2020-07-12 06:27:10 <Cogitri> amom: https://gitlab.alpinelinux.org/alpine/aports/-/issues/11717
2020-07-12 06:27:25 <Cogitri> Seems like it only happens on aarch64 though, never had git crash for me on x86_64
2020-07-12 06:27:51 <Cogitri> Could you please install git-dbg and musl-dbg and then get a stacktrace of git crashing?
2020-07-12 06:28:28 <amom> Cogitri: sure give me a second
2020-07-12 06:30:52 <amom> Im just happy someone else has this bug. It was driving me up a wall
2020-07-12 06:38:17 <amom> Cogitri: sorry, Im a linux brainlet. How do I cat all text output to a file?
2020-07-12 06:38:24 <amom> it's not catting some of the text
2020-07-12 06:38:27 <amom> I tried > and 2>
2020-07-12 06:39:07 <afontain_> > file 2>&1
2020-07-12 06:39:23 <afontain_> the second part means "and two will send to the same place as 1"
2020-07-12 06:39:31 <afontain_> it must come afterwards
2020-07-12 06:39:39 <amom> ohhhh
2020-07-12 06:39:42 <amom> thank you afontain_
2020-07-12 06:40:11 <afontain_> the bash shortcurt is &> btw
2020-07-12 06:40:17 <afontain_> (there's |& too)
2020-07-12 06:40:34 <amom> thanks. Im gonna add all that to my personal notes
2020-07-12 06:40:37 <amom> super helpful
2020-07-12 06:40:46 <afontain_> (which is different from &|)
2020-07-12 06:43:54 <amom> Cogitri: https://paste.sr.ht/~anjan/ec77277704f6dd61fc7396b4cc5c8812090b0cce
2020-07-12 06:45:09 <afontain_> amom, cogitri: https://gitlab.alpinelinux.org/alpine/aports/-/issues/11717
2020-07-12 06:48:30 <Cogitri> Uh, that's the strace, I wanted the backtrace (so attach gdb to the process)
2020-07-12 06:50:49 <amom> Cogitri: sorry. Im not familiar with that
2020-07-12 06:51:30 <ikke> gdb --args git log
2020-07-12 06:51:31 <ikke> run
2020-07-12 06:52:37 <amom> I installed the git-dbg package
2020-07-12 06:52:45 <amom> but i still get (No debugging symbols found in git)
2020-07-12 06:54:14 <amom> https://paste.sr.ht/~anjan/1c2ef470dc16e87c1ce3960b0fa508c713f5908a
2020-07-12 06:54:23 <amom> that is the backtrace I did anyways...
2020-07-12 06:54:26 <amom> Cogitri:
2020-07-12 06:55:40 <Cogitri> Not really useful without the debugging info...hmm, why does it not find it
2020-07-12 06:59:36 <amom> Cogitri: Im gonna watch a movie, feel free to ping me. git-send-email is being buggy for me as a result of this i think and it's annoying for the project maintainer.
2020-07-12 07:02:19 <Cogitri> I'll see if I can reproduce this later today
2020-07-12 07:22:02 <ikke> amom: is this happening with a specific repository only, or in general?
2020-07-12 07:22:03 <mps> amom: Cogitri: git works fine on my few aarch64 installation, 3.12 and edge, native install and in lxc
2020-07-12 07:26:44 <nsz> at least disas around $pc and print the regs
2020-07-12 07:27:32 <nsz> or provide the binary so somebody else can look at the instructions
2020-07-12 07:28:12 <nsz> although  we would need a info proc map at least
2020-07-12 07:29:32 <amom> ikke: I am using alpine edge
2020-07-12 07:29:42 <amom> aarch64
2020-07-12 07:29:56 <amom> I cannot use stable because Im using sxmo on pinephone
2020-07-12 07:30:02 <ikke> amom: does it happen with git for alpine aports as well, or only self-built version?
2020-07-12 07:30:49 <ikke> amom: and with repository, I meant git repository
2020-07-12 07:32:22 <amom> ikke: I have tried compiling git from source using aports
2020-07-12 07:32:27 <mps> only once noticed problem but that was with not full upgraded system
2020-07-12 07:32:32 <amom> and installing from alpine's repos
2020-07-12 07:33:00 <amom> git passes all checks when compiling from source btw
2020-07-12 07:36:01 <ikke> So that means this issue doesn't happen in general
2020-07-12 07:36:13 <ikke> Did you try in another git repo?
2020-07-12 07:36:30 <afontain_> In the issue I linked above, it appeared after quite a lot of work
2020-07-12 07:36:41 <afontain_> like "print every commit in aports"
2020-07-12 07:36:52 <afontain_> but it's a bit random
2020-07-12 07:36:58 <afontain_> I could not reproduce btw
2020-07-12 07:41:14 <amom> ikke: yes, ive tried it in another git repo
2020-07-12 07:41:29 <amom> afontain_: certain git commands dont work no matter the repo or how many times I try
2020-07-12 07:41:43 <amom> ik for some ppl it sometimes segfaults and other times it doesnt
2020-07-12 07:42:03 <afontain_> hmm
2020-07-12 07:46:43 <nsz> so run it under gdb and print out the state at the segfault
2020-07-12 09:01:44 <opal> h
2020-07-12 09:02:10 <opal> git crashing for someone? odd
2020-07-12 09:10:24 <ikke> Yup
2020-07-12 10:44:53 <Hello71> amom: what is the rest of the gdb output
2020-07-12 10:45:42 <amom> Hello71: that's literally all the gdb output
2020-07-12 10:45:55 <Hello71> these addresses look kinda funny, but I guess that's how arm works?
2020-07-12 10:46:17 <Hello71> amom: I mean going up, not down
2020-07-12 10:46:53 <amom> ya, going up too
2020-07-12 10:54:24 <Hello71> so... you don't have GNU gdb (GDB) 9.2, Copyright (C) 2020 Free Software Foundation, Inc., License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>...
2020-07-12 10:55:51 <amom> ahttps://paste.sr.ht/~anjan/a76cf4deb9b727b53577be98f32127c314806b1a
2020-07-12 10:56:03 <amom> https://paste.sr.ht/~anjan/a76cf4deb9b727b53577be98f32127c314806b1a
2020-07-12 10:56:40 <opal> Hello71 you dont invoke with `gdb -q`?
2020-07-12 10:56:42 <opal> :^)
2020-07-12 10:57:55 <Hello71> amom: do you have /usr/lib/debug/usr/bin/git.debug?
2020-07-12 10:58:17 <amom> Hello71: yes
2020-07-12 10:58:49 <Hello71> try gdb /usr/libexec/git-core/git-log, and then symbol-file /usr/lib/debug/usr/libexec/git-core/git-log.debug
2020-07-12 10:59:46 <amom> Hello71: https://paste.sr.ht/~anjan/b2f31e458f701dbe9264ed3558d292aa4b89479e
2020-07-12 11:05:44 <Hello71> strange but alright
2020-07-12 11:06:00 <Hello71> Cogitri, nsz: ^
2020-07-12 11:07:44 <ikke> So something to do with notes
2020-07-12 11:07:50 <ikke> amom: does that repo has any notes recorded?
2020-07-12 11:08:45 <Hello71> also what version of alpine are youusing
2020-07-12 11:12:16 <ikke> Hello71: "I'm using alpine edge"
2020-07-12 11:12:34 <Hello71> missed it, only saw mps
2020-07-12 11:13:37 <Cogitri> Good that you have a backtrace now, I guess it's time to report it usptream then
2020-07-12 11:13:46 <Cogitri> Unfortunately I'm too busy to do that
2020-07-12 11:14:19 <amom> ikke: this happens on any repo I use
2020-07-12 11:14:38 <ikke> hmm
2020-07-12 11:14:43 <amom> doesnt matter what I do. Running git log on a bare repo, I get a segfault. Running git status, I get a segfault
2020-07-12 11:14:51 <ikke> ok
2020-07-12 11:15:35 <ikke> What version of git?
2020-07-12 11:15:45 <amom> the latest version from alpine edge
2020-07-12 11:15:52 <amom> sorry. ssh isnt working rn
2020-07-12 11:16:00 <amom> oh it's cause im using crust
2020-07-12 11:16:03 <amom> 1 sec
2020-07-12 11:16:39 <ikke> Did you try older versions to see if it happens therE?
2020-07-12 11:16:50 <ikke> 2.27.0 would that be
2020-07-12 11:16:52 <amom> git version 2.27.0
2020-07-12 11:16:57 <amom> I am currently using
2020-07-12 11:17:05 <amom> lemme see if I have an older version in my cache
2020-07-12 11:17:27 <amom> I only have 2.27.0
2020-07-12 11:17:44 <ikke> last change to notes.c was in 2.26.0
2020-07-12 11:17:51 <ikke> but a very minor chnage
2020-07-12 11:18:40 <amom> ikke: it's 4:30 am here
2020-07-12 11:18:43 <ikke> ah ok
2020-07-12 11:18:43 <amom> Im gonna go to sleep
2020-07-12 11:18:57 <ikke> https://tpaste.us/wn0P
2020-07-12 11:19:01 <amom> I will hop on irc tomorrow. If you find something, please post in the gitlab issue
2020-07-12 11:19:19 <amom> Im happy to help test any patches or troubleshoot more
2020-07-12 11:19:22 <Hello71> odd that the backtrace would stop at the {, given the assert
2020-07-12 11:19:40 <Hello71> and even if NDEBUG is set, display_notes_trees isn't null, because sb != sb@entry
2020-07-12 11:19:52 <Hello71> wait
2020-07-12 11:20:09 <ikke> Hello71: Yes, I found that odd to
2020-07-12 11:20:16 <ikke> hence verifying the version
2020-07-12 11:20:37 <Hello71> how does sb=0xffffffffe818
2020-07-12 11:21:20 <ikke> ?
2020-07-12 11:30:12 <Weeltin> Hello, anyone installed "bind-dev" and has it running as a recursive?
2020-07-12 11:42:53 <ikke> bind-dev are build-time dependencies
2020-07-12 11:43:22 <ikke> https://pkgs.alpinelinux.org/contents?branch=edge&name=bind-dev&arch=x86_64&repo=main
2020-07-12 13:15:54 <nsz> why do ppl only print a backtrace after segfault
2020-07-12 13:16:43 <nsz> usually you need to see the surrounding code (either the source or the instructions, the latter is always available)
2020-07-12 13:17:10 <nsz> and preferably the register state too as well as the file mappings (so you know what is where and how memory is protected)
2020-07-12 13:17:11 <ikke> nsz: because that's where gdb stops :)
2020-07-12 13:17:45 <ikke> and for most people, that's about all they know when using gdb
2020-07-12 13:17:53 <mps> :)
2020-07-12 13:18:06 <mps> right
2020-07-12 13:18:30 <mps> not that I know a lot more, tbh
2020-07-12 13:18:41 <ikke> me neither
2020-07-12 13:19:36 <mps> and most upstream instructs such: press 'r' and when it stops type 'bt'
2020-07-12 13:21:30 <nsz> i would add 'info proc map; info reg; disas $pc-40,+80'
2020-07-12 13:21:38 <nsz> ($rip on x86)
2020-07-12 13:22:35 <ikke> $rip instead of $pc?
2020-07-12 13:25:43 <mps> makes sense to have rip in report
2020-07-12 13:26:22 <nsz> ikke: yes
2020-07-12 13:26:30 <ikke> ok
2020-07-12 13:26:41 <nsz> e.g. such info usually makes it clear if it's a stackoverflow
2020-07-12 13:27:01 <nsz> because you see sp in the regs and the stack access at the segfaulting instruction
2020-07-12 13:28:06 <ikke> Trying to debug a sigbus error now on armhf, see what info I can get
2020-07-12 13:28:43 <ikke> probably some unaligned access
2020-07-12 14:20:52 <nponaraHAa> does this work on alpine? --> MOZ_WEBRENDER=1
2020-07-12 14:21:11 <nponaraHAa> from the article --> https://www.phoronix.com/scan.php?page=news_item&px=WebRender-For-Some-Linux
2020-07-12 14:22:00 <Cogitri> I guess you could try, but keep in mind that webrenderer is still experimental on Linux
2020-07-12 14:28:30 <maxice8> I am using it on GNOME Wayland and don't notice much difference
2020-07-12 14:28:42 <maxice8> about:support shows me I'm using WebRender
2020-07-12 14:29:47 <nponaraHAa> maxice8: i might give another try at sway/wayland since I've switched to edge
2020-07-12 14:30:09 <maxice8> Use GNOME Wayland, it is great
2020-07-12 14:30:14 <maxice8> Switched from sway :D
2020-07-12 14:30:36 <nponaraHAa> maxice8: in stable sway never was able to show any menu
2020-07-12 14:30:44 <nponaraHAa> I could only move the mouse
2020-07-12 14:31:01 <nponaraHAa> nothing else, no right/left clicks, no mouse
2020-07-12 14:31:09 <maxice8> huh
2020-07-12 14:35:22 <mckaygerhard> ikke ncopa clandmeter all the documentation about setup install modes are gone from the wiki page.. many users modify too much the install page.. and seem are focused for windosers users!
2020-07-12 14:37:35 <mckaygerhard> oh no the user Sb1 modified so many that page and now imortant informatio are gone.. i mean.. if any info will be changed the previous must be in some wehere!
2020-07-12 17:16:18 <picmyk> i installed the kernel headers and tried to compile v4l2loopback but got an error saying "/lib/modules/5.4.51-0-lts/build: No such file or directory.", will I have to compile the kernel manually for the headers?
2020-07-12 17:17:57 <afontain_> you might have to reboot
2020-07-12 17:18:44 <afontain_> after upgrading your kernel, the /lib/modules directory doesn't contain `uname -r` anymore
2020-07-12 17:18:54 <afontain_> no idea if it's supposed to work though
2020-07-12 17:18:56 <picmyk> i rebooted already
2020-07-12 17:19:58 <afontain_> actually, isn't that something that uses DKMS?
2020-07-12 17:20:28 <picmyk> i'm not sure
2020-07-12 17:20:34 <picmyk> but DKMS isn't in the alpine repos
2020-07-12 17:20:36 <afontain_> because last time I looked, DKMS wasn't supported on Alpine
2020-07-12 17:20:38 <picmyk> is it not supported?
2020-07-12 17:20:42 <picmyk> that sucks
2020-07-12 17:20:43 <afontain_> and likely still isn't
2020-07-12 17:20:45 <picmyk> why is that?
2020-07-12 17:20:49 <picmyk> just curious
2020-07-12 17:21:18 <afontain_> because it's complicated maybe?
2020-07-12 17:21:46 <afontain_> or because an important part of DKMS users are things like proprietary NVidia drivers that won't work anyway?
2020-07-12 17:22:04 <picmyk> ahh
2020-07-12 17:22:04 <picmyk> well
2020-07-12 17:22:12 <picmyk> i don't know really
2020-07-12 17:22:26 <picmyk> maybe this isn't the channel to ask this, but
2020-07-12 17:22:30 <picmyk> what about anbox?
2020-07-12 17:22:36 <picmyk> it refuses to run
2020-07-12 17:22:42 <picmyk> i'll go and ask on their channel
2020-07-12 17:24:56 <Hello71> picmyk: did you install linux-lts-dev
2020-07-12 17:25:18 <afontain_> hello on both sides btw
2020-07-12 17:25:26 <picmyk> yes
2020-07-12 17:25:35 <picmyk> well actually i had not installed linux-lts-dev
2020-07-12 17:41:48 <picmyk> that worked Hello71, thank you
2020-07-12 19:10:53 <eydaimon> I'm looking at doing `zfs set sharenfs=on ...` and getting an error.  https://github.com/openzfs/zfs/issues/4534  based on this people fix it by doing `apt install nfs-kernel-server`. what's the alpine linux equivalent ?
2020-07-12 19:11:25 <eydaimon> nfs-utils ?
2020-07-12 19:13:37 <eydaimon> oh, looks like it was
2020-07-12 19:13:42 <eydaimon> I asked too soon
2020-07-12 19:14:57 <ikke> indeed
2020-07-12 19:15:02 <ikke> https://packages.ubuntu.com/bionic/amd64/nfs-kernel-server/filelist
2020-07-12 19:15:09 <ikke> https://pkgs.alpinelinux.org/contents?branch=edge&name=nfs-utils&arch=x86_64&repo=main
2020-07-12 19:20:08 <eydaimon> anything else I need to do after besides `service nfs-share start` ?
2020-07-12 19:20:22 <eydaimon> I'm getting an error when running showmount
2020-07-12 19:27:38 <eydaimon> oops
2020-07-12 19:27:55 <eydaimon> if someone answered, kiwi asked me to refresh, and then I had to login again 🤦
2020-07-12 19:28:06 <eydaimon> so I missed any answer
2020-07-12 19:28:12 <ikke> no
2020-07-12 19:29:57 <eydaimon> any idea ?
2020-07-12 19:30:21 <ikke> did you export something?
2020-07-12 19:31:00 <ikke> there is also a rpc service
2020-07-12 19:32:20 <eydaimon> https://wiki.alpinelinux.org/wiki/Setting_up_a_nfs-server  🙌
2020-07-12 19:32:42 <eydaimon> and I had been searching for zfs sharing ...
2020-07-12 20:04:37 <eyda|mon> I don't suppose there's a zfs expert around... looks like the zfsonlinux implementation is a bit different than the one from freebsd. For freebsd, I would run: zfs set sharenfs="-mapall=daniel -network 192.168.0.0 -mask=255.255.0.0" titan/foobar, for example, but if I do that on alpine, I get: cannot set property for 'titan/foobar': 'sharenfs' cannot be set to invalid options
2020-07-12 20:34:32 <eyda|mon> how do I get all man-pages installed? apk add man-pages and I still do things like man exports, and there's no entry
2020-07-12 20:34:42 <ikke> apk add docs
2020-07-12 20:34:48 <eyda|mon> gracias
2020-07-12 20:35:50 <ikke> If you just wanted it for nfs-utils, you could also install just nfs-utils-doc
2020-07-12 20:36:05 <eyda|mon> ohh. good to know. thank you :)
2020-07-12 20:36:28 <eyda|mon> does add docs install it only for installed packages ?
2020-07-12 20:36:35 <eyda|mon> it looks like it
2020-07-12 20:36:35 <ikke> yes
2020-07-12 20:36:42 <eyda|mon> neat
2020-07-12 20:41:29 <eyda|mon> oh. most is only available in edge? 🤔
2020-07-12 20:41:44 <ikke> most of what?
2020-07-12 20:43:47 <eyda|mon> https://pkgs.alpinelinux.org/package/edge/testing/x86_64/most
2020-07-12 20:43:53 <eyda|mon> it's my favorite PAGER
2020-07-12 20:44:02 <ikke> aha :P
2020-07-12 20:44:12 <ikke> a who
2020-07-12 20:44:15 <eyda|mon> more → less → most :)
2020-07-12 20:44:24 <ikke> a who's on first kind of situation
2020-07-12 20:44:38 <ikke> seems like yes
2020-07-12 20:44:39 <eyda|mon> oh hehe
2020-07-12 20:44:43 <ikke> it's in testing
2020-07-12 20:44:50 <eyda|mon> just added edge temporarilyh and added it
2020-07-12 20:45:03 <ikke> so it would need to be moved to community before it gets part of stable releases
2020-07-12 20:45:12 <ikke> (mixsing is officially not supported :))
2020-07-12 20:45:22 <eyda|mon> understood :)
2020-07-12 20:46:28 <bougyman> damn how is the alpine container so small?
2020-07-12 20:46:51 <ikke> busybox, splitting out lots of things
2020-07-12 20:47:11 <ikke> (ie, docs, language files, etc)
2020-07-12 20:47:14 <bougyman> yeah but libc is 4MB lighter weight than libc in void-musl
2020-07-12 20:47:22 <bougyman> I've killed all of that other stuff
2020-07-12 20:47:31 <ikke> No idea there
2020-07-12 20:47:37 <maxice8> compiling with -Os instead of -O2 ?
2020-07-12 20:48:00 <bougyman> maybe... but 4MB?
2020-07-12 20:48:08 <bougyman> (out of 4.5MB)
2020-07-12 20:48:10 <bougyman> that's a lot
2020-07-12 20:48:28 <bougyman> alpine: -rwxr-xr-x    1 root     root        596528 May 22 17:57 ld-musl-x86_64.so.1
2020-07-12 20:48:56 <bougyman> void: -rwxr-xr-x 1 root root 4475024 Jun  8 08:36 libc.so
2020-07-12 20:49:10 <bougyman> (libc on the musl image is a symlink to ld-musl)
2020-07-12 20:49:14 <leah2> is it stripped?
2020-07-12 20:49:15 <bougyman> er on the alpine image
2020-07-12 20:53:37 <ikke> bougyman: just compiled on alpine: 4006840 Jul 12 20:52 libc.so
2020-07-12 20:54:15 <ikke> stripped:  596416 Jul 12 20:53 libc-stripped.so
2020-07-12 20:54:17 <ikke> so yes
2020-07-12 20:54:23 <leah2> yup
2020-07-12 20:54:23 <ikke> it's symbol stripping
2020-07-12 20:55:03 <bougyman> gotcha
2020-07-12 20:55:07 <bougyman> thanks!
2020-07-12 20:55:15 <bougyman> so that plus busybox does it.
2020-07-12 21:44:36 <opal> lol #philosophy bougyman?
2020-07-12 22:24:23 <picmyk> do android virtual devices work?
2020-07-12 22:52:34 <eyda|mon> in case someone is super curious about my earlier zfs sharenfs question, the answer is something like this: zfs set sharenfs="rw=@192.168.0.0/16,sync,all_squash,anonuid=1000,anongid=10,no_subtree_check" rpool/titan/zster
2020-07-12 23:05:34 <parazyd> How can I clean up some unnecessary packages? I'm unable to delete certain things, yet they don't have any dependencies: http://ix.io/2rtQ
2020-07-12 23:14:59 <lunaphyte> apk del py3-flask-login py3-flask
2020-07-12 23:17:34 <parazyd> lunaphyte: World updated, but the following packages are not removed due to: py3-flask: py3-flask-login
2020-07-12 23:17:41 <parazyd> It just doesn't work.
2020-07-12 23:35:16 <eyda|mon> weird, just an empty dependency list
2020-07-12 23:36:50 <parazyd> Yup
2020-07-12 23:37:04 <eyda|mon> phew. finally got automount working on macos. There seems to be some difference in the NFS protocol between linux and freebsd. I had to add resvport to /etc/auto_master on the /net line
2020-07-13 00:55:01 <lp2skyline> Hi, I have existing LVM on my laptop, with separate home. How would I go about installing this without wiping and redoing my partitioning?
2020-07-13 01:46:49 <adu> lp2skyline: I've heard it's possible, probably have to learn some tools
2020-07-13 01:50:47 <adu> lp2skyline: probably something like $(lvreduce home)
2020-07-13 03:14:42 <nacelle> imo... if the data is valuable enough that you dont want it destroyed, make a backup.  outside of that, be very careful about what you're doing with that.  preserving a partition while installing around it is never really a fun time even when it does go ok.
2020-07-13 03:15:32 <nacelle> and if you have lvm with /home outside of lvm, I would seriously rethink why you're using lvm in the first place :-)
2020-07-13 03:34:27 <lp2skyline> I meant I have 2 partitions under the LVM. root and home
2020-07-13 03:35:02 <lp2skyline> Thank you for the informations though. I'll work on this tomorrow. I'll probably have more questions before I install.
2020-07-13 06:58:10 <js_msi> hello, is there a command to recreate the apk database? i'm on edge and after an update of a few packages i get a FDB format error when i run any apk command
2020-07-13 07:01:35 <js_msi> i can just reinstall as I am just testing it out on a vm, no big deal, I checked the wiki but couldnt find an option like "-f" to force the database recreation when using apk update
2020-07-13 07:07:38 <ikke> perhaps `apk fix`, but no clue if that works
2020-07-13 07:09:54 <js_msi> https://pasteboard.co/JhqmO8U.jpg
2020-07-13 07:10:00 <js_msi> same error
2020-07-13 07:11:17 <js_msi> the dir /var/lib/apk is empty, is that normal?
2020-07-13 07:13:07 <ikke> it's empty for me is well
2020-07-13 07:14:57 <ikke> js_msi: the db is in /lib/apk/db
2020-07-13 07:26:43 <amom> Cogitri: did you get a chance to reproduce my git issue?
2020-07-13 07:37:40 <Cogitri> Unfortunately no, pretty busy right now :/
2020-07-13 07:40:22 <amom> No worries! Cogitri
2020-07-13 07:40:42 <ikke> Is it only pinephone where this happens?
2020-07-13 07:44:29 <afontain_> I tried on my braveheart without success, at least
2020-07-13 07:45:28 <ikke> reproducing without success?
2020-07-13 07:46:53 <afontain_> yeah, it just worked
2020-07-13 07:48:04 <Cogitri> Maybe this really just is caused by the DRAM overclocking on the Pinephone?
2020-07-13 07:55:52 <afontain_> I was wondering that too, that's why I did send a suggestion for a second test where ag is asked to scan every file on the system
2020-07-13 07:57:10 <afontain_> Cogitri: do note however, that the new version of uboot does not overclock anymore
2020-07-13 07:57:11 <afontain_> hmm
2020-07-13 07:57:29 <afontain_> amom: how old is your install?
2020-07-13 07:58:43 <amom> afontain_: how can I check that>
2020-07-13 08:00:29 <afontain_> hmm
2020-07-13 08:01:01 <afontain_> `busybox ls -l /etc/hostname`?
2020-07-13 08:01:46 <afontain_> look for the modification date
2020-07-13 08:02:22 <amom> 10 Jun  2
2020-07-13 08:02:30 <afontain_> 2020?
2020-07-13 08:03:33 <afontain_> yup, too old
2020-07-13 08:05:00 <afontain_> should be flashed with a `dd if=u-boot-sunxi-with-spl.bin of=/dev/sdX bs=8k seek=1`
2020-07-13 08:05:30 <amom> afontain_: is that file in / ?
2020-07-13 08:05:41 <amom> I did install crust on this install
2020-07-13 08:05:59 <afontain_> /usr/share/u-boot/pine64-pinephone?
2020-07-13 08:06:03 <amom> so I did end up flashing uboot. however, the git problem existed before I installed crust
2020-07-13 08:06:12 <afontain_> weird
2020-07-13 08:06:35 <afontain_> I'd advise to try flashing the latest version of uboot
2020-07-13 08:06:44 <afontain_> though it may not help
2020-07-13 08:07:08 <afontain_> `apk upgrade -a` will drop your locally built version
2020-07-13 08:07:10 <amom> dd if=/usr/share/u-boot/pine64-pinephone/u-boot-sunxi-with-spl.bin of=/dev/mmcblk2 bs=8k seek=1
2020-07-13 08:07:19 <amom> for eemc right?
2020-07-13 08:07:27 <afontain_> looks reasonable
2020-07-13 08:09:28 <amom> that fixed it
2020-07-13 08:09:45 <amom> wait no
2020-07-13 08:09:48 <amom> it fixed git log
2020-07-13 08:09:53 <amom> git show segfaults
2020-07-13 08:10:18 <amom> maybe I just need to refresh git's index
2020-07-13 08:10:21 <amom> 1 sec
2020-07-13 08:11:12 <amom> still segfaulting git show =(
2020-07-13 08:11:20 <amom> and git log is now segfaulting too
2020-07-13 08:11:44 <afontain_> latest version?
2020-07-13 08:11:48 <afontain_> did you reboot?
2020-07-13 08:12:00 <afontain_> which revision of the pinephone?
2020-07-13 08:12:30 <amom> afontain_: braveheart
2020-07-13 08:12:32 <amom> I did reboot
2020-07-13 08:12:39 <amom> latest version of everything
2020-07-13 08:12:57 <amom> for git: 2.2.27
2020-07-13 08:13:03 <amom> I mean 2.27.0
2020-07-13 08:14:04 <afontain_> can you try to run ag as in the issue?
2020-07-13 08:15:34 <amom> afontain_: Im sorry, I dont have the link to the issue handy
2020-07-13 08:15:51 <amom> I cant find it in the recent emails or recent issues on gitlab
2020-07-13 08:16:15 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues/11717
2020-07-13 08:16:54 <amom> much appreciated ikke
2020-07-13 08:17:11 <ikke> I still had it in my IRC history :)
2020-07-13 08:18:12 <amom> afontain_: I am getting 1 error on apk
2020-07-13 08:18:17 <amom> ERROR: File not found: /usr/share/dtb/allwinner/sun50i-a64-pinephone-1.1 allwinner/sun50i-a64-pinephone-1.2.dtb
2020-07-13 08:18:22 <amom> ERROR: postmarketos-mkinitfs-0.12.3-r0.trigger: script exited with error 1
2020-07-13 08:18:27 <amom> when I run doas apk fix
2020-07-13 08:18:43 <amom> could that be related to this?
2020-07-13 08:18:59 <afontain_> in theory, 1.2 is not your revision
2020-07-13 08:19:10 <afontain_> but did you upgrade -a?
2020-07-13 08:19:38 <amom> 1 sec
2020-07-13 08:20:51 <amom> ok done. Should I reboot?
2020-07-13 08:51:24 <afontain_> I'd reflash before rebooting
2020-07-13 08:56:34 <amom> afontain_: no luck
2020-07-13 11:06:45 <rapha> hi all
2020-07-13 11:07:37 <rapha> what's the lightest (in terms of configuration and disk space) way of allowing an alpine box to send email via a MTA in its own network (and to which it has open relay access)? postfix, or is there something lighter?
2020-07-13 11:12:19 <mps> rapha: ssmtp?
2020-07-13 11:12:43 <mps> depends what you actually need
2020-07-13 11:12:52 <mps> postfix is also small
2020-07-13 11:12:56 <rapha> mps: just to send the occasional cron mail to my backup address
2020-07-13 11:13:08 <rapha> ssmtp looks perfect, thank you!
2020-07-13 11:13:36 <rapha> also nice archlinux wiki page, that's always a boon
2020-07-13 11:13:44 <mps> for such things I also use ssmtp
2020-07-13 11:16:55 <rapha> hmm ... for testing from the command line, is there a package other than sendmail that has mail(1)?
2020-07-13 11:17:43 <rapha> ah, mailutils perhaps
2020-07-13 11:18:10 <mps> ssmtp contains sendmail
2020-07-13 11:22:44 <rapha> mps: unfortunately it seems to put question marks around the hostname it uses for HELO and is being rejected because of that. is there a way to get it to not do that?
2020-07-13 11:25:06 <rapha> ooooh
2020-07-13 11:25:17 <rapha> okay, you MUST not put quotes around the hostname in the config file
2020-07-13 11:25:19 <rapha> intricate
2020-07-13 11:26:22 <mps> man ssmtp will tell you difference with the 'normal' sendmail
2020-07-13 11:40:10 <rapha> kind of liking the one from mailutils
2020-07-13 11:40:15 <rapha> it has -A and -s
2020-07-13 11:40:38 <rapha> `mail`, that is; not `sendmail`
2020-07-13 15:26:04 <codebam> the last firefox update removed the .desktop file for me? or at least rofi isn't finding it anymore
2020-07-13 15:26:29 <maxice8> No, it added StartupWMClass=firefox
2020-07-13 15:26:42 <maxice8> so Wayland can correctly associate the Window with the Icon
2020-07-13 15:30:04 <codebam> oh okay I see
2020-07-13 15:31:16 <codebam> like idk it's only showing these now https://0x0.st/it06.png (the flatpak packages)
2020-07-13 15:31:39 <codebam> is the flatpak taking prevalence over the regular package now because of startupwmclass?
2020-07-13 15:32:10 <codebam> I can still start it from the command line so not a big issue
2020-07-13 17:15:42 <albr_> R&[d]: @bovis inside a chroot Executing dbus-1.10.28-r0.post-install
2020-07-13 17:15:43 <albr_> R&[d]: Failed to generate UUID: Could not open /dev/urandom: No such file or directory. So that’s one of your problems, no random number generator dev inside an Alpine chroot. https://discordapp.com/channels/661832166946373642/661834182934265883/731681148538716271
2020-07-13 17:33:39 <albr_> unowe[d]: they can't see Discord links from IRC
2020-07-13 17:56:27 <albr_> female[d]: when will the bridge be kill
2020-07-13 20:19:50 <finnwww78> How do I run alpine linux under Crostini?
2020-07-13 20:21:08 <ikke> What is Crostini?
2020-07-13 20:21:33 <finnwww78> Chrome OS's virtual machine
2020-07-13 20:21:45 <finnwww78> ikke ^
2020-07-13 20:22:45 <ikke> How do you run other OS' on Crostini?
2020-07-13 20:23:28 <finnwww78> I'm having a hard time with that myself. I can only figure out how to run the default Debian.
2020-07-13 20:23:43 <finnwww78> If I can figure this out, I'll write a wiki page :)
2020-07-13 20:24:12 <ikke> https://wiki.archlinux.org/index.php/Chrome_OS_devices/Crostini
2020-07-13 20:24:16 <ikke> this is for Archlinux
2020-07-13 20:24:48 <finnwww78> Do I just replace everything with alpine?
2020-07-13 20:24:53 <ikke> Apparently these are just containers, not VMs?
2020-07-13 20:25:06 <finnwww78> Ohh
2020-07-13 20:26:02 <ikke> The link on that wiki to linuxcontainers.org also contains alpine images
2020-07-13 20:26:05 <mps> 
ACTION runs alpine on chromebooks natively :)

2020-07-13 20:26:18 <finnwww78> What's mps?
2020-07-13 20:26:27 <finnwww78> Or who?
2020-07-13 20:26:33 <ikke> mps is mps :)
2020-07-13 20:26:36 <mps> me, I think
2020-07-13 20:26:48 <finnwww78> Do you run alpine under Crostini?
2020-07-13 20:26:55 <mps> no
2020-07-13 20:27:00 <finnwww78> Oh
2020-07-13 20:27:06 <ikke> no ChromeOS, just Alpine
2020-07-13 20:27:22 <mps> I installed alpine on few chromebooks
2020-07-13 20:27:33 <finnwww78> Oh okay. Yeah I'm looking to install it inside ChromeOS
2020-07-13 20:27:43 <mps> i.e. removed cromeos
2020-07-13 20:30:56 <mps> there were some guides on net how to run some linux distros in chromeos in chroot but I didn't tried any of these
2020-07-13 20:52:35 <finnwww7> I'm following the guide to install Arch Linux on Crostinil, except with Alpine (https://wiki.archlinux.org/index.php/Chrome_OS_devices/Crostini#Replacing_the_default_Debian_Linux_container_with_Arch_Linux).
2020-07-13 20:53:03 <finnwww7> So I got to the step where I needed to create a container:
2020-07-13 20:53:08 <finnwww7> vmc container termina arch https://us.images.linuxcontainers.org archlinux/current
2020-07-13 20:53:25 <finnwww7> I replaced that with alpine and:
2020-07-13 20:53:26 <finnwww7> Error: routine at frontends/vmc.rs:477 `container_create(vm_name,user_id_hash,container_name,source)` failed: failed to create container: `FAILED`: failed to get alias: Alias 'images:alpine/3.12' doesn't exist
2020-07-13 20:53:47 <finnwww7> The command I ran was:
2020-07-13 20:54:05 <finnwww7> vmc container termina alpine https://us.images.linuxcontainers.org images:alpine/3.12
2020-07-13 20:54:56 <finnwww7> Why is it not working >:(
2020-07-13 22:28:55 <codebam> I have alpine on my phone too lol
2020-07-13 22:28:59 <codebam> in a chroot
2020-07-13 22:29:08 <maxice8> Nice
2020-07-13 22:29:34 <codebam> I think alpine is my new go to distro for everything tbhj
2020-07-13 22:29:36 <codebam> *tbh
2020-07-13 22:29:50 <maxice8> Nice++
2020-07-13 22:30:41 <codebam> :)
2020-07-13 23:02:06 <finnwww15> The Arch Linux Wiki has a page on installing Alpine Linux under Crostini (https://wiki.archlinux.org/index.php/Chrome_OS_devices/Crostini#Replacing_the_default_Debian_Linux_container_with_Arch_Linux), and I think it would be pretty great if the Alpine Linux wiki did too, because I struggled to install Alpine Linux, and I instead installed Arch
2020-07-13 23:02:06 <finnwww15> under Crostini because it was the most well documented option.
2020-07-13 23:02:41 <finnwww15> Is there a place I can open an issue? I already did on the Gitlab documentation project
2020-07-14 01:45:07 <albr_> R&[d]: http://uk.alpinelinux.or
2020-07-14 01:45:08 <albr_> R&[d]: g/alpine/v3.8/main/x86/python3-3.6.9-r1.apk appears to have BAD signature.
2020-07-14 01:45:28 <albr_> R&[d]: http://uk.alpinelinux.org/alpine/v3.8/main/x86/python3-3.6.9-r1.apk appears to have BAD signature.
2020-07-14 01:54:03 <albr_> R&[d]: http://uk.alpinelinux.org/alpine/v3.8/main/x86/python3-3.6.9-r1.apk appears to have BAD signature. Plus I can’t get more than 31% downloaded after trying all day.
2020-07-14 01:59:02 <c705> maybe try a different mirror
2020-07-14 04:12:47 <albr_> R&[d]: > maybe try a different mirror
2020-07-14 04:12:48 <albr_> R&[d]: @c705#0000  you don’t know me, but I’ve been ‘doing’ Linux for a very very very long time. Such a cookie cutter solution is on the verge of being morally inappropriate; in fact, it’s right along the same lines as, ‘Why not just “wait” in another queue’. Absolutely discounting the irrelevant fact that I already HAD copies from several other ‘mirrors’ which also reported BAD signatures, a SOMEWHAT bette
2020-07-14 04:12:51 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732449490497175642/image0.png
2020-07-14 06:21:05 <iggy> the fact that you posted the same thing 3 times in varying formats really sells that you've been "doing" linux for a "very very very" long time
2020-07-14 06:31:47 <TBB> it has to be said tho that I've rarely encountered similar problems with any other distribution
2020-07-14 06:32:14 <TBB> not that it would've prevented me personally from ever working with Alpine
2020-07-14 09:31:41 <das_ofce> Hey guys, any idea if I should this report to alpine bugs? https://pastebin.com/B3sefvTq .. I ofcourse found https://github.com/gliderlabs/docker-alpine/issues/191 but we are not using any authentication on proxy
2020-07-14 12:01:42 <albr_> R&[d]: # mount -t proc proc /proc
2020-07-14 12:01:43 <albr_> R&[d]: ~ # mount -t sysfs sys ./sys -o nosuid,nodev,noexe
2020-07-14 12:01:44 <albr_> R&[d]: c,ro
2020-07-14 12:01:45 <albr_> R&[d]: mount: mounting sys on ./sys failed: Invalid argument
2020-07-14 12:01:46 <albr_> R&[d]: ~ #
2020-07-14 12:01:48 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732567495042400286/image0.png
2020-07-14 12:05:01 <albr_> R&[d]: # mount -t proc proc /proc
2020-07-14 12:05:01 <albr_> R&[d]: ~ # mount -t sysfs sys ./sys -o nosuid,nodev,noexec,ro
2020-07-14 12:05:03 <albr_> R&[d]: mount: mounting sys on ./sys failed: Invalid argument
2020-07-14 12:05:04 <albr_> R&[d]: ~ #
2020-07-14 12:05:05 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732567495042400286/image0.png
2020-07-14 12:06:36 <albr_> R&[d]: # mount -t proc proc /proc
2020-07-14 12:06:36 <albr_> R&[d]: ~ # mount -t sysfs sys ./sys -o nosuid,nodev,noexec,ro
2020-07-14 12:06:38 <albr_> R&[d]: mount: mounting sys on ./sys failed: Invalid argument
2020-07-14 12:06:39 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732567495042400286/image0.png
2020-07-14 12:08:17 <ikke> R& does ./sys exist?
2020-07-14 12:09:08 <ikke> and it seems you need to do mount -t sysfs none ./sys ..
2020-07-14 12:12:01 <albr_> R&[d]: ~ # mount -t sysfs none ./sys
2020-07-14 12:12:01 <albr_> R&[d]: mount: mounting none on ./sys failed: Invalid argument
2020-07-14 12:12:03 <albr_> R&[d]: ~ # ls ./|grep sys
2020-07-14 12:12:04 <albr_> R&[d]: sys
2020-07-14 12:14:37 <albr_> R&[d]: ~ # mount -t sysfs none ./sys
2020-07-14 12:14:37 <albr_> R&[d]: mount: mounting none on ./sys failed: Invalid argument
2020-07-14 12:14:39 <albr_> R&[d]: ~ # ls ./|grep sys
2020-07-14 12:14:40 <albr_> R&[d]: sys
2020-07-14 12:14:41 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732570746005618788/image0.png
2020-07-14 12:17:16 <ikke> from `man sysfs`: 'mount -t sysfs sysfs /sys
2020-07-14 12:17:17 <ikke> '
2020-07-14 12:40:48 <albr_> R&[d]: RtFm, said ikke.
2020-07-14 12:40:49 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732577334791766026/image0.png
2020-07-14 12:41:34 <ikke> R& Hey, I'm trying to help you
2020-07-14 12:41:41 <ikke> but I will stop as you are not appreciating it
2020-07-14 12:42:44 <albr_> R&[d]: RtFm, said ikke.
2020-07-14 12:42:45 <albr_> R&[d]: ~ # mount -t sysfs sysfs /sys
2020-07-14 12:42:46 <albr_> R&[d]: mount: mounting sysfs on /sys failed: Invalid argument
2020-07-14 12:42:47 <albr_> R&[d]: https://cdn.discordapp.com/attachments/661834182934265883/732577334791766026/image0.png
2020-07-14 12:45:43 <albr_> R&[d]: Your augmentation of the command from ./sys to /sys produces identical commands, and yet I’m the one being shown the manual. Don’t doubt that I was fully ready to cast doubt on all my own intelligence.
2020-07-14 12:47:48 <albr_> R&[d]: Your augmentation of the command from ./sys to /sys produces identical commands, and yet I’m the one being shown the manual. Don’t doubt that I was fully ready to cast doubt on all my own intelligence, considering that iggy had already swung a haymaker at it -which didn’t connect.
2020-07-14 15:01:29 <eyda|mon> is there a way to get pkg info on something that's not installed ? like homebrew for macos ?
2020-07-14 15:01:43 <eyda|mon> from terminal, that is
2020-07-14 15:01:49 <eyda|mon> ( apk )
2020-07-14 15:02:24 <ikke> apk info <pkgname>
2020-07-14 15:10:25 <eyda|mon> I tried that on a package i don't have installed, and it showed nothing
2020-07-14 15:10:43 <eyda|mon> apk info ufw, for example
2020-07-14 15:11:02 <eyda|mon> oh 🤦
2020-07-14 15:11:06 <eyda|mon> that's not the right name
2020-07-14 15:11:45 <eyda|mon> oh, it is the right name, it's only on edge
2020-07-14 15:14:40 <ikke> Indeed
2020-07-14 17:25:45 <malade_mental> I don't find resize2fs in e2fsprogs in alpine install live, is this normal?
2020-07-14 17:26:02 <maxice8> e2fsprogs-extra
2020-07-14 17:26:11 <malade_mental> you rock maxice8
2020-07-14 18:10:22 <eyda|mon> malade_mental: a way to find the answer for questions like this I've been using is to go to pkgs.alpinelinux.org and searching contents to find the package it belongs to
2020-07-14 18:11:08 <eyda|mon> malade_mental:https://pkgs.alpinelinux.org/contents?file=resize2fs&path=&name=&branch=v3.12
2020-07-14 18:56:31 <gjabell> did the recent qt upgrade break qutebrowser's UI for anyone else?
2020-07-14 19:05:37 <Cogitri> PureTryOut[m]: ^
2020-07-14 19:09:17 <eyda|mon> Cogitri: are you good at ufw ?
2020-07-14 19:10:04 <Cogitri> Good enough to start it and allow ssh and HTTPS apparently :D
2020-07-14 19:10:44 <PureTryOut[m]> I don't use Qutebrowser so I wouldn't know. Break in what way? Does it not launch at all? Could you check the logs gjabell?
2020-07-14 19:13:36 <gjabell> it launches fine but the UI is a bit broken (status bar is too short, tabs titles don't show up at all)
2020-07-14 19:13:43 <gjabell> doesn't seem to be any issues in the logs
2020-07-14 19:14:03 <gjabell> I tried downgrading from 1.13.0 => 1.12.0 and it didn't fix it so I *think* it's a qt issue and not a qutebrowser issue
2020-07-14 19:15:32 <eyda|mon> Cogitri: looks like by default it disallows ssh but it allows me to do a curl request and get a response on some random port which is weird.  I'm wanting to deny traffic from everywhere to all ports, then enable them for specific ones. I thought ufw deny in from 0.0.0.0/32 to 192.168.0.0/16 would work, but i can still run curl http://192.168.0.20:7878
2020-07-14 19:16:22 <guest68> Hi, I'm looking for information on procmail in alpine 3.12. It looks like it was removed. Google didn't help me to find current information, could someone please point me to the removal info or is it an oversight?
2020-07-14 19:17:24 <eyda|mon> ufw default deny incoming  There we go
2020-07-14 19:18:26 <eyda|mon> errr.... no, I can still curl
2020-07-14 19:18:27 <eyda|mon> wtf
2020-07-14 19:19:12 <Cogitri> Maybe it needs a restart after that?
2020-07-14 19:20:37 <Cogitri> guest68: 52d979af44a8cda5d06b478bde496b0b91d7670a
2020-07-14 19:20:47 <eyda|mon> yeah, maybe
2020-07-14 19:21:33 <eyda|mon> nope
2020-07-14 19:22:39 <eyda|mon> I'm thinknig it's because there's an established connection already
2020-07-14 19:22:44 <eyda|mon> because it was open in my browser
2020-07-14 19:23:12 <eyda|mon> because when I tried using: python3 -m http.server 8000, I couldn't connect to 8000
2020-07-14 19:27:38 <guest68> Cogitri, awesome, thanks!
2020-07-14 19:37:18 <eyda|mon> uhhh
2020-07-14 19:38:06 <eyda|mon> Cogitri: I have another clue. I am running a docker service on 7878. If I stop the service on docker, use python3 -m http.server 7878, then the port is blocked. But then I stop that and restart the docker service, then the port is accessible again.
2020-07-14 19:38:15 <eyda|mon> Does docker have some magic powers that can override the firewall??
2020-07-14 19:38:56 <eyda|mon> between stopping the python service which was inaccessible, and starting the docker service, I did nothing to the firewall
2020-07-14 19:39:10 <eyda|mon> I made sure the python service was accessble while the firewall was down
2020-07-14 19:40:16 <Cogitri> I think docker sets iptables rules too, yes
2020-07-14 19:40:27 <ikke> it does
2020-07-14 19:41:49 <eyda|mon> ahh. mystery solved
2020-07-14 19:43:14 <eyda|mon> so iptables and ufw have access to the same .. erh, stuff ?
2020-07-14 19:43:52 <ikke> ufw uses iptables
2020-07-14 19:44:14 <ikke> it's a front-end for iptables
2020-07-14 19:44:21 <eyda|mon> "On Linux, Docker manipulates iptables rules to provide network isolation" --  https://docs.docker.com/network/iptables/
2020-07-14 19:44:22 <eyda|mon> confirmed
2020-07-14 19:45:04 <ikke> note that it should only set rules for forwarding / nat
2020-07-14 19:45:27 <eyda|mon> given that I can access it from another host, I'd say it does more
2020-07-14 19:46:08 <eyda|mon> "It is possible to set the iptables key to false in the Docker engine’s configuration file at /etc/docker/daemon.json"
2020-07-14 19:46:11 <eyda|mon> I shall try and see
2020-07-14 19:50:00 <eyda|mon> well, lunch is over. it will have to wait
2020-07-14 20:23:42 <chlor> hi, I got wrong version /etc/os-release - did I had a failure on upgrade?
2020-07-14 20:24:14 <chlor> root@alpine:/# egrep ^[^#]+[0-9]\. /etc/issue /etc/alpine-release /etc/os-release /etc/apk/repositories
2020-07-14 20:24:16 <chlor> /etc/issue:Welcome to Alpine Linux 3.12
2020-07-14 20:24:18 <chlor> /etc/alpine-release:3.11.0
2020-07-14 20:24:20 <chlor> /etc/os-release:VERSION_ID=3.11.0
2020-07-14 20:24:22 <chlor> /etc/os-release:PRETTY_NAME="Alpine Linux v3.11"
2020-07-14 20:24:24 <chlor> /etc/apk/repositories:http://dl-cdn.alpinelinux.org/alpine/v3.12/main
2020-07-14 20:25:02 <chlor> I had version 3.11 and made a "apk upgrade --awailable"
2020-07-14 20:26:48 <ikke> what does apk version alpine-base return?
2020-07-14 20:29:40 <chlor> alpine-base-3.12.0-r0                   = 3.12.0-r0
2020-07-14 20:30:18 <ikke> does /etc/alpine-release.apk-new exist?
2020-07-14 20:30:23 <chlor> OK, seems like some packages just missing an update (like /etc/os-release and /etc/alpine-release)
2020-07-14 20:30:42 <ikke> well, it's alpine-base that contains those files
2020-07-14 20:31:16 <chlor> root@alpine:/# cat /etc/alpine-release.apk-new
2020-07-14 20:31:17 <chlor> 3.12.0
2020-07-14 20:31:25 <ikke> rgith
2020-07-14 20:31:33 <ikke> so you probably touched those files
2020-07-14 20:31:37 <ikke> so apk didn't overwrite them
2020-07-14 20:31:52 <ikke> Run update-conf to merge them
2020-07-14 20:31:57 <chlor> didnt touch anything
2020-07-14 20:33:05 <chlor> update-conf just say I changed /etc/hostname (yes I did)
2020-07-14 20:42:05 <chlor> ah, now I got it. I should just figure out update-conf is working. Everything is fine now.
2020-07-14 20:42:47 <chlor> Thanks!
2020-07-14 20:44:34 <ikke> yw
2020-07-15 08:00:22 <danci1973> Good morning...
2020-07-15 08:01:54 <danci1973> Tonight one of my Alpine 3.12 based routers had a 'network hiccup' - and this seems to have caused it: e1000e 0000:01:00.1 eth6: Reset adapter unexpectedly
2020-07-15 08:02:36 <danci1973> Now this interface is a part of a bond interface and 4 hours later, the other interface in that bond (also e1000e) did the same...
2020-07-15 08:04:47 <danci1973> I've looked up several possible causes and workarounds, and one of them was to disable 'pcie_aspm' (kernel boot parameter pcie_aspm=off) - however, the 'dmesg' output on my router doesn't mention ASPM anywhere (unlike the examples I've found) - does Alpine kernel 5.4.43-1-lts perhaps has this off by deafult?
2020-07-15 08:09:22 <ikke> danci1973: zgrep ASPM /proc/config.gz
2020-07-15 08:10:19 <danci1973> There is no /proc/config.gz  ...
2020-07-15 08:10:45 <ikke> Hmm..
2020-07-15 08:11:02 <danci1973> This is a USB / ramdisk install, maybe that's why?
2020-07-15 08:16:33 <mps> modprobe configs
2020-07-15 08:31:40 <danci1973> Thanks, didn't think that would be a loadable module. :)
2020-07-15 08:32:24 <danci1973> Seems like ASPM is enabled by default (CONFIG_PCIEASPM=y CONFIG_PCIEASPM_DEFAULT=y).
2020-07-15 11:38:17 <Hello71> did your dmesg overflow
2020-07-15 12:07:10 <dustfinger> Does apk list -t only show triggers of installed packages?
2020-07-15 12:11:32 <Hello71> probably
2020-07-15 12:26:43 <mps> dustfinger: you mean 'apk info -t', not list
2020-07-15 12:27:50 <mps> 'apk info -t pkgname'
2020-07-15 12:41:31 <dustfinger> Yes, that is what I meant. Sorry. I was curious because I created a list of all installed packages in one file and a list of all packages not installed in another. I grepped the trigger URI from both sets and found that the set of all not installed packages did not yield any triggers. So I concluded that either I have all the packages with triggers installed, or apk info -t does not show triggers for packages that are not installed.
2020-07-15 12:41:41 <dustfinger> I was wondering which was the case.
2020-07-15 12:46:28 <mps> hm, maybe: apk info -t `apk info`
2020-07-15 12:52:23 <dustfinger> Actually, I made a mistake when I ran that command pipe. I am going to fix and try again. I don't think my list of not installed packages was complete.
2020-07-15 13:13:10 <dustfinger> Yah, for the set of all trigger URIs, I have a package that contains a trigger that references each of those URIs. https://termbin.com/6dy4. There actually only 21 URIs being watched by triggers, so it is not that astounding. I was just surprised that there was not some others. I was trying to determine if there existed a package with a trigger for a specific URI, so I wanted to look at all the ones that I did not already have
2020-07-15 13:13:10 <dustfinger> installed.
2020-07-15 17:12:28 <rah> which packages are included in the "Extended" install over the "Standard"?
2020-07-15 17:12:39 <rah> I can't seem to find that information on either the website or the wiki
2020-07-15 17:20:04 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh#L33
2020-07-15 19:20:51 <gjabell> is nheko's package broken for anyone else?
2020-07-15 19:21:51 <ikke> broken in what sense?
2020-07-15 19:22:18 <gjabell> seems to be missing a shared library (libmatrix_client.so.0.3.0)
2020-07-15 19:22:30 <gjabell> although nothing seems to provide that, which is odd
2020-07-15 19:22:55 <ikke> Maybe a package that got removed?
2020-07-15 19:23:48 <gjabell> hmm could be
2020-07-15 19:23:53 <maxice8> mtxclient-0.3.1
2020-07-15 19:24:06 <mps> apk search so:libmatrix_clien
2020-07-15 19:24:15 <mps> ah, :)
2020-07-15 19:24:37 <gjabell> ahh, it got bumped to 0.3.1
2020-07-15 19:24:41 <gjabell> that'd explain it haha
2020-07-15 19:25:06 <gjabell> does nheko just need a rebuild then?
2020-07-15 19:25:08 <ikke> packages in testing do not always get bumped
2020-07-15 19:25:12 <ikke> yes, probably
2020-07-15 19:25:13 <maxice8> doubt it
2020-07-15 19:25:24 <maxice8> mtxclient was updated  on Jun 17
2020-07-15 19:25:24 <ikke> doubt what?
2020-07-15 19:25:49 <ikke> nheko April 25th
2020-07-15 19:25:58 <maxice8> huh
2020-07-15 19:26:34 <maxice8> then yeah a build is needed again
2020-07-15 19:27:35 <gjabell> should I open an issue or can someone kick off a build
2020-07-15 19:27:42 <maxice8> Might also want to find out why their soname is .so.X.Y.Z and not .so.X
2020-07-15 19:28:25 <ikke> would it make sense to call it mtxclient-libs btw?
2020-07-15 19:28:32 <ikke> mtxclient sounds like it's an application
2020-07-15 19:31:51 <mps> iirc dalias made something with that or similar name, mail client with DANE check
2020-07-15 19:33:23 <mps> ah no, https://github.com/richfelker/mxclient
2020-07-15 20:03:47 <Orcs53> Hi guys, does anyone know how to install a self-signed SSL certificate into Alpine Linux?
2020-07-15 20:05:46 <c705> Orcs53: probably dump it in /etc/ssl/certs
2020-07-15 20:06:02 <c705> depends on your application using TLS
2020-07-15 20:08:11 <Orcs53> I will look into that
2020-07-16 06:59:17 <the6543> @ikke it looks like nobody has taken a look at https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests?scope=all&utf8=%E2%9C%93&state=opened&search=%5D+community%2Fgo
2020-07-16 06:59:38 <the6543> doesnt the both MR show up in https://gitlab.alpinelinux.org/alpine/aports/-/issues/11743 ?
2020-07-16 09:31:00 <sonicpp> hello, how long does it usually takes to be package rebuilt and available for download after new commit in aports git?
2020-07-16 09:31:57 <mps> 15-30 minutes till it appears on mirrors
2020-07-16 09:32:26 <mps> but depends how big is build queue
2020-07-16 09:33:29 <mps> can take hours sometimes
2020-07-16 09:35:44 <sonicpp> Hmm I see. There is new commit for sasl2 since yesterday. New version is availabel for armv7 for example (https://pkgs.alpinelinux.org/package/edge/main/armv7/cyrus-sasl) but aarch64 has still old version (https://pkgs.alpinelinux.org/package/edge/main/aarch64/cyrus-sasl)
2020-07-16 09:38:11 <sonicpp> It is more than 24 hours since commit
2020-07-16 09:38:28 <mps> these builders are 'out-of-luck' now
2020-07-16 09:38:45 <sonicpp> Ok, thanks for answer
2020-07-16 09:39:02 <mps> currently rearrangement is in progress
2020-07-16 09:39:37 <ikke> aarch64 builder is hanging now though
2020-07-16 11:43:52 <radiomark> Just discovered Alpine Linux, I think it could be good for our servers to replace CentOS .. I am automating the building of install images
2020-07-16 11:44:04 <radiomark> Is there a way to enable network/SSH on boot of the install ISO?
2020-07-16 12:11:19 <MuffinPimp> setup-sshd should do that
2020-07-16 12:12:49 <MuffinPimp> And setup-interfaces for networking.
2020-07-16 12:13:45 <MuffinPimp> Not sure how scriptable they are though
2020-07-16 12:14:11 <radiomark> And those look like interactive commands
2020-07-16 12:24:30 <Hello71> https://git.alpinelinux.org/alpine-conf/tree/setup-sshd.in
2020-07-16 12:25:48 <Cogitri> Alternatively you can make a custom install image by editing mkimage: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimage.sh
2020-07-16 12:26:11 <ikke> You would create a custom profile
2020-07-16 12:34:02 <radiomark> Is mkimage.sh part of the scripts to build the distribution .iso ?
2020-07-16 12:39:13 <radiomark> The packer example at https://wiki.alpinelinux.org/wiki/Packer_installation recommends to manually set up the interface
2020-07-16 12:39:22 <radiomark> Which I have now got working
2020-07-16 12:39:55 <radiomark> I am left with a final problem which is how to do ACPI shutdown of the 'host' ?   The "halt" command just stops but doesn't switch off
2020-07-16 12:41:16 <Cogitri> There should be a poweroff command I think
2020-07-16 12:41:23 <radiomark> aha
2020-07-16 12:41:26 <radiomark> That could be it
2020-07-16 12:41:28 <radiomark> Thanks
2020-07-16 14:34:33 <sonicpp> aarch64 build is ready, thanks!
2020-07-16 14:35:11 <ikke> sonicpp: np
2020-07-16 15:40:55 <rah> 
ACTION has installed alpine

2020-07-16 15:40:57 <rah> https://www.youtube.com/watch?v=o6F6u5TIUbI
2020-07-16 16:01:55 <eyda|mon> root@nami /mnt# mount -t nfs luffy:/titan /mnt/titan
2020-07-16 16:01:57 <eyda|mon> mount.nfs: rpc.statd is not running but is required for remote locking
2020-07-16 16:01:59 <eyda|mon> mount.nfs: Either use '-o nolock' to keep locks local, or start statd.
2020-07-16 16:02:01 <eyda|mon> mount.nfs: Protocol not supported
2020-07-16 16:02:03 <eyda|mon> sorry for spam
2020-07-16 16:02:08 <eyda|mon> any idea how to deal with this ?
2020-07-16 16:02:21 <eyda|mon> I can't find statd in any package
2020-07-16 16:02:42 <radiomark> I am only just getting to grips with Alpine distribution, but usually this is 'rpcbind' service
2020-07-16 16:02:54 <eyda|mon> thank you radiomark  :)
2020-07-16 16:03:09 <eyda|mon> it's already installed
2020-07-16 16:03:25 <mps> apk search cmd:rpc.statd
2020-07-16 16:03:46 <mps> nfs-utils pkg
2020-07-16 16:04:03 <eyda|mon> service rpc.statd start
2020-07-16 16:04:04 <eyda|mon> there we go
2020-07-16 16:04:18 <eyda|mon> apk search cmd!
2020-07-16 16:04:29 <eyda|mon> that's a nice one either way
2020-07-16 16:10:16 <Austin> hey, does anyone know if I can specify an existing apkovl instead of using the genapkovl scripts when using the aports mkimage scripts?
2020-07-16 16:51:25 <Guest69> hi, gnome-session shows this error:  Failed to create proxy for timedated
2020-07-16 16:51:34 <Guest69> then it exit
2020-07-16 16:53:54 <ikke> (on WSL)
2020-07-16 16:54:13 <Guest69> yes :)
2020-07-16 16:55:29 <Guest69> it is a common thing, happened with other distros i tried
2020-07-16 17:02:10 <Guest69> I have seen gnome-session can use a lot of  systemd related stuff (timedated is in systemd). I need some option to pass to gnome-session , so that it is not dependent on systemd related stuff
2020-07-16 17:04:13 <Guest69> I started service dbus and udev as root, then as user I started gnome-session
2020-07-16 17:16:02 <Guest69> xfce works perfectly on wsl, the problem with  all the distros i tried is the dependence of gnome on systemd stuff . Now gnome in Alpine works without systemd , so systemd is not necessary for gnome.  But when i start it typing: gnome-session, it use a 'systemd mode', searching for systemd stuff.   mode
2020-07-16 17:16:55 <Guest69> searching for systemd stuff
2020-07-16 17:17:59 <Guest69> I should start gnome with gdm, i have read that in the documentation. But gdm doesn't work in WSL
2020-07-16 17:18:29 <mps> alpine is not best choice for these systemd 'oriented' software
2020-07-16 17:19:12 <Guest69> but gnome works in alpine without WSL ?
2020-07-16 17:19:35 <mps> I think it works, though never tried
2020-07-16 17:20:30 <mps> (don't see reason to use small and simple distro to run bloat on it)
2020-07-16 17:31:35 <Guest69> If it works gdm should automagically  pass the right options to gnome-session. There are other distros, as Devuan (which I haven't tried) that have not systemd and have gnome. Tried alpine WSL today, very easy to get started, very simple and clean. Has not systemd and this is a bonus for me.  apk is the fastest and cleanest package manager  i
2020-07-16 17:31:36 <Guest69> tried, similar to pacman but better. I love also abuild
2020-07-16 17:34:09 <mps> nice to hear :)
2020-07-16 17:34:59 <c705> is the fellow I was chatting with about KVM/libvirtd in here?
2020-07-16 17:35:31 <c705> Anyways status update: it's been stable for the last week or so but I don;t want to get my hopes up. Gotta do some more testing
2020-07-16 17:41:47 <Guest69> if i want try gnome, and then remove gnome is simple as typing apk add gnome, apk del gnome. Install and removes all the packages in  5 minutes! The only problem, related to WSL is that i can't use gdm to start gnome, and i don't know the right options gdm use to start gnome-session without searching for systemd. So i can't start it, as with other
2020-07-16 17:41:47 <Guest69> WSL distros
2020-07-16 17:43:04 <Guest69> gnome-session has a --session option, don't know what gdm uses
2020-07-16 17:45:09 <gray_-_wolf> Dunno if it will help in this particular case, but Arch's documentation is usually really great, section on manual starting gnome: https://wiki.archlinux.org/index.php/GNOME#Manually
2020-07-16 17:55:38 <Guest69> the problem with arch, which is a nice distro,  is that is has systemd (the documentation reflects that). So even with these exports gnome rely on systemd  and doesn't start.  The alpine way should be to start it with gdm, which passes the right options. If gnome works with Alpine, it should not depend on systemd and this is a very valuable fact,
2020-07-16 17:55:39 <Guest69> that is important even for the other distro  How can i find the options gdm uses ?
2020-07-16 18:31:58 <Guest69> Anyone has tried gnome in Alpine? It works ?
2020-07-16 18:32:43 <ajin> Guest69: I'm running Gnome 3 on my Pinebook Pro (Postmarket OS)
2020-07-16 18:33:39 <Guest69> You start it with gdm?
2020-07-16 18:34:01 <ajin> I think so though I'm not sure lol
2020-07-16 18:34:51 <ajin> I downloaded an image and it was all set up
2020-07-16 18:38:02 <Guest69> How is this Alpine related ?
2020-07-16 18:39:35 <Guest69> Ah... postmarket OS is Alpine for phones!
2020-07-16 18:39:43 <Guest69> Interesting!
2020-07-16 18:47:05 <Cogitri> I'm using GNOME on all of my setups with Alpine and it works for me (but I'm the maintainer, so it'd be weird if it didn't work for me :)
2020-07-16 18:52:19 <Austin> is anyone super familiar with the mkimage script in aports?
2020-07-16 18:53:14 <Guest69> wow, so i may help me.  How can i start gnome via command line and not using gdm?
2020-07-16 18:53:56 <Guest69> you may help me
2020-07-16 18:54:20 <Cogitri> https://wiki.archlinux.org/index.php/GNOME#Manually this should work I think, but I only ever used GDM
2020-07-16 18:55:56 <Guest69> how can i find the options gdm use in Alpine to start gnome? I think it is different from arch, which is systemd based
2020-07-16 18:56:27 <Cogitri> Sorry?
2020-07-16 18:58:15 <Guest69> arch uses systemd, alpine doesn't use systemd , I tried starting gnome with this guide and doesn't work
2020-07-16 18:59:16 <Cogitri> The manually section of that wiki article should be the same for both Arch and Alpine
2020-07-16 18:59:22 <Guest69> tried gnome-session --builtin
2020-07-16 19:00:12 <Guest69> error in gnome initial setup
2020-07-16 19:00:33 <Guest69>   Failed to create proxy for timedated
2020-07-16 19:04:47 <Cogitri> Hm, did you install the gnome package or just gnome-shell?
2020-07-16 19:05:03 <Guest69> installed gnome
2020-07-16 19:05:08 <Cogitri> The GNOME package has a dep on openrc-settingsd and that should provide that interface
2020-07-16 19:13:12 <Guest69> when i installed gnome didn't installed that dependency. It it is necessary maybe should be included. It shows same error
2020-07-16 19:14:11 <Cogitri> Well, it gets installed with the gnome metapackage
2020-07-16 19:14:37 <Guest69> are you sure?
2020-07-16 19:14:54 <Guest69> are you sure?
2020-07-16 19:14:55 <Cogitri> And shouldn't be necessary for running, it might be a warning but certainly isn't fatal
2020-07-16 19:15:24 <Guest69> ok, i didn't have that
2020-07-16 19:17:00 <Guest69> anyway, i installed it, exported the variables as says the arch wiki, then gnome-session
2020-07-16 19:17:50 <Guest69> the error is the same
2020-07-16 19:19:22 <Guest69> can i disabela gnome-initial-setup?
2020-07-16 19:19:32 <Guest69> can i disable gnome-initial-setup?
2020-07-16 19:22:42 <Cogitri> I think there was some kernel parameter to disable it
2020-07-16 19:38:06 <Guest69> ah, ok, I can't try that for now.   Tried Alpine on WSL, I have to partition and install the real thing. First impression is that it is very amazing. Simple and clean.  Thank you ! .
2020-07-16 21:12:59 <radiomark> Should I be using syslinux for VM hosts if I am aiming for simplicity? Or is it just really for the installer
2020-07-16 21:18:15 <mps> radiomark: whatever works and what you prefer
2020-07-16 21:21:54 <radiomark> GRUB just feels a bit unweildy
2020-07-16 21:24:12 <radiomark> Thought I'd give it a go as I don't need menus or other kernels in a VM environment
2020-07-16 21:29:10 <tartar> i need `festival` but can't find it under `apk`, any ideas?
2020-07-16 21:59:10 <radiomark> Great, I'm booted. Is there any way to disable the setting of the console mode? I just want to continue with the old-school BIOS console mode
2020-07-16 21:59:21 <radiomark> Tried "vga=none" doesn't seem to have an effect
2020-07-16 22:08:36 <c705> tartar: build it
2020-07-16 22:09:13 <c705> radiomark: you mean the consolefont? not sure what you mean by "old-school BIOS console mode"
2020-07-16 22:09:48 <radiomark> No, not the font. During bootup, syslinux doesn't set the mode, it follows the bios. Half way through boot, the kernel or similar switches screen mode
2020-07-16 22:09:57 <radiomark> (this is with the console btw., no X)
2020-07-16 22:12:10 <zmc>  radiomark: I think I know what you're talking about. This happens to me when the gfx card driver loads. You could find the module that's doing it and blacklist it.  If that works, there is probably an option to do what you want
2020-07-16 22:12:33 <c705> radiomark: nomodeset on the kernel cmdline?
2020-07-16 22:12:49 <c705> or modeset..I forget which one switches on BIOS control of VGA
2020-07-16 22:12:54 <zmc> radiomark: in my case it's the nouveau module
2020-07-16 22:13:06 <radiomark> I think you're right, it's actually later in /etc/modprobe.d/kms.conf
2020-07-16 22:13:17 <radiomark> I don't actually want this; can I trace which package installed that file?
2020-07-16 22:14:03 <radiomark> Looks like 'nomodest' also achieves it
2020-07-16 22:14:11 <radiomark> *nomodeset
2020-07-16 22:17:20 <c705> try it and see
2020-07-16 22:17:31 <radiomark> Yes, that's what I mean -- nomodeset works
2020-07-16 22:26:22 <radiomark> Thanks
2020-07-16 22:45:32 <tartar> c705 as if it was this simple. there's no gmake.
2020-07-16 22:46:33 <c705> tartar: yes there is
2020-07-16 22:46:34 <c705> https://pkgs.alpinelinux.org/package/edge/main/x86/make
2020-07-16 22:47:55 <tartar> apk add make && gmake: -ash: gmake: not found
2020-07-16 22:48:27 <tartar> i give up. idk why i even bothered with alpine.
2020-07-16 22:48:43 <c705> see ya
2020-07-16 22:49:26 <hyiltiz> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=956431
2020-07-16 22:49:37 <hyiltiz> I installed grub-imageboot and put the iso under /boot/images/ then did update-grub
2020-07-16 22:49:37 <hyiltiz> It shows up in the grub menu but doesn't boot; any ideas?
2020-07-17 01:48:46 <eyda|mon> trying to mount my zfs created drive on macos but I get this: https://i.postimg.cc/SsB1DBt9/Screen-Shot-2020-07-16-at-18-47-15.png    anyone know how to correctly get one to mount on both macos as well as linux ?
2020-07-17 02:20:27 <Hello71> it seems pretty self explanatory
2020-07-17 02:21:32 <eyda|mon> and what's your wise answer ?
2020-07-17 02:22:04 <eyda|mon> would I ask if I knew ?
2020-07-17 02:23:29 <eyda|mon> I can try to make it without the attributes, but there it looks like it should work to at least mount read-only
2020-07-17 02:23:36 <eyda|mon> note the input/output error however
2020-07-17 05:32:05 <Cogitri> eyda|mon: The ZFS version you created the pool with is too new
2020-07-17 05:32:32 <Cogitri> You have to create it on the host that has the lowest version (so supports the least features)
2020-07-17 05:32:55 <Cogitri> And make sure you don't upgrade the pool on any host but the oldest host, otherwise you'll be in the same situation again
2020-07-17 12:26:56 <Hello71> regarding the i/o error you should ask the zfs on mac people
2020-07-17 13:22:59 <JonasRC> ahoy
2020-07-17 13:24:44 <Cogitri> 👋
2020-07-17 13:27:48 <JonasRC> I haven't used irc for a looong time :>
2020-07-17 13:59:07 <JonasRC> Has anyone managed to build glibc-2.31 on alpine ?
2020-07-17 13:59:13 <Guest83> I have a system running Alpine 3.12 that has 3 interfaces: wan, lan0, and lan1. I wish to bridge lan0 and lan1 into br0, and give br0 an address from DHCP. However, both lan0 and lan0 still give themselves addresses that I cannot seem to remove, and I can't seem to give br0 a DHCP address (so I can't use it to connect to the network unless I set it
2020-07-17 13:59:13 <Guest83> statically). Any ideas for this? See: https://www.reddit.com/r/linuxquestions/comments/hsg8dd/bridged_networking_on_alpine_linux/ for more details
2020-07-17 14:48:38 <eyda|mon> running: `dd if=/dev/random of=data1.bin bs=1M count=10`  is there a reason this takes forever ? I tried with both zfs encrypted and ext4
2020-07-17 14:49:39 <ikke> JonasRC: hello, welcome
2020-07-17 14:49:56 <JonasRC> Thank you
2020-07-17 14:49:58 <ikke> eyda|mon: lack of entropy
2020-07-17 14:50:40 <ikke> JonasRC: oneinsect is working on a glibc port of alpine
2020-07-17 14:50:47 <ikke> sorry
2020-07-17 14:50:55 <JonasRC> cool :)
2020-07-17 14:50:58 <ikke> That was meant for Guest83
2020-07-17 14:51:01 <ikke> oh no
2020-07-17 14:51:04 <ikke> still you :P
2020-07-17 14:51:10 <JonasRC> hahaha :)
2020-07-17 14:51:20 <MY-R> eyda|mon, use urandom and/or install "haveged" and start it as a service
2020-07-17 14:51:28 <eyda|mon> MY-R:thank you
2020-07-17 14:52:49 <eyda|mon> dd for linux doesn't output how fast it wrote ?
2020-07-17 14:53:05 <Lord> not by default
2020-07-17 14:53:19 <MY-R> not busybox one
2020-07-17 14:53:24 <Lord> you may use pv (pipe viewer)
2020-07-17 14:54:13 <MY-R> dd from coreutils got something like "status=progress"
2020-07-17 14:54:44 <eyda|mon> thank you both
2020-07-17 14:55:30 <eyda|mon> I need to do some benchamrking because the performance i"m seeing isn't what I'm expecting
2020-07-17 14:55:46 <eyda|mon> trying to figure out if it's a case of encrypted zfs or if this machine is old and crappy
2020-07-17 14:56:25 <eyda|mon> so want to figure out a nice standard test I can run on the same machine under both circumstances
2020-07-17 14:56:37 <eyda|mon> geli + zfs vs encrypted zfs
2020-07-17 14:56:48 <ikke> Don't forget that linux can cache things
2020-07-17 14:56:56 <eyda|mon> (geli is freebsd version of luks)
2020-07-17 14:57:07 <eyda|mon> well, it's equivalent, not version
2020-07-17 14:57:24 <eyda|mon> ikke: how can I account for that ?
2020-07-17 14:58:30 <ikke> coreutils dd has oflag=direct
2020-07-17 14:58:39 <ikke> I believe that should skip any caching
2020-07-17 14:59:11 <mps> yes, I use it to measure raw speed of block devices
2020-07-17 15:00:21 <eyda|mon> ikke: sweet thank you
2020-07-17 15:06:24 <eyda|mon> hm, so it's 3x speed on linux with encrypted zfs using dd and urandom
2020-07-17 15:23:14 <oneinsect> JonasRC: the port is fairly complete just waiting for one more final build before putting everything out in open
2020-07-17 15:23:21 <oneinsect> may be 10 more days
2020-07-17 15:23:37 <oneinsect> thanks ikke: for remembering me
2020-07-17 15:24:24 <ikke> Ofcourse :-)
2020-07-17 15:24:44 <JonasRC> Awesome! :)
2020-07-17 15:24:54 <JonasRC> Any issues ?
2020-07-17 15:25:09 <JonasRC> I tried building it but i got some errors.
2020-07-17 15:25:58 <oneinsect> lol i have been building since like eternity, the main problem is with glibc itself
2020-07-17 15:26:29 <oneinsect> each iteration they bring something new like recently they are phasing out libcrypt so i was forced to use libxcrypt
2020-07-17 15:26:49 <komnpomat> hei
2020-07-17 15:27:26 <oneinsect> so much more better with musl
2020-07-17 15:27:35 <komnpomat> w00t mame in alpine
2020-07-17 15:27:39 <oneinsect> but then glibc is needed till musl takes over the world
2020-07-17 15:27:46 <komnpomat> is anybody familiar with tyrian game?
2020-07-17 15:27:55 <oneinsect> no
2020-07-17 15:27:59 <oneinsect> what is that game
2020-07-17 15:28:06 <oneinsect> do you want to package it for alpine?
2020-07-17 15:28:22 <komnpomat> oneinsect: those are some ambitious goals for an insect
2020-07-17 15:28:28 <oneinsect> yes
2020-07-17 15:28:34 <oneinsect> damn bugs everywhere
2020-07-17 15:28:44 <komnpomat> oneinsect: it's a fairly standarnd game in any distro
2020-07-17 15:28:49 <oneinsect> lol
2020-07-17 15:29:36 <komnpomat> oneinsect: one interesting thing about comparing musl and glibc is the mentality of each project.
2020-07-17 15:29:53 <oneinsect> 1995 game from Epic (my previous employeer)
2020-07-17 15:30:02 <oneinsect> yes indeed
2020-07-17 15:31:15 <oneinsect> i may port railroad typcoon one day to alpine
2020-07-17 15:31:32 <oneinsect> tycoon*
2020-07-17 15:31:45 <oneinsect> any fans here?
2020-07-17 15:32:08 <oneinsect> The Railroad_Tycoon_(series)?
2020-07-17 15:34:54 <JonasRC> me!
2020-07-17 15:34:57 <JonasRC> haha
2020-07-17 15:35:01 <oneinsect> :-P
2020-07-17 15:42:22 <komnpomat> not in alpine either, another standard game in distros
2020-07-17 15:42:41 <oneinsect> hmmmm
2020-07-17 15:43:17 <komnpomat> oneinsect: if gaming isn't sorted out on alpine it will never take over the world
2020-07-17 15:43:40 <oneinsect> it will be sorted this August
2020-07-17 15:44:05 <oneinsect> i will work on it once glibc port is complete
2020-07-17 15:44:15 <rah> what's special about august?
2020-07-17 15:44:56 <oneinsect> lol...i will get time thats all
2020-07-17 15:45:25 <rah> I see
2020-07-17 15:45:38 <komnpomat> wishlist: dolphin, gngeo, psx2, desmume
2020-07-17 15:46:00 <JonasRC> I use alpine for containers :)
2020-07-17 15:46:04 <oneinsect> will be granted
2020-07-17 15:46:50 <komnpomat> emulators are extremely important as they benchmark any lib C implementation.
2020-07-17 15:46:53 <rah> what's involved in the glibc port?
2020-07-17 15:47:19 <rah> does the whole package repository need to be rebuilt?
2020-07-17 15:47:25 <oneinsect> yes it has to be
2020-07-17 15:47:33 <oneinsect> there more than 10,000 packages
2020-07-17 15:47:38 <rah> is this going to be a hard fork of alpine?
2020-07-17 15:48:07 <rah> or will there be Alpine Classic(tm) and Alpine Glibc(tm)
2020-07-17 15:48:16 <komnpomat> uh?
2020-07-17 15:48:24 <komnpomat> are you kidding me?
2020-07-17 15:48:27 <oneinsect> abuild, apk-tools etc will remain same but the APKBUILDS of packages will need modification
2020-07-17 15:48:46 <oneinsect> as most have patches which needs to be removed and new ones added
2020-07-17 15:48:57 <komnpomat> oneinsect: wait, did you mean glibc port as porting alpine to glibc on not the other way around?
2020-07-17 15:49:37 <rah> oneinsect: will the APKBUILDs be shared between alpine musl and alpine glibc?
2020-07-17 15:49:40 <oneinsect> i am build a fresh alpine os from scratch where every package is build against glibc
2020-07-17 15:49:43 <oneinsect> pure glibc only
2020-07-17 15:49:46 <oneinsect> no musl at all
2020-07-17 15:49:51 <oneinsect> building*
2020-07-17 15:49:53 <komnpomat> lol
2020-07-17 15:50:02 <komnpomat> oneinsect: what's the motivation?
2020-07-17 15:50:11 <komnpomat> oneinsect: you might as well add systemd
2020-07-17 15:50:19 <oneinsect> noooo
2020-07-17 15:50:23 <oneinsect> neverrrrr
2020-07-17 15:50:31 <oneinsect> i hate systemd
2020-07-17 15:50:39 <oneinsect> i needed it for a project of mine
2020-07-17 15:50:41 <komnpomat> oneinsect: you want to infect alpine?
2020-07-17 15:50:45 <oneinsect> deep learning project
2020-07-17 15:50:58 <oneinsect> no there is an unofficial port
2020-07-17 15:51:01 <komnpomat> oohh.. got it
2020-07-17 15:51:02 <oneinsect> nothing to do with Alpine
2020-07-17 15:51:07 <oneinsect> this is an*
2020-07-17 15:51:09 <komnpomat> pet project
2020-07-17 15:51:15 <oneinsect> exactly!!!
2020-07-17 15:51:44 <komnpomat> oneinsect: but, wouldn't it be more interesting port something like linuxmint, devuan, opensuse to musl?
2020-07-17 15:52:07 <oneinsect> rah: APKBUILDS will not be probably shared, i mean they are not compiling as is
2020-07-17 15:52:44 <oneinsect> they have 1000s of packages
2020-07-17 15:53:03 <oneinsect> the human effort needed is enormous
2020-07-17 15:53:33 <oneinsect> besides companies NVIDIA etc are never supporting musl
2020-07-17 15:53:42 <oneinsect> I wish they could have
2020-07-17 15:53:52 <oneinsect> but this is only temporary
2020-07-17 15:54:15 <komnpomat> oneinsect: have musl devs approached these companies for some sort of support?
2020-07-17 15:54:38 <oneinsect> I believe they do now and then
2020-07-17 15:54:42 <oneinsect> the motivation isnt there
2020-07-17 15:54:51 <oneinsect> by companies like NVIDIA
2020-07-17 15:54:56 <rah> so this is a hard fork of alpine which, for now, uses the same tools as alpine and has a package repository which is a hard fork of alpine's package repository
2020-07-17 15:55:34 <oneinsect> yes correct
2020-07-17 15:55:42 <rah> good luck
2020-07-17 15:56:07 <komnpomat> ha ha ha
2020-07-17 15:56:13 <oneinsect> the tools are same and they will be in sync but only packages will be rebuilt
2020-07-17 15:56:20 <oneinsect> lol
2020-07-17 15:56:25 <oneinsect> thanks
2020-07-17 15:56:27 <komnpomat> oneinsect: such a waste of a wonderfull August
2020-07-17 15:56:42 <oneinsect> indeed
2020-07-17 15:56:44 <oneinsect> :P
2020-07-17 15:57:11 <JonasRC> Haha :)
2020-07-17 15:57:30 <rah> oneinsect: just out of curiosity, what kind of code compiles on glibc but doesn't compile on musl?
2020-07-17 15:57:38 <oneinsect> for example M4
2020-07-17 15:57:46 <oneinsect> requires a patch to compile on glibc
2020-07-17 15:57:55 <oneinsect> and gcc 10.1
2020-07-17 15:57:58 <rah> O_o
2020-07-17 15:58:09 <rah> M4 as in the macro processor used by autoconf et al?
2020-07-17 15:58:13 <oneinsect> yes
2020-07-17 15:58:33 <rah> requires a patch to compile on glibc or requires a patch to compile on musl?
2020-07-17 15:58:48 <oneinsect> requires a patch to compile on glibc
2020-07-17 15:58:53 <oneinsect> with musl its fine
2020-07-17 15:58:56 <oneinsect> m4-1.4.18-glibc-change-work-around.patch
2020-07-17 15:59:00 <oneinsect> google that
2020-07-17 15:59:26 <oneinsect> update for glibc libio.h removal in 2.28+
2020-07-17 15:59:31 <rah> I take it M4 isn't a GNU program then
2020-07-17 15:59:33 <JonasRC> I need glibc to run factorio server
2020-07-17 15:59:39 <oneinsect> musl is perfectly fine
2020-07-17 16:00:00 <ikke> JonasRC: You can also look into something like flatpack
2020-07-17 16:00:04 <oneinsect> you shall soon have it
2020-07-17 16:00:14 <oneinsect> yes flatpack is a very good alternative
2020-07-17 16:00:28 <JonasRC> flatpack?
2020-07-17 16:00:31 <oneinsect> M4 is a GNU
2020-07-17 16:00:43 <ikke> flatpak I mean
2020-07-17 16:00:48 <ikke> https://www.flatpak.org/
2020-07-17 16:01:09 <maxice8> rah: code that use Glibc-isms
2020-07-17 16:01:15 <maxice8> flatpak is amazing :D
2020-07-17 16:01:41 <rah> maxice8: REALLY?!!!111
2020-07-17 16:01:52 <maxice8> yes
2020-07-17 16:02:15 <rah> odd that a GNU program doesn't compile with glibc
2020-07-17 16:02:44 <rah> presumably that's temporary until the next release
2020-07-17 16:03:44 <rah> oneinsect: gcc 10 was only released recently, how come that can't compile on glibc?
2020-07-17 16:04:00 <oneinsect> some issues rah:
2020-07-17 16:04:09 <oneinsect> there have some changes in gcc 10
2020-07-17 16:04:25 <oneinsect> let me give you an example
2020-07-17 16:06:46 <rah> please do
2020-07-17 16:07:02 <rah> (while I wait for my "apk add xonotic extremetuxracer quakespasm supertuxkart" to download everything)
2020-07-17 16:09:15 <oneinsect> take gdbm package as an example i am forced to e will need to inject -fcommon into CFLAGS via /etc/abuild.conf before compiling in gcc 10.1
2020-07-17 16:09:31 <oneinsect> forced to inject*
2020-07-17 16:10:00 <komnpomat>  oneinsect: LOLOLLO
2020-07-17 16:10:01 <oneinsect> https://bugs.gentoo.org/705898
2020-07-17 16:10:08 <oneinsect> gdbm-1.18.1-gcc-10.patch
2020-07-17 16:10:19 <oneinsect> so the mess is way bigger
2020-07-17 16:10:20 <komnpomat> that got absolutely nothing to do with glibc
2020-07-17 16:10:38 <oneinsect> i mean not with glibc ofcourse M4 had an issue with glibc
2020-07-17 16:10:43 <oneinsect> just giving an example
2020-07-17 16:10:51 <oneinsect> there are issues with both glibc and gcc 10.1
2020-07-17 16:11:15 <komnpomat> oneinsect: fcommon is a non issue and not musl/glibc dependent
2020-07-17 16:12:05 <komnpomat> oneinsect: read this --> https://wiki.gentoo.org/wiki/Gcc_10_porting_notes/fno_common
2020-07-17 16:12:39 <komnpomat> I don't code, but from what I understand is about having duplicate headers on your project
2020-07-17 16:12:54 <oneinsect> but it compiles fine on gcc 9.3
2020-07-17 16:13:20 <komnpomat> oneinsect: add -fnocommon with gcc, same error
2020-07-17 16:13:21 <oneinsect> they flipped the default in 10.1
2020-07-17 16:13:28 <oneinsect> yes indeed
2020-07-17 16:14:12 <oneinsect> but i have complaints with gcc 10.1
2020-07-17 16:14:12 <rah> oneinsect: that's a problem in compiling other programs with gcc 10, not a problem with gcc 10 itself
2020-07-17 16:14:22 <oneinsect> have NO*
2020-07-17 16:14:24 <rah> the gdbm issue
2020-07-17 16:14:39 <oneinsect> there are other issues i documented
2020-07-17 16:15:00 <oneinsect> its glibc which is messy
2020-07-17 16:15:05 <oneinsect> gcc 10.1 is fine
2020-07-17 16:15:13 <oneinsect> i mean not as clean as musl
2020-07-17 16:15:29 <komnpomat> troll!!
2020-07-17 16:15:35 <oneinsect> hence i cannot directly use the same APKBUILDs
2020-07-17 16:15:38 <oneinsect> as musl
2020-07-17 16:15:40 <komnpomat> troll alert!!
2020-07-17 16:16:30 <oneinsect> lol did you mean me komnpomat:
2020-07-17 16:17:09 <oneinsect> oneinsect -> troll -> (in folklore) an ugly creature depicted as either a giant or a dwarf.
2020-07-17 16:17:11 <komnpomat> you're not making a lot of sense
2020-07-17 16:17:19 <oneinsect> its okie
2020-07-17 16:17:32 <oneinsect> may be i am not making writing clearly
2020-07-17 16:18:21 <JonasRC> When did they mess up glibc?
2020-07-17 16:18:51 <rah> wait, there was a time when glibc wasn't a mess? :-)
2020-07-17 16:18:58 <oneinsect> i honestly dont mind, didnt take it heart, all i am trying to do is just compile those packages against glibc and gcc 10.1
2020-07-17 16:19:05 <oneinsect> thats all
2020-07-17 16:20:44 <rah> cool
2020-07-17 16:20:47 <rah> good luck
2020-07-17 16:21:23 <oneinsect> may be problem is not the right word
2020-07-17 16:21:33 <oneinsect> or messy or clean
2020-07-17 16:21:45 <oneinsect> may be those are not right words to express the issues i am facing
2020-07-17 16:21:49 <oneinsect> sorry for that
2020-07-17 16:23:11 <rah> I would suggest to be on the look out for opportunities to permit APKBUILDs to be shared
2020-07-17 16:23:31 <oneinsect> i will upload them to github in a few days
2020-07-17 16:24:27 <rah> I will offer some unsolicited advice
2020-07-17 16:24:49 <rah> creating software is great, and sharing software on the Net is great too
2020-07-17 16:24:52 <oneinsect> :D thank you!
2020-07-17 16:25:29 <rah> but much, much greater than creating and sharing software is creating software that can be maintained
2020-07-17 16:25:40 <oneinsect> indeed
2020-07-17 16:25:45 <oneinsect> 100% agree
2020-07-17 16:26:31 <rah> in this instance, what that means is that creating a hard fork of alpine is great but changing alpine so that it can cope with either musl or glibc is orders of magnitude greater
2020-07-17 16:27:27 <oneinsect> yes rightly put!
2020-07-17 18:18:27 <vpseg> I am attempting to bridge two of three network interfaces on my Alpine 3.12 system: br0 enslaves lan0 and lan1. I want lan0 and lan1 to have no address, and br0 to have an address from DHCP. Currently the config I have that matches the documentation on the topic gives lan0 and lan1 addresses from DHCP or APIPA, and br0 nothing (unless statically
2020-07-17 18:18:27 <vpseg> set). Any ideas? https://www.reddit.com/r/linuxquestions/comments/hsg8dd/bridged_networking_on_alpine_linux/
2020-07-17 18:25:56 <komnpomat> vpseg: lan0_net="null" in /etc/conf.d/net
2020-07-17 18:28:26 <vpseg> komnpomat, I still get an IP on lan0 after that entry
2020-07-17 18:31:22 <c705> webstrand: hi. were we chatting about libvirt last week?
2020-07-17 18:32:05 <nacelle> vpseg: getting two ips on two different nics via dhcp is not a trivial thing to do.  why do you want dhcp on the inside?
2020-07-17 18:32:55 <nacelle> (normally, it should be a static IP on the inside and it -serves- dhcp to a subnet)
2020-07-17 18:33:03 <vpseg> nacelle, I am getting DHCP on lan0 and nor br0. I need DHCP to set the address for br0, NOT either lan interface.
2020-07-17 18:33:08 <nacelle> or a static IP and you have another DHCP server
2020-07-17 18:33:16 <nacelle> no, why are you setting br0 with it?
2020-07-17 18:33:38 <nacelle> i'm ignoring your lower level issue and focusing on the "what in the heck are you doing here?" thing
2020-07-17 18:33:48 <vpseg> My DHCP server is not running on this box
2020-07-17 18:33:55 <nacelle> ok?
2020-07-17 18:34:03 <nacelle> can you do a static on br0?
2020-07-17 18:34:48 <vpseg> I _can_ but I want DHCP for the bridge. This isn't and x y  problem, I have this working on another distro but Alpine is being difficult
2020-07-17 18:34:56 <nacelle> ah ok
2020-07-17 18:39:40 <nacelle> and your box does what with two default routes?
2020-07-17 18:40:44 <nacelle> i suspect on the box that you have it working that you have two dhclient confs, one of them custom that ignores the gateway, but ... why are you here asking if you knew that? eh?
2020-07-17 18:48:29 <c705> i'm having libvirt stability issues. last night it worked, now it's not working again. nothing in the logs at all. if i strace virsh commands it hits the libvirtd socket and hangs. is anyone else having libvirtd issues? I want to blame either upstream or openrc for this, but i'm really not sure. 5.4.52-0-lts, thanks
2020-07-17 18:48:50 <Cogitri> It's upstream
2020-07-17 18:48:55 <c705> I swear nothing changed between last night and now except packages in edge, haven't looked at the changelog yet
2020-07-17 18:48:58 <c705> Cogitri: for sure?
2020-07-17 18:49:09 <Cogitri> https://gitlab.alpinelinux.org/alpine/aports/-/issues/11602#note_102715
2020-07-17 18:49:23 <c705> Cogitri: thank you so much, this has been driving me nuts for 2 weks or so now
2020-07-17 19:27:59 <Hello71> vpseg: try installing iproute2
2020-07-17 19:28:07 <Hello71> and I guess bridge-utils while you're at it
2020-07-17 19:38:04 <komnpomat> br0_net="null"
2020-07-17 19:38:19 <komnpomat> oh wait.
2020-07-17 19:38:33 <komnpomat> br0_net="dhcpd"
2020-07-17 19:38:53 <Hello71> wat
2020-07-17 19:41:32 <komnpomat> my bad, module_br0="udhcp" config_lan0="null"  config_lan1="null"
2020-07-17 19:43:13 <komnpomat> vpseg: are you still here?
2020-07-17 19:43:54 <komnpomat> c705: hanging is a symptom of borked libs
2020-07-17 19:44:36 <komnpomat> unable to initialize properly, since libvirt depends on so many things as  long as one needed is borked will prolly hang
2020-07-17 19:44:53 <komnpomat> c705: how to troubleshoot? convert your libvirt to command line
2020-07-17 19:45:18 <c705> komnpomat: I did, and already stated that strace shows the socket hanging
2020-07-17 19:45:52 <c705> i'm not debugging upstream, once I saw the bug on upstream, it's their problem to fix
2020-07-17 19:46:14 <c705> Will read more tonight, but now I have #52 on libvirt
2020-07-17 19:46:22 <komnpomat> c705: sorry I didn't read above.
2020-07-17 19:46:53 <c705> now libvirtd and musl are fighting about it
2020-07-17 20:07:21 <c705> Q: if I run mkimage.sh from aports/scripts/mkimage.sh, am I compiling the kernel during that process? My guess is yes
2020-07-17 20:07:55 <ikke> I would not expect so
2020-07-17 20:08:32 <c705> Hmm..OK
2020-07-17 20:08:48 <c705> One of these days, I'm gonna have to hack that script
2020-07-17 20:21:00 <vpseg> komnpomat Thank you, I'll check that
2020-07-17 20:39:57 <komnpomat> vpseg: do you have docs installed? then check /usr/share/doc/netifrc
2020-07-17 20:40:32 <vpseg> Okay, thank you
2020-07-18 13:20:13 <maxwilliamson> Hey guys, static build of ffmpeg (https://www.johnvansickle.com/ffmpeg/) can't seem to resolve dns addresses in alpine. Can anyone test this?
2020-07-18 13:21:17 <maxwilliamson> Outputs "[tcp @ 0x6770240] Failed to resolve hostname bitdash-a.akamaihd.net: System error"
2020-07-18 13:23:04 <maxwilliamson> seems to be linked with this error https://sourceware.org/bugzilla/show_bug.cgi?id=10652
2020-07-18 13:29:13 <Cogitri> That's a bug report for glibc and not musl though?
2020-07-18 13:47:17 <maxwilliamson> yeah, because it's a static build, so it does not link to musl. What I was wondering was why does this happen only on alpine
2020-07-18 13:52:29 <Cogitri> I don't think static builds against glibc are truly static, so binaries might not function correctly if you don't use glibc in your host system
2020-07-18 13:59:02 <mps> Cogitri: you are right
2020-07-18 14:00:34 <mps> that is reason why devs build true static binaries with musl
2020-07-18 14:09:14 <maxwilliamson> ugh that's a shame
2020-07-18 14:09:43 <maxwilliamson> had to use debian docker image instead, and it's pita
2020-07-18 14:10:09 <maxwilliamson> alpine simplifies things magically, you feel it when you use other distro images
2020-07-18 17:02:33 <komnpomat> what's the equivalent in busybox of 'sort -h'??
2020-07-18 17:02:46 <ikke> there is no equivalent
2020-07-18 17:07:21 <komnpomat> sooo
2020-07-18 17:07:25 <komnpomat> what then?
2020-07-18 17:09:13 <ikke> If you need it, install coreutils
2020-07-18 17:09:23 <ikke> That contains the full sort program
2020-07-18 17:10:51 <komnpomat> crazy
2020-07-18 18:16:23 <eydaimon> So if I install alpine with zfs on root encrypted, it works fine. But if I add new HDDs, then it refuses to boot saying it can't fine the rpool for alpine. I can disconnect the drives for it to prompt for zfs password and boot as before. Any ideas how to fix this ?
2020-07-18 18:18:16 <komnpomat> eydaimon: which tool do you use to create your intrd image?
2020-07-18 18:18:52 <komnpomat> eydaimon: booting zfs with multiple pools works just fine, either extra zfs or non-zfs drive
2020-07-18 18:19:25 <eydaimon> it does, but not when I add extra devices later
2020-07-18 18:19:26 <eydaimon> https://wiki.alpinelinux.org/wiki/Alpine_Linux_with_root_on_ZFS_with_native_encryption
2020-07-18 18:19:29 <eydaimon> I followed the instructions here
2020-07-18 18:26:17 <eydaimon> komnpomat  I suspect if I reinstall with those devices connected, I will not have that problem, but I would like to understand the issue incase I need to connect more devices later
2020-07-18 18:27:02 <komnpomat> technically there shouldn't be
2020-07-18 18:27:23 <eydaimon> I can reboot and show the exact error if that's helpful
2020-07-18 18:29:53 <komnpomat> eydaimon: update zfs cache perhaps?
2020-07-18 18:30:22 <komnpomat> eydaimon: either way zfs+encryption you need to rebuild your intitrd
2020-07-18 18:32:26 <eydaimon> komnpomat why would I need to rebuild my initrd when it boots fine without the additional devices ?
2020-07-18 18:32:36 <eydaimon> how does one update zfs cache ?
2020-07-18 18:33:01 <eydaimon> https://www.dropbox.com/s/ak4du3dosnunmqv/2020-07-18%2011.30.14.jpg?dl=0
2020-07-18 18:33:04 <eydaimon> that's the error
2020-07-18 18:33:20 <komnpomat> eydaimon: the additional devices DO make changes in zfs
2020-07-18 18:33:40 <eydaimon> so every time I add a new device, I need to rebuild my initrd ?
2020-07-18 18:33:43 <komnpomat> eydaimon: zfs isn't easily hotpluggable
2020-07-18 18:34:33 <komnpomat> eydaimon: wow that sucks, I will keep this in mind whenever I try encryption
2020-07-18 18:34:37 <eydaimon> eh ? that's the point of zfs. that  it's replacable on the fly. I can connect them while it's running after booting with them disconnected, and import them just fine. Am I misunderstanding ?
2020-07-18 18:34:54 <eydaimon> zfs is very easily hotpluggable
2020-07-18 18:38:15 <eydaimon> also, initrd isn't available on the /boot filesystem, it's available on root, which is encrypted zfs. So it doens't make sense to tamper with initrd
2020-07-18 18:38:37 <eydaimon> it would  be something on /boot that changes
2020-07-18 18:40:14 <eydaimon> well, since no one knows i"ll just reinstal. not like it takes long
2020-07-18 18:43:21 <eydaimon> actually, I'll wait a bit more to see if someone else knows Cogitri perhaps ?
2020-07-18 18:43:37 <eydaimon> ikke is always resourceful too :)
2020-07-18 18:44:07 <ikke> my knowledge around zfs is basically 0 :)
2020-07-18 18:44:46 <eydaimon> 😱
2020-07-18 18:45:16 <eydaimon> any idea who would know this type of thing ?
2020-07-18 18:45:36 <eydaimon> I had the same problem with my experimental machine, and reinstalling with the disk connected solved the issue
2020-07-18 18:46:51 <komnpomat> eydaimon: that got to do with zfs cache
2020-07-18 18:47:17 <komnpomat> eydaimon: or actually your kernel command line
2020-07-18 18:47:32 <komnpomat> but again I am not familiar with the encrytion set up
2020-07-18 18:50:44 <eydaimon> so how do I fix the zfs cache ?
2020-07-18 18:50:45 <Cogitri> <eydaimon "actually, I'll wait a bit more t"> Ah?
2020-07-18 18:51:28 <eydaimon> Cogitri I was going to reinstall to fix the issue, but I deicded to wait more  because I want to understand the issue
2020-07-18 18:51:50 <eydaimon> Cogitri do you have any idea ?
2020-07-18 18:51:59 <Cogitri> Fix ZFS cache? 🤔
2020-07-18 18:52:14 <Cogitri> You mean to import the pool?
2020-07-18 18:52:15 <komnpomat> eydaimon: the thing is that I am not familiar with encryption so I don't know which kernel command line are needed
2020-07-18 18:52:22 <eydaimon> Cogitri I added some HDD's with zfs on after rebooting, and it seems the loader gets confused. https://www.dropbox.com/s/ak4du3dosnunmqv/2020-07-18%2011.30.14.jpg?dl=0
2020-07-18 18:52:43 <komnpomat> eydaimon: most likely the extra drives is upsetting some probing and throwing it out of order.
2020-07-18 18:52:52 <eydaimon> komnpomat the reinstall on the experimental machine that works now doesn't show any extra commands tha tI can tell but I'm not sure where to look
2020-07-18 18:53:01 <eydaimon> komnpomat that's what I think too, but I don't know how to fix
2020-07-18 18:53:30 <Cogitri> You can import by disk id too, no need to use the cache
2020-07-18 18:53:41 <komnpomat> eydaimon: there are various ways of setting zfs kernel command line, some more losely some more strict.
2020-07-18 18:56:04 <eydaimon> Cogitri yeah, but it's confusing it on boot. I can't import during boot time
2020-07-18 18:56:16 <eydaimon> Cogitri did you see the screenshot ?
2020-07-18 18:57:02 <komnpomat> eydaimon: if you unplug the drive boots?
2020-07-18 18:57:25 <eydaimon> komnpomat indeed
2020-07-18 18:58:07 <eydaimon> " I can disconnect the drives for it to prompt for zfs password and boot as before."
2020-07-18 18:58:26 <eydaimon> https://forums.freebsd.org/threads/gptzfsboot-error-128-after-adding-new-disks.65677/
2020-07-18 18:59:33 <eydaimon> "So after a lot of experimenting I figured out what the problem was. It looks like FreeBSD was probing all disks and was throwing the error when one of them didn't have a valid partition table. "  This may be my issue. One of the disks is encrypted with FreeBSD's GELI
2020-07-18 19:01:30 <eydaimon> nope. no good
2020-07-18 19:03:10 <eydaimon> although it does work if I disconnect the drive that has zfs on only
2020-07-18 19:27:30 <eydaimon> `exlinux --install /boot` also failed
2020-07-18 19:46:51 <komnpomat> eydaimon: wipefs -a
2020-07-18 19:54:54 <eydaimon> on what device ? the boot device or the other ZFS device ?
2020-07-18 19:55:58 <eydaimon> wipefs: error: /dev/sda: probing initialization failed: Resource busy
2020-07-18 19:57:40 <komnpomat> eydaimon: wait!
2020-07-18 19:58:02 <komnpomat> eydaimon: in the formatted disk that's giving you issues
2020-07-18 19:58:20 <komnpomat> wipefs erases ALL signatures of a hdd
2020-07-18 19:58:22 <eydaimon> there's no formatted disk that's giving me issues. The one that's giving me issues is a second drive with zfs
2020-07-18 19:58:41 <eydaimon> it wasn't connected when I did the original install
2020-07-18 19:59:36 <komnpomat> eydaimon: this is what I would do, boot without the drive, plug it and import it, the rebuild your initrd/boot/grup whatever.
2020-07-18 19:59:39 <komnpomat> then reboot
2020-07-18 20:00:05 <eydaimon> komnpomat I mentioned this earlier, but the initrd file isn't even accessible at mount time, so how can that play a part ?
2020-07-18 20:00:54 <komnpomat> eydaimon: redo that wiki guide about zfs+encrytion.
2020-07-18 20:01:06 <eydaimon> oh nvm
2020-07-18 20:01:10 <komnpomat> in other words, repeat the guide but with the drive added this time
2020-07-18 20:01:14 <eydaimon> you mean /dev/initrd ?
2020-07-18 20:01:40 <komnpomat> dunno, whatever the guide says.
2020-07-18 20:02:05 <eydaimon> komnpomat I know it will work if I do that. I said so earlier. I appreciate your help, but I'm getting the impression you aren't reading what I write
2020-07-18 20:02:40 <komnpomat> eydaimon: I think the zfs cache is added to the initrd image, so when you added the new drive the zfs cache info in the initrd is missing that.
2020-07-18 20:03:12 <eydaimon> can I not update the initrd some other way ?
2020-07-18 20:03:31 <eydaimon> komnpomat also, I wrote the guide based  😅
2020-07-18 20:04:01 <komnpomat> eydaimon: with dracut
2020-07-18 20:04:09 <eydaimon> dracut ?
2020-07-18 20:04:58 <eydaimon> apologies about your comments about initrd. I thought it was in /etc 😳😳😳
2020-07-18 20:05:27 <komnpomat> w00t
2020-07-18 20:05:41 <komnpomat> the initrd is what the kernel uses to boot up
2020-07-18 20:06:16 <komnpomat> early boot files, I am most certain for encrytion is needed, and for zfs for that matter.
2020-07-18 20:07:21 <komnpomat> kernel -> initrd (with early boot files) -> system
2020-07-18 20:07:49 <eydaimon> if initrd is the issue, can't I regenerate it ?
2020-07-18 20:08:15 <komnpomat> eydaimon: exactly, that's what I was trying to do, so it'll pick up the updated zfs cache.
2020-07-18 20:09:59 <eydaimon> Did I miss a step ? How do I do that ?
2020-07-18 20:10:12 <komnpomat> eydaimon: initrd packs early boot files, extra modules and so on.
2020-07-18 20:11:04 <komnpomat> eydaimon: dracut. but bear in mind each guide got their own way of doing it. So make sure you adapt all the encrytion bits to dracut
2020-07-18 20:11:32 <komnpomat> eydaimon: or use whatever alpine uses which I don't know how it builds an initrd
2020-07-18 20:12:08 <eydaimon> what is dracut ?
2020-07-18 20:14:29 <komnpomat> anybody else?
2020-07-18 20:14:47 <eydaimon> ls -l /dev/initrdls: cannot access '/dev/initrd': No such file or directory
2020-07-18 20:15:05 <eydaimon> it doesn't even exist
2020-07-18 20:15:31 <komnpomat> eydaimon: I haven't delved deeper into alpine, I use stock alpine settings without anything fancy.
2020-07-18 20:15:55 <eydaimon> komnpomat follow my guide :)  it was easy. just adding another drive later is confusing
2020-07-18 20:16:07 <komnpomat> eydaimon: ls /boot
2020-07-18 20:17:06 <eydaimon> System.map-lts  config-lts     initramfs-lts  ldlinux.sys   libutil.c32  mboot.c32  vesamenu.c32 boot@           extlinux.conf  ldlinux.c32    libcom32.c32  lost+found/  menu.c32   vmlinuz-lts
2020-07-18 20:19:19 <artok> mkinitfs is to create init ramdisk
2020-07-18 20:19:41 <eydaimon> I was looking at the manpage for that after seeing that in setup-disk
2020-07-18 20:21:51 <eydaimon> but I don't know where to go with that
2020-07-18 20:22:58 <eydaimon> I reckon setup-disk has to do something to get things to work
2020-07-18 20:27:10 <artok> setup-disk adds persistent thingies
2020-07-18 20:30:27 <eydaimon> well, I'm just going to reinstall
2020-07-18 20:30:36 <eydaimon> seems no one knows the answer
2020-07-18 20:36:11 <artok> google might, and I don't even know your original question
2020-07-18 20:38:18 <eydaimon> So if I install alpine with zfs on root encrypted, it works fine. But if I add new HDDs, then it refuses to boot saying it can't fine the rpool for alpine. I can disconnect the drives for it to prompt for zfs password and boot as before. Any ideas how to fix this ?
2020-07-18 20:38:25 <eydaimon> ☝️ that
2020-07-18 20:38:50 <eydaimon> https://www.dropbox.com/s/ak4du3dosnunmqv/2020-07-18%2011.30.14.jpg?dl=0  ← the error
2020-07-18 20:40:23 <nacelle> have you tried doing something like moving the new disk into a different position on the motherboard?  (maybe its coming up earlier?)
2020-07-18 20:41:39 <nacelle> normally something thats coming up in a scan order on a modern system is either positionally inserted wrong for the expected order, or the new disk has an existing filesystem on it thats being picked up
2020-07-18 20:41:42 <eydaimon> it's /dev/sda   isn't that the first disk ?
2020-07-18 20:42:00 <nacelle> that is, thats what you dont want, right?
2020-07-18 20:42:05 <nacelle> you want your existing disks to be first
2020-07-18 20:42:05 <eydaimon> it is what I want
2020-07-18 20:42:10 <nacelle> the new disk shouldnt be
2020-07-18 20:42:12 <eydaimon> that's where the OS is installed
2020-07-18 20:42:16 <eydaimon> no, the new disk is /dev/sdd
2020-07-18 20:42:30 <nacelle> but thats after boot order
2020-07-18 20:42:35 <nacelle> not boot loader stuff
2020-07-18 20:43:24 <nacelle> so if its not positional (which isnt very likely in 2020, but eh, i'm troubleshooting blindly), do you need anything on the disk you're inserting? have you wiped it if not?  (cat /dev/zero > /dev/sdd ) , etc.
2020-07-18 20:43:26 <eydaimon> I can try moving it, but that doesn't feel like a real fix
2020-07-18 20:43:43 <eydaimon> I need everything on the disk I'm inserting
2020-07-18 20:43:59 <nacelle> its already zfs?
2020-07-18 20:44:14 <artok> how about using disk by id ?
2020-07-18 20:44:18 <eydaimon> it's already zfs
2020-07-18 20:44:24 <artok> or by partuuid
2020-07-18 20:44:28 <nacelle> then this is proabbly a zfs specific issue
2020-07-18 20:44:34 <eydaimon> both are zfs
2020-07-18 20:44:36 <nacelle> i would check in a zfs channel, not here
2020-07-18 20:44:42 <eydaimon> the root drive, and the new drive
2020-07-18 20:44:52 <eydaimon> OK
2020-07-18 20:45:14 <nacelle> i would also ask yourself why zfs if its like this, but ymmv :-)
2020-07-18 20:45:34 <artok> zfs is hip
2020-07-18 20:45:45 <nacelle> zfs was hip a decade ago
2020-07-18 20:47:06 <eydaimon> zfs is still hip
2020-07-18 20:47:27 <eydaimon> because it's an awesome fs
2020-07-18 20:47:34 <Cogitri> At least for servers/NASes it's still pretty widespread
2020-07-18 20:48:02 <eydaimon> and because who hasn't had a drive fail? being able to just add a new one when one fails in a sliver is SUPER useful
2020-07-18 20:48:17 <eydaimon> s/sliver/mirror
2020-07-18 20:49:01 <eydaimon> granted, I don't need to have zfs on root. But when I lookd at LUKS on root, that looked more complicated than encrypted ZFS
2020-07-18 20:49:03 <eydaimon> to set up
2020-07-18 20:54:49 <eydaimon> well crap. reinstalling didn't work. it's still getting confused
2020-07-18 20:55:27 <artok> did it by id ?
2020-07-18 20:55:53 <eydaimon> what do you mean did it by id ?
2020-07-18 20:55:57 <artok> not by /dev/sd?
2020-07-18 20:56:40 <eydaimon> artok: I follow this https://wiki.alpinelinux.org/wiki/Alpine_Linux_with_root_on_ZFS_with_native_encryption
2020-07-18 20:57:25 <eydaimon> I guess one alternative is to remove the drives every time I boot 😅
2020-07-18 20:57:28 <artok> use uuid instead of /dev/sda2 for example
2020-07-18 20:57:41 <eydaimon> artok does your nick mean "heart" ?
2020-07-18 20:57:55 <artok> no
2020-07-18 20:58:02 <eydaimon> artok not sure that makes sense to do given the installation instructions there
2020-07-18 20:58:04 <artok> Arto is my real name
2020-07-18 20:58:21 <eydaimon> ah
2020-07-18 20:58:30 <artok> last name starts with K
2020-07-18 20:58:55 <eydaimon> nice name. I like short names
2020-07-18 20:58:55 <artok> but using UUID everywhere, fstab etc is recommended anyways
2020-07-18 20:59:46 <eydaimon> you mean instead of `/dev/sda` here I would use echo -e "/dev/sda1: start=1M,size=100M,bootable\n/dev/sda2: start=101M" | sfdisk --quiet --label dos <ID> ?
2020-07-18 20:59:47 <eydaimon> for example
2020-07-18 20:59:52 <eydaimon> given that manual
2020-07-18 21:00:06 <eydaimon> er, well, replace it everywhere with id
2020-07-18 21:00:11 <eydaimon> or where exactyl would I use the ID ?
2020-07-18 21:00:23 <artok> when creating the pool?
2020-07-18 21:01:12 <artok> just guessing, but that usually does the job
2020-07-18 21:02:19 <artok> for example if I have bunch of USB memory attached on my rpi4, using UUID is needed in the kernel commandline to be able to mount correct partition to root
2020-07-18 21:02:23 <artok> and others also
2020-07-18 21:02:42 <artok> makes possible to have your disk in any order
2020-07-18 21:16:56 <eydaimon> interesting
2020-07-18 21:18:17 <eydaimon> arto how do I get the id ?
2020-07-18 21:18:25 <eydaimon> artok
2020-07-18 21:19:16 <Cogitri> blkid
2020-07-18 21:19:29 <eydaimon> thanks
2020-07-18 21:19:31 <eydaimon> I'll give it a shot
2020-07-18 21:19:42 <eydaimon> alpine is luckily a breeze to install :)
2020-07-18 21:19:46 <Cogitri> Just deleting /etc/zfs/cache and importing by if should just work though
2020-07-18 21:19:53 <Cogitri> s/if/id/
2020-07-18 21:20:17 <eydaimon> Cogitri I'll try that first
2020-07-18 21:21:29 <eydaimon> can I use that ID anywhere in place of /dev/sda then ?
2020-07-18 21:22:14 <eydaimon> I'm also curious that it worked on my experimental setup first.
2020-07-18 21:23:22 <eydaimon> OK, changing drive location did work
2020-07-18 21:24:11 <Cogitri> Just do `zpool import -d /dev/disk/by-id $poolname`
2020-07-18 21:24:14 <eydaimon> so that would explain that
2020-07-18 21:24:32 <Cogitri> Afterwards your pool will always be imported by drive ID instead of /dev/sd* name
2020-07-18 21:24:34 <eydaimon> Cogitri it works even when it's mounted ? cool
2020-07-18 21:25:07 <Cogitri> I think you need to export and then import with the that command
2020-07-18 21:26:06 <eydaimon> ok
2020-07-18 21:28:07 <eydaimon> should I use that in /boot/extlinux.conf too then ?
2020-07-18 21:28:22 <eydaimon> or maybe it's enough to just specify the ID there ?
2020-07-18 21:28:37 <eydaimon> right now it's: APPEND root=rpool/ROOT/alpine modules=sd-mod,usb-storage,zfs nomodeset quiet rootfstype=zfs
2020-07-18 21:28:45 <eydaimon> so APPEND root=32e1ba5d/ROOT/alpine modules=sd-mod,usb-storage,zfs nomodeset quiet rootfstype=zfs
2020-07-18 21:28:47 <eydaimon> instead ?
2020-07-18 21:29:29 <eydaimon> actually, not sure that makes sense
2020-07-18 21:36:47 <eydaimon> Cogitri the disk isn't listed in /dev/   and /dev/disk doesn't exist
2020-07-18 21:38:16 <eydaimon> not when booting from the alpine iso anyway
2020-07-18 21:40:33 <eydaimon> blkid doesn't even give an id
2020-07-18 21:48:02 <eydaimon> what's the `sh: active: unknown operand`   mean ?
2020-07-18 22:01:15 <eydaimon> I take back what I said about it working when changing drive location. that bay wasn't connected
2020-07-18 22:02:06 <eydaimon> I'm stuck
2020-07-18 22:02:24 <eydaimon> luckily it's easy to remove the drive from the bay, and I never reboot
2020-07-18 22:02:37 <eydaimon> so I guess if I have to reboot, I can just remove it
2020-07-18 22:02:46 <eydaimon> or them, rather
2020-07-18 22:29:35 <lp2skyline> Is there anyway to reliably install Nvidia drivers? I want to use Alpine on my laptop, although alpine might not be the right distribution for me.
2020-07-18 22:33:12 <Hello71> eydaimon: personally I would recommend just using a fat/ext4 /boot
2020-07-18 22:33:19 <Hello71> er
2020-07-18 22:33:23 <Hello71> sorry, ignore that
2020-07-18 22:33:40 <Hello71> got you confused with someone else
2020-07-18 22:40:29 <eydaimon> sometimes I'm someone else 😂
2020-07-18 22:55:01 <hodapp> is https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi still pretty up-to-date if I want to migrate a copy of an existing Ras Pi installation to a persistent one?
2020-07-18 22:55:26 <hodapp> pretty much all I'm trying to do is make a dev environment where I can build things with golang & GCC and run them on the diskless Ras Pi install
2020-07-18 23:20:39 <c705> hodapp: i would take it with a grain of salt
2020-07-19 00:54:10 <eydaimon>  /filter add joinquit * irc_join,irc_part,irc_quit *
2020-07-19 00:54:11 <eydaimon> oops
2020-07-19 02:54:09 <LACampbell> lp2skyline: I'm considering installing alpine on my laptop too. I really like it as a web and dev server. not sure how many limitations I'll run into with a desktop though
2020-07-19 02:54:28 <LACampbell> all I know is I am sick to death of debian and its awful package management
2020-07-19 02:54:41 <tehcloud> LACampbell, I moved from Debian to Alpine
2020-07-19 02:54:52 <tehcloud> now I'm about 7 months in
2020-07-19 02:55:17 <tehcloud> I use it on a really cheap Chromebook type laptop and due to its minimalism it runs very well
2020-07-19 02:57:06 <LACampbell> tehcloud: anything that was more difficult? the usual shit - wifi drivers/trackpad/sound/video
2020-07-19 02:57:36 <tehcloud> I had an issue with my trackpad but it was fixed by someone modifying the default kernel build options to support it
2020-07-19 02:58:04 <tehcloud> if you're a relatively experienced Linux operator you should be fine
2020-07-19 02:58:26 <tehcloud> it may remind you of many things Debian hid from you
2020-07-19 02:58:39 <tehcloud> but it's nothing insurmountable
2020-07-19 02:59:12 <LACampbell> if I compare alpine on the server vs debian on the server, i don't feel like debian hides anything. I feel like it just badly papers over stuff and puts lipstick on a few pigs
2020-07-19 02:59:19 <LACampbell> but desktop is usually a different story
2020-07-19 02:59:24 <LACampbell> good to hear though
2020-07-19 02:59:44 <tehcloud> it depends on what you want out of a desktop OS really
2020-07-19 02:59:54 <tehcloud> I'm happy with a tiling window manager and a terminal as my file manager
2020-07-19 03:00:11 <tehcloud> getting GNOME working might take a little extra effort
2020-07-19 03:00:20 <tehcloud> or that monstrosity they call KDE
2020-07-19 03:00:23 <LACampbell> nah, I'll probably just use openbox or something
2020-07-19 03:00:33 <LACampbell> but you can run gtk and qt apps, right?
2020-07-19 03:00:38 <tehcloud> yeah
2020-07-19 03:00:57 <tehcloud> are you a developer?
2020-07-19 03:01:08 <LACampbell> yes. so I will have to bite the bullet and learn vim properly :D
2020-07-19 03:01:12 <tehcloud> reason I ask is because I chose Alpine partially for musl
2020-07-19 03:01:35 <tehcloud> developing software against musl can often reveal how shitty of a programmer you are
2020-07-19 03:01:39 <LACampbell> lol
2020-07-19 03:01:43 <tehcloud> makes your software stronger :p
2020-07-19 03:02:02 <LACampbell> I'm a web dev, so it's all pretty much invisbile to me. if a browser can run on it, I'm good
2020-07-19 03:02:32 <LACampbell> I already use it to run some node.js monstrosity for work, in WSL
2020-07-19 03:03:01 <tehcloud> how is WSL?
2020-07-19 03:03:32 <LACampbell> mostly fine. but it can only run in its little window
2020-07-19 03:03:38 <tehcloud> I wrote multimedia oriented application recently that has some issues on Windows and I'm thinking about just seeing how WSL does
2020-07-19 03:03:46 <LACampbell> I honestly miss linux on the desktop. windows 10 feels so limited in comparison
2020-07-19 03:04:14 <tehcloud> it runs perfectly on Linux and the 4 most popular BSDs but not Windows :p
2020-07-19 03:04:33 <LACampbell> full disclaimer, I don't do any cross compiling
2020-07-19 03:04:37 <LACampbell> not actually sure how that works
2020-07-19 03:04:49 <LACampbell> like if you can compile a windows EXE from WSL
2020-07-19 03:05:03 <LACampbell> this is a C++ or C thing right?
2020-07-19 03:05:10 <tehcloud> ISO C99
2020-07-19 03:05:23 <tehcloud> 2 external deps, SDL and epoxy
2020-07-19 03:05:37 <tehcloud> my problems may be SDL problems
2020-07-19 03:07:09 <LACampbell> I used to do cross platform native dev with qt. I can't even remember how it worked
2020-07-19 03:08:44 <tehcloud> I've done some Qt stuff in the past, can't say I had fun :p
2020-07-19 03:08:55 <tehcloud> I'm kind of a C religious extremist
2020-07-19 03:09:11 <LACampbell> ha!
2020-07-19 03:09:17 <LACampbell> yeah Qt was... interesteing
2020-07-19 03:10:13 <LACampbell> I mean C++ is complex enough by itself, it could get a bit overwhelming with that + Qts own object system that uses macros
2020-07-19 03:10:58 <tehcloud> I find most of these big libs suck
2020-07-19 03:11:13 <tehcloud> I'm using SDL which sucks, but it's the best option for what I need
2020-07-19 03:11:24 <tehcloud> if GLFW3 can give me force feedback I'll switch
2020-07-19 03:13:13 <LACampbell> haha me too re: big libs. I suppose it's no coincidence we both find ourselves using alpine
2020-07-19 03:13:21 <LACampbell> it's that kind of aesthetic
2020-07-19 03:13:45 <LACampbell> the thing for me is - I love a good abstraction. but it has to be a really good one. if it's a shit one, it's worse than no abstraction
2020-07-19 03:13:47 <tehcloud> yeah, there are many different tastes in technology
2020-07-19 03:14:09 <tehcloud> some people think bloated and complex is a remarkable achievement, while others think it is shit
2020-07-19 03:14:23 <tehcloud> some people view minimalism as more elegant
2020-07-19 03:14:59 <LACampbell> have you grown to kind of hate software and just want less of it? less code, less libraries
2020-07-19 03:15:04 <tehcloud> some people legimately prefer the Corolla to the Escalade
2020-07-19 03:15:16 <tehcloud> lol, oh yeah... extreme minimalist in every area of my life
2020-07-19 03:15:27 <tehcloud> current goal is to reduce my belongings to nothing but a computer and cooking utensils
2020-07-19 03:15:39 <LACampbell> haha
2020-07-19 03:15:53 <LACampbell> well I am not quite that minimalist :)
2020-07-19 03:16:12 <LACampbell> the thing is I do front end dev. not sure if you're familiar of it, but it's a horrible mess of worthless bloated abstractions
2020-07-19 03:16:41 <LACampbell> are you in embedded or systems dev?
2020-07-19 03:17:24 <tehcloud> I'm a sysadmin by profession, C programmer by hobby, mostly because nobody will hire me for programming jobs because C programming jobs are hard to come by :p
2020-07-19 03:17:41 <tehcloud> and yes, very familiar with web software stack :(
2020-07-19 03:18:56 <LACampbell> yeah I guess even embedded programming is a lot of C these days
2020-07-19 03:19:08 <tehcloud> a lot of it is also C++
2020-07-19 03:19:38 <LACampbell> sysadmin! I miss those. in my market they all want developers doing "dev ops"
2020-07-19 03:19:47 <tehcloud> well, that is actually what I do :p
2020-07-19 03:19:50 <tehcloud> sysadmin evolved into that
2020-07-19 03:20:07 <LACampbell> that's why I moved out of back end dev
2020-07-19 03:20:12 <tehcloud> basically sysadmins had to learn to program
2020-07-19 03:20:13 <LACampbell> I respect sysadmins enough to know I am not one
2020-07-19 03:20:30 <LACampbell> I feel like sysadmins learning programming is better than vice versa
2020-07-19 03:20:37 <tehcloud> lol.... yes
2020-07-19 03:21:11 <tehcloud> but since I do both I understand the pitfalls of getting stuck in one or the other paradigm
2020-07-19 03:21:37 <tehcloud> most admins don't appreciate the level of skill required to write good code
2020-07-19 03:21:49 <tehcloud> most programmer don't appreciate the skill required to manage infrastructure
2020-07-19 03:22:12 <LACampbell> alpine linux is part of my 'sysadmin' strategy, if you could call it that. it's so primitive
2020-07-19 03:22:23 <LACampbell> to deploy, it's one rsync command
2020-07-19 03:22:25 <LACampbell> I love it
2020-07-19 03:22:27 <tehcloud> it will teach you a lot
2020-07-19 03:22:35 <tehcloud> so will Gentoo :p
2020-07-19 03:22:38 <LACampbell> I tried to get into docker and AWS and...
2020-07-19 03:25:19 <tehcloud> I spent about 5 years in AWS daily
2020-07-19 03:25:32 <tehcloud> can't say I like where this industry is going
2020-07-19 03:26:37 <LACampbell> do you think serverless is a fad?
2020-07-19 03:26:38 <tehcloud> my main issue is that I worry about relying on these cloud vendors to dictate everything based on business goals, and it becoming harder to do things independent of them
2020-07-19 03:27:03 <tehcloud> you pretty much have to use AWS or one of the lesser competitors these days
2020-07-19 03:27:09 <tehcloud> serverless is not a fad
2020-07-19 03:31:03 <LACampbell> then I've made the right decision to leave backend :)
2020-07-19 03:33:29 <tehcloud> Frontend devs get overpaid right now, it's like "data scientists" a few years back
2020-07-19 03:33:35 <tehcloud> anything that overpays is a great bet
2020-07-19 03:33:52 <tehcloud> ride that wave and go to the next overpaying job
2020-07-19 03:34:27 <tehcloud> now the definition of data scientist is getting more stringent, but before it was anyone who could write Python and took a stats course :p
2020-07-19 03:37:41 <LACampbell> it really depends on the market. front end devs dont seem to get any more than other devs
2020-07-19 03:37:58 <LACampbell> I'm self employed now though, so it's less about what I do technically and more about the problems I can solve
2020-07-19 03:37:58 <tehcloud> but as you know, their work is significantly easier
2020-07-19 03:38:09 <LACampbell> than what?
2020-07-19 03:38:14 <tehcloud> although, they do deal with far more pain with the shitty JS libs that exist
2020-07-19 03:38:15 <LACampbell> I've done desktop, backend, and front end dev
2020-07-19 03:39:00 <tehcloud> well, I guess I'm talking about what I've seen personally but every area of dev has varying levels of complexity
2020-07-19 03:39:20 <LACampbell> I'd say a lot of front end dev complexity is self imposed
2020-07-19 03:39:32 <tehcloud> I've done some WebGL stuff so I've seen how ugly things can get there :p
2020-07-19 03:40:23 <LACampbell> I felt like win forms desktop apps were probably the easiest
2020-07-19 03:40:50 <tehcloud> a lot of it depends on what you're actually writing too
2020-07-19 03:41:19 <LACampbell> yeah. I'm doing web apps, so there's not much focus on design for me
2020-07-19 03:41:35 <tehcloud> I used to work with GTK a lot
2020-07-19 03:41:46 <tehcloud> but then I revolted and said no more UI toolkits again
2020-07-19 03:42:10 <LACampbell> gtk3 got me into using so many command line programs, because all my favourite gtk2 apps either disappeared or looked like shit
2020-07-19 03:42:38 <tehcloud> yep, I'm going to learn curses soon
2020-07-19 03:42:45 <tehcloud> that's the closest I'll go to a UI toolkit
2020-07-19 03:43:08 <LACampbell> wxwidgets seems alright. but I'm not sure what the market for desktops apps still is in 2020
2020-07-19 03:43:12 <tehcloud> plus curses apps make you look leet
2020-07-19 03:43:58 <LACampbell> I don't want to go around the electron route. I've got no problem with html/css/js, but I do have a problem with starting a new chrome instance for an app. I mean chrome is practically an OS
2020-07-19 03:44:16 <tehcloud> well, it's going in that direction... everything in the browser
2020-07-19 03:44:22 <tehcloud> I don't think electron is a long term thing
2020-07-19 03:44:25 <LACampbell> I honestly think embedded and systems and programmers would make the best front end devs
2020-07-19 03:44:29 <tehcloud> it's like the transition
2020-07-19 03:44:34 <LACampbell> I think back end devs kind of ruined it :p
2020-07-19 03:45:18 <tehcloud> I realized a long time ago that Windows 3.11 was a faster and smoother experience than what I was using at that time
2020-07-19 03:45:32 <tehcloud> and that is still the case if you make that comparison at many points in time
2020-07-19 03:45:49 <LACampbell> what do you mean?
2020-07-19 03:46:04 <tehcloud> basically there seems to be this threshold we are willing to handle for application startup and fluidity
2020-07-19 03:46:52 <tehcloud> this is kind of how I ended up with Alpine
2020-07-19 03:46:59 <LACampbell> ahh
2020-07-19 03:47:06 <tehcloud> just the basics, no bullshit running in the background
2020-07-19 03:47:13 <LACampbell> there are only a few areas where this matters anymore
2020-07-19 03:47:15 <tehcloud> I can make it feel lightning fast
2020-07-19 03:48:13 <tehcloud> in many ways I think we're making everything worse on the web these days too
2020-07-19 03:48:18 <tehcloud> "WE USE COOKIES"
2020-07-19 03:48:31 <tehcloud> yes, yes... I know :p
2020-07-19 03:56:56 <LACampbell> haha, I have a plugin for that
2020-07-19 03:57:00 <LACampbell> "I don't care about cookies"
2020-07-19 03:57:36 <LACampbell> the weird thing about the web is - there's so much data showing that slow websites negetiavely impact bounce rate, which then has an effect on sales
2020-07-19 03:57:44 <LACampbell> but website performance isn't a big topic
2020-07-19 03:58:36 <tehcloud> if it's not there in a second or two, then it is assumed down
2020-07-19 04:01:20 <LACampbell> or people just get frustrated then go to your competitor
2020-07-19 04:01:29 <LACampbell> but how do you make sites fast? it's not doing what facebook does...
2020-07-19 07:34:24 <opal> guess i'll bump my raspi to alpine edge
2020-07-19 07:34:37 <opal> i want to use it "in production" but its just my personal shit, im not hosting any other users on it
2020-07-19 07:34:47 <opal> if something goes wrong then it isnt the end of the world
2020-07-19 07:43:46 <mps> tehcloud: there are also notcurses https://nick-black.com/dankwiki/index.php/Notcurses
2020-07-19 07:43:58 <mps> looks interesting
2020-07-19 12:02:53 <matthewcroughan> Hey guys! I can't believe it's true, but Alpine is absolutely the best choice out of any distribution of anything in the whole world to run on a crappy 32 bit laptop from 2005..
2020-07-19 12:03:10 <matthewcroughan> I've tried NetBSD, FreeBSD, Void, NixOS. None compare.
2020-07-19 12:04:06 <matthewcroughan> NixOS wouldn't boot because of the 256MB of ram. FreeBSD chugs and can't do Wayland. NetBSD's packaging is a nightmare, and it uses the most ram out of all of them. Void is probably going to die in the not too distant future.
2020-07-19 12:04:19 <opal> tried openbsd? :p
2020-07-19 12:04:34 <opal> ive actually been planning to get an x200 soon and install openbsd
2020-07-19 12:04:50 <matthewcroughan> Well, at least FreeBSD had some immense issues with libdrm and the intelGPU, the BSDs just aren't an optino.
2020-07-19 12:04:53 <matthewcroughan> option*
2020-07-19 12:05:10 <matthewcroughan> Can OpenBSD do wayland stuff opal ?
2020-07-19 12:05:18 <opal> and yes ive not liked void in my experience. runit is a plus but the package manager isnt great, maintainership isnt great, they do all their work on github
2020-07-19 12:05:25 <opal> poor support for server packages
2020-07-19 12:05:31 <opal> wayland isnt a priority for me
2020-07-19 12:05:35 <opal> so idk
2020-07-19 12:05:42 <matthewcroughan> Well that's the least of its issues. It's really crazy how fast they're moving and how many packages they support.
2020-07-19 12:05:52 <matthewcroughan> I just don't think it's going to be around for very long.
2020-07-19 12:05:53 <opal> ive used wayland and it has some stoppers in my use case before im able to switch and commit to it
2020-07-19 12:06:19 <matthewcroughan> Ya but this is on a laptop from 2005. I want it to run Sway and just one application :D
2020-07-19 12:06:29 <leah2> matthewcroughan: news to me, thanks :p
2020-07-19 12:06:37 <opal> which application, emacs?
2020-07-19 12:06:59 <matthewcroughan> Universal GCODE Sender
2020-07-19 12:07:08 <opal> never heard of it
2020-07-19 12:07:16 <matthewcroughan> Sends GCODE to CNC Routers and machinery
2020-07-19 12:07:44 <opal> leah2 ah i remember you being a void proponent
2020-07-19 12:07:45 <matthewcroughan> I'm happy to do things from a tty too lmao
2020-07-19 12:07:55 <opal> >G-code (also RS-274), which has many variants, is the common name for the most widely used computer numerical control (CNC) programming language.
2020-07-19 12:07:55 <opal> this?
2020-07-19 12:07:59 <leah2> no, i'm just a maintainer :p
2020-07-19 12:08:18 <matthewcroughan> I was so surprised to find a package for Hikari in there
2020-07-19 12:08:27 <opal> you werent a maintainer on any of the packages i had issues with then, lucky you
2020-07-19 12:08:32 <leah2> ;)
2020-07-19 12:09:16 <matthewcroughan> Oh lol on 32 bit I saw some crazy warnings with xbps regarding elf binaries.
2020-07-19 12:09:35 <matthewcroughan> There's' no doubt in my mind that Void is like someone doing heroin who's going to die soon.
2020-07-19 12:09:57 <matthewcroughan> You can't maintain that many packages, that fast, so close to upstream. That's what NixOS is for :D
2020-07-19 12:10:06 <leah2> i dont see why we are more likely to die now than 3 years ago, but perhaps you know something i dont
2020-07-19 12:10:15 <matthewcroughan> The traditional distro model where you have maintainers and central repos like that just can't function very well IMO
2020-07-19 12:10:29 <opal> come on we all know any distro without a proper mailing list isnt worth its salt :>
2020-07-19 12:10:31 <matthewcroughan> At least not whilst remaining secure and quality. Just my opinion.
2020-07-19 12:10:32 <leah2> every distro in existance is like that :p
2020-07-19 12:10:47 <opal> exherbo isnt
2020-07-19 12:10:53 <leah2> yes, exherbo is dead
2020-07-19 12:10:58 <opal> exherbo has a small main set and designed to be used with overlays
2020-07-19 12:11:06 <matthewcroughan> Yes, but your quality and security must go down with quantity, unless you're decentralized in the way nixpkgs are.
2020-07-19 12:11:17 <leah2> nixpkgs is not decentralized
2020-07-19 12:11:22 <matthewcroughan> very much so it is.
2020-07-19 12:11:24 <leah2> void accepts lots of PR too
2020-07-19 12:11:27 <leah2> there is no difference
2020-07-19 12:11:43 <matthewcroughan> https://github.com/colemickens/nixpkgs-wayland
2020-07-19 12:11:47 <matthewcroughan> No, there's a big difference.
2020-07-19 12:11:58 <matthewcroughan> The difference is that I can just go to this guy's wonderful collection of wayland apps and use them.
2020-07-19 12:12:03 <matthewcroughan> I need not rely on any one distribution of LInux.
2020-07-19 12:12:12 <opal> you can add custom repositories with anything
2020-07-19 12:12:23 <opal> i do it all the time
2020-07-19 12:12:28 <matthewcroughan> Yes, but that relies upon a package manager
2020-07-19 12:12:33 <matthewcroughan> it relies upon a distro of linux, this doesn't
2020-07-19 12:12:34 <opal> is nix not one?
2020-07-19 12:12:38 <leah2> you can also do your own void repo, which is not as easy as it should be, admittedly
2020-07-19 12:12:39 <matthewcroughan> NixOS is.
2020-07-19 12:12:44 <leah2> but it's still work to maintain that
2020-07-19 12:12:46 <opal> no, nix is the package manager
2020-07-19 12:12:50 <matthewcroughan> nix, the package manager, doesn't care what distribution you're on.
2020-07-19 12:12:56 <opal> nor does apk
2020-07-19 12:13:00 <opal> nor does dpkg
2020-07-19 12:13:02 <matthewcroughan> yes it does, nothing is going to work
2020-07-19 12:13:07 <matthewcroughan> yes they do, nothing is going to work
2020-07-19 12:13:13 <opal> you have no clue what youre talking about and im just going to /ignore you now
2020-07-19 12:13:23 <matthewcroughan> that's quite rude :(
2020-07-19 12:14:29 <matthewcroughan> If you care about the concept there is a paper on it which I read 8 pages of earlier, since I wanted to start understanding what I was talking about.
2020-07-19 12:15:20 <maerwald> have there been any attempts to bring ghc package to 32bit?
2020-07-19 12:15:21 <mps> I think this discussion doesn't belong here
2020-07-19 12:15:41 <matthewcroughan> Ya, I didn't come here to talk about that :D
2020-07-19 12:16:10 <matthewcroughan> I was actually just going to ask how I can modify `setup-alpine` to use a different FS for the rootfs
2020-07-19 12:16:34 <matthewcroughan> I'm looking through the scripts and they dereference a whole lot, so does anybody know what's actually happening in them?
2020-07-19 12:16:41 <ikke> matthewcroughan: the common way to do that is to not let it setup a disk, then setup your root fs, and then run setup-disk
2020-07-19 12:17:02 <matthewcroughan> I think the installer should ask what fs the user wants to use, maybe I could implement that
2020-07-19 12:17:52 <matthewcroughan> I couldn't find documentation about doing this sort of thing on the wiki, maybe it's best not to modify the script and just detail how to do this?>
2020-07-19 12:18:35 <Cogitri> maerwald: Not to x86 if you mean that, but to aarch64
2020-07-19 12:18:51 <matthewcroughan> I'll put things more simply too. If I `grep -r 'mke2fs /sbin/` I will not find anything useful. Yet the scripts call it somehow.
2020-07-19 12:18:58 <matthewcroughan> I'm really confused as to how it ends up getting called.
2020-07-19 12:19:17 <Cogitri> But currently it's kinda stale because bootstrapping ghc isn't fun and I currently don't have time to pull it off
2020-07-19 12:19:30 <matthewcroughan> Are there any tools that'll help me figure this out? Since the scripts in /sbin/ absolutely call mke2fs somehow, but it's two layers deep, and nothing does it directly.
2020-07-19 12:19:50 <matthewcroughan> For when grep isn't enough, there must be a tool that assists with this, out of complete curiosity!
2020-07-19 12:20:24 <maerwald> Cogitri: cross compiling from 64bit might work
2020-07-19 12:20:56 <maerwald> so no need to bootstrap, but there were a few bugs for ghc cross compilers afair
2020-07-19 12:21:06 <maerwald> (template haskell mostly)
2020-07-19 12:25:03 <maerwald> another idea is using https://github.com/redneb/ghc-alt-libc
2020-07-19 12:25:24 <maerwald> which are linked against musl, but built on gentoo, so there are those ncurses issues
2020-07-19 12:25:43 <maerwald> and I've experienced segfaults with them trying to build ghc
2020-07-19 12:26:52 <Cogitri> Unfortunately I have basically 0 knowledge of Haskell, not sure if someone in the dev team has the expertise/time to set it up
2020-07-19 12:28:00 <matthewcroughan> ikke that seems not to work
2020-07-19 12:28:13 <matthewcroughan> I can't "use the disk" without "deleting it" and formatting it with ext4 which is the default for setup-disk
2020-07-19 12:28:29 <matthewcroughan> I've partitioned and formatted my disks to my liking, setup-disk seems not to offer anything important, am I misunderstanding?
2020-07-19 12:28:32 <maerwald> Cogitri: is there a way to install 8.0.2 from v3.6 on edge?
2020-07-19 12:28:57 <maerwald> then I could build a i386 cross compiler for 8.0.2 and bootstrap the rest from there
2020-07-19 12:31:37 <matthewcroughan> AH. I keep using sys like a dummy. Getting data and sys mixed up/
2020-07-19 12:33:22 <Cogitri> maerwald: I guess if no solibs have changed you can just install it via apk add $archive
2020-07-19 12:38:07 <maerwald> https://pkgs.alpinelinux.org/package/v3.6/community/x86_64/ghc doesn't show the url to the archive
2020-07-19 12:38:27 <matthewcroughan> ikke No. This did not seem to work. `setup-disk` simply formats your disk as ext4 no matter what you do.
2020-07-19 12:38:43 <matthewcroughan> Whether I have partitioned or made an f2fs filesystem prior to running `setup-disk` does not matter. Can you help with this?
2020-07-19 12:41:11 <matthewcroughan> >f2fs is not suitable for partitioning
2020-07-19 12:41:20 <matthewcroughan> `setup-disk -s 0 ROOTFS=f2fs`
2020-07-19 12:42:23 <ikke> It looks it up under /sys/block/*/device
2020-07-19 12:42:39 <mps> setup-disk works fine
2020-07-19 12:42:48 <mps> I tested it few times
2020-07-19 12:42:54 <matthewcroughan> Ya I just didn't understand what the help was telling me
2020-07-19 12:43:18 <matthewcroughan> When it says "If BOOTFS is provided" it's not immediately obvious to people who aren't that technical that you mean to state BOOTFS=
2020-07-19 12:43:34 <mps> there is wiki guide how to use setup-disk
2020-07-19 12:43:47 <matthewcroughan> I'm looking at it now, it's verbose and directed at different circumstances
2020-07-19 12:43:52 <ikke> matthewcroughan: that it's written in capitals means it's an environment variable
2020-07-19 12:43:58 <matthewcroughan> it does not accommodate what I want to do, which is use f2fs.
2020-07-19 12:44:05 <Cogitri> maerwald: Currently you have to go to https://alpine.global.ssl.fastly.net/alpine/v3.6/ and then go to the right dir to download them directly
2020-07-19 12:44:15 <matthewcroughan> ikke Semantics like that aren't obvious to most people, even people that have been using Linux for 2-3 years like me.
2020-07-19 12:44:26 <matthewcroughan> I got it, but I don't think people under my exp level would get it.
2020-07-19 12:44:27 <mps> sure it worked for me with f2fs rootFS
2020-07-19 12:44:40 <ikke> matthewcroughan: ROOTFS=ext4 setup-disk
2020-07-19 12:44:41 <matthewcroughan> mps Really? It's telling me f2fs isn't a valid FS type.
2020-07-19 12:44:46 <ikke> matthewcroughan: ROOTFS=f2fs setup-disk
2020-07-19 12:45:05 <matthewcroughan> Oh I'm a complete and utter dummy.
2020-07-19 12:45:15 <matthewcroughan> See, if you said environment variable, I'd have done that. I thought it was an argument..
2020-07-19 12:45:24 <matthewcroughan> When it says "PROVIDE THIS" you do not assume envvar...
2020-07-19 12:45:55 <matthewcroughan> The help states "If BOOTFS, ROOTFS, VARFS are specified" it could be anything. It does not necessarily mean envvar
2020-07-19 12:46:29 <ikke> It can certainly be improved, but it's just a very common convention under linux programs
2020-07-19 12:46:35 <matthewcroughan> There are things that expect it on the other side as an argument, rather than something parsed by the shell. I understand this, I just think it would be better stated as calling it an envvar
2020-07-19 12:47:01 <matthewcroughan> I can submit a PR
2020-07-19 12:49:37 <matthewcroughan> ikke Can I do `ROOTFS=f2fs setup-alpine` and therefore not have to run `setup-disk` separately?
2020-07-19 12:49:53 <ikke> matthewcroughan: Yes, I think so
2020-07-19 12:50:08 <matthewcroughan> But I cannot provide `-s` to rnot have swap :D
2020-07-19 12:50:13 <ikke> correct
2020-07-19 12:50:35 <matthewcroughan> Everything should be envvars then, no parameters.
2020-07-19 12:51:40 <ikke> Or ask the user during setup-disk
2020-07-19 12:52:00 <matthewcroughan> Ya probably worthwhile. I'll look into it, I'm going to do a lot of installation so I'll look at the scripts properly eventually
2020-07-19 12:52:04 <mps> export SWAP_SIZE=0
2020-07-19 12:52:16 <matthewcroughan> Oh there is a var for it, thank you. Guess nothing needs to be done then.
2020-07-19 12:52:20 <mps> then swap will not be created
2020-07-19 12:52:26 <matthewcroughan> Really all that needs to be done is documenting these use cases properly, is all.
2020-07-19 12:52:43 <matthewcroughan> Since you can find that out by reading the help for setup-disk but not on the wiki, where I'm reading prior to even reading the help./
2020-07-19 12:53:05 <matthewcroughan> mps or did you find that on the wiki?
2020-07-19 12:53:45 <maerwald> seems alpine has no cross toolchains
2020-07-19 12:54:00 <mps> reading source, it is shell at the end
2020-07-19 12:54:27 <mps> maerwald: no, alpine build native
2020-07-19 12:54:41 <mps> ok, in lxc
2020-07-19 12:54:54 <maerwald> I can't parse that sentence
2020-07-19 12:54:58 <ikke> Lxc is still native
2020-07-19 12:55:28 <mps> ikke: yes ofc, but wanted to give more info
2020-07-19 13:03:02 <matthewcroughan> mps when I do `ROOTFS=f2fs setup-disk` I end up with mounting /dev/sdb2 on /mnt failed invalid argument
2020-07-19 13:03:27 <matthewcroughan> -v doesn't provide more info
2020-07-19 13:05:48 <mps> I think f2fs still have bug for dependencies loading
2020-07-19 13:06:12 <matthewcroughan> What does that mean? It ran fine on Void, for example. What is the bug you're referring to?
2020-07-19 13:06:44 <mps> I did install by creating partition, making fs with mkfs.f2fs, mounting and runnig setup-disk
2020-07-19 13:07:24 <matthewcroughan> it runs mkfs just fine, it just does not mount something at some stage
2020-07-19 13:07:38 <mps> I never posted fix for f2fs to upstream and actually not sure if I still have it somewhere
2020-07-19 13:09:09 <matthewcroughan> mps where do I need to mount things?
2020-07-19 13:10:14 <matthewcroughan> ah mkfs.f2fs required -f O.o
2020-07-19 13:12:41 <mps> I usually mount under /mnt
2020-07-19 13:13:06 <matthewcroughan> I did not need to mount it. If I mounted it, then I wouldn't be able to use setup-disk on it, because it wouldn't be available in the menu.
2020-07-19 13:13:37 <matthewcroughan> All I needed to do was manually make the FS, since setup-disk doesn't provide`mkfs.f2fs -f` only `mkfs.f2fs` which fails, therefore `mount` has nothing to mount.
2020-07-19 13:14:26 <matthewcroughan> I also wonder if Alpine takes into account MBR vs GPT
2020-07-19 13:14:33 <matthewcroughan> And DOS
2020-07-19 13:14:35 <mps> setup-disk require mounted FS
2020-07-19 13:14:45 <matthewcroughan> no, setup-disk mounted the FS for me
2020-07-19 13:14:53 <mps> ok
2020-07-19 13:15:04 <albr_> Watynecc[d]: Hello
2020-07-19 13:15:16 <komnpomat> matthewcroughan: for which purpose?
2020-07-19 13:15:50 <matthewcroughan> komnpomat Well I had a disk that was formatted previously as GPT, and if Alpine does nothing that disk is going to not boot, but the installation will appear succesful
2020-07-19 13:16:14 <matthewcroughan> mps actually, you're right. Now that Alpine has booted, there is this error
2020-07-19 13:16:21 <matthewcroughan> Filesystems couldn't be fixed, rc: Aborting!
2020-07-19 13:17:16 <komnpomat> weird error from a brand new install
2020-07-19 13:17:24 <matthewcroughan> komnpomat I'm using f2fs
2020-07-19 13:17:40 <komnpomat> oh I see, some exotic filesystem
2020-07-19 13:17:44 <matthewcroughan> mps mentioned that there is a "bug for dependencies loading", which lines up with what I've just done
2020-07-19 13:17:56 <matthewcroughan> no it's not exotic, it improves perf massively on flash drives
2020-07-19 13:18:04 <matthewcroughan> it's used on samsung phones for example
2020-07-19 13:18:17 <matthewcroughan> it stands for flash friendly filesystem
2020-07-19 13:18:34 <komnpomat> matthewcroughan: since it's your root filesystem, why don't you include it in the kernel rather than a module?
2020-07-19 13:18:55 <matthewcroughan> That would mean I have to compile my own kernel.
2020-07-19 13:19:06 <matthewcroughan> Alpine doesn't compile it in. I guess that's the issue. Whereas Void Linux does?
2020-07-19 13:19:26 <komnpomat> matthewcroughan: ok then in command line set it to load before anything else
2020-07-19 13:19:42 <matthewcroughan> how do I do that? Initramfs stuff right?
2020-07-19 13:19:44 <ikke> It's quite common to build all these filesystems as modules
2020-07-19 13:19:47 <matthewcroughan> I've only dabbled once or twice with that
2020-07-19 13:20:11 <mps> need to put 'f2fs' in /etc/mkinitfs/mkinitfs.conf and rebuild initramfs, probably
2020-07-19 13:20:54 <matthewcroughan> In arch it's mkinitcpio.conf in /etc
2020-07-19 13:20:56 <matthewcroughan> what is it in alpine?
2020-07-19 13:21:01 <komnpomat> matthewcroughan: the syntax is weird, hold on
2020-07-19 13:21:04 <matthewcroughan> Oh sorry didn't read mps, thanks!
2020-07-19 13:21:26 <matthewcroughan> that's really not too different from arch hah!
2020-07-19 13:21:43 <matthewcroughan> mps f2fs is already in there.
2020-07-19 13:21:55 <mps> you can try to add f2fs at APPEND line in boot loader in modules list
2020-07-19 13:22:11 <matthewcroughan> the contents of that file on my system are features="ata base ide scsi usb virtio f2fs"
2020-07-19 13:22:39 <mps> then something else is the issue
2020-07-19 13:23:13 <matthewcroughan> my /boot is empty
2020-07-19 13:23:15 <matthewcroughan> :thinking:
2020-07-19 13:23:26 <matthewcroughan> is that irregular? How the hell did I boot haha
2020-07-19 13:24:19 <matthewcroughan> ah no it's to do with the environment I'm in, strange
2020-07-19 13:24:24 <matthewcroughan>   APPEND root=UUID=8d5496bd-e1a2-4488-9a62-49f07ff3ec01 modules=sd-mod,usb-storage,f2fs nomodeset quiet rootfstype=f2fs
2020-07-19 13:24:34 <matthewcroughan> nope, rootfstype=f2fs is in there right, along with f2fs
2020-07-19 13:24:43 <matthewcroughan> who knows what the issue is
2020-07-19 13:25:11 <ikke> sounds like fsck is the issue
2020-07-19 13:25:18 <mps> if the /boot empty maybe you forgot to install boot loader
2020-07-19 13:25:33 <ikke> References to "dmvpn-1.2.1-r0                          = 1.2.1-r0
2020-07-19 13:25:35 <ikke> arg
2020-07-19 13:25:38 <matthewcroughan> no, it's not empty, the shell I'm put in after failed boot prohibits me from using /boot/ in any way
2020-07-19 13:25:44 <matthewcroughan> I can't mount /dev/sda1 for example
2020-07-19 13:26:05 <matthewcroughan> ya fsck is the issue ikke
2020-07-19 13:26:23 <matthewcroughan> "Filesystems couldn't be fixed" is the first actual error
2020-07-19 13:26:50 <ikke> So might mean you need f2fs-tools in the initramfs?
2020-07-19 13:27:04 <komnpomat> matthewcroughan: perhaps add some delay, is the usb slow?
2020-07-19 13:27:15 <matthewcroughan> No way, like I said, I ran void linux just fine with f2fs on root
2020-07-19 13:28:56 <ikke> try echo '/usr/sbin/f2fs.fsck' >>/etc/mkinitfs/features.d/f2fs.files
2020-07-19 13:29:02 <ikke> and than regenerate initramfs
2020-07-19 13:29:16 <matthewcroughan> how do you regen the initramfs on alpine? o.O
2020-07-19 13:29:23 <matthewcroughan> sorry if I'm too much of a noob here
2020-07-19 13:29:45 <ikke> mkinitfs -o /boot/<kernel>
2020-07-19 13:29:54 <matthewcroughan>  not too different from arch again, nice :D
2020-07-19 13:30:07 <matthewcroughan> hahaha
2020-07-19 13:30:14 <matthewcroughan> I modified /etc/fstab and disabled fsck and it works fine
2020-07-19 13:30:16 <matthewcroughan> funny
2020-07-19 13:30:27 <ikke> But you probably need to include dependencies for f2fs.fsck too
2020-07-19 13:30:41 <ikke> https://pkgs.alpinelinux.org/contents?branch=edge&name=f2fs-tools-libs&arch=x86_64&repo=main
2020-07-19 13:32:13 <komnpomat> matthewcroughan: I suspected it had to do about fsck
2020-07-19 13:32:27 <ikke> matthewcroughan: I don't think it makes sense
2020-07-19 13:32:31 <ikke> what I said
2020-07-19 13:32:36 <komnpomat> ikke: it does
2020-07-19 13:32:44 <ikke> I mean, it being part of initramfs
2020-07-19 13:32:46 <komnpomat> i've seen this before.
2020-07-19 13:33:24 <matthewcroughan> So what does f2fs.fsck represent?
2020-07-19 13:33:38 <komnpomat> ikke: what's the option in fstab to tell to do the fsck? the numbers right at the end? "0 0" or "0 1"
2020-07-19 13:33:44 <matthewcroughan> because /usr/sbin/f2fs only matches 3 files. f2fs_io f2fscrypt f2fstat
2020-07-19 13:33:50 <matthewcroughan> there is no .fsck
2020-07-19 13:34:05 <matthewcroughan> komnpomat 0 0 == no fsck
2020-07-19 13:34:12 <matthewcroughan> 0 1 of course means fsck
2020-07-19 13:34:26 <ikke> matthewcroughan: https://pkgs.alpinelinux.org/contents?branch=edge&name=f2fs-tools&arch=x86_64&repo=main
2020-07-19 13:34:26 <komnpomat> matthewcroughan: perhaps doesn't find fsck
2020-07-19 13:34:50 <ikke> matthewcroughan: fsck.f2fs
2020-07-19 13:35:00 <ikke> I swapped the order
2020-07-19 13:35:02 <matthewcroughan> Ah I have that package but do not have that ool
2020-07-19 13:35:06 <mps> usr/sbin/fsck.f2fs
2020-07-19 13:35:12 <matthewcroughan> Yes, I do not have that
2020-07-19 13:35:18 <matthewcroughan> Probably because I'm on i686 :D
2020-07-19 13:35:39 <komnpomat> matthewcroughan: which type of flash drive? regular usb sticks?
2020-07-19 13:35:46 <matthewcroughan> Yes, doesn't matter
2020-07-19 13:36:04 <ikke> matthewcroughan: if you include fsck.f2fs in initramfs, you need /usr/lib/libf2fs.so.7 as well
2020-07-19 13:36:07 <matthewcroughan> ikke I do not have /usr/sbin/f2fs.fsck despite having f2fs-tools installed
2020-07-19 13:36:14 <komnpomat> matthewcroughan: better than ext4, xfs, zfs? (on flash)
2020-07-19 13:36:25 <ikke> it's fsck.f2fs
2020-07-19 13:37:14 <komnpomat> ikke: it's not there
2020-07-19 13:37:26 <matthewcroughan> komnpomat very much so yes
2020-07-19 13:37:34 <ikke> -rwxr-xr-x    1 root     root        157912 Oct  1  2019 /usr/sbin/fsck.f2fs
2020-07-19 13:37:40 <ikke> I just installed f2fs-tools
2020-07-19 13:37:48 <komnpomat> I am on i686 too, and apk search fsck doesn't show f2fs
2020-07-19 13:37:58 <matthewcroughan> You said earlier the other way around
2020-07-19 13:38:06 <matthewcroughan> I just corrected it, that's the problem/
2020-07-19 13:38:08 <ikke> matthewcroughan: yes, I noticed later
2020-07-19 13:38:20 <ikke> I also mentioned that I said it wrong
2020-07-19 13:38:43 <matthewcroughan> Ah ok
2020-07-19 13:38:54 <matthewcroughan> So it's in there now, I'll regen my initramfs, and that should be it?
2020-07-19 13:39:05 <ikke> also include the lib I mentioned
2020-07-19 13:39:14 <matthewcroughan> when you say include, what do you mean?
2020-07-19 13:39:16 <matthewcroughan> install it?
2020-07-19 13:39:23 <ikke> In f2fs.files
2020-07-19 13:39:23 <matthewcroughan> It's installed, but what part of the lib do I need to put in the initramfs?
2020-07-19 13:39:34 <matthewcroughan> which particular thing?
2020-07-19 13:39:48 <ikke> /usr/lib/libf2fs.so.7
2020-07-19 13:41:37 <ikke> maybe that is not necessary
2020-07-19 13:41:51 <ikke> I don't see libraries explicitly installed for anything else
2020-07-19 13:42:18 <matthewcroughan> Learning a lot just by doing this, thanks for the help guys :)
2020-07-19 13:43:11 <matthewcroughan> Oh no hahaha it's not booting anymore
2020-07-19 13:43:22 <ikke> oink
2020-07-19 13:43:25 <matthewcroughan>  was vmlinuz-lts the wrong thing to put at the end of -o?
2020-07-19 13:43:34 <ikke> yes
2020-07-19 13:43:39 <matthewcroughan> lol that was THE kernel
2020-07-19 13:43:41 <ikke> yes
2020-07-19 13:43:44 <matthewcroughan> so where am I supposed to put it?
2020-07-19 13:44:24 <matthewcroughan> so much fun :D
2020-07-19 13:44:44 <ikke> mkinitfs -o /boot/initramfs-lts
2020-07-19 13:45:00 <matthewcroughan> I'm an extra special kind of braindead.
2020-07-19 13:45:09 <matthewcroughan> At least I have an opportunity to change it before booting into the system/
2020-07-19 14:00:01 <matthewcroughan> ikke Neither of your solutions worked, so I'm going to give up and run without fsck for now haha
2020-07-19 14:00:10 <ikke> sadface
2020-07-19 14:00:31 <ikke> But the fact that fsck is controlled by /etc/fstab kind of already means it has nothing to do with initfs
2020-07-19 14:00:56 <matthewcroughan> oh, interesting
2020-07-19 14:01:46 <matthewcroughan> you think it's as simple as symlinking /usr/sbin/fsck.fs2fs -> /usr/bin/fsck
2020-07-19 14:01:52 <ikke> no
2020-07-19 14:02:22 <ikke> fsck defers to these fsck.* programs
2020-07-19 14:02:58 <matthewcroughan> God damn
2020-07-19 14:03:07 <matthewcroughan> I'm just marvelling at that 21M ram usage on this 2005 laptop
2020-07-19 14:03:41 <matthewcroughan> of which HTOP is taking up the most ram on system hahaha
2020-07-19 14:04:03 <matthewcroughan> sshd second most. Perhaps lower with dropbear? O.o
2020-07-19 14:06:26 <matthewcroughan> I'll bet musl helps a lot on this old thing, more than I think.
2020-07-19 19:55:21 <eydaimon> komnpomat: did you see my comment about not being able to import the zpool using -d ?  the boot ISO doesn't allow it
2020-07-19 20:23:22 <eydaimon> like even with the full install, `/dev/disk` is empty
2020-07-19 20:23:25 <eydaimon> should it be ?
2020-07-19 20:24:27 <eydaimon> ikke: is that populated for you ?
2020-07-19 20:25:00 <ikke> ls: /dev/disk: No such file or directory
2020-07-19 20:26:56 <eydaimon> 🤔
2020-07-19 20:27:11 <eydaimon> https://www.reddit.com/r/zfs/comments/hsed8q/zfs_pool_degraded_smart_data_disagrees/fy9weon/   see they say the same thing here
2020-07-19 20:28:55 <eydaimon> since /dev/sda can change depending on reboot, is there anything static in alpine ?
2020-07-19 20:29:02 <ikke> uuid
2020-07-19 20:29:47 <eydaimon> how can I use that ?
2020-07-19 20:30:07 <ikke> in what context/
2020-07-19 20:30:11 <ikke> I'm not familiar with zfs
2020-07-19 20:30:24 <eydaimon> I need to do this:
2020-07-19 20:30:26 <eydaimon> sudo zpool export poolname
2020-07-19 20:30:28 <eydaimon> sudo zpool import poolname -d /dev/disk/by-id
2020-07-19 20:30:58 <eydaimon> So I use blkid to get the uuid, but how do I specify a drive using it ?
2020-07-19 21:44:32 <komnpomat> eydaimon: do you mean a LiveCD?
2020-07-19 21:46:21 <komnpomat>  /dev/disk empty? quite impossible
2020-07-19 21:48:12 <eydaimon> komnpomat: it's empty for me and ikke
2020-07-19 21:50:09 <ikke> I believe it's udev that creates those entries, which is not installed by default
2020-07-19 21:50:33 <mps> mdev I think
2020-07-19 21:52:51 <mps> hmm, udev, ikke is right
2020-07-19 22:01:26 <Hello71> shouldn't zfs support UUID directly
2020-07-19 22:02:07 <komnpomat> eydaimon: what does 'zpool status' say?
2020-07-19 22:02:25 <komnpomat> eydaimon: use a livecd with proper  zfs support
2020-07-19 22:18:58 <albr_> Watynecc[d]: hello?
2020-07-19 22:32:18 <eydaimon> komnpomat: sda2      ONLINE       0     0     0
2020-07-19 22:32:44 <eydaimon> ikke: what do I install to get those entries ?
2020-07-19 22:33:03 <eydaimon> komnpomat: if I boot with like ubuntu and use those, will it work with alpine after ?
2020-07-19 22:36:31 <eydaimon> zfs-udev ?
2020-07-20 07:41:34 <komnpomat> eydaimon: eudev actually
2020-07-20 07:42:26 <komnpomat> eydaimon: possibly zfs-udev too
2020-07-20 09:16:59 <albr_> Watynecc[d]: komnpomat: hello
2020-07-20 09:52:38 <komnpomat> albr_: hei
2020-07-20 09:56:18 <maerwald> It appears the GHC package has a missing dependency on binutils-gold. If you run `ghc --info` you will see '("ld command","ld.gold")' and then get link errors in certain circumstances (like bootstrapping another ghc)
2020-07-20 10:45:26 <komnpomat> is there something like pretend in apk installer?
2020-07-20 10:46:03 <ikke> what would it do?
2020-07-20 10:46:13 <komnpomat> show what deps will install
2020-07-20 10:46:20 <ikke> ah, dry run
2020-07-20 10:46:28 <komnpomat> I am concern installing mpv will pull trillion deps
2020-07-20 10:46:40 <ikke> apk add --simulate <package>
2020-07-20 10:47:20 <komnpomat> geez, libdvd libcdio
2020-07-20 10:47:40 <komnpomat> I wish there was a minimal mpv without dvd/cd legacy stuff
2020-07-20 10:49:20 <komnpomat> it wants to install 18 packages
2020-07-20 10:51:22 <mps> yes, I also dislike so much deps
2020-07-20 10:51:56 <ikke> Any dep you don't need is one too many, but if a dep is missing, you are sol
2020-07-20 10:53:48 <komnpomat> libxscrnsaver libcdio libcdio-paranoia libdvdcss libdvdread libdvdnav lua5 libpulse
2020-07-20 10:54:06 <komnpomat> all those could be removed for a no-frills basic mpv
2020-07-20 10:54:21 <komnpomat> and call it mpv-minimal
2020-07-20 10:54:30 <mps> or, any dep more which are not useful today is more bloat :)
2020-07-20 10:55:00 <mps> komnpomat: yes, I thought to do that but didn't found time
2020-07-20 10:55:27 <mps> and, wanted to have mpv and mpv-full
2020-07-20 10:55:53 <komnpomat> ha ha, great I am not the only one
2020-07-20 10:56:32 <mps> and there are more pkgs which I think to make base and full
2020-07-20 10:56:37 <mps> vim for example
2020-07-20 10:57:55 <Cogitri> maerwald: Could you please open an issue for that?
2020-07-20 10:58:24 <maerwald> Cogitri: but then again... ld.gold produces broken binaries with ghc on musl
2020-07-20 10:58:50 <maerwald> https://gitlab.haskell.org/ghc/ghc/-/issues/17508
2020-07-20 10:58:56 <maerwald> https://sourceware.org/bugzilla/show_bug.cgi?id=23856
2020-07-20 11:02:02 <Cogitri> Ah right, the bug I commented on but then forgot about again :D
2020-07-20 11:02:16 <Cogitri> I think we had something similiar with ldc
2020-07-20 11:02:31 <Cogitri> Can we configure ghc to use bfd/lld by default?
2020-07-20 11:02:39 <maerwald> yes
2020-07-20 11:03:09 <maerwald> LD=ld.bfd ./configure --disable-ld-override
2020-07-20 11:03:47 <maerwald> and additionally you may need "LD_STAGE0 = ld.bfd" in mk/build.mk
2020-07-20 11:04:26 <maerwald> then you can check after installation what `ghc --info` reports
2020-07-20 11:04:51 <Cogitri> Ah, nice. Mind opening an issue/MR with that info? :)
2020-07-20 11:04:53 <maerwald> I find this automagic ld.gold selection terrible
2020-07-20 11:05:10 <maerwald> good build systems should always use ld only
2020-07-20 11:05:21 <maerwald> unless the user says something different
2020-07-20 11:06:24 <maerwald> I still have to do some tests
2020-07-20 11:58:58 <SomeoneNew> Hey,
2020-07-20 11:58:58 <SomeoneNew> I couldn't find any references online
2020-07-20 11:58:58 <SomeoneNew> Just wanted to know which compilation options (hardening) Alpine Linux employs when building packages e.g -fstack-protector-strong vs -fstack-protector-all
2020-07-20 12:02:16 <Cogitri> TL;DR: stack-protector-all has a pretty huge performance impact in comparison to -strong and isn't useful most of the time
2020-07-20 12:03:05 <Cogitri> Since -strong already covers most cases where a stack overflow might happen
2020-07-20 12:13:46 <SomeoneNew> Ok, thanks. I guess it's pretty safe for me to assume Alpine cranks up all the hardening options
2020-07-20 12:16:26 <mps> not all but reasonable ones
2020-07-20 14:15:28 <eydaimon> komnpomat: thanks
2020-07-20 14:30:23 <webstrand> c705: Nope, I was asking about PAM support in the busybox package
2020-07-20 14:33:53 <webstrand> Speaking of, has anyone had the time to look at my ticket for busybox-pam? Or did I put it in the wrong issues list?
2020-07-20 14:42:02 <Cogitri> I thought you were working on it, so I didn't take a second look
2020-07-20 15:13:51 <PureTryOut[m]> What do I need to do to make postgreSQL accept connections from the local network instead of just the local machine? StackOverflow and such suggests editing `pg_hba.conf` and `postgresql.conf` which I did in multiple locations, but none of those files seems to make any difference, it just keeps listening on 127.0.0.1 only
2020-07-20 15:14:14 <rehwanne1> did you restart postgres?
2020-07-20 15:23:20 <mps> PureTryOut[m]: you have to enable listening interfaces in postgresql.conf
2020-07-20 15:24:15 <mps> and access rights for users/groups/networks/auth_methods in pg_hba.conf, then restart
2020-07-20 15:25:50 <mps> stgresql/pg_ident.conf'  # ident configuration file
2020-07-20 15:25:54 <mps> stgresql/pg_ident.conf'  # ident configuration file
2020-07-20 15:26:01 <mps> hmm
2020-07-20 15:26:30 <mps> listen_addresses and port
2020-07-20 15:31:16 <eydaimon> there's no telnet package ? How do I get telnet ? I use this for debugging all the time.
2020-07-20 15:31:57 <mps> eydaimon: netcat-openbsd or busybox-extra
2020-07-20 15:32:31 <mps> though few weeks ago I enabled telnet option in screen also
2020-07-20 15:32:48 <mps> didn't tested it much
2020-07-20 15:32:50 <eydaimon> mps: they give me the telnet command ? I didn't see it when searching pkgs
2020-07-20 15:33:25 <mps> only busybox-extra have telnet cmd
2020-07-20 15:33:40 <mps> others can be used in telnet mode
2020-07-20 15:34:01 <eydaimon> oh? I never tried netcat in telnet mode
2020-07-20 15:34:07 <eydaimon> how do I do that ?
2020-07-20 15:34:18 <eydaimon> gfuess i'll install and man :)
2020-07-20 15:34:26 <mps> nc host 23
2020-07-20 15:35:48 <eydaimon> doesn't give me a prompt to type when connecting to http
2020-07-20 15:35:55 <eydaimon> for example
2020-07-20 15:36:16 <eydaimon> apk add busybox-extra → ERROR: unsatisfiable constraints: busybox-extra (missing): required by: world[busybox-extra]
2020-07-20 15:36:58 <eydaimon> I've never ran into a distro that doens't have a telnet package. Can we add that?
2020-07-20 15:37:37 <MY-R> busybox-extras
2020-07-20 15:38:23 <eydaimon> MY-R: see my line above
2020-07-20 15:38:31 <MY-R> busybox-extras
2020-07-20 15:38:36 <eydaimon> oh now it worked
2020-07-20 15:38:43 <eydaimon> missing 's' 😅
2020-07-20 15:38:55 <eydaimon> thank you
2020-07-20 15:39:06 <MY-R> :]
2020-07-20 15:39:24 <eydaimon> copy/paste ftw
2020-07-20 15:40:17 <eydaimon> omg, i forgot the name of my favorite portscanner
2020-07-20 15:40:53 <eydaimon> nmpap, that's the one
2020-07-20 16:34:12 <helby> hello ladies, gentlemen and others
2020-07-20 16:34:33 <helby> Is there any alternative to sway for a wayland out there? something not so bloated :)
2020-07-20 16:35:18 <afontain_> trick question
2020-07-20 16:35:44 <afontain_> cage? (disclaimer: not what you likely want)
2020-07-20 16:36:03 <afontain_> waymonad? (if you're into haskell)
2020-07-20 16:36:13 <afontain_> sway but without default config?
2020-07-20 17:12:23 <helby> I stripped config to a minimum, but overal is not very lighweight as simple tilling wm
2020-07-20 17:12:44 <helby> also little thing like swaybg running for nothing, on some distros it's optional
2020-07-20 17:13:01 <helby> or if I remember well, I did not have it somewhere
2020-07-20 17:13:33 <helby> fuck put background to tilling wm is nonsense, and put some package for it ...
2020-07-20 17:21:13 <artok> if there is dwm thingie for wayland, i might give it a shot
2020-07-20 17:22:39 <helby> there is some kind of dwl :) dwm for wayland
2020-07-20 17:23:06 <helby> i was okay with sway, however was few days on openbsd and used cwm and found it probably more practical
2020-07-20 17:24:09 <helby> but all that fakery with xorg after long time, it was killing me
2020-07-20 17:26:14 <afontain_> > also little thing like swaybg running for nothing, on some distros it's optional
2020-07-20 17:26:14 <afontain_> well, if you don't have it in your config, it won't run
2020-07-20 17:26:24 <afontain_> (whoops matrix replies)
2020-07-20 17:27:13 <afontain_> helby: what do you mean by "fakery"?
2020-07-20 17:28:23 <helby> yes, it runs
2020-07-20 17:28:32 <helby> eve I don't have anything related to it in config
2020-07-20 17:29:12 <helby> so yes, maybe it was not optional on some distro, probably just did not run
2020-07-20 17:29:21 <helby> but now I am looking and processes and it's not running
2020-07-20 17:29:35 <helby> so maybe when I checked it last time, I did some shit what activated that crap
2020-07-20 17:30:14 <helby> afontain_: fakery, I mean fuckering - tweaking all that Xresource .xsession xterm xinitrc .... crap!
2020-07-20 17:30:32 <helby> fuck and .... xterm
2020-07-20 17:31:06 <afontain_> ah, yes
2020-07-20 17:31:09 <afontain_> understandable
2020-07-20 17:32:01 <helby> oh now I see what activated swaybg
2020-07-20 17:32:15 <helby> 'sway reload' activates that crap
2020-07-20 17:32:38 <helby> but sitll I did not realized how sway is big compared to what id does
2020-07-20 17:32:43 <helby> * it does
2020-07-20 17:48:04 <maerwald> error: collect2: fatal error: cannot find 'ld' -- is that something known?
2020-07-20 17:48:44 <maerwald> https://bsd.to/n9GQ/raw
2020-07-20 18:10:23 <Cogitri> I don't think so, but I don't use use Haskell
2020-07-20 18:10:39 <Cogitri> But we built cabal with it, so I guess ghc is functional?
2020-07-20 18:17:31 <maerwald> this is with an official bindist
2020-07-20 18:17:55 <maerwald> I can't figure out what's going on there... ld definitely is available
2020-07-20 20:11:27 <opal> oh actually someone made an s6-overlay thats neat
2020-07-20 20:15:07 <PureTryOut[m]> mps: I did that, made no change
2020-07-20 20:30:05 <mps> PureTryOut[m]: can you telnet to host and port
2020-07-20 20:33:39 <linux-pine> hello
2020-07-20 20:33:54 <linux-pine> how to configure keyboard layout once installed?
2020-07-20 20:36:33 <linux-pine>  /sbin/setup-keymap: cd: line 17: can't cd to //usr/share/bkeymaps: No such file or directory
2020-07-20 20:40:40 <mps> linux-pine: can you check with 'apk info -e kbd-bkeymaps' is keymaps pkg installed
2020-07-20 20:45:43 <linux-pine> installing it
2020-07-20 20:46:28 <linux-pine> oh, it's dns issue
2020-07-20 20:46:31 <linux-pine> let me fix that first
2020-07-20 20:57:23 <linux-pine> it works, thanks
2020-07-20 20:57:31 <linux-pine> although I'm not finding my keyboard layout in the list
2020-07-20 21:00:29 <mps> something rare?
2020-07-20 21:00:37 <mps> or special?
2020-07-20 21:05:38 <linux-pine> mps http://0x0.st/iwzV.png
2020-07-20 21:07:10 <mps> I'm not sure what key layout it is
2020-07-20 21:07:49 <linux-pine> pt-br
2020-07-20 21:09:09 <mps> maybe just 'br'
2020-07-20 21:37:03 <linux-pine> well, that wouldn't make sense
2020-07-20 21:37:20 <linux-pine> but now some keys work, thanks anyway
2020-07-20 22:19:19 <maerwald> Cogitri: I've successfully bootstrapped 32bit GHCs, based on the gentoo musl ones. The 8.0.2 was even without libtinfo split
2020-07-20 22:19:50 <maerwald> bindists that is... not alpine pkgs
2020-07-20 22:20:03 <maerwald> you could use them to build alpine packages
2020-07-20 23:03:19 <afontain_> funnily enough, we discussed the topic of bootstraping ghc a few hours ago
2020-07-20 23:30:24 <j`ey> anyone know of a guide to using apk on another distro?
2020-07-20 23:30:55 <maldridge> sure, install Adelie and use apk there :P
2020-07-20 23:31:04 <j`ey> heh
2020-07-20 23:31:16 <maldridge> assuming though you meant with the apk repos from alpine, perhaps you could flesh out your question
2020-07-20 23:31:45 <j`ey> I want to download some .apks from another system
2020-07-20 23:31:58 <j`ey> (so apk fetch)
2020-07-20 23:32:45 <maldridge> I believe you can just install and configure it and it will work
2020-07-20 23:34:08 <j`ey> maybe I need to try that in a chroot then
2020-07-20 23:35:00 <j`ey> might just be faster to run a vm with alpine actually
2020-07-20 23:35:27 <leah2> or a docker container *cough*
2020-07-20 23:36:56 <maldridge> or use wget if you don't need index aware retrieval
2020-07-20 23:37:22 <j`ey> I wanted the full tree of deps too
2020-07-20 23:43:09 <linux-pine> ERROR: Cannot open TUN/TAP dev /dev/net/tun: No such file or directory (errno=2)
2020-07-20 23:43:52 <linux-pine> ¯\_(ツ)_/¯
2020-07-20 23:44:07 <afontain_> isn't tun at /dev/tun on Alpine?
2020-07-20 23:46:04 <linux-pine> no such file or directory, afontain_
2020-07-20 23:47:09 <afontain_> you may need to modprobe it?
2020-07-20 23:47:34 <afontain_> (litterally `modprobe tun`)
2020-07-20 23:48:11 <linux-pine> that did the trick, yes
2020-07-20 23:48:12 <linux-pine> thanks
2020-07-20 23:54:01 <linux-pine> Options error: --up script fails with '/etc/openvpn/update-resolv-conf': No such file or directory (errno=2)
2020-07-20 23:57:09 <linux-pine> this looks like a missing dependency
2020-07-20 23:57:24 <linux-pine> openvpn needs something related to resolvconf
2020-07-20 23:57:33 <linux-pine> and it isn't finding it
2020-07-21 00:05:29 <linux-pine> meh, just commented out the resolvconf part on the ovpn config file, and it worked
2020-07-21 00:09:54 <linux-pine> but now I don't have any name resolution working
2020-07-21 00:15:56 <linux-pine> https://forums.openvpn.net/viewtopic.php?p=61374&sid=04f1230321d127045c2bd9958a1afe0c#p61374
2020-07-21 00:16:06 <linux-pine> apparently Alpine doesn't provide a "decent repo"
2020-07-21 00:25:36 <linux-pine> Ok, this is how I fixed it:
2020-07-21 00:25:44 <linux-pine> first, you need to install busybox
2020-07-21 00:25:49 <linux-pine> apk add bash
2020-07-21 00:26:23 <linux-pine> then do "nano /etc/openvpn/update-resolv-conf"
2020-07-21 00:27:45 <linux-pine> paste this: https://dpaste.com/DGP22SA8W
2020-07-21 00:28:14 <linux-pine> and finally make it executable: "chmod +x /etc/openvpn/update-resolv-conf"
2020-07-21 00:30:22 <linux-pine> and that's how you make openvpn work in Alpine
2020-07-21 09:39:57 <maerwald> afontain_: you can basically use ghcup now to bootstrap a 32bit package
2020-07-21 09:40:03 <hiro> 10:37 pocke ah fuck
2020-07-21 09:40:03 <hiro> 10:37 pocke i just realised i didnt turn off caching on chromium
2020-07-21 09:40:03 <hiro> 10:38 pocke bundle.js is like 80% of the total size
2020-07-21 09:40:03 <hiro> 10:37 pocke so it had already had bundle.js
2020-07-21 09:40:04 <hiro> 10:37 pocke that explains the speed difference lol
2020-07-21 09:40:04 <hiro> 10:38 spark ah, well... there we go? all solved?
2020-07-21 09:40:04 <hiro> 10:38 pocke yeah
2020-07-21 09:40:05 <hiro> 10:38 spark not much we can do about a lack of edge caching. at least we arent on 10s page loads anymore
2020-07-21 09:40:05 <hiro> 10:39 spark used to be that way as there was no bundling at all...
2020-07-21 09:40:06 <hiro> 10:39 pocke haha yeah
2020-07-21 09:40:06 <hiro> 10:39 pocke you know how sites like the new reddit overhaul
2020-07-21 09:40:07 <hiro> 10:39 pocke loads a sort of blank, yet partially theme'd/shaped page before it actually starts downloading the content
2020-07-21 09:40:07 <hiro> 10:40 pocke that's the sort of thing i was thinking in terms of loading the forms first before the other heavy content
2020-07-21 09:40:08 <hiro> 10:40 pocke so the user can immediately start filling out a field even though the page isnt fully loaded
2020-07-21 09:42:00 <hiro> sorry for that accident
2020-07-21 09:42:01 <mps> hiro: ?
2020-07-21 09:42:14 <mps> ah. ok
2020-07-21 09:42:20 <mps> happens
2020-07-21 09:42:26 <hiro> i /wc 'ed when i realised i pasted a huge thing accidentally
2020-07-21 09:42:36 <hiro> :D
2020-07-21 12:27:17 <bndeter> so this wiki entry mentions putting stuff in /etc/crypttab > https://wiki.alpinelinux.org/wiki/LVM_on_LUKS
2020-07-21 12:28:08 <bndeter> but thats just wrong? crypttab is never used in normal setup? its some systemd stuff and the 'real' way to do it is adding entrys to /etc/conf.d/dmcrypt and add dmcrypt to boot?
2020-07-21 12:29:57 <bndeter> also the procedure described in "Syslinux with BIOS" doesn't work, as one is supposed to chroot into a directory, which doens't have /dev bind mounted so calling 'update-extlinux' just fails.
2020-07-21 12:31:30 <bndeter> gentoo wiki comes to the rescue. but its sad, that the only resource to make fde on alpine is proned with so many errors.
2020-07-21 12:35:45 <bndeter> also the entry for installing a bootloader > https://wiki.alpinelinux.org/wiki/Bootloaders#Installing_Syslinux
2020-07-21 12:36:54 <bndeter> it mentions writing gptmbr.bin to the mbr. but doesn't say a thing about adding the legacy bootable flag to the '/boot' partition. so it also just doesn't work.
2020-07-21 12:37:04 <bndeter> syslinux/gentoo wiki to the rescue again.
2020-07-21 12:38:15 <bndeter> i think it may be better to just redirect people to the real resources instead of having really misleading, untested and by the time frustrating wiki articles.
2020-07-21 12:40:00 <mps> your observations are right but I think wiki is not official guides/docs
2020-07-21 12:40:53 <mps> and wiki can be edited/added/fixed by anyone who are ready to do that
2020-07-21 12:42:04 <mps> so, would be nice if you add your notes or write correct guide, ofc if you have time and will
2020-07-21 12:42:26 <bndeter> sorry for sounding a bit 'ranty'. what are the official guides/docs? with much of the stuff inhereted from other distros (gentoo, debian) its somehow not very clear to me where to look.
2020-07-21 12:42:54 <mps> alpine doesn't have much of official docs
2020-07-21 12:43:12 <mps> only this https://docs.alpinelinux.org/user-handbook/0.1a/index.html
2020-07-21 12:44:37 <bndeter> thanks. didn't know there were any official docs :)
2020-07-21 12:45:34 <bndeter> they seem pretty barebone. repeating what the autoinstaller does and one page about openrc and apk.
2020-07-21 12:46:48 <mps> https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_partitioning_your_disk
2020-07-21 12:47:14 <mps> and, yes, it is not 'exhaustive' guide
2020-07-21 12:48:22 <mps> but alpine is mostly used by developers and experienced admins/users (not important but wanted to note that)
2020-07-21 12:50:13 <mps> alpine doesn't have so much develepors and contributors like big distros, and most are concentrated on improving software and less about docs
2020-07-21 12:51:09 <Cogitri> The docs should get better eventually, but someone has to dedicate the time to do so
2020-07-21 12:54:37 <bndeter> well then maybe i can't dodge and have to write an improved fde article. hope i can carve out the time in the next few days before i forget eveything again.
2020-07-21 12:56:34 <bndeter> rest of the distro seems rock solid so i was very sad about misleading docs, which hopefully may excuse my rant a bit.
2020-07-21 12:58:58 <mps> don't worry about your rant, we know that docs are 'not so good' :)
2020-07-21 13:16:10 <helby> hello ladies, gentlemen and others
2020-07-21 13:16:47 <helby> Any idea what can cause firefox crash during watching video https://dpaste.org/h3OW/slim
2020-07-21 13:17:36 <helby> that crash is not just random, it's all the time I watch a video in a browser for a minute approx
2020-07-21 13:17:38 <Cogitri> Is that DRM video?
2020-07-21 13:17:46 <helby> no idea
2020-07-21 13:17:58 <helby> can be that?
2020-07-21 13:18:08 <Cogitri> We've had crashes with widevine enabled, but I haven't ever had FF crash on me with other media (or at all really)
2020-07-21 13:18:35 <helby> I have disabled widevine, but tried also with fresh profile and it crash the same
2020-07-21 13:19:14 <helby> I think it did not happen on the same system with sway/wayland
2020-07-21 13:19:28 <helby> now I am on X/cwm
2020-07-21 13:19:37 <helby> and it happens :)
2020-07-21 13:20:58 <helby> hard to watch pornhub then :/
2020-07-21 13:30:04 <helby> I have relatively lean system and after change to X from wayland something maybe missing, no idea what is related to
2020-07-21 13:37:28 <helby> testing it now in sway/wayland and it does not happen
2020-07-21 13:38:36 <Cogitri> Huh
2020-07-21 13:41:32 <helby> I get that second error NS_ERROR_DOM_MEDIA_METADATA_ERR but nothing crashing or any visible issue
2020-07-21 13:41:37 <helby> on sway/wayland
2020-07-21 13:42:05 <helby> I have no idea what these errors even mean and what it is related to, so cahnging window manager can do something?
2020-07-21 13:42:34 <helby> bc. still cwm is from openbsd, there can be some 'security' magic :D
2020-07-21 13:42:43 <helby> will try another wm
2020-07-21 13:53:17 <opal> openbsd isnt the pinnacle of software security just to let you know
2020-07-21 13:53:30 <opal> they do focus on it but they arent immune to bugs
2020-07-21 13:53:34 <opal> or poor design
2020-07-21 13:54:33 <helby> I know
2020-07-21 13:55:38 <helby> their 'security' many times fuck up usage of something
2020-07-21 14:26:38 <helby> now that shit does not happen after tried in sway and other X wm ...
2020-07-21 14:26:57 <helby> I don't know, uninstalled all and back to cwm but it did not happen
2020-07-21 14:28:48 <helby> maybe I miss some lib package for X or whatever ...
2020-07-21 17:22:52 <komnpomat> we gonna see how alpine perfroms in the real worldz
2020-07-21 17:24:45 <komnpomat> mm.. standard or extended?
2020-07-21 17:27:39 <komnpomat> no information or details --> Most common used packages included.
2020-07-21 17:46:09 <helby> so after some long time I am again back to xorg and all that X fakery is here :)
2020-07-21 17:46:36 <helby> I am little bit lost with xf86-video-intel/modesetting mesa-gallium mesa-classic crap
2020-07-21 17:47:12 <helby> I would not fuckering with that much and install just bare minimum if system run, but somehow have crashes in firefox during video play
2020-07-21 17:47:33 <ikke> komnpomat: extended is for when you need extra packages before you can get online
2020-07-21 17:47:34 <helby> crashes/freezing that does not happen in wayland
2020-07-21 18:10:16 <komnpomat> helby: rock solid here
2020-07-21 18:10:56 <helby> komnpomat: I don't know, maybe I miss some Xsomething
2020-07-21 18:11:03 <eydaimon> ikke: do i need to reboot or something to get the devs populated in /dev/disk after installing udev ? I've started the service but I see nothing there
2020-07-21 18:11:23 <helby> komnpomat: the same system, issue on xorg, but not wayland
2020-07-21 18:12:01 <helby> komnpomat: that fucking firefox freeze the way, can't even close it from wm, just to kill it
2020-07-21 18:12:31 <helby> komnpomat: I don't know why I came back to X again ;)
2020-07-21 18:13:01 <helby> komnpomat: oh I know, sway is 5MB! wm
2020-07-21 18:14:21 <komnpomat> interesting
2020-07-21 18:14:31 <komnpomat> sway didn't work for me, never
2020-07-21 18:15:48 <helby> for me without a touch, that I like on sway/wayland, not any fakery in setting, driver or anything, started sway and done
2020-07-21 18:17:04 <afontain_> I've never had an issue with sway
2020-07-21 18:17:15 <afontain_> I don't have NVidia drivers, so maybe that's why
2020-07-21 18:17:42 <helby> well, just mouse cursor sometimes missing there :)
2020-07-21 18:17:53 <helby> but only in videos
2020-07-21 18:18:17 <afontain_> ah, yes, the mouse cursor thing
2020-07-21 18:18:28 <afontain_> I've had some issues with swayidle though
2020-07-21 18:18:48 <afontain_> no, swaylock actually
2020-07-21 18:21:39 <ikke> eydaimon: might be that you need config that creates those entries
2020-07-21 18:21:48 <ikke> eydaimon: https://unix.stackexchange.com/a/86775/10643
2020-07-21 18:21:59 <eydaimon> ikke: thanks much
2020-07-21 18:30:14 <komnpomat> helby: which package did you add? I will try right now
2020-07-21 18:34:44 <komnpomat> helby: which commands installs sway flawlessly?
2020-07-21 18:40:13 <helby> komnpomat: try 'apk add sway' ;)
2020-07-21 18:41:24 <helby> btw. you don't need to leave your X
2020-07-21 18:41:41 <helby> u can open sway from other tty, not fakering like in X ;)
2020-07-21 18:41:58 <komnpomat> helby: sway pulls all the other deps?
2020-07-21 18:42:04 <helby> well actually u can even open in inside your X
2020-07-21 18:42:20 <komnpomat> helby: what about starting from scratch?
2020-07-21 18:42:23 <helby> komnpomat: something yes, probably not your gpu driver /mesa
2020-07-21 18:42:30 <komnpomat> does it need a menu, terminal?
2020-07-21 18:42:39 <helby> komnpomat: do u need terminal? :)
2020-07-21 18:42:55 <helby> komnpomat: some font, terminal is useful
2020-07-21 18:43:01 <komnpomat> helby: you said 'sway that's it'
2020-07-21 18:43:13 <komnpomat> I am concerned about missing pieces
2020-07-21 18:43:34 <helby> u will see :)
2020-07-21 18:45:13 <helby> komnpomat: if u install font and some terminal, ideally alacritty that it works even without your config then you are okay and it runs
2020-07-21 18:46:00 <helby> komnpomat: just remember sway is 50 times bigger ehan some small wm :)
2020-07-21 18:46:06 <helby> it's monster like i3
2020-07-21 18:46:24 <komnpomat> :(
2020-07-21 18:46:43 <komnpomat> helby: how to tweak grub command line?
2020-07-21 18:46:54 <helby> what u want to tweak there?
2020-07-21 18:46:58 <komnpomat> do I need to run /etc/mkinitrd?
2020-07-21 18:47:00 <helby> delete whole fucking grub
2020-07-21 18:47:06 <komnpomat> add some stuf
2020-07-21 18:47:14 <helby> add staff to grub?
2020-07-21 18:47:43 <komnpomat> things
2020-07-21 18:47:53 <helby> example?
2020-07-21 18:48:27 <helby> there is some shit /etc/default/grub
2020-07-21 18:48:55 <helby> and /etc/grub.d
2020-07-21 18:49:06 <komnpomat> how to redo the initrd?
2020-07-21 18:49:09 <helby> custom section for people who wanna some bullshit
2020-07-21 18:49:26 <komnpomat> eydaimon: hei, where you the one making initrd?
2020-07-21 18:49:42 <helby> mkinitfs
2020-07-21 18:50:06 <helby> just run that crap as root
2020-07-21 18:50:17 <helby> I bet you will not boot again
2020-07-21 19:06:01 <komnpomat> I found it, it was in the backlog
2020-07-21 19:06:19 <komnpomat> is it me, or alpine's boot isn't that much verbose?
2020-07-21 19:08:24 <helby> in grub u have 'quiet' optin, so u don't see tons of messages
2020-07-21 19:10:50 <komnpomat> good call
2020-07-21 19:10:51 <komnpomat> thanks
2020-07-21 19:20:27 <helby> fucking google
2020-07-21 19:20:46 <helby> again mark my emails as spam with even warnign that similar email was used to steal info bla bla
2020-07-21 19:21:09 <helby> can't do shit about it, server and all headers are top notch
2020-07-21 19:28:56 <komnpomat> a header can top notch?
2020-07-21 19:29:01 <komnpomat> how?
2020-07-21 19:31:40 <helby> I meant my mail sever is well set up
2020-07-21 19:32:01 <helby> with all dkim spf dmarc, headers return-path everythins is done the way, that can't improve it more
2020-07-21 19:32:24 <eydaimon> komnpomat: nope. I had the zfs boot issues
2020-07-21 19:32:31 <eydaimon> komnpomat: or have rather
2020-07-21 20:08:23 <helby> btw. fuck X I am back to wayland :)
2020-07-21 21:18:43 <eydaimon> shouldn't `apk search cmd:telnet` list telnet under busybox-extras ?
2020-07-21 21:24:50 <adu> apk info --who-owns $(which telnet)
2020-07-21 21:25:40 <eydaimon> adu: does that work with packages you don't have installed too ?
2020-07-21 21:25:54 <mps> eydaimon: looks like it doesn't work for bb applets
2020-07-21 21:26:52 <adu> eydaimon: no
2020-07-21 21:27:38 <mps> eydaimon: yes
2020-07-21 21:28:55 <mps> I don't have vis, but  apk search cmd:vis shows vis, beside other pkgs which have cmd beginning with 'vis'
2020-07-21 21:30:34 <mps> and it works for coreutils 'applets' but not for busybox
2020-07-21 21:30:38 <mps> interesting
2020-07-21 21:32:27 <eydaimon> mps: I was specifically asking about busybox. If ikke hadn't told me telnet was part of busybox-extras, I wouldn't have a way of knowing
2020-07-21 21:32:46 <eydaimon> I think it wold be useful if busybox commands were indexed too
2020-07-21 21:33:17 <mps> I think I told you that
2020-07-21 21:33:28 <TBB> they're a bit of a special case aren't they
2020-07-21 21:33:53 <mps> yes, bb applets are 'instantiad' on install
2020-07-21 21:35:22 <mps> heh, I was right '2020-07-20 17:33............. mps| only busybox-extra have telnet cmd'
2020-07-21 21:35:37 <eydaimon> mps: oh, perhaps. thank you for that :)
2020-07-21 21:35:47 <eydaimon> so without you, I would have been lost
2020-07-21 21:36:07 <mps> np, I just checking how my memory works :)
2020-07-21 21:36:22 <eydaimon> TBB: yes, special cases, but doesn't mean it can't be done. telnet is somethign people will look for
2020-07-21 21:38:32 <mps> I could add one 'advice', if you can't find command type 'busybox | grep cmdname'
2020-07-21 22:52:36 <eydaimon> mps: thanks
2020-07-21 23:04:37 <simon> Hello!
2020-07-21 23:04:55 <c705> welcome
2020-07-21 23:05:14 <simon> I'm using Alpine 3.12 which comes with clang v10, but actually I'd like to be using clang v9 which Alpine 3.11 used...
2020-07-21 23:06:07 <simon> Is there a way to install clang v9 with Alpine 3.12 without recompiling clang etc?
2020-07-21 23:22:39 <simon> https://superuser.com/questions/1055060/how-to-install-a-specific-package-version-in-alpine
2020-07-21 23:23:09 <simon> this seems to work for installing clang v9 from Alpine 3.11 into Alpine 3.12 ...
2020-07-21 23:29:48 <eydaimon> is it a package? I believe you can do apk add clang=9 .. whatever version
2020-07-21 23:29:53 <eydaimon> or maybe apk add clang<10 will work
2020-07-21 23:30:13 <eydaimon> simon: ☝️
2020-07-21 23:31:01 <simon> yep, that's what the link suggests: 1. add previous Alpine repo. 2. apk add 'clang<10' :-)
2020-07-21 23:54:16 <Hello71> when it breaks you get to pick up the pieces
2020-07-21 23:59:10 <simon> :-)
2020-07-22 03:57:17 <eydaimon> ikke: I didn't learn anything from the stackexchange on udev :/ i still don't know how to access by disk
2020-07-22 06:01:45 <komnpomat> lol --> floppy:x:11:root
2020-07-22 06:01:58 <komnpomat> why still around?
2020-07-22 06:54:57 <komnpomat> wow, segfaulting hard
2020-07-22 10:32:23 <ncopa> komnpomat: probably because nobody removed it
2020-07-22 10:39:25 <Arahael> Historical fact: The floppy inspired today's "save icon", often confused with the representation of a vending machine.  Infact, they were a thin, bendable "floppy" magnetic disk that was encased in a harder plastic cover.
2020-07-22 10:40:36 <ikke> They were called floppies because they used to be quite floppy. But the later smaller ones had a harder case, and hence, not that floppy
2020-07-22 10:42:02 <Arahael> Right, but the material inside was always floppy.  The case on the larger 5" and even 8" ones were particularly floppy, though, indeed.
2020-07-22 10:49:35 <jos> Hi, I'm probably missing something terribly stupid but with the update to erlang 23 in 53aebeddaa4ae6901b1c40652c1cdd21b57106cb most of the subpackages have been removed. I've been looking for information that all these packages appear to have been merged into plain "erlang". Is there some documentation out there I have missed? Maybe some info in
2020-07-22 10:49:36 <jos> the upgrade notes for 3.12?
2020-07-22 11:01:25 <ikke> It's something that possibly could have been mentioned in the release notes, but u
2020-07-22 11:01:36 <ikke> it's not easy to capture all significant changes
2020-07-22 11:02:46 <ikke> The libraries are still present, so it's more like a logistical change than a functional one
2020-07-22 11:06:56 <albr> Count Geiger[d]: How do I get a startup service to actually wait for the network to come online? I have added `need "net network-online"` to the service, but as soon as udhcp times out, it starts the service, and then I get my IP adress a few seconds later.
2020-07-22 11:08:08 <ikke> https://github.com/OpenRC/openrc/blob/master/service-script-guide.md#be-wary-of-need-net-dependencies
2020-07-22 11:11:16 <mps> komnpomat: ncopa: floppy driver got some improvement in a latest kernels
2020-07-22 11:15:00 <albr> Count Geiger[d]: Thanks for the link, but as I said, the problem is that dhcpd times out, forks to background and then says network is started, which makes my service start as well. I have tried using network-online, but i don't seem to be able to get it to run at all, regardless of whether I add it to a runlevel or not. Does net-online (the rc-script that provides network-online, from what I can tell) need to be invoked in a diff
2020-07-22 11:18:57 <ikke> It's called net-online, not network-online
2020-07-22 11:22:14 <albr> Count Geiger[d]: I have tried both, but in the net-online-script it has a `provide network-online` which I assume is the keyword I should be basing my dependency on? Either way, I have tried both, but the service doesn't get started regardless of which runlevels I try to put it in.
2020-07-22 13:42:18 <abenz> I see that thunderbird is available in edge branch, testing repository
2020-07-22 13:42:31 <abenz> is it safe to assume this will be included in the next version of alpine?
2020-07-22 13:42:45 <abenz> its not available in 3.12
2020-07-22 13:42:46 <ikke> No, as long as it's in testing, it won't be part of a stable release
2020-07-22 13:42:57 <abenz> I see
2020-07-22 13:43:15 <abenz> is there a good GUI email client that I can use? I'm open to using any
2020-07-22 13:43:25 <abenz> .. that is included in 3.12
2020-07-22 13:44:54 <Cogitri> I use Evolution on all of my machines
2020-07-22 13:47:44 <abenz> Cogitri: that looks good enough for my needs, thanks
2020-07-22 14:17:06 <mps> abenz: claws also is gui and is simple
2020-07-22 14:48:29 <abenz> mps: claws seems like a better fit for an openbox user
2020-07-22 14:48:31 <abenz> cheers
2020-07-22 14:53:05 <mps> yes, and it not resources hungry like other gui clients
2020-07-22 14:55:04 <linux-pine> I just came to say THANKS, because crontab's @reboot works!
2020-07-22 14:55:12 <linux-pine> THANK YOU VERY MUCH for Alpine Linux
2020-07-22 14:55:17 <linux-pine> <3
2020-07-22 14:55:23 <ikke> :D
2020-07-22 14:55:56 <linux-pine> I love its nobullshit approach
2020-07-22 14:56:10 <linux-pine> got my stuff running in no time, and learned a lot in the process
2020-07-22 14:58:34 <ikke> nice to hear
2020-07-22 16:25:15 <wsinatra> anyone here have a working nagios setup on Alpine? I for the life of me can't get lighttpd to serve it correctly, just end up with 403 forbidden errors..
2020-07-22 16:38:22 <dazoe> So, i have a networking question... I'm trying to share my server's ipv6 with my home using wireguard and routing. When I try to ping out i see the packets going from my home all the way out the internet but packets don't come back
2020-07-22 16:53:38 <marc5000> Hello, is there anyone able to help me with dcron on Alpine? It doesn't create cron.d in /etc/cron.d -- so doesn't conform to the man page. It also doesn't run any cron job inside a manual made /etc/cron.d location. There also isn't any log output at /var/log. Has anyone got any experience with this?
2020-07-22 16:54:04 <marc5000> To confirm, the service is running
2020-07-22 17:01:48 <komnpomat> the day alpine adds support to systemd I live stream my hands getting guillotined
2020-07-22 17:03:09 <komnpomat> big letdown with qemu here
2020-07-22 17:03:35 <komnpomat> whereas in gentoo musl it works flawlessly
2020-07-22 21:20:44 <c705> Cogitri + firneds: random question, if I notice a new longterm linux-lts dropped on kernel.org, should i submit an MR bumping the linux-lts package or is there some automated process that covers that?
2020-07-22 21:24:33 <mps> c705: this is on ncopa to decide when and which kernel will be introduced to alpine
2020-07-22 21:25:04 <c705> understood
2020-07-22 21:25:42 <mps> we expect that the new LTS will be released in september-october
2020-07-22 21:25:57 <c705> oh, i meant for edge
2020-07-22 21:26:21 <mps> yes, I'm talking about edge
2020-07-22 21:27:06 <mps> https://www.kernel.org/category/releases.html
2020-07-22 21:27:07 <c705> i mean 5.4.53 was released today, but the packages are refing to 5.4.52. We'll wait until September to bump that?
2020-07-22 21:27:22 <mps> ah, that
2020-07-22 21:27:37 <mps> this is upgrade
2020-07-22 21:27:46 <c705> not like i'm in a rush, just seeing if I can be helpful
2020-07-22 21:28:12 <mps> this is usually done by ncopa or someone from kernel team
2020-07-22 21:28:37 <mps> and we are fast at that :)
2020-07-22 21:28:45 <codebam> should I reboot for musl 1.1.24-r9 -> 1.2.0-r0 ?
2020-07-22 21:29:04 <c705> mps: sweetness
2020-07-22 21:29:06 <mps> I already noticed that I have to upgrade linux-edge tomorrow
2020-07-22 21:29:22 <c705> yeah you guys are usually within 24h I find
2020-07-22 21:29:26 <mps> codebam: not needed, but you can
2020-07-22 21:29:37 <c705> i just wait to build my stuff until the new upgrade releases
2020-07-22 21:29:38 <codebam> okay, cool :)
2020-07-22 21:30:14 <codebam> if I install linux-edge will that just add another boot entry to grub?
2020-07-22 21:30:46 <mps> it should, but I'm not 100% sure
2020-07-22 21:30:54 <codebam> I'm gonna find out, brb :)
2020-07-22 21:31:28 <c705> i don't see linux-edge in any repos. where is it?
2020-07-22 21:31:54 <mps> c705: testing
2020-07-22 21:31:55 <c705> ah, i see it, nevermind
2020-07-22 21:31:58 <c705> yup
2020-07-22 21:33:16 <codebam> yep just added another entry, awesome
2020-07-22 21:34:07 <codebam> for some reason the tty font is different on edge than lts. doesn't bother me, just something I noticed. did they change the default font in the kernel recently?
2020-07-22 21:34:09 <bndeter_> i have encountered a very wierd phenomenon. virsh/libvirt just hang and do nothing. libvirtd is definitely started. /var/log/libvirt/libvirtd.log says nothing.
2020-07-22 21:34:34 <c705> bndeter_: yeah, that's a known bug. let me link you the issue
2020-07-22 21:35:11 <mps> codebam: they should be same
2020-07-22 21:35:12 <c705> bndeter_: https://gitlab.alpinelinux.org/alpine/aports/-/issues/11602
2020-07-22 21:35:35 <c705> workaround is to kill the extra libvirtd process that spawns off of the openrc service start
2020-07-22 21:35:48 <codebam> idk then. the one in edge is less blocky and looks more like a display font
2020-07-22 21:36:37 <bndeter_> c705: thanks. was very confused.
2020-07-22 21:37:09 <c705> bndeter_: yeah, no kidding. i spent 3 weeks trying to figure it out myself
2020-07-22 21:37:11 <bndeter_> i think i tried everything. and somehow this issue dodged my googling.
2020-07-22 21:37:54 <mps> codebam: fonts have different settings :)
2020-07-22 21:38:41 <bndeter_> c705: did you automate the process killing by any chance? :D
2020-07-22 21:38:48 <c705> yep
2020-07-22 21:39:27 <c705> works like a charm. unfortunately upstream doesn't see this as a bug, so might have to be doing that for a while
2020-07-22 21:39:37 <c705> until libvirt wants to conform to POSIX again
2020-07-22 21:41:22 <bndeter_> seems confusing. just reading the issue log.
2020-07-22 21:41:45 <c705> bndeter_: https://gitlab.com/libvirt/libvirt/-/issues/52 is the upstream bug
2020-07-22 21:42:01 <c705> dalias put it quite well, though most of it goes over my head
2020-07-22 21:43:49 <c705> the way i see it, long story short, they write non conforming code that violates POSIX, but works somehow in glibc. musl has problems with their non conforming code. they don;t se an issue because "they don't target musl systems"
2020-07-22 21:49:02 <mps> c705: heh, we are fast kernel, just pushed linux-edge-5.7.10 :)
2020-07-22 21:49:23 <mps> didn't wanted to wait tomorrow
2020-07-22 21:49:36 <bndeter_> c705: i'm at the iptables part of the issue. somehow i'm really confused this is not a bigger/common issue.
2020-07-22 21:50:01 <c705> neat-o. i think i'll start using edge kernels now that I know there's a package
2020-07-22 21:50:23 <c705> bndeter_: as far as i'm concerned, it is a big issue for musl based systems. keep on reading
2020-07-22 22:06:05 <mps> c705: I forgot to tell that you can create MR for linux-lts but then you should try add all dependent packages and linux-rpi
2020-07-22 22:06:54 <c705> mps: yeah, i'm too dumb for that right now. i'll leave it up to the professionals. I'd like to get more involved at some point though
2020-07-22 22:07:17 <mps> ok, np
2020-07-22 22:52:17 <bndeter_> c705: spicy issue thread. nice read. makes me consider avoiding libvirt alltogether and qemu the stuff myself.
2020-07-22 22:53:08 <c705> bndeter_: i've half thought about it
2020-07-22 22:53:18 <c705> i'm hoping someone bigger than us complains and they get their shit together
2020-07-22 22:54:16 <bndeter_> yeah that would be nice. but also sad, that it would take that. would not increase my confidence in the tool.
2020-07-22 22:56:20 <bndeter_> i think i understood enough of the technical part to be concerned about the maintainers attitude.
2020-07-22 22:56:50 <c705> upstream is usually like that
2020-07-22 22:57:42 <c705> the guy who amde xscreensaver code bombed debian back in like 2015 and left a nasty message for us all about how our operating system was shit
2020-07-22 22:57:52 <c705> f_cking developers man..
2020-07-22 23:05:46 <bndeter_> yeah there is some xscreensaver popcorn action in my memory.
2020-07-22 23:06:58 <bndeter_> c705: any chance i can copy+paste your workaround? or are you just using this? > https://gitlab.alpinelinux.org/alpine/aports/-/issues/11602#note_97838
2020-07-22 23:07:56 <c705> yeah, hold on a sec
2020-07-22 23:09:24 <c705> https://termbin.com/f097
2020-07-22 23:09:38 <c705> it ain't pretty but it works
2020-07-22 23:09:55 <c705> first command find the pid of the exta process, second kills it
2020-07-22 23:11:53 <bndeter_> c705: thanks a lot :) that regains me a few minutes back after i tried debugging for hours :D
2020-07-22 23:12:35 <c705> bndeter_: yeah, believe me, i spent so much time trying to debug this myself. I thought it was something I was doing because generally alpines maintainers are decent
2020-07-22 23:12:53 <maldridge> c705: that works in glibc because glibc specifically has partial account for AS-unsafe code in MT forks.  Its not something that should be relied upon because its a side effect of other unspecified bugs in glibc
2020-07-22 23:13:10 <c705> maldridge: yeah, totally agree
2020-07-22 23:13:55 <maldridge> as to why libvirt has this bug, I'd say obstinance from developers that don't want to publicly admit they drew a line in the sand without fully understanding the implications of that line
2020-07-22 23:14:02 <maldridge> (and then tripping over it a few lines later)
2020-07-22 23:14:43 <c705> maldridge: it's even in their own documentation according to what dalias posted
2020-07-22 23:15:04 <c705> i can't pretend I fully understand it, but if they break posix, then what did they expect?
2020-07-22 23:15:32 <bndeter_> this line in the send looks pretty freshly drawn ;)
2020-07-22 23:17:44 <maldridge> I'd discuss more in -offtopic, but fwiw I think jwz was fully justified in their unmaintained splash screen notice
2020-07-22 23:19:08 <c705> maldridge: i don't but we'll leave it at that
2020-07-22 23:20:59 <bndeter_> c705: before i copy the commands into another script. whats the language/tool you used in the workaround? it doesn't look familiar to me
2020-07-22 23:21:25 <maldridge> that's ansible
2020-07-22 23:21:33 <c705> bndeter_: oh, this is an ansible yaml file. you can just copy the shell commands if you'd like
2020-07-22 23:24:41 <bndeter_> okey thanks. was just curious an obviously didn't use it before.
2020-07-22 23:25:55 <maldridge> https://ansible.com has more information if you're curious
2020-07-22 23:26:33 <bndeter_> i will just paste in into a service, thats depends on libvirt before i open the magic box of another automation tool :D
2020-07-22 23:26:57 <maldridge> ansible is incredibly powerful, but like all foot-cannons its best aimed at someone else
2020-07-22 23:28:07 <bndeter_> maldridge: i
2020-07-22 23:28:52 <bndeter_> i will probably not be able to dodge it alltogether, but i try to delay it as far as i can :D
2020-07-23 00:10:11 <eydaimon> if i have this rule in iptables: -A sshguard -s 112.85.42.0/24 -j DROP,  why would I see this in my /var/log/messages just a few minutes ago? Failed password for root from 112.85.42.87
2020-07-23 00:17:39 <c705> eydaimon: do you have an accept rule before the drop?
2020-07-23 00:17:52 <c705> might want to share in a pastebin `iptables -nvL` output
2020-07-23 00:20:45 <eydaimon> c705: thanks
2020-07-23 00:20:52 <eydaimon> can I send it in private to you ?
2020-07-23 00:22:11 <c705> sure
2020-07-23 00:23:35 <c705> eydaimon: you don;t have a jump rule to sshguard
2020-07-23 00:24:04 <eydaimon> c705: what's a jump rule? Sorry I've never used iptables before
2020-07-23 00:24:16 <eydaimon> I know ipfw decently :P
2020-07-23 00:24:42 <c705> eydaimon: oh..right. I don't know the firewall you're using, but it uses iptables in the backend
2020-07-23 00:25:02 <eydaimon> c705: ufw
2020-07-23 00:25:05 <c705> it looks like you created a ruleset called `sshguard` but you are not directing traffic anywhere to that ruleset
2020-07-23 00:25:10 <eydaimon> c705: well, sshguard uses iptables directly
2020-07-23 00:25:32 <eydaimon> https://sshguard.net/ is good stuff, FYI
2020-07-23 00:25:38 <c705> oh, i see. you need to configure ufw to pass traffic into that ruleset
2020-07-23 00:25:47 <c705> you know what's even safer? not running ssh!
2020-07-23 00:26:34 <eydaimon> c705: ok, let's say I stop using ufw. then what do I do ?
2020-07-23 00:26:57 <c705> then you'd have to configure iptables
2020-07-23 00:27:28 <c705> https://wiki.archlinux.org/index.php/Iptables
2020-07-23 00:27:28 <eydaimon> oh, iptables -A INPUT -j sshguard
2020-07-23 00:27:48 <c705> -A will append the rule. it needs to come before the drop rule
2020-07-23 00:27:51 <eydaimon> i freaking read this in the docs when I was setting up the other day
2020-07-23 00:27:56 <eydaimon> but I guess I missed it
2020-07-23 00:28:02 <c705> well, there ya go
2020-07-23 00:28:17 <c705> -j is a jump FYI
2020-07-23 00:28:19 <eydaimon> thanks for looking
2020-07-23 00:28:22 <c705> np
2020-07-23 00:28:25 <nacelle> you werent born with iptables skills burnt into your grey matter??!?!
2020-07-23 00:28:29 <eydaimon> yupp, I rember that in the man-page :)
2020-07-23 00:28:36 <eydaimon> nacelle: 🤣🤣🤣
2020-07-23 00:29:30 <c705> mps: i don't know if you're still around, but any idea why xtables-addons-lts exists but not xtables-addons-edge
2020-07-23 00:30:09 <helby> if I play some crap with ffplay, does it should open a new window by default if option -nodisp is not present?
2020-07-23 00:30:17 <c705> i mean, can I just add xtables in userspace?
2020-07-23 00:30:37 <c705> helby: no idea, but you can ask ##linux as well
2020-07-23 00:30:54 <helby> can't get how to play video with it in wayland
2020-07-23 00:31:01 <helby> not sure if it's related to wayland ;)
2020-07-23 00:31:10 <eydaimon> wayland works now ?
2020-07-23 00:31:31 <c705> wayland always worked. but it has some pretty significant features missing you'd might expect
2020-07-23 00:31:51 <helby> anything better than fakery with xorg
2020-07-23 00:31:58 <eydaimon> always is a stretch, I remember when it was announced ...
2020-07-23 00:32:16 <helby> I can't get this thing with ffplay
2020-07-23 00:32:22 <c705> recently, anyways
2020-07-23 00:32:27 <helby> it looks like it playing video but somehow in background somewhere
2020-07-23 00:32:30 <eydaimon> I remember when ubuntu first got it as part of the distro ... etc
2020-07-23 00:32:35 <helby> does not open that extra window with it
2020-07-23 00:32:44 <c705> helby: it's not an alpine related question, feel free to ask ##linux that has more traffic
2020-07-23 00:32:59 <helby> c705: no thank you :)
2020-07-23 00:33:13 <c705> eydaimon: ubuntu introducing it was stupid. I don;t think it's stable enough to introduce it in a main distro
2020-07-23 00:34:15 <eydaimon> c705: yeah, perhaps. I waited for it for years hoping it would make linux have a really decent desktop comperable with macos or something
2020-07-23 00:35:06 <c705> does macOS include the screen isolation as well?
2020-07-23 00:35:12 <c705> I kind of doubt that
2020-07-23 00:36:28 <eydaimon> doesn't matter to me. it's a proper window manager. Even android ... but it's been awhile since I ran X on a desktop
2020-07-23 00:37:06 <c705> wayland isn't a window manager though
2020-07-23 00:37:16 <eydaimon> window server then
2020-07-23 00:37:44 <eydaimon> last time I used  X professionally was back in 2007 on FreeBSD. I thought my machine was starting to run a bit sluggish. Turns out I only had 256mb of ram 😂
2020-07-23 00:38:21 <eydaimon> still getting: Failed password for root from 112.85.42.87
2020-07-23 00:38:40 <eydaimon> and that'a fter adding that: iptables -A INPUT -j sshguard
2020-07-23 06:06:47 <eydaimon> is there a way to flush all the rules and start clean  with iptables ?
2020-07-23 06:29:52 <mps> --flush   -F [chain]          Delete all rules in  chain or all chains
2020-07-23 08:15:58 <ikke> Don't forget to reset policies before that
2020-07-23 08:16:07 <ikke> if the default policy is DROP, you lock yourself out
2020-07-23 08:20:11 <ikke> (-F does not reset the default policies of each chain)
2020-07-23 08:32:54 <Hund> Hello! I cannot get bitlbee to use any user but "nobody" in Alpine. What am I doing wrong?
2020-07-23 08:33:42 <ikke> How are you specifying another user to use?
2020-07-23 08:34:47 <ikke> nobody:nobody seems hardcoded in the initd file
2020-07-23 08:35:08 <Hund> "User = bitlbee" in the config.
2020-07-23 08:35:12 <Hund> As always.
2020-07-23 08:35:16 <Hund> Hm.
2020-07-23 08:35:36 <ikke> that requires it to start as root
2020-07-23 08:35:38 <ikke> which it doesn't
2020-07-23 08:35:51 <Hund> It doesn't?
2020-07-23 08:35:53 <ikke> no
2020-07-23 08:35:57 <ikke> it starts as nobody
2020-07-23 08:36:10 <Hund> Why did Alpine change that?
2020-07-23 08:36:31 <ikke> It's a best practice to not start daemons as root if they don't require it
2020-07-23 08:36:47 <Hund> Is it safe to run bitlbee as nobody?
2020-07-23 08:37:51 <ikke> Yes, that's kind of it's purpose
2020-07-23 08:39:04 <Hund> Thanks!
2020-07-23 08:39:12 <ikke> Hmm, this does give a different perspective though: https://askubuntu.com/a/674397/1733
2020-07-23 08:40:18 <Hund> I read about that before. It didn't sound safe to run bitlbee as nobody, but I trusted you. :P
2020-07-23 08:41:12 <ikke> Well, it's not unsafe to use the user nobody
2020-07-23 08:42:07 <Hund> Okey. :)
2020-07-23 08:42:29 <ikke> it's basically just another user on the system
2020-07-23 08:43:36 <ikke> That answer says basically that it's usefulness diminished when you run more than one daemon as the same user
2020-07-23 08:43:45 <ikke> disminishes*
2020-07-23 08:44:06 <Hund> I'm only using WeeChat. It's a tiny client with like two things running on it.
2020-07-23 08:45:04 <ikke> I would not worry too much about it
2020-07-23 09:22:17 <Hund> I changed my cache folder to an external disk with setup-apkcache. How do I delete the old read only folder?
2020-07-23 09:22:46 <ikke> the cache folder should not be read-only
2020-07-23 09:23:18 <Hund> Oh
2020-07-23 09:23:23 <Hund> It is for some reason.
2020-07-23 09:24:39 <Hund> It's set to "ro" in fstab.
2020-07-23 09:25:19 <ikke> On what disk / device?
2020-07-23 09:25:21 <Hund> `/dev/mmcblk0p1` to be precise.
2020-07-23 09:29:05 <Hund> The disk is full and the old cache is about 200 MB in size, so it would be nice if I could remove it.
2020-07-23 09:29:28 <ikke> Where is that disk mounted?
2020-07-23 09:30:33 <Hund> /dev/mmcblk0p1           29.7G    139.2M     29.6G   0% /media/mmcblk0p1
2020-07-23 09:32:25 <ikke> ok, you can do: mount -oremount,rw /media/mmcblk0p1
2020-07-23 09:32:30 <ikke> that should mount it writable again
2020-07-23 09:33:35 <Hund> Nope.
2020-07-23 09:35:09 <TBB> flick the switch on your mmc card perhaps :)
2020-07-23 09:35:43 <Hund> I don't tink thos tiny cards have one.
2020-07-23 09:35:45 <Hund> those
2020-07-23 09:36:02 <Hund> I'm pretty sure I wouldn't be able to save anything to it then either.
2020-07-23 09:39:28 <ikke> But how can you write anything to it if it's mounted ro?
2020-07-23 09:40:22 <Hund> How do I know? I'm not the one who created Alpine. :P
2020-07-23 09:41:18 <ikke> rootfs is tmpfs, that's not persisted
2020-07-23 09:41:51 <ikke> I assume this a run-from-ram system, given that you are using sd cards
2020-07-23 09:42:00 <Hund> Yes
2020-07-23 09:42:48 <ikke> what does blkid /dev/mmcblk0p1 return?
2020-07-23 09:43:54 <Hund> /dev/mmcblk0p1: UUID="D7BE-67A6" TYPE="vfat"
2020-07-23 09:44:09 <ikke> ok, vfat should be writable at least
2020-07-23 09:44:12 <ikke> and dmesg | tail?
2020-07-23 09:44:33 <Hund> [10239.046428] FAT-fs (mmcblk0p1): error, fat_free_clusters: deleting FAT entry beyond EOF
2020-07-23 09:45:28 <ikke> aha, so it's mounted ro due to a filesystem error
2020-07-23 09:45:28 <Hund> [10239.046444] FAT-fs (mmcblk0p1): Filesystem has been set read-only
2020-07-23 09:45:33 <ikke> yes
2020-07-23 09:45:37 <Hund> Yep.
2020-07-23 09:46:01 <ikke> So you need to fix that first
2020-07-23 09:46:10 <ikke> if possible
2020-07-23 09:46:55 <ikke> make sure dosfstools is installed
2020-07-23 09:47:07 <ikke> then try to run fsck /dev/mmcblk0p1
2020-07-23 09:48:42 <oliv3r[m]> Hey Guys, I'm seeing a lot of BAD SIGNATURE on the arm repo `ERROR: py3-dbusmock-0.19-r0: BAD signature` for example, but also others. amd64 seems to be fine; any idea when this is resolved (and where the problem lies?)
2020-07-23 09:49:02 <oliv3r[m]>  * Hey Guys, I'm seeing a lot of BAD SIGNATURE on the armv7 repo `ERROR: py3-dbusmock-0.19-r0: BAD signature` for example, but also others. amd64 seems to be fine; any idea when this is resolved (and where the problem lies?)
2020-07-23 09:49:26 <ikke> oliv3r[m]: https://lists.alpinelinux.org/~alpine/devel/%3C20200721171650.48fa63a4%40ncopa-desktop.copa.dup.pw%3E
2020-07-23 09:51:19 <ikke> to get around it, either use a different mirror (our cdn mirror cached the old packages), or wait for us to have purged them
2020-07-23 09:51:53 <Hund> ikke: It fixed some things, but I can't mount it as rw.
2020-07-23 09:52:49 <oliv3r[m]> thanks ikke; so different mirror is a good idea; how long will the purging of the CDN take you reccon?
2020-07-23 09:53:38 <ikke> oliv3r[m]: puring itself doesn't take long, but because package are still being rebuild (after they have been fixed), it will have to be done regularly
2020-07-23 09:54:12 <oliv3r[m]> ok; so a few days at least to be running 'normally' again
2020-07-23 09:54:45 <Hund> ikke: Running it two more times seems to have fixed it.
2020-07-23 09:56:57 <Hund> I'm not sure how to fix the full tmpfs though?
2020-07-23 09:57:55 <Hund> So I can save the changes to the 'disk'.
2020-07-23 09:58:14 <ikke> lbu commit
2020-07-23 09:59:09 <Hund> I know, but something is full so I can't save the changes.
2020-07-23 09:59:24 <Hund> gzip: short write: No space left on device
2020-07-23 09:59:26 <Hund> tar: write error: Broken pipe
2020-07-23 09:59:28 <Hund> Problems creating archive. aborting
2020-07-23 09:59:36 <ikke> Then you would need to make space first :-/
2020-07-23 09:59:42 <Hund> Where?
2020-07-23 09:59:51 <ikke> I don't know
2020-07-23 10:00:00 <ikke> verify what is on the sdcard
2020-07-23 10:00:00 <Hund> I have 29.6G free space on my sdcard.
2020-07-23 10:00:41 <ikke> yes, so something must have taken that space
2020-07-23 10:01:11 <Hund> df shows me that I have 29.6G free to use.
2020-07-23 10:04:04 <Hund> https://dpaste.org/Z2Uo/slim
2020-07-23 10:18:31 <Hund> I'm relly close to scrapping this Pie and renting a VPN so I can use something properly.
2020-07-23 10:25:51 <ikke> Hund: can you check /etc/lbu/lbu.conf?
2020-07-23 10:26:47 <Hund> I found the issue.
2020-07-23 10:26:54 <Hund> It was me. Who would have guessed. :P
2020-07-23 10:27:49 <Hund> I added the folder /home to be included and then forgot to exclude it when I moved /home to the external disk, which have about 90GB of data it wanted to compress on the 32GB card.
2020-07-23 10:35:37 <ikke> Hund: aha, that explains
2020-07-23 10:35:45 <Hund> :)
2020-07-23 13:38:02 <opal> how would i e.g. add wireguard-rpi kernel module? i can bindmount /boot but do i have to manually create a squashfs and replace the modloop file?
2020-07-23 13:41:28 <abenz> mps: I'm transferring all my email accounts over to claws, thanks again for the recommendation. Do you happen to know a good irc client that follows the same minimalist philosophy? I'm using hexchat atm
2020-07-23 13:50:43 <mps> abenz: no, I use only text mode user clients whenever possible, currently my IRC client is irssi, though I used weechat for some short time before switched to irssi
2020-07-23 13:51:31 <mps> of the gui IRC clients I tried hexchat looks best, in my experience
2020-07-23 13:51:36 <abenz> I see
2020-07-23 13:52:04 <MY-R> hexchat ye, is best if talk about gui, very configurable
2020-07-23 13:54:40 <telmich> A question regarding firefox+drm on alpine: it seems watching netflix is not possible / the DRM plugin crashes on 3 unrelated alpine notebooks when I try to play a movie - does anyone know the background to this?
2020-07-23 13:55:37 <shiroyasha> 
ACTION gave up trying to play DRM video on Alpine Linux.

2020-07-23 13:56:10 <MY-R> probably widevine and musl
2020-07-23 13:59:12 <qman__> abenz: I use quassel IRC, it isn't minimalist, but it the UI is typical classic design and it has a split core/UI design, so you can use it like a built in bouncer
2020-07-23 13:59:57 <qman__> I'm also using claws mail and it's clunky at best, though I'm guessing it works better if you have less email than I do
2020-07-23 14:00:19 <Cogitri> telmich: I think Ariadne was looking into that
2020-07-23 14:00:23 <qman__> one thing that's not obvious from the start is that claws mail can only do one thing at a time
2020-07-23 14:00:42 <opal> hell does "minimalist" mean here anyway
2020-07-23 14:01:15 <ikke> not using all ram :P
2020-07-23 14:01:20 <qman__> well, to me that means irssi
2020-07-23 14:01:44 <Ariadne> i use quassel as well
2020-07-23 14:02:19 <qman__> unfortunately there aren't many options for GUI email clients on alpine
2020-07-23 14:02:41 <telmich> Ariadne: how far did you come in terms of DRM/firefox?
2020-07-23 14:02:48 <opal> 285M  0:30.30 claws-mail
2020-07-23 14:02:54 <opal> its used up to 700mb
2020-07-23 14:02:55 <telmich> <qman__ "unfortunately there aren't many "> I have to say mu4e is one of the best mail clients I have been using so far
2020-07-23 14:03:15 <Ariadne> telmich: need libxul debug symbols to dig further ;/
2020-07-23 14:03:17 <telmich> (which is available in alpine)
2020-07-23 14:04:52 <MY-R> found file which provide binaries/plugins with widevine, so I doubt that can do anything without sources: https://github.com/mozilla/gecko-dev/blob/master/toolkit/content/gmp-sources/widevinecdm.json
2020-07-23 14:07:29 <telmich> Ahh, there are blobs involved that are potentially linked against glibc?
2020-07-23 14:07:47 <MY-R> yep
2020-07-23 14:08:43 <telmich> ok. In this regard I appreciate being on musl to see the effects, even if it stops me from procrastinating
2020-07-23 14:16:42 <abenz> qman__: thunderbird is very bloated
2020-07-23 14:17:10 <abenz> so claws mail, when compared to thunderbird, looks/feels very minimalist and light on resources
2020-07-23 14:17:18 <opal> yeah compared to tbird its great
2020-07-23 14:17:20 <opal> If you are running from a RAM disk you can't modify the modloop.
2020-07-23 14:17:20 <opal> You can get around it by unpacking the modloop, mount the unpacked modules folder and then installing wireguard.
2020-07-23 14:17:26 <opal> ah, thats the only way?
2020-07-23 14:17:46 <opal> now im starting to feel the pain of diskless
2020-07-23 14:20:13 <afontain_> nico: it works well for games too. It doesn't work so well for IRC/Matrix though.
2020-07-23 14:21:13 <telmich> afontain: true, for matrix I also use a browser - I have it open for other web stuff [tm] anyway
2020-07-23 14:21:22 <opal> oh telmich is nico
2020-07-23 14:21:25 <opal> oof
2020-07-23 14:34:41 <Latrina> folks, a quick question. Will alpine with syslog-ng work as netconsole receiver when the sender is a Debian host?
2020-07-23 14:35:00 <Latrina> do I need to build the kernel module or anything else for it to work?
2020-07-23 14:39:18 <eydaimon> hm, I tried reinstalling iptables and cleaning out everything, but if I do that it complains saying i need to "First create some rules", but I didn't need to do that on first run. What's different ?
2020-07-23 14:40:05 <opal> Latrina it should speak the same stuff
2020-07-23 14:40:13 <opal> netconsole is pretty simple iirc
2020-07-23 14:40:31 <Latrina> cool opal, thanks :)
2020-07-23 15:10:00 <Latrina> I will give it a go
2020-07-23 15:43:07 <telmich__> opal: sorry for the confusion, I'm still connected via irc + matrix, did not yet disconnect from IRC
2020-07-23 16:09:59 <c705> hi friends. any idea why xtables-addons-edge does not exist when xtables-addons-lts does? the script i'm using (mkimage.sh) is trying to add xtables add ons but since xtables-addons-edge doesn;t exist, it's failing
2020-07-23 16:10:24 <c705> furthermore, can I just add xtables in userspace instead of with the kernel? what impact will this have?
2020-07-23 16:14:46 <komnpomat> which package to add for missing characters?
2020-07-23 16:14:56 <komnpomat> a lot of site showing missing characters
2020-07-23 16:20:47 <mps> c705: because I don't want to make third party drivers or modules for linux-edge, and no one asked with elaborate write why they are needed
2020-07-23 16:23:03 <c705> mps: what am I going to lose if I skip xtables-addons?
2020-07-23 16:24:13 <mps> if you didn't used it in your config you lose nothing
2020-07-23 16:24:29 <mps> firewall config, I mean
2020-07-23 16:25:02 <c705> use what exactly? i'm using iptables
2020-07-23 16:25:41 <mps> yes, iptables with userspace extensions in xtables addons
2020-07-23 16:26:10 <mps> I'm also using iptables but never needed xtables
2020-07-23 16:26:28 <c705> ok, i'll just delete it and give it a shot I guess
2020-07-23 16:26:49 <c705> I imgine I can just add xtables as a standalone package and modprobe it later if I needed it, no?
2020-07-23 16:27:53 <mps> 'apk info -L xtables-addons' will tell you which extensions it contains
2020-07-23 16:28:04 <c705> thanks
2020-07-23 16:28:27 <mps> btw, linux-edge is not replacement for linux-lts
2020-07-23 16:29:15 <mps> linux-edge is more for testing new features and for users who need new kernels with these new features
2020-07-23 16:30:05 <mps> and it is optimized more for performance (workstations) than linux-lts which is for broad use cases
2020-07-23 16:30:22 <c705> yeah, good point. I would like the kernel availiable in my boot though
2020-07-23 16:30:26 <helby> mps: why not replacement? it's stable kernel
2020-07-23 16:30:37 <mps> I also use linux-lts on most servers, except those where I need new features
2020-07-23 16:31:02 <c705> and for my workstation. my servers will still run lts
2020-07-23 16:31:47 <mps> helby: because alpine have to backport fixes to kernel for two years at least, and LTS is best suited for this
2020-07-23 16:32:12 <helby> mps: but u don't need any backport for edge I suppose
2020-07-23 16:32:44 <mps> with non-LTS which disappears fast this would be hard and time consuming
2020-07-23 16:32:53 <helby> how disappears?
2020-07-23 16:33:03 <helby> it will be just upgraded
2020-07-23 16:33:04 <mps> helby: edge will next stable
2020-07-23 16:33:33 <helby> edge is already stable
2020-07-23 16:33:55 <mps> heh
2020-07-23 16:34:03 <mps> helby: look here https://www.kernel.org/
2020-07-23 16:34:55 <mps> there are no 5.5 and 5.6, soon 5.7 will disapper when 5.9 arrive
2020-07-23 16:35:11 <helby> it's not yes, but edge in alpine will be upgraded not?
2020-07-23 16:35:35 <mps> linux-edge or alpine edge?
2020-07-23 16:35:47 <helby> linux-edge on alpine edge :)
2020-07-23 16:36:16 <helby> so 5,7 will become 5,8 and then 5,9 or not?
2020-07-23 16:36:24 <mps> yes, that is plan, whenever new stable is released it should be soon on alpine edge
2020-07-23 16:36:31 <helby> so
2020-07-23 16:36:41 <helby> where is a problem with edge then?
2020-07-23 16:37:05 <mps> terminology, edge is alpine development version
2020-07-23 16:37:14 <helby> like long term is supported for some time, edge is infinite supported ;)
2020-07-23 16:37:32 <mps> and linux-edge is kernel which will be always in alpine edge, never in stable
2020-07-23 16:38:42 <mps> but at the same time linux-lts will be also always in alpine edge as next kernel for next alpine stable
2020-07-23 16:39:28 <helby> that was annoying during installation, that no edge installer
2020-07-23 16:39:39 <helby> had to install stable old crap, then update
2020-07-23 16:39:54 <helby> or did not find a way how to do directly edge
2020-07-23 16:42:01 <mps> that could be done by creating chroot of egde repos directly, but I think install stable then upgrade is easier and less error prone
2020-07-23 16:42:45 <helby> yes, chroot is not so cool, especially with only one machine available
2020-07-23 16:43:21 <helby> last time I tried to chroot and remove fucking grub, I had to reinstall :)
2020-07-23 16:44:34 <helby> only grub can crate so many folder in /boot
2020-07-23 16:44:37 <helby> folders
2020-07-23 16:44:54 <helby> fuck grub should be banned!
2020-07-23 16:46:08 <opal> just dont use it
2020-07-23 16:46:32 <opal> pretty simple
2020-07-23 16:47:35 <afontain_> tbh, I switched back to grub
2020-07-23 16:47:54 <opal> interesting, how come
2020-07-23 16:48:00 <opal> im happy with syslinux
2020-07-23 16:50:02 <afontain_> I found I had less trouble getting it to work with efi
2020-07-23 16:50:41 <afontain_> maybe it's because I'm more used to it
2020-07-23 16:55:27 <helby> the issue is, that I had to change also efi partition to be just /boot not stupid /boot/efi
2020-07-23 16:55:41 <helby> that I tried last time I was not able to boot :)
2020-07-23 16:56:15 <helby> not sure how to do that a proper way? just format as vfat and change fstab is enough?
2020-07-23 16:56:43 <helby> then I would like just add an entry with efibootmgr obviously without any bootloader
2020-07-23 16:57:16 <helby> but how will linux/mkinitfs know where to put these entries
2020-07-23 16:58:27 <helby> I mean thse ssytem.map, config, initramfs and vmlinuz crap
2020-07-23 17:17:13 <helby> I go try :)
2020-07-23 17:17:45 <helby> just not sure if without changing fstab and right know keep boot at boot/efi I should put 'efi/ in path while in efibootmgr
2020-07-23 17:18:34 <helby> or probably copy right know these shits to /boot/efi
2020-07-23 17:20:00 <helby> see u after reboot, I hope :)
2020-07-23 17:33:09 <helby> the issue is ... nothign is creating System.map config and vmlinuz in efi
2020-07-23 17:33:20 <helby> mkinitfs does just initramfs
2020-07-23 17:52:32 <helby> okay fuck it for now, no good time to risk noboot ;)
2020-07-23 17:53:08 <helby> did not find a word in alpine docs about efistub boot
2020-07-23 17:54:41 <helby> for fuck sake, 90% of wiki should be deleted already
2020-07-23 19:25:08 <bndeter_> which package provides pdf printing for lp? i can't seem to find it :|
2020-07-23 19:47:45 <bndeter_> seems i have a bigger problem :D 'lp: Unsupported document-format "text/plain"'
2020-07-23 19:48:45 <bndeter_> i may have to add some stuff in addition to the ppd file i extracted from the debian package for my printer.
2020-07-23 19:49:29 <bndeter_> has anyone expirience with those brother printer drivers? they provide .deb package.
2020-07-23 19:58:36 <bndeter_> do i really have to install these packages somehow > https://support.brother.com/g/b/downloadlist.aspx?c=us_ot&lang=en&prod=hll8250cdn_all&os=128
2020-07-23 20:01:15 <bndeter_> the alternative seems to try su execute brothers wierd installer tool.
2020-07-23 20:01:36 <bndeter_> which will probably not work with musl.
2020-07-23 21:28:22 <helby> of course I had to reinstall bc. somehow fucked up that boot without grub
2020-07-23 21:29:50 <helby> chroot is so complicated on alpine, that of course I don't remember what everything and how to mount
2020-07-23 21:29:53 <helby> so ...
2020-07-23 21:31:22 <helby> right now I have to somehow survive with 17MB useless bootloader
2020-07-23 21:53:08 <helby> I have this shit in tty, not sure how it appears there now ..."Kernel·has·no·file·descriptor·comparison·support:·Function·not·implemented"
2020-07-23 22:04:16 <linux-pine> Hello!
2020-07-23 22:04:20 <linux-pine> https://wiki.alpinelinux.org/wiki/Installing_Oracle_Java
2020-07-23 22:04:24 <linux-pine> this doesn't work?
2020-07-23 22:04:53 <linux-pine> "If nothing comes up install it: apk add paxctl"
2020-07-23 22:05:29 <linux-pine> ERROR: unsatisfiable constraints:
2020-07-23 22:05:29 <linux-pine>     required by: world[paxctl]
2020-07-23 22:05:29 <linux-pine>   paxctl (missing):
2020-07-23 22:12:30 <linux-pine> it looks like this documentation is obsolete
2020-07-23 22:12:38 <linux-pine> any idea how to run java on current versions?
2020-07-23 22:18:32 <mps> apk search openjdk, and select version to install
2020-07-23 22:19:58 <linux-pine> mps no result
2020-07-23 22:20:49 <mps> did you enabled community repo
2020-07-23 22:21:01 <linux-pine> no
2020-07-23 22:21:33 <mps> read wiki about installation and settings extra repos
2020-07-23 22:22:23 <mps> linux-pine: btw, do you run alpine on some of the pinebook arm64 boxes
2020-07-23 22:22:55 <linux-pine> no, mps
2020-07-23 22:32:44 <linux-pine> it worked, mps, thanks!
2020-07-23 22:37:48 <mps> np
2020-07-23 22:57:34 <bndeter_> helby: apk add arch-install-scripts and arch-chroot if you dont want to manually bind mount everything.
2020-07-23 23:04:32 <eydaimon> root@luffy /t/docker (docker) [1]# apk upgrade
2020-07-23 23:04:34 <eydaimon> 1 error; 1910 MiB in 504 packages
2020-07-23 23:04:39 <eydaimon> I keep seeing this erorry. how can I find out what this is ?
2020-07-23 23:06:09 <ikke> try apk fix
2020-07-23 23:06:13 <bndeter_> i'm no expert, but woult try apk fix -s first :)
2020-07-23 23:08:02 <eydaimon> ok. apk fix didn't solve it, but I will try fix -s
2020-07-23 23:08:26 <eydaimon> (1/1) Reinstalling groff-doc (1.22.4-r0) 1 error; 1910 MiB in 504 packages
2020-07-23 23:08:39 <eydaimon> if without the -s it just keeps trying to reinstall groff-doc
2020-07-23 23:09:05 <ikke> -s is simulate, it wouldn't do anything
2020-07-23 23:09:14 <ikke> except maybe reporting
2020-07-23 23:09:22 <eydaimon> how do i force delete a package ?
2020-07-23 23:09:35 <eydaimon> apk del --force hasn't worked
2020-07-23 23:24:15 <helby> bndeter_: oh shit, if I knew that
2020-07-23 23:24:46 <helby> that will mount all crap and chroot inside just with 'arch-chroot' then I can fuck it up how I want? :d
2020-07-23 23:25:20 <ikke> bndeter_: nice btw, wasn't aware that we had arch-chroot
2020-07-23 23:25:47 <helby> bndeter_: btw. install it where? on rescue usb?
2020-07-23 23:29:14 <helby> for fuck sake, it's archlinux installing script
2020-07-23 23:31:18 <ikke> arch-chroot isn't
2020-07-23 23:31:55 <helby> arch-chroot is not as package
2020-07-23 23:33:39 <helby> ... if that wiki is not so useless ;)
2020-07-23 23:33:50 <bndeter_> ikke: was also pleasently suprised when i found it.
2020-07-23 23:35:04 <bndeter_> helby: i can't/didn't really follow what you do/try, so am not able to answer your question.
2020-07-23 23:36:44 <helby> bndeter_: already reinstalled but before fucked up bootloader /boot and shit around, so wanted chroot from install usb of alpine and repair it
2020-07-23 23:37:08 <helby> bndeter_: but that requires more than just mount efi and ext4 boot
2020-07-23 23:37:27 <helby> so as it's complicated I did not know what to do, so reinstalled
2020-07-23 23:38:55 <helby> my issue with fucking up boot came , that I wanted to get rid of grub, then mount efi as /boot instead of default /boot/efi put there all that vmlinuz initram crap, add entry to uefi with efibootmgr but during boot got error that can't boot /root and showed me sh console where keyboard did not work at all
2020-07-23 23:39:25 <helby> that's my story and it's only bc. nobody helped me here!
2020-07-23 23:40:27 <helby> so I blame these 400 unhelpful users here :)
2020-07-23 23:41:01 <helby> and of course stupid outdated wrong wiki
2020-07-23 23:41:04 <bndeter_> if you want to repair you main system you have to 'apk add arch-install-scripts' on the live stick. and then 'apk add -P /mnt/ bash'(or wherevery you mounted the main os. then arch-chroot /mnt and begin repair as normal.
2020-07-23 23:41:39 <helby> archlinux chroot works on alpine too?
2020-07-23 23:41:53 <ikke> yes, it's just a helper script
2020-07-23 23:41:56 <helby> oh
2020-07-23 23:41:57 <bndeter_> its just a script.
2020-07-23 23:42:12 <helby> I bet first it will cry about not bash in a system
2020-07-23 23:42:22 <bndeter_> it makes some bind mounts in addition to chrooting, so you can access /dev/sd* and install bootloader stuff.
2020-07-23 23:42:27 <helby> and 20 other gnu tools?
2020-07-23 23:42:55 <bndeter_> only bash, as you can see in my previous message.
2020-07-23 23:43:01 <helby> good to know that it works, bc. chroot on arch was okay for me all the time
2020-07-23 23:43:10 <Thermi> arch-chroot is king
2020-07-23 23:43:27 <helby> will be I able to install that bash in that broken system?
2020-07-23 23:43:30 <Thermi> Much better for newbies, too, because trying to use chroot directly just asks for pain and suffering
2020-07-23 23:43:37 <ikke> you can also use /bin/sh instead
2020-07-23 23:43:42 <helby> oh okay
2020-07-23 23:43:58 <bndeter_> yes you could also define another shell if you like.
2020-07-23 23:44:17 <ikke> You can start any program in the chroot you want
2020-07-23 23:44:22 <helby> so I will just run script after intallation form that usb and it will mount all sys proc and other crap ye?
2020-07-23 23:44:37 <helby> I was already in chroot, but mounted just efi and root
2020-07-23 23:44:49 <helby> so it was not enough
2020-07-23 23:45:17 <Thermi> arch-chroot mounts, dev, proc, sys (maybe tmp, too)
2020-07-23 23:45:19 <helby> I was able to install something, but grub always ends with error bc. something was not mounted obviously
2020-07-23 23:45:24 <helby> cool
2020-07-23 23:45:55 <helby> and how arch-chroot know about alpine shits?
2020-07-23 23:46:06 <ikke> it doesn't
2020-07-23 23:46:07 <Thermi> It doesn't and doesn't need to because it doesn't do distro specific shits
2020-07-23 23:46:09 <ikke> and it doesn't need to
2020-07-23 23:46:10 <Thermi> generally
2020-07-23 23:46:27 <Thermi> And yeah, it doesn't need to
2020-07-23 23:47:01 <Thermi> Alpine <> Arch is basically just userspace
2020-07-23 23:48:45 <helby> it still does not look a way
2020-07-23 23:48:58 <helby> it even requires some gnu coreutils
2020-07-23 23:49:12 <helby> h it will install it
2020-07-23 23:49:16 <helby> okay
2020-07-23 23:49:27 <helby> n
2020-07-23 23:49:54 <helby> so now is maybe time for second atempt to get rid of grub
2020-07-23 23:50:35 <helby> the main issue is probably that /dev/...1 is mounted at /boot/efi
2020-07-23 23:50:46 <Thermi> why?
2020-07-23 23:50:48 <helby> it did not help if I just changed it in fstab for /boot
2020-07-23 23:51:10 <Thermi> The files need to be in the right place, not just for EFI but also where the package maintainer expect them to be
2020-07-23 23:51:13 <helby> bc. mkinifs and linux is installed on /boot and it sound logical to use that
2020-07-23 23:51:41 <Thermi> EFI directory doesn't need to contain linux and the initramfs
2020-07-23 23:52:02 <Thermi> Just needs to contain the EFI bootloader you want to use and it then needs to find the boot dir
2020-07-23 23:52:08 <Thermi> or boot partition
2020-07-23 23:52:09 <helby> so I changed in fstab mount /dev/nvme0n1p1 crap on /boot then installed linux and everything was there, all that system.map vmlinuz and initrd
2020-07-23 23:52:34 <helby> but I wanted to use directly uefi / efistub
2020-07-23 23:52:46 <Thermi> You don't want that, really
2020-07-23 23:52:46 <helby> so just added entry with efibootmgr directly to uefi
2020-07-23 23:53:02 <helby> why not? I never needed any bootloader
2020-07-23 23:53:04 <helby> it's useless
2020-07-23 23:53:20 <helby> kernel include efistub, uefi is able to load kernel directly
2020-07-23 23:53:28 <helby> why use 17MB grub?
2020-07-23 23:53:37 <helby> there is no one reason, just to be idiot
2020-07-23 23:53:52 <Thermi> If the kernel is broken you won't be able to just boot another one. You'll have to either use external tools and boot from an external source to change the efi config to boot another one or permanently have another stub setup
2020-07-23 23:54:27 <helby> I always have just one kernel
2020-07-23 23:54:37 <helby> so grub will not help here anyway
2020-07-23 23:54:37 <Thermi> No native support for fallback initramfs or other stuff to fall back to. You'll need to hand weave that into your efi boot order
2020-07-23 23:54:52 <helby> default alpine did not any fallback
2020-07-23 23:55:01 <helby> or any other option during grub boot
2020-07-23 23:55:01 <Thermi> Then that's your own downfall. I have had broken kernels several times in the last years already.
2020-07-23 23:55:13 <helby> it's just add stupid bootloader which complicate things
2020-07-23 23:55:19 <Thermi> I had to boot linux-lts then and revert linux back to an old version
2020-07-23 23:55:30 <helby> never happened to me
2020-07-23 23:55:38 <Thermi> I hope you'll never have.
2020-07-23 23:55:48 <helby> how can be kernel broken?
2020-07-23 23:56:14 <Thermi> Fails to prepare devices, initialize or use devices correctly such that the system can not continue booting or the init system fails initializing the system
2020-07-23 23:56:40 <helby> but for that case, there is chroot probably
2020-07-23 23:56:41 <Thermi> Fails to boot before control is passed to the user. No shell, no UI
2020-07-23 23:56:56 <Thermi> For which you will need an external source to boot from
2020-07-23 23:57:12 <helby> yes, get it, u right
2020-07-23 23:57:20 <Thermi> If you have a different, known good kernel, to boot from, you don't need that external source in that specific case.
2020-07-23 23:57:37 <Thermi> *a different, known good, kernel, ...
2020-07-23 23:57:40 <helby> well, that's possible with efistub too
2020-07-23 23:57:44 <helby> just adding two entries
2020-07-23 23:57:55 <helby> not any bloated bootloader needed
2020-07-23 23:57:56 <Thermi> You'll have to manually create it is what I wrote
2020-07-23 23:58:07 <helby> yes with one command
2020-07-23 23:58:14 <Thermi> With complex syntax
2020-07-23 23:58:28 <helby> that's true, still don't remember it
2020-07-23 23:58:29 <Thermi> Don't take me for a fool, I did such things some time ago
2020-07-23 23:58:46 <helby> there is something like just efibootmgr -c but somehow it never worked bc. I don't know it requires shity setup
2020-07-23 23:58:50 <Thermi> I reverted back to using GRUB when there was more trouble than it was worth
2020-07-23 23:58:59 <helby> every trouble is bc. of grub
2020-07-23 23:59:03 <Thermi> Ah, no
2020-07-23 23:59:09 <helby> yes :)
2020-07-23 23:59:18 <helby> 17MB bootloader for fuck sake?
2020-07-23 23:59:38 <Thermi> 17 MB on nonvolatile storage in size of Gigabytes?
2020-07-23 23:59:45 <Thermi> Peanuts.
2020-07-23 23:59:53 <helby> storage is not important
2020-07-23 23:59:57 <Thermi> Not worth my time to save just 17 MB
2020-07-24 00:00:00 <helby> but what a bunch of code is it
2020-07-24 00:00:12 <helby> not about size, but code what do shit, just boot
2020-07-24 00:00:31 <helby> and add some second to boot time
2020-07-24 00:00:49 <helby> of course also it's another point what can break in a system
2020-07-24 00:01:04 <helby> bc. with direct boot, u have zero problem
2020-07-24 00:01:08 <Thermi> You got to decide what your time is worth to you. I have to get back to beating ZFS
2020-07-24 00:01:23 <helby> oh yeah, grub, zfs and encryption
2020-07-24 00:01:27 <helby> there are problems
2020-07-24 00:01:41 <helby> complicated things ...
2020-07-24 00:01:43 <Thermi> Depends on how bad the EFI is (pretty bad, all the 6-8 EFI versions I have had the fun to deal with)
2020-07-24 00:01:51 <Thermi> hf
2020-07-24 00:01:52 <Thermi> good effort
2020-07-24 00:02:40 <helby> shit, before I tried without changing crap in /etc/fstab and it booted well
2020-07-24 00:02:58 <helby> then I removed grub, mounted /boot instead /boot/efi and it was gone
2020-07-24 00:03:07 <helby> if I know why it failed
2020-07-24 00:03:45 <helby> if I keep it mounted how it is in /boot/efi and manualy copy these entries from /boot it will work, but I will once forget to copy them and will need that script ;0
2020-07-24 00:09:42 <helby> I will do it later, some day when nothing to do, bc. I am not sure how all these works
2020-07-24 03:48:36 <codebam> do we have wireguard in linux-edge?
2020-07-24 03:49:07 <codebam> I have wireguard-lts but I don't think that works with linux-edge?
2020-07-24 04:00:15 <eydaimon> so how do i deal with the apk errors ?
2020-07-24 04:00:49 <eydaimon> oh derp. I had edge uncommented
2020-07-24 04:01:26 <eydaimon> probably didn't help things
2020-07-24 04:05:15 <eydaimon> all fixed ;)
2020-07-24 06:30:50 <mps> codebam: wireguard driver is in kernel as of 5.6 version, so not needed external wireguard-edge
2020-07-24 08:06:55 <mtneug> I'm having trouble with networking on Alpine edge after an update a couple of days ago and also on completely new systems
2020-07-24 08:06:59 <mtneug> on boot eth0 remains down and udhcpc fails to get a lease
2020-07-24 08:07:04 <mtneug> /etc/network/interfaces has a very simple auto eth0 iface eth0 inet dhcp
2020-07-24 08:07:09 <mtneug> `ifup eth0` does not enable the interface and thus udhcpc fails
2020-07-24 08:07:12 <mtneug> I have to manually run `ip link set eth0 up` and only after that `ifup eth0` can successfully run DHCP
2020-07-24 08:38:08 <EF> Hello all
2020-07-24 09:02:21 <shiroyasha> mtneug: Not sure but maybe you can add `pre-up ip link set eth0 up` to your eth0 settings in `/etc/network/interfaces`.
2020-07-24 09:07:11 <shiroyasha> But I don't know if you can do that with DHCP.
2020-07-24 09:12:02 <mtneug> shiroyasha: mmh this seems to work. thx.
2020-07-24 09:12:31 <mtneug> still wondering what change actually caused this
2020-07-24 09:19:35 <bndeter_> what package am i missing when there are no pretty icons/symbols in gui applications?
2020-07-24 09:20:19 <ikke> some icon package
2020-07-24 09:20:44 <bndeter_> just a random one? or does it need configuration?
2020-07-24 09:20:50 <bndeter_> running sway atm.
2020-07-24 09:22:29 <helby> bndeter_: probably adwaita-icon-theme
2020-07-24 09:22:35 <bndeter_> ikke: can you recommend a basic package for the average slowpoke?
2020-07-24 09:22:58 <bndeter_> helby: thanks that sounds familiar. is it default somewhere? xfce?
2020-07-24 09:23:57 <helby> bndeter_: on many distros it's even dependency of gtk+3
2020-07-24 09:24:16 <helby> bndeter_: default for gnome I think
2020-07-24 09:27:33 <bndeter_> fancy it just plugs in. no wm restart. not even application restart.
2020-07-24 09:54:31 <telmich> A  lot of prometheus tools are in Alpine, however the blackbox exporter isn't. Is this just "nobody has done it yet" or is there a specific reason why the blackbox exporter is missing?
2020-07-24 09:55:09 <ikke> telmich: hard to tell if someone tried it or not
2020-07-24 09:55:38 <telmich> I was just puzzled, because the prometheus ecosystem looks 90% complete
2020-07-24 12:39:45 <EF> Trying to set up a server for use by trusted people only but providing Nextcloud. Wireguard, xmpp server on two multiplexed ports of my choosing.
2020-07-24 12:40:09 <EF> Is Docker necessary/wise?
2020-07-24 12:40:47 <EF> Or is a there a less involved way? Any recommendations welcome
2020-07-24 12:45:51 <wsinatra> docker is probably the less involved way to do that
2020-07-24 12:46:25 <wsinatra> you could install each service on the server itself though, there's no reason you can't manually configure those services and have them resident on the same box
2020-07-24 12:49:33 <Cogitri> Yup, I just have them setup manually right now
2020-07-24 12:50:08 <Cogitri> Although I should switch to using Ansible or something in the future, setting this machine up again in case of a reinstall will be a royal pain otherwise
2020-07-24 12:51:19 <EF> So it not being a mission critical machine, bare metal should be adequate?
2020-07-24 12:55:23 <maerwald> Ansible playbooks have an appropriate name, because they're mostly just that: play stuff :p
2020-07-24 12:55:31 <abenz> you could use something like virt-manager to create VMs, if your system is recent then it most likely has hardware virt which has negligible difference on performance
2020-07-24 12:56:05 <abenz> the advantage is you can set things up, clone the VM, take snapshots etc
2020-07-24 12:56:17 <abenz> few clicks on GUI, no docker or friends needed
2020-07-24 12:56:55 <Cogitri> Hm, I'd say just bare metal with some FS which allows snapshots (e.g. BTRFS/ZFS) will be fine
2020-07-24 12:57:11 <Cogitri> virtualization is yet another thing you have to setup that might break
2020-07-24 12:57:44 <abenz> no its a safety net, that allows rapid redeployment if needed
2020-07-24 12:58:38 <abenz> but hey, this like a vi vs nano thing, you will get different opinions
2020-07-24 13:00:38 <wsinatra> RAID the important data, using some level of configuration management to ensure the state of the box
2020-07-24 13:01:03 <wsinatra> ansible is probably the most accessible there, but shell scripts would be fine too. More than one way to skin a cat
2020-07-24 13:01:08 <EF> My box is RAID10
2020-07-24 13:01:14 <wsinatra> +10 in that case
2020-07-24 13:02:12 <wsinatra> Personal take, if I was building something like this I would use some form of containerization or virtualization, to isolate the services from the host
2020-07-24 13:02:36 <wsinatra> But, it isn't necessary in the slightest if you don't want to bother learning yet another thing to throw up a nextcloud instance for your friends
2020-07-24 13:06:54 <EF> Thanks for the advice. May start bare metal then slowly migrate to Docker. It seems you can use the configs, dbs etc., stored on the host anyhow. Everything sensitive will pnly be accessible via Wireguard anyhow.
2020-07-24 13:08:38 <Cogitri> Docker probably won't make things more secure, but it sure makes deploying easier (and more reproducible)
2020-07-24 13:10:13 <wsinatra> docker'd services are essentially just nat'd hosts. So your host system is acting as a nat gateway, and routing traffic directed that way into the docker nat
2020-07-24 13:11:16 <wsinatra> which means network traffic has to ingress into your system before reaching the containers, which isn't exactly secure. Setup something like F2B to reject stuff at mangle and you've got an OK setup, but don't rely on docker as secure by design
2020-07-24 13:16:40 <EF> I'm moving from Synology so learning lots but wish to ensure the box is secure first and the additional work to learn Docker doesn't seem worth the time at the moment despite its deployment benefits.
2020-07-24 13:17:49 <wsinatra> I made the same move, though for me it was from a Synology to a RHEL8 hypervisor that got turned into a glorified NFS server/build system, and just recently rebuilt again as Alpine
2020-07-24 13:18:11 <wsinatra> Don't fret over docker, make it a goal to learn when you want to learn it's great industry knowledge
2020-07-24 13:20:02 <EF> It took me a week to get Alpine installed as raid10 doesn't work (I coundn't make it0 so have a hybrid Raid1/10 system.
2020-07-24 13:22:20 <EF> IT isn't my industry but I'm persistent
2020-07-24 13:24:58 <oneinsect> pray anyone kindly tell me why kbd-bkeymaps has xkeyboard-config as makedepends?
2020-07-24 13:25:51 <oneinsect> xkeyboard-config then pulls xkbcomp and others since it has them as makedepends and then xkbcomp pulls libx11-dev etc
2020-07-24 13:25:56 <oneinsect> ooh my
2020-07-24 13:26:12 <wsinatra> That's a good mentality to have. I actually wrote about rebuilding my nas, which is here http://lambdacreate.com/ if you're curious.
2020-07-24 13:26:27 <wsinatra> Though I didn't setup nextcloud or anything like that, just a simple Samba + raid type deal
2020-07-24 13:28:12 <oneinsect> can i avoid xkeyboard-config from kbd-bkeymaps APKBUILD
2020-07-24 13:29:01 <wsinatra> oneinsect: you could try and rebuild kbd-bkeymaps without xkeyboard-config on your glibc implementation, but that seems like extra effort
2020-07-24 13:29:17 <wsinatra> maybe a .post-install to remove the package and then see what breaks when you do so?
2020-07-24 13:29:38 <oneinsect> yeaaa
2020-07-24 13:30:00 <oneinsect> Bash, Binutils, Bison, Check, Coreutils, Flex, GCC, Gettext, Glibc, Gzip, Make, Patch, and Sed should be enough to build kbd-bkeymaps
2020-07-24 13:30:18 <oneinsect> but still trying to understand the rational behind including xkeyboard-config
2020-07-24 13:30:27 <oneinsect> wsinatra: thanks i will try to do that
2020-07-24 13:30:27 <wsinatra> Pull the APKBUILD and try and rebuild it without the xkeyboard-config
2020-07-24 13:30:31 <oneinsect> yes
2020-07-24 13:30:37 <oneinsect> i will do it
2020-07-24 13:31:01 <wsinatra> speaking from experience, sometimes I throw stuff in my packages that don't belong because it just worked, and I equated package building as a sign that whatever I did was necessary
2020-07-24 13:31:08 <wsinatra> I imagine there are others doing the same :)
2020-07-24 13:31:42 <oneinsect> hmmmm
2020-07-24 13:31:42 <EF> wsinatra: couple of typos. Code section one, is should be mdadm and code section three, "should" be >> when scanning into /etc/mdadm.conf to save the dafault file content
2020-07-24 13:31:43 <oneinsect> https://www.archlinux.org/packages/core/x86_64/kbd/
2020-07-24 13:32:41 <wsinatra> EF: thanks for pointing that out! I think the find command is off too, pretty sure that should be -exec instead of exec
2020-07-24 13:32:48 <wsinatra> need to be more diligent when I write
2020-07-24 13:34:04 <oneinsect> even gentoo doesnt list that as dependency https://gitweb.gentoo.org/repo/gentoo.git/tree/sys-apps/kbd/kbd-2.3.0-r1.ebuild
2020-07-24 13:34:06 <oneinsect> but anyway
2020-07-24 13:34:10 <oneinsect> sorry for the long rant
2020-07-24 13:34:18 <oneinsect> i will follow your advise wsinatra:
2020-07-24 13:35:34 <wsinatra> oneinsect: I'm curious to see the results, let me know what you find!
2020-07-24 13:35:47 <EF> wsinatra: Think the mdadm monitoring stuff is included in the default mdadm.conf and just requires uncommenting/modifying two lines
2020-07-24 13:37:08 <wsinatra> Really? I didn't notice anything when I set this up, but I have a tendency to over engineer things
2020-07-24 13:37:14 <oneinsect> phew! yeaa i will share soon
2020-07-24 13:37:26 <oneinsect> going package by package now to establish the core set of packages
2020-07-24 13:38:56 <oneinsect> now now
2020-07-24 13:38:57 <oneinsect> https://src.fedoraproject.org/rpms/kbd/blob/master/f/kbd.spec
2020-07-24 13:39:10 <oneinsect> fedora wants xkeyboard-config
2020-07-24 13:39:13 <oneinsect> same as alpine
2020-07-24 13:39:20 <oneinsect> strange
2020-07-24 13:39:29 <oneinsect> for building kbd
2020-07-24 13:39:34 <wsinatra> EF: right you are, since I overwrote my mdadm.conf with the --scan info I didn't realize there was configuration settings for it
2020-07-24 13:39:58 <wsinatra> I'll have to edit the post with that information, that's a much more direct way to handle things
2020-07-24 13:40:41 <wsinatra> oneinsect: if it's packaged like that upstream of RHEL, and it stays that way back into RHEL, then there's likely a reason for it
2020-07-24 13:41:17 <oneinsect> wsinatra: meaning? i mean this is fedora right?
2020-07-24 13:42:50 <wsinatra> RHEL's the defacto enterprise distro, dunno what implications it has for this specific package, but they pay a significantly large team to vet/test this stuff
2020-07-24 13:43:03 <oneinsect> hmmmm
2020-07-24 14:26:13 <mps> oneinsect: thanks for rant. we should check this
2020-07-24 14:30:43 <Cogitri> b2ed890e7f876a42229e6181268828c2493c972c
2020-07-24 14:31:07 <Cogitri> > main/kbd: generate (binary) keymaps from X.org keyboard-config
2020-07-24 14:31:16 <Cogitri> That added the dep
2020-07-24 14:42:21 <oneinsect> main/kbd: generate (binary) keymaps from X.org keyboard-config
2020-07-24 14:42:30 <oneinsect> why doesnt some other distro do that?
2020-07-24 14:42:57 <oneinsect> i mean keymaps dependency pulls in significant packages
2020-07-24 14:44:54 <Cogitri> I guess the X keymaps could be split off into another okg
2020-07-24 14:45:46 <oneinsect> Cogitri sire would you be kind enough to do the honours of splitting it
2020-07-24 14:47:16 <oneinsect> kind sire, kindest of the kind sires
2020-07-24 14:47:22 <oneinsect> :D
2020-07-24 14:58:39 <Cogitri> Ah, currently kinda busy due to GUADEC, I might be able to take a look at the weekend
2020-07-24 14:59:34 <oneinsect> thanks
2020-07-24 15:06:00 <oneinsect> https://www.happyassassin.net/posts/2013/11/23/keyboard-layouts-in-fedora-20-and-previously/
2020-07-24 15:06:16 <oneinsect> more on console kbd package and xkb package
2020-07-24 15:06:33 <oneinsect> keyboard layout maps
2020-07-24 15:12:08 <omni> Cogitri: hi, carg-zsh-completion installs to usr/share/zsh/site-functions/_cargo/_cargo instead of usr/share/zsh/site-functions/_cargo (I've mislaid my gitlab.a.o credentials, why I write here)
2020-07-24 15:12:37 <omni> *cargo-zsh-completion
2020-07-24 21:04:40 <amom> is there any program I can use on alpine linux to check services/programs that need to be restarted
2020-07-24 21:04:53 <amom> similar to `checkrestart` on ubuntu
2020-07-24 21:05:37 <amom> http://manpages.ubuntu.com/manpages/trusty/man1/checkrestart.1.html
2020-07-24 21:08:01 <ikke> not as polished, but I use this: 'lsof | grep -e DEL | grep lib | sort -u -k1,1' (requires full lsof, not bb lsof)
2020-07-24 21:16:33 <Hello71> that doesn't include if the program itself is updated
2020-07-24 21:16:39 <Hello71> that's (deleted)
2020-07-24 21:27:14 <ikke> Filter on lib and bin then
2020-07-24 21:27:21 <ikke> though
2020-07-24 21:27:36 <ikke> not sure if replaced files are considered deleted
2020-07-24 21:33:16 <Hello71> that's the point of install(1)
2020-07-25 07:52:52 <Unode> hi everyone. I'm trying to install py3-biopython. The package is listed in https://pkgs.alpinelinux.org/packages?name=*biopython*&branch=edge but apk tells me it's not available.
2020-07-25 07:53:26 <Unode> is there something I'm missing to be able to install it?
2020-07-25 07:53:34 <Unode> py3-scikit-learn is in the same situation
2020-07-25 07:56:49 <MY-R> Unode, you need to enable testing repository in /etc/apk/repository
2020-07-25 07:57:02 <Unode> MY-R: ah, thanks
2020-07-25 07:57:08 <MY-R> /etc/apk/repositories
2020-07-25 07:57:19 <Unode> I'm already using alpine:edge in docker so I was confused on that.
2020-07-25 08:00:00 <Unode> MY-R:  thanks all working now.
2020-07-25 08:00:18 <MY-R> ye can imagine, if packages in testing are stable enough and well maintained then can create pull request or issue on gitlab to ask maintainer if can move it to "community" repo
2020-07-25 08:00:55 <MY-R> Unode, you are welcome :)
2020-07-25 09:55:42 <albr> unowe[d]: hi all, what provides `man` now?  the FAQ says to add `man`, which seemingly does not exist anymore (https://pkgs.alpinelinux.org/packages?name=man&branch=edge&arch=x86_64) `ERROR: unsatisfiable constraints: man (missing): required by: world[man]`
2020-07-25 09:56:33 <albr> unowe[d]: I have the edge and 3.12 repos available on my system as well, neither have `man`
2020-07-25 09:58:06 <ikke> https://pkgs.alpinelinux.org/contents?file=man&path=%2Fusr%2Fbin&name=&branch=edge&arch=x86_64
2020-07-25 09:58:20 <ikke> mandoc or man-db
2020-07-25 09:58:49 <mps> mandoc is better, we should remove man-db
2020-07-25 10:03:29 <albr> unowe[d]: ah, thanks. didn't realise it moved to `mandoc`
2020-07-25 10:42:55 <gjabell> I have two drives with luks keyfiles that should be encrypted _after_ the system is booted (since the boot drive is decrypted with a password on boot)
2020-07-25 10:43:14 <gjabell> are there any additional steps other than adding those drives as crypttab entries?
2020-07-25 10:43:42 <gjabell> so far they aren't being decrypted automatically, not sure if it's because the system is trying to decrypt them before the base system has been mounted, since the keyfile is in /etc
2020-07-25 10:43:52 <Latrina> Guys I have been wondering. Was Alpinelinux originally forked from gentoo? Or am I completely wrong?
2020-07-25 10:44:15 <Arahael> What makes you think that?
2020-07-25 10:44:40 <Latrina> I seem to remember someone told me it awhile ago, but than I look at it and uses busybox
2020-07-25 10:44:52 <Latrina> so I can't find much correlation between the two outside of OpenRC
2020-07-25 10:44:56 <ikke> No, it's not a fork
2020-07-25 10:45:22 <Latrina> I mean I know it isn't anymore. But was it ever?
2020-07-25 10:45:33 <ikke> https://web.archive.org/web/20160514053441/https://osdir.com/ml/linux.leaf.devel/2005-08/msg00039.html
2020-07-25 10:45:40 <ikke> That was the initial anouncement
2020-07-25 10:46:01 <Latrina> thanks
2020-07-25 10:46:25 <Arahael> Latrina: I'm just curious what made you think it was, or ever was. :)  There's usually a reason. ;)
2020-07-25 10:46:35 <Latrina> btw, I have been using alpine for the past 4 years, but for some reason I always kept on wondering about this
2020-07-25 10:46:50 <Latrina> Arahael, OpenRC I think...
2020-07-25 10:47:05 <Arahael> Latrina: The init system?
2020-07-25 10:48:29 <Latrina> yes
2020-07-25 10:48:37 <Latrina> that I think is what fooled me
2020-07-25 10:48:52 <MY-R> gjabell, add "dmcrypt" to start at boot time (rc-update add dmcrypt boot) and check /etc/conf.d/dmcrypt basically need 3 lines, target=data, source='UUID=12345-1235-etc', key='/yourkeyfilepath'
2020-07-25 10:49:36 <Arahael> Latrina: Ah, fair enough.  Linux distributions in general do copy from each other quite a lot.  If you look at the confluence page for OpenRC, you'll note that gentoo and alpine aren't the only systems that use it.
2020-07-25 10:51:51 <Latrina> true that
2020-07-25 10:54:38 <mps> yes, I remember leaf on floppy :)
2020-07-25 10:55:08 <gjabell> MY-R thanks, that worked perfectly!
2020-07-25 10:55:20 <mps> I tested it after it was 'released' for some time but stopped and forgot
2020-07-25 10:56:06 <mps> and now I use it exclusively, how the life is unpredictable :)
2020-07-25 10:56:33 <MY-R> gjabell, good job, you are welcome :)
2020-07-25 11:00:22 <MY-R> gjabell, and dont need crypttab file (not even sure if anything read it) just line in fstab
2020-07-25 11:00:41 <gjabell> hmm ok
2020-07-25 11:00:58 <gjabell> does the encrypted boot drive refer to crypttab?
2020-07-25 11:02:01 <MY-R> nope
2020-07-25 11:02:29 <gjabell> ok, guess it's just extra in the wiki haha
2020-07-25 11:02:51 <MY-R> in different distributions that file is put in initrd, Alpine not using it
2020-07-25 11:03:05 <gjabell> ah gotcha
2020-07-25 11:03:41 <gjabell> another question since you might know; what's the proper way to autoimport zfs pools after boot?
2020-07-25 11:03:55 <gjabell> in this case I have the two luks volumes mirror in a zfs pool but not as the boot device
2020-07-25 11:04:51 <MY-R> I have no idea, my zfs knowledge is equal zero in Linuxes :)
2020-07-25 11:05:02 <gjabell> no worries, thanks anyways :)
2020-07-25 11:05:18 <MY-R> np
2020-07-25 11:06:46 <MY-R> gjabell, and remember to use disks/partitions UUIDs as much possible
2020-07-25 11:09:24 <gjabell> yeah think I have UUIDs set everywhere I can set them
2020-07-25 11:09:32 <gjabell> don't want to get bitten from that again :P
2020-07-25 11:10:19 <MY-R> heh :)
2020-07-25 11:35:18 <rah> specing: is your Ada steam-a-like up on the net anywhere
2020-07-25 11:35:25 <rah> (continuing an old conversation from #freegamer)
2020-07-25 11:35:30 <rah> ?
2020-07-25 11:35:36 <specing> rah: nope
2020-07-25 11:36:27 <specing> it's an unfinished gui mockup in GTK Glade + minimal Ada to load that gui, and it exists on my disks only
2020-07-25 11:43:47 <fyr> is nfs been messed up for anyone else on edge?
2020-07-25 11:44:10 <fyr> since I upgraded yesterday (and not for awhile) clients just hang on mount
2020-07-25 11:45:58 <rah> specing: ok
2020-07-25 13:28:29 <feuerteufel> Hi @ all, I'm looking for procmail. Can't find it, am I to blind?
2020-07-25 13:29:01 <mps> feuerteufel: I think it is removed some time ago
2020-07-25 13:29:06 <solar> procmail hasn't been updated in 20 years
2020-07-25 13:37:15 <feuerteufel> I know, but it works great ...
2020-07-25 13:41:05 <feuerteufel> But anyway, Thx
2020-07-25 14:02:19 <mps> maybe it could be added from some distro where it is maintained
2020-07-25 14:02:35 <mps> people ask for it from time to time
2020-07-25 14:05:29 <solar> the latest patch i see floating around is in the linux from scratch patch set, it appears to just be an aggregation of patches from debian and fedora bug trackers though
2020-07-25 14:05:31 <feuerteufel> Also I can't fint an alternative
2020-07-25 14:05:36 <solar> ~2018
2020-07-25 14:05:46 <solar> only 7k of changes
2020-07-25 14:05:59 <solar> they are CVE fixes though
2020-07-25 14:30:09 <feuerteufel> It still was in 3.11
2020-07-25 14:33:09 <feuerteufel> Is there a way to install it from 3.11 into my 3.12 installation?
2020-07-25 14:34:15 <ikke> yes, but it might not work
2020-07-25 14:38:39 <mps> feuerteufel: I think it will work
2020-07-25 14:38:49 <mps> only deps is musl
2020-07-25 14:39:18 <feuerteufel> Yes, that's what I yust found
2020-07-25 14:41:06 <feuerteufel> So yust open up the 3.11 entry in the /etc/apk/ repositories?
2020-07-25 14:41:23 <ikke> You want it pinned
2020-07-25 14:41:59 <mps> I would rather download apk from alpine and install it from apk, and not mess with pinning
2020-07-25 14:54:41 <feuerteufel> Is there a special directory where the "procmail-3.22-r4.apk" has to be?
2020-07-25 14:56:19 <mps> no
2020-07-25 14:56:53 <mps> apk add procmail-3.22-r4.apk, that's all, whenever you put it
2020-07-25 14:59:19 <feuerteufel> Yea, thx! got it!
2020-07-25 15:01:18 <mps> does it work?
2020-07-25 15:02:07 <feuerteufel> At least "procmail -v" work's
2020-07-25 15:02:43 <feuerteufel> Now I have to configure und install the rest I need
2020-07-25 15:03:38 <mps> nice, maybe you should fill bug/pkg request on gitlab.a.o
2020-07-25 15:04:03 <feuerteufel> solar is right:
2020-07-25 15:04:16 <feuerteufel> procmail v3.23pre 2001/09/13
2020-07-25 15:04:21 <feuerteufel> ;)
2020-07-25 15:04:38 <solar> ok, almost 20 years :D not quite
2020-07-25 15:05:02 <mps> yes, it is old, but if it works and don't have sec issues I don't see why it can't be packaged
2020-07-25 15:05:30 <mps> especially if users find it useful
2020-07-25 15:06:42 <feuerteufel> hehe, I'm over 60 years old and still working :-))
2020-07-25 15:08:06 <solar> the known CVEs have patches available, i would just worry about any future bugs going unsolved, not having an active maintainer
2020-07-25 15:08:50 <mps> feuerteufel: heh, also I'm :)
2020-07-25 15:11:40 <feuerteufel> solar: But everywhere you looking some stuff about filtering mails (wikis, howtos) it say's " use procmail ... :-))
2020-07-25 15:12:35 <solar> it would be good if somebody took up maintaining it. maybe modernizing the code :)
2020-07-25 16:09:55 <feuerteufel> One more today ... I'm trying tu install a Printer like it is descriped here:
2020-07-25 16:10:10 <feuerteufel> https://wiki.alpinelinux.org/wiki/Printer_Setup
2020-07-25 16:11:05 <feuerteufel> I'm not able to install cups-pdf ...
2020-07-25 16:12:39 <feuerteufel> The Wiki say's to use the one out of testing, bzut I can't get zhis to work, any hint's?
2020-07-25 16:13:14 <feuerteufel> WARNING: The repository tag for world dependency 'cups-pdf@testing' does not exist
2020-07-25 16:13:14 <feuerteufel> apk add cups-pdf@testing
2020-07-25 16:13:14 <feuerteufel> ERROR: Not committing changes due to missing repository tags. Use --force-broken-world to override.
2020-07-25 16:15:38 <feuerteufel> .
2020-07-25 16:16:05 <ikke> Did you add a pinned repo to /etc/apk/repositories?
2020-07-25 16:16:14 <ikke> @test http://...
2020-07-25 16:16:35 <feuerteufel> No
2020-07-25 16:17:01 <ikke> @testing http://dl-cdn.alpinelinux.org/alpine/edge/testing
2020-07-25 16:17:17 <feuerteufel> I'll try
2020-07-25 16:20:28 <feuerteufel> Thx, it works
2020-07-25 17:58:24 <mps> just had kernel problem on one cloud provider of xen VM, using x86_64 linux-virt alpine stable
2020-07-25 17:59:00 <mps> solved with linux-virt from edge
2020-07-25 17:59:53 <mps> didn't had time to look deeply at problem because server needs to work
2020-07-25 18:00:40 <mps> writing this here if someone have/could have same problem
2020-07-25 19:57:21 <gjabell> anyone know why the amdgpu drivers don't show up in the linux-virt modules?
2020-07-25 19:57:39 <gjabell> do I need to switch the VM to lts?
2020-07-25 20:03:11 <mps> linux-virt is minimal for VM, there is no much drivers
2020-07-25 20:03:54 <mps> gjabell: and yes, you can use linux-lts in VM if you need some drivers or features from it
2020-07-25 20:04:19 <gjabell> ah that seemed to fix it, thanks :)
2020-07-25 22:05:55 <gjabell> hmm, rsync seems to be broken atm
2020-07-25 22:06:03 <gjabell> does anyone else get an error trying to rsync anything?
2020-07-25 22:15:27 <Lohmar> hello
2020-07-25 22:15:41 <Lohmar> anyone care to help a lost fellow ?
2020-07-25 22:16:04 <Lohmar> I'm trying to install virtualbox guest addition from the iso
2020-07-25 22:16:17 <Lohmar> (the one from the repos is too old
2020-07-25 22:16:21 <Lohmar> any hints ?
2020-07-26 00:31:39 <winny> is it possible to replace busybox syslogd with rsyslog?
2020-07-26 00:35:58 <c705> winny: yes
2020-07-26 01:04:59 <nwm> does anyone remember the name of that small templating engine like gomplate except not 38MB?
2020-07-26 07:21:32 <mouldysammich> hey, a patch i sent is failing on x86 and armv7 with the same error, i dont understand it though would anyone mind taking a look? https://gitlab.alpinelinux.org/amcgrath/aports/-/jobs/172089
2020-07-26 07:24:06 <azarus_> I take it that if you have very new hardware only supported by, say linux 5.7, the best course of action is compiling linux yourself?
2020-07-26 07:30:58 <Cogitri> mouldysammich: Edge is currently being rebuilt for musl 1.2, so the repos are a bit incomplete on 32-bit arches, sorry for the inconvenience
2020-07-26 07:31:24 <ikke> azarus_: we have linux-edge in the testing repo, which is 5.7
2020-07-26 07:33:23 <azarus_> ikke: oh, very nice. thanks!
2020-07-26 07:35:08 <mouldysammich> Cogitri: Ah cheers!
2020-07-26 11:51:21 <dott> Hi, i'm ne to alpine. Is it possible that there is a bug on x86 edge right now? I got a system with cryptsetup and lvm and it does not boot correctly after upgrading to edge. root get mounteted as ro and run does not get mountet because superuser privileges are missing.
2020-07-26 11:51:34 <dott> downgrading to main fixed everything
2020-07-26 11:51:44 <afontain_> quite possible
2020-07-26 11:52:19 <afontain_> Alpine just switched to musl-1.2 which caused breakages on 32 bits arch
2020-07-26 11:52:22 <ikke> We upgraded edge musl to 1.2, which requires us to rebuild all packages, (and fix any breakage)
2020-07-26 11:53:42 <dott> ah ok, so i'll stay away from edge right now. Thanks :)
2020-07-26 14:47:09 <winny> hmm, libvirtd service seems to never stop
2020-07-26 14:48:43 <Cogitri> https://gitlab.alpinelinux.org/alpine/aports/-/issues?scope=all&utf8=%E2%9C%93&state=opened&search=libvirt
2020-07-26 14:50:38 <winny> ah nice
2020-07-26 14:56:22 <winny> upstream sure is not nice about bugs in their code
2020-07-26 14:56:40 <winny> https://gitlab.com/libvirt/libvirt/-/issues/52
2020-07-26 15:07:05 <Cogitri> :/
2020-07-26 15:31:11 <eydaimon> how up to date are packages typically in alpine ?
2020-07-26 15:31:37 <ikke> edge is reasonably up-to-date
2020-07-26 15:31:57 <eydaimon> is there a reason not to use edge ?
2020-07-26 15:32:11 <ikke> it can be unstable from time to time
2020-07-26 15:32:54 <eydaimon> how far behind would stable be? weeks ? days ? depends totally on the package ?
2020-07-26 15:33:15 <eydaimon> I notice nnn is behind by two weeks, but it's not really a prominent program
2020-07-26 15:33:27 <ikke> well, stable will only receive bugfixes / security fixes
2020-07-26 15:34:11 <eydaimon> in what way does edge mainest as being unstable ?
2020-07-26 15:35:22 <ikke> from bugs in the newest versions to temporary library mismatches
2020-07-26 15:35:49 <ikke> right now, we've upgraded to musl-1.2, which affects all 32-bits arches in the sense that they need to be rebuilt and bugs fixed
2020-07-26 15:36:10 <ikke> so lots of packages might be missing still
2020-07-26 15:40:16 <eydaimon> thanks for answering :)
2020-07-26 15:43:22 <eydaimon> ikke: are there any goals to get packages updated faster in stable? like macos does a pretty good job of keeping things I use up to date?  ANd finally, how can I help ?
2020-07-26 15:43:51 <ikke> it's a matter of policy
2020-07-26 15:44:36 <Cogitri> Stable is stable because it doesn't get (feature) updates
2020-07-26 15:44:53 <Cogitri> bug fix updates are backported of course, but not much else
2020-07-26 15:46:08 <eydaimon> hm, but userland programs wouldn't make the OS unstable. Like keeping a program like nnn wouldn't mess with the stability of th eOS
2020-07-26 15:46:37 <eydaimon> but there is no distinction there ?
2020-07-26 15:46:56 <ikke> no
2020-07-26 15:47:01 <eydaimon> ok :)
2020-07-26 15:47:18 <Cogitri> Yes, but userland programs suddenly crashing on stable isn't what stable is for
2020-07-26 15:48:18 <eydaimon> not judging, just wanting to understand the distro's policy and reasoning
2020-07-26 15:48:31 <eydaimon> this is for my NAS so stable is good for my use-case
2020-07-26 15:50:07 <ikke> If you really need the latest version of a certain program, it's relatively easy enough to build that yourself against a stable release
2020-07-26 15:50:12 <Hello71> what is "the OS"
2020-07-26 15:50:15 <Hello71> how do you define it
2020-07-26 15:50:22 <Hello71> linux is a bazaar
2020-07-26 17:07:44 <omni> eydaimon: I run stable but include a few non-critical packages from @edge and @testing through repository pinning https://wiki.alpinelinux.org/wiki/Alpine_Linux_package_management#Repository_pinning
2020-07-26 17:08:05 <omni> so it is possible to mix, but you may be on your own if you break things this way
2020-07-26 17:08:07 <eydaimon> oo, thank you
2020-07-26 17:08:13 <eydaimon> I have just one package from edge I think
2020-07-26 17:08:43 <eydaimon> but this is a useful doc
2020-07-26 17:09:06 <eydaimon> very cool
2020-07-26 17:10:44 <Cogitri> Keep in mind that mixing repos is unsupported and will break at some point (when we do incompatible changes in edge vs your release, e.g. a major version bump in one of the libs)
2020-07-26 17:12:08 <eydaimon> Cogitri: right, but in this case it looks package specific. So for example, I'm using `most` from edge since it's not even available elsewhere. At least then I can have it for now
2020-07-26 17:12:30 <eydaimon> Cogitri: and I would only do it for a few packages I really need to be up to date
2020-07-26 17:12:36 <eydaimon> which I think should be pretty safe
2020-07-26 17:13:45 <Cogitri> Sure, just be careful
2020-07-26 17:14:01 <eydaimon> aye aye capt'n :)
2020-07-26 17:19:53 <eydaimon> doesn't work exactly like how i imagined. I was thinking I could tag specific apps to a repo. But I ran apk list -u, and it showed updates for other apps too
2020-07-26 17:22:02 <eydaimon> which means that I don't really understand the pinning :/
2020-07-26 17:23:14 <eydaimon> oh, I had left edge on by accident in the past ... I wonder if that made a difference
2020-07-26 17:28:01 <eydaimon> oh but it doesn't upgrade
2020-07-26 17:29:40 <eydaimon> although list -u doesn't seem to distinguish is all. would be useful if list -u included the mask so it was possible to see
2020-07-27 10:18:50 <komnpomat> mm.. no soulseek??!!!
2020-07-27 11:13:16 <[twisti]> is ... is alpine unable to use ping ? i find it hard to believe, but "docker run --rm -it alpine ping www.google.de" says "ping: bad address 'www.google.de'". what gives ?
2020-07-27 11:13:48 <solar> sounds like dns isn't working inside your ct
2020-07-27 11:14:00 <solar> or network connectivity
2020-07-27 11:14:07 <solar> have you checked yet?
2020-07-27 11:16:30 <[twisti]> i dont know what a ct is. i can ping from the host perfectly fine
2020-07-27 11:17:13 <solar> container, the docker guest
2020-07-27 11:19:05 <[twisti]> ah, yeah. i agree, it looks that way. but my guest is just a bare alpine
2020-07-27 11:19:35 <solar> can you ping an ip address?
2020-07-27 11:20:28 <[twisti]> yes
2020-07-27 11:20:41 <ikke> [twisti]: https://tpaste.us/0Pg6
2020-07-27 11:21:21 <solar> so presumably your alpine guest has misconfigured or unconfigured dns
2020-07-27 11:21:32 <ikke> more like your host
2020-07-27 11:21:32 <[twisti]> ikke: could you share (if it contains no secret) the output of "docker run --rm -it alpine cat /etc/resolv.conf" ? theres an issue that mentions a relationship
2020-07-27 11:21:44 <solar> ikke does docker just pass thru dns to the host ?
2020-07-27 11:21:49 <solar> i thought it had a bridged virtual network
2020-07-27 11:22:27 <ikke> [twisti]: nameserver 8.8.8.8\n nameserver 8.8.4.4
2020-07-27 11:22:44 <[twisti]> ikke: same here :|
2020-07-27 11:22:45 <ikke> solar: docker reuses the dns from the host
2020-07-27 11:23:04 <ikke> you can also override it through parameters to the docker daemon
2020-07-27 11:23:19 <ikke> [twisti]: can you ping 8.8.8.8/8.8.4.4 ?
2020-07-27 11:23:21 <solar> but you could still replace resolv.conf in the ct to override that, it's just writing the file out at ct  start yes ?
2020-07-27 11:23:31 <[twisti]> nslookup shows some interesting additional info: docker run --rm -it alpine nslookup www.google.de -> nslookup: write to '8.8.8.8': Host is unreachable
2020-07-27 11:23:46 <ikke> solar: resolv.conf is read-only
2020-07-27 11:23:50 <[twisti]> yes, i can ping 8.8.8.8 fine
2020-07-27 11:24:27 <ikke> n/m that. you can modify resolv.conf
2020-07-27 11:27:23 <ikke> [twisti]: seems some kind of network issue.
2020-07-27 12:44:28 <abenz> hey guys
2020-07-27 12:44:51 <abenz> I keep getting "filesystem unmounted properly...etc" on my alpine usb
2020-07-27 12:45:00 <abenz> do I need to do something special? diskless mode btw
2020-07-27 12:45:18 <abenz> I simply issue reboot and unplug the usb before the system reboots into its other OS
2020-07-27 12:45:51 <solar> but after alpine has already shut down?
2020-07-27 12:47:49 <abenz> I see openrc logs flashing by quickly with a couple red ones (I think those are the fs unmounts), its at this point (when I see the system cycling) I unplug the usb
2020-07-27 12:47:58 <abenz> so I don't unplug it then issue reboot
2020-07-27 12:49:36 <Cogitri> Do you have the mount-ro service enabled?
2020-07-27 12:50:01 <Cogitri> Maybe it doesn't remount the rootfs as ro before shutting down and as such the fs gets dirtt
2020-07-27 12:50:08 <Cogitri> s/tt/tt/
2020-07-27 12:50:14 <Cogitri> ty* I can't type :D
2020-07-27 12:50:34 <ws2k3> is it correct that postfix does not some with qshape on alpine linux?
2020-07-27 12:51:40 <abenz> Cogitri: it says status: stopped
2020-07-27 12:52:01 <Cogitri> What does rc-update say?
2020-07-27 12:52:04 <abenz> I'm using plain diskless mode, I wasnt aware I needed to make any tweaks/modifications
2020-07-27 12:52:25 <Cogitri> Well, I'm just guessing here, I don't use diskless on any of my machines :)
2020-07-27 12:52:38 <abenz> mount-ro ------------- shutdown
2020-07-27 12:52:45 <Cogitri> If that does turn out to be the error it's a bug we should fix
2020-07-27 12:52:49 <Cogitri> ah hm 🤔
2020-07-27 12:53:09 <Cogitri> so it is run during shutdown, not sure how the FS ends up getting dirty then
2020-07-27 12:53:57 <abenz> I see
2020-07-27 12:54:14 <Cogitri> ws2k3: Dunno, please make a bug report at https://gitlab.alpinelinux.org/alpine/aports/issues/new and go into more detail, then we can discuss it with the postfix maintainer
2020-07-27 12:54:25 <abenz> would an abrupt shutdown weeks ago (eg due to system getting stuck) cause the fs to be dirty?
2020-07-27 12:54:49 <abenz> I mean would it persist all this time and I need to run fsck on the partition on a different system (since I cant do it on a mounted system)
2020-07-27 12:55:40 <ikke> ws2k3: seems not
2020-07-27 12:55:48 <ikke> https://pkgs.alpinelinux.org/contents?file=qshape&path=&name=postfix&branch=edge&repo=main&arch=x86_64
2020-07-27 12:58:29 <Cogitri> abenz: Ah yes, you'd have to run fsck once to make sure the FS is OK again
2020-07-27 12:59:44 <ws2k3> i see ikke i will create a gitlab issue about it
2020-07-27 14:57:24 <skyroveRR> Heya guys, I'm building rust for my own distro based on alpine's rust build recipe https://git.alpinelinux.org/aports/tree/community/rust/APKBUILD however I'm getting this error: http://pktsurf.in/p/rust.error.txt . Prior to the build I created the common glibc-based symlinks also. Any ideas?
2020-07-27 15:00:42 <Cogitri> I guess the tarballs you downloaded for bootstrapping are built against glibc instead of musl
2020-07-27 15:01:12 <skyroveRR> Cogitri: so the symlinking will not work?
2020-07-27 15:01:33 <Cogitri> Well no, symlinking musl to what glibc supports doesn't make it glibc
2020-07-27 15:01:43 <skyroveRR> I googled the same errors, and it seems they fixed that for the x86_64 builds.
2020-07-27 15:04:31 <ikke> fixed what exactly?
2020-07-27 15:06:59 <skyroveRR> ikke: this one: identical error, https://stackoverflow.com/questions/58766740/how-can-i-target-my-rust-application-and-its-dependencies-to-musl and this: https://github.com/rust-lang/rust/pull/58575
2020-07-27 15:08:55 <Cogitri> You're using the -gnu tarballs linked against glibc though
2020-07-27 15:09:06 <Cogitri> You have to use the -musl ones (or Alpine's tarballs)
2020-07-27 15:18:03 <skyroveRR> Cogitri: ok, I attempted to fix this issue by correcting the configure options, and it appears to be downloading the rust-std tarball correctly at least, but fails to download rustc and cargo musl tarballs because they don't exist on the server: http://pktsurf.in/p/rust.error.1.txt and here's their directory: https://static.rust-lang.org/dist/2020-05-07 ... wonder how alpine is building for aarch64 in
2020-07-27 15:18:06 <skyroveRR> this case.
2020-07-27 15:19:51 <Cogitri> We have the rust-bootstrap package
2020-07-27 15:20:04 <Cogitri> That practically is just the rust package itself, so we bootstrap the rust package from the rust package
2020-07-27 15:20:44 <skyroveRR> Can you be kind enough to provide a link to it? I can't find it anywhere in git.alpinelinux.org.
2020-07-27 15:21:14 <Cogitri> It's just in the rust APKBUILD via `provides="rust-bootstrap"`
2020-07-27 15:22:07 <skyroveRR> Oh.
2020-07-27 15:22:26 <Cogitri> See the Rust apks from https://alpine.global.ssl.fastly.net/alpine/edge/community/aarch64/ for what packages we bootstrap from
2020-07-27 15:23:53 <ikke> Cogitri: fyi: https://dl-cdn.alpinelinux.org now works
2020-07-27 15:24:09 <ikke> no need to use that fasly domain name anymore
2020-07-27 15:24:58 <Cogitri> Oh, I just grabbed that from mirrors.alpinelinux.org
2020-07-27 15:24:59 <Cogitri> I guess the URL should be updated there
2020-07-27 15:25:08 <Cogitri> Nice that it works now :)
2020-07-27 15:26:40 <ikke> fastly supports custom domains now via lets encrypt
2020-07-27 15:26:43 <ikke> so that's nice
2020-07-27 15:26:54 <ikke> (not sure how long they support it, may have been a while already)
2020-07-27 15:35:46 <skyroveRR> Cogitri: I can't seem to figure out how to do this bootstrapping thing. Is there a separate build recipe for this? Cos I'm running almost the exact same commands as your build recipes, and this thing is getting over my head.
2020-07-27 15:36:47 <skyroveRR> Cogitri: this is my build script: http://pktsurf.in/p/rust.SMBuild.txt
2020-07-27 15:37:21 <Cogitri> Everything we do is contained in the rust APKBUILD
2020-07-27 15:37:37 <Cogitri> We just use the system's rust package, we don't bootstrap from tarballs
2020-07-27 15:37:49 <Cogitri> Sorry, I don't have much time to help due to GUADEC right now
2020-07-27 15:39:03 <ikke> skyroveRR: The initial bootstrap was done a long time ago. Nowadays we bootstrap by using an existing rust compiler package from the alpine repositories
2020-07-27 15:40:17 <skyroveRR> ikke: oh, so the starting point will be grabbing the rust-1.44.1-r0.apk from https://alpine.global.ssl.fastly.net/alpine/edge/community/aarch64/ ? And then.... ?
2020-07-27 15:40:32 <ikke> you need to install that somehow
2020-07-27 15:41:02 <ikke> if you don't use apk, you probably need to untar it to /
2020-07-27 15:41:05 <Cogitri> 6c8381ba967cd3842484965f3f7c1200a42a5539 is how I bootstrapped Rust on Alpine
2020-07-27 15:41:17 <ikke> 
ACTION pokes algitbot

2020-07-27 15:42:04 <Cogitri> I think algitbot has been disabled in #alpine-linux a few months ago
2020-07-27 15:42:15 <ikke> oh
2020-07-27 15:42:32 <ikke> clandmeter: ^ on purpose?
2020-07-27 15:44:36 <gunix> hello
2020-07-27 15:45:06 <ikke> skyroveRR: https://gitlab.alpinelinux.org/alpine/aports/-/commit/6c8381ba967cd3842484965f3f7c1200a42a5539
2020-07-27 15:45:29 <gunix> i am confused. When using a go binary to make curl requests (by actually using "net/http"), the binary seems to find CA certificates within alpine:edge. However this happens even if "ca-certificates" are not installed. I see installing curl installs ca-certificates.
2020-07-27 15:45:39 <gunix> are there any root certificates shipped by default with alpine:edge?
2020-07-27 15:50:02 <ikke> /etc/ssl/cert.pem, which is owned by ca-certificates-cacert
2020-07-27 15:50:47 <ikke> Oh, that's old :/
2020-07-27 15:50:58 <gunix> ikke: i was about to ask about `/etc/ssl/cert.pem`
2020-07-27 15:51:12 <gunix> it works, though
2020-07-27 15:51:35 <ikke> yeah, that file works, just not the package that owns it
2020-07-27 15:52:00 <ikke> in edge it's called ca-certificates-bundle
2020-07-27 15:53:36 <gunix> ikke: so what does `ca-certificates` add?
2020-07-27 15:54:41 <gunix> bash-5.0# apk list -I | grep -i cert
2020-07-27 15:54:44 <gunix> ca-certificates-cacert-20190108-r0 x86_64 {ca-certificates} (MPL-2.0 GPL-2.0-or-later) [installed]
2020-07-27 15:54:46 <gunix> i see now
2020-07-27 15:54:47 <gunix> this comes by default
2020-07-27 15:55:20 <skyroveRR> ikke: alright. I'll have a look. Thanks for the assistance thus far.
2020-07-27 15:59:05 <gunix> ikke: how did you find the package that owns the file?
2020-07-27 16:03:29 <ikke> apk info -W /etc/ssl/certs/cert.pem
2020-07-27 16:03:43 <ikke> ca-certificates.pem
2020-07-27 16:24:46 <clandmeter> ikke: nope
2020-07-27 16:25:18 <clandmeter> algitbot: ping
2020-07-27 16:26:03 <clandmeter> Don't know
2020-07-27 16:26:09 <clandmeter> Maybe holidays
2020-07-27 16:26:36 <clandmeter> The script is not enabled I guess for this channel
2020-07-27 16:26:54 <clandmeter> You can symlink it
2020-07-27 17:02:25 <gunix> ikke: thank you. `apk info -W` will help me in many situations
2020-07-27 17:04:10 <abenz> Cogitri: I see. will do that, thanks
2020-07-27 17:50:25 <wsinatra> anyone know if openrc has a way to handle running init scripts on user login?
2020-07-27 17:50:36 <Cogitri> It doesn't
2020-07-27 17:51:19 <wsinatra> didn't think so based on my failure to find anything indicating otherwise, but it was worth asking :)
2020-07-27 17:51:23 <ikke> it's a system init service
2020-07-27 17:51:35 <wsinatra> probably best way to handle that stuff is profile?
2020-07-27 17:51:38 <Cogitri> Would be super great to have something like systemd-user though
2020-07-27 17:51:51 <Cogitri> It's certainly by far not the best thing, but the easiest
2020-07-27 17:52:13 <Cogitri> The best would be having a proper service manager, you could use something like runit or s6-rc, but that'll be rather annoying to setup
2020-07-27 17:52:36 <wsinatra> that's sounds painful, I like my openrc :(
2020-07-27 17:53:42 <wsinatra> might have to just suck it up and go with it. My use case is to run a modified version of anbox-container-manager which points to different user android images
2020-07-27 18:57:36 <afontain_> <wsinatra "might have to just suck it up an"> the Anbox container manager is a system service, run as root
2020-07-27 18:57:45 <afontain_> well, on Alpine at least
2020-07-27 18:58:13 <afontain_> sudoedit /etc/init.d/anbox-container-manager
2020-07-27 22:08:58 <mike18> hi how does it select the arch here? https://hub.docker.com/_/alpine
2020-07-27 22:09:24 <mike18> -> Supported architectures: (more info) amd64, arm32v6, arm32v7, arm64v8, i386, ppc64le, s390x
2020-07-27 22:26:16 <fireside> Hello
2020-07-27 22:26:37 <fireside> which build do you recommend for ? RK3328 (NanoPi R2S)
2020-07-28 00:06:41 <picmyk> ,
2020-07-28 05:00:37 <PhoenixMage> Anyone used the edge cloud-init package?
2020-07-28 06:06:51 <PhoenixMage> What does stack a runlevel mean?
2020-07-28 06:12:56 <PhoenixMage> Looking at the rc-update help
2020-07-28 06:14:37 <mps> 'runlevel stack'?
2020-07-28 06:16:21 <PhoenixMage> Is like a grouping?
2020-07-28 06:16:58 <PhoenixMage> I have never come across the term before
2020-07-28 06:52:24 <wener[m]> Hi, I want to use setup-disk to install on an root dir, but don't want to copy the host's apk world
2020-07-28 06:52:54 <wener[m]> https://github.com/alpinelinux/alpine-conf/blob/master/setup-disk.in I don't find where host's world was passed to the rootfs
2020-07-28 06:53:07 <wener[m]> Is this possible ?
2020-07-28 07:38:08 <wener[m]> Figured out, this is caused by lbu package, use miniroot as apkvol solved the problem
2020-07-28 08:30:33 <mtneug> PhoenixMage: I use the cloud-init package in an OpenStack cloud environment
2020-07-28 08:32:07 <mtneug> though network configuration is deactivated
2020-07-28 08:32:28 <mtneug> I can't recall the problem, but my VM image is configured to use DHCP anyway so I didn't need cloud-init to manage networking
2020-07-28 09:08:43 <uaa> hello, I've searched trying to reset root password but didn't find a solution. how should I do it on Alpine? I've booted into single mode but what ever I do there wouldn't work. everything mounted rw
2020-07-28 09:11:54 <PhoenixMage> mtneug: Did you set all the init scripts to run at boot level?
2020-07-28 09:12:48 <mtneug> yes
2020-07-28 09:12:50 <PhoenixMage> uaa: Single user still asks for the root password doesnt it?
2020-07-28 09:13:06 <uaa> no
2020-07-28 09:13:25 <ikke> What doesn't work?
2020-07-28 09:13:38 <PhoenixMage> mtneug: Cool thanks, thats what I was trying to find out since I couldnt find any documentation
2020-07-28 09:13:51 <ikke> I typically use init=/bin/sh
2020-07-28 09:13:52 <mtneug> PhoenixMage: you may need to install cloud-utils-growpart separately
2020-07-28 09:14:46 <PhoenixMage> mtneug: I am using f2fs as its a raspberry pi and it doesnt support resizing through cloud-init that I could see
2020-07-28 09:15:52 <PhoenixMage> Thanks for the heads up though
2020-07-28 09:16:50 <PhoenixMage> uaa: like ikke I use init=/bin/sh and I usually run sync to make sure its written to disk before I reboot
2020-07-28 09:19:31 <uaa> do I need to remount / ?
2020-07-28 09:20:11 <ikke> You did not tell us yet what error you are getting
2020-07-28 09:20:18 <ikke> So we have no clue
2020-07-28 09:20:45 <uaa> I'm unable to reset the root password
2020-07-28 09:21:00 <uaa> it's seems like it's empty. like disabled or something
2020-07-28 09:23:35 <PhoenixMage> did they just leave? lol I have parts turned off
2020-07-28 09:23:44 <ikke> yes, they did
2020-07-28 09:23:50 <ikke> And now they are back
2020-07-28 09:24:11 <uaa> a bug in my system forced me out, sorry
2020-07-28 09:24:26 <uaa> hope no one said something bad while I'm away
2020-07-28 09:24:39 <ikke> lots of gozzip
2020-07-28 09:24:52 <uaa> :-X
2020-07-28 09:25:01 <PhoenixMage> uaa: You boot to single user mode and type "passwd" put in your new password and then reboot but the password doesnt work?
2020-07-28 09:25:14 <PhoenixMage> WE need more info
2020-07-28 09:25:20 <uaa> I type passwd root
2020-07-28 09:25:34 <uaa> I open /etc/password I see change
2020-07-28 09:25:41 <uaa> but after reboot nothing there
2020-07-28 09:25:59 <PhoenixMage> /etc/passwd doesnt store passwords
2020-07-28 09:26:11 <ikke> Do you use a run-from-ram system?
2020-07-28 09:26:16 <uaa> no
2020-07-28 09:26:29 <uaa> yeah I guess it's in /etc/shadow
2020-07-28 09:26:43 <uaa> not ram
2020-07-28 09:26:52 <uaa> unless I missed up something while installing
2020-07-28 09:27:15 <PhoenixMage> You shouldnt need to specify a user but it shouldnt hurt either
2020-07-28 09:27:44 <PhoenixMage> I'd try again and run a sync after running passwd
2020-07-28 09:28:01 <uaa> do I have to remount / < this
2020-07-28 09:28:11 <PhoenixMage> Just out of curiousity, how do you boot to single user?
2020-07-28 09:28:11 <uaa> I think I'm using LVM not sure if it's related
2020-07-28 09:28:19 <ikke> no, that should not matter
2020-07-28 09:28:36 <uaa> I edit grub and add single or init=/bin/sh
2020-07-28 09:28:54 <PhoenixMage> ok, one of those is single user mode the other isnt
2020-07-28 09:29:12 <uaa> which one?
2020-07-28 09:29:46 <PhoenixMage> When you add single you get single user mode, when you add init=/bin/sh you are essentially bypassing booting
2020-07-28 09:30:32 <ikke> bypassing the normal init system
2020-07-28 09:30:54 <uaa> I get this message "sh: can't access tty; job control turned off" not sure if it's related to this
2020-07-28 09:31:01 <uaa> no I"m on single again. what should I do?
2020-07-28 09:31:06 <uaa> remount something?
2020-07-28 09:31:12 <uaa> or just change the password?
2020-07-28 09:31:13 <ikke> Only if things are mounted ro
2020-07-28 09:31:26 <uaa> I don't see things mounted ro
2020-07-28 09:31:27 <ikke> but if passwd doesn't give an error message, then that was not the case
2020-07-28 09:31:41 <uaa> no error
2020-07-28 09:32:09 <uaa> apparently there is not shadow file so it's stored in /etc/passwd right?
2020-07-28 09:32:40 <wener[m]> <uaa "I get this message "sh: can't ac"> most time, this is caused by rootfs not mount
2020-07-28 09:32:57 <ikke> uaa: there should be a shadow file
2020-07-28 09:34:18 <uaa> I think the installation might be corrupt. I'll re install,, what do you think?
2020-07-28 09:35:23 <uaa> https://imgur.com/a/YqE7jhZ
2020-07-28 09:38:26 <ikke> try apk fix alpine-baselayout
2020-07-28 09:40:10 <uaa> error: unable to lock database no such file or directory
2020-07-28 09:40:32 <uaa> error: failed to open apk database: similar error
2020-07-28 09:41:05 <ikke> sounds like something is corrupted indeed
2020-07-28 09:41:13 <ikke> did anything happen to this system?
2020-07-28 09:41:42 <uaa> not sure, I installed it a month ago then forgot it
2020-07-28 09:42:06 <uaa> sorry for bothering I'll try to create new one
2020-07-28 09:59:08 <PhoenixMage> Anyone know where the patches at https://dev.alpinelinux.org/archive/rpi-patches/ come from?
2020-07-28 10:05:52 <ikke> They are somehow based on https://github.com/raspberrypi/linux iirc, but not sure if there is any customization
2020-07-28 10:08:15 <mps> this patches are created by genpatch() in linux-rpi APKBUILD, I think
2020-07-28 10:09:58 <ikke> is there a reason we use patches rather than just release tarbals?
2020-07-28 10:23:00 <pablonut> Q about packages: I'd like to contribute a package to community i think i've got the APKBUILD file and patches done should i simply attach these to the `feat` request ?
2020-07-28 10:23:17 <ikke> pablonut: best would be to make a merge request
2020-07-28 10:23:26 <ikke> pablonut: but note that packages initially start in testing
2020-07-28 10:23:43 <ikke> after you or someone confirmed the package is working stable, it can be moved to community
2020-07-28 10:24:23 <pablonut> perfect - which repo would that need to go on ?
2020-07-28 10:24:34 <pablonut> guessing abuild
2020-07-28 10:24:37 <ikke> aports
2020-07-28 10:24:44 <ikke> https://gitlab.alpinelinux.org/alpine/aports
2020-07-28 10:25:02 <ikke> abuild is just the tool to build packages :)
2020-07-28 10:25:14 <pablonut> ahh yeah sorry my bad :P
2020-07-28 10:25:23 <ikke> no worry
2020-07-28 10:28:52 <mps> ikke: I don't know, ncopa can give answer to this question
2020-07-28 10:29:16 <mps> but I hope with linux 5.9 mainline will work with RPis
2020-07-28 10:30:30 <ncopa> PhoenixMage: rpi patches are diffs between https://github.com/raspberrypi/linux and mainline release
2020-07-28 10:31:07 <ncopa> sometimes i merge v5.4.y into latest rpi and push a upload a diff
2020-07-28 10:31:34 <ncopa> that is when i want push new mainline kernel release but rpi has not yet updated
2020-07-28 12:49:31 <albr> av[d]: Has anyone gotten f2fs working as their thier root partition?
2020-07-28 12:50:38 <mps> yes
2020-07-28 12:51:23 <albr> av[d]: What kind of install did you have?
2020-07-28 12:51:46 <mps> manual, on arm32 SBCs
2020-07-28 12:52:15 <albr> av[d]: I'm currently livebooting of a USB with persistence using `lbu` and I'm having trouble figuring out how to switch my root partition to f2fs
2020-07-28 12:52:42 <mps> add f2fs to /etc/mkinitfs/mkinitfs.conf
2020-07-28 12:52:54 <mps> for example, 'features="ata base ext4 kms mmc raid usb f2fs"'
2020-07-28 12:53:35 <mps> I'm not sure how to do that for run-from-ram installs
2020-07-28 12:54:07 <albr> av[d]: I don't even have a `/etc/mkinitfs/mkinitfs.conf`
2020-07-28 12:54:49 <ikke> run-from-ram uses tmpfs :)
2020-07-28 12:58:56 <albr> av[d]: Is there a way to change the tempfs configuration to use f2fs?
2020-07-28 12:59:20 <ikke> That does not make sense
2020-07-28 12:59:23 <ikke> tmpfs is in ram
2020-07-28 12:59:35 <albr> av[d]: Oh, tru
2020-07-28 13:00:42 <albr> av[d]: Idk what's happening then. My home partition is already using f2fs but I'm getting major lock-ups indicating my USB drive's endure taking a hit from all the random read/writes.
2020-07-28 13:00:54 <albr> av[d]: endurance*
2020-07-28 13:54:46 <wener[m]> luks with correct key, device mapper not show up
2020-07-28 13:55:11 <wener[m]> APPEND is. modules=sd-mod,usb-storage,ext4,encrypt nomodeset quiet cryptroot=UUID=99e003a0-1418-4e97-9f8f-24fb52ee8250 cryptdm=cryptroot
2020-07-28 13:56:37 <wener[m]> entered emergency shell, manul cryptsetup open and mount is working
2020-07-28 14:09:12 <wener[m]> need help https://unix.stackexchange.com/questions/600743/luks-device-mapper-not-show-up-after-entered-password
2020-07-28 14:10:15 <Cogitri> I can check in a bit what boot args I have set on my laptop
2020-07-28 14:16:41 <winny> anybody have a script handy to manage apk-new files? maybe something like dispatch-conf on gentoo? or sysmerge on freebsd?
2020-07-28 14:17:20 <ikke> update-conf
2020-07-28 14:17:59 <winny> wener[m]— it smells like cryptsetup hasn't unlocked the disk
2020-07-28 14:23:39 <winny> wener[m]— can you manually run cryptsetup in the rescue shell?
2020-07-28 14:24:07 <winny> also did you edit /etc/mkinitfs/mkinitfs.conf
2020-07-28 14:24:17 <wener[m]> winny: I can manually cryptsetup open, mount ad /sysroot and boot
2020-07-28 14:24:37 <wener[m]> yes, add cryptsetup in /etc/mkinitfs/mkinitfs.conf
2020-07-28 14:25:04 <winny> so you ran 'cryptsetup luksOpen /dev/disk/by-uuid/99e003a0-1418-4e97-9f8f-24fb52ee8250 cryptroot' (i think that's how the mkinitramfs initramfs will do it)
2020-07-28 14:26:23 <wener[m]> yes, here is how I manual open https://i.stack.imgur.com/BalR9.png
2020-07-28 14:26:33 <winny> ikke— cheers, for what it's worth, is there some documentation i missed?
2020-07-28 14:26:47 <winny> could you try the above
2020-07-28 14:26:47 <wener[m]> I think this should be handled by initramfs or something
2020-07-28 14:27:05 <winny> the initramfs doesn't know about a sda2, it knows about a uuid
2020-07-28 14:27:24 <ikke> winny: No idea, i've also learned it from someone else
2020-07-28 14:27:25 <wener[m]> I use UUID=
2020-07-28 14:27:38 <wener[m]> dont have udev that mount /dev/disk/by-uuid
2020-07-28 14:28:25 <wener[m]> maybe this is the problem
2020-07-28 14:29:42 <winny> oh maybe it's using nlplug-findfs
2020-07-28 14:30:16 <wener[m]>                / # cryptsetup open UUID="99e003a0-1418-4e97-9f8f-24fb52ee8250" cryptroot
2020-07-28 14:30:16 <wener[m]>                Device /dev/disk/by-uuid/99e003a0-1418-4e97-9f8f-24fb52ee8250 does not exist or
2020-07-28 14:30:16 <wener[m]>                access denied.
2020-07-28 14:31:41 <wener[m]> But if initram can not found by uuid, is should not ask for password in the first place
2020-07-28 14:32:53 <winny> https://github.com/alpinelinux/mkinitfs/blob/961726b6aeb8e12176009675f22ed0ffc2b26e14/initramfs-init.in#L443-L482
2020-07-28 14:34:08 <winny> for what it's worth here is one of my alpine boxes's default_kernel_opts in update-extlinux.conf: default_kernel_opts="nomodeset quiet rootfstype=ext4 cryptroot=UUID=3cb7aacf-4975-466b-bb6d-92a13ad60496 cryptdm=rootCrypt cryptdiscards cryptkey"
2020-07-28 14:35:07 <wener[m]> thanks
2020-07-28 14:38:46 <winny> you could try passing debug_init=1 seeing if that gives you any better hints
2020-07-28 14:40:16 <wener[m]> After add cryptdiscards , the dm showed up
2020-07-28 14:40:29 <cvicpp> hi all, I am using alpine linux in an open source side project of mine, called https://github.com/chrisvel/wreeto_official
2020-07-28 14:40:55 <cvicpp> There's an issue with rubygems and IPv6 which is open for some time and I can't find a solution https://github.com/chrisvel/wreeto_official/issues/5
2020-07-28 14:41:23 <winny> strange, that should have no bearing on its available wener[m], but glad it worked, maybe another step you did helped too?
2020-07-28 14:41:56 <cvicpp> unfortunately there is no gai.conf in alpine, so I have been searching for an alternative and couldn't find something useful. Any help will be appreciated
2020-07-28 14:44:30 <wener[m]> <winny "strange, that should have no bea"> trying to build a reproducable setup image https://github.com/wenerme/alpine-image
2020-07-28 14:44:50 <wener[m]> If this is caused by other tweak, will find out
2020-07-28 14:45:37 <winny> sounds like you are good hands, if it is related to cryptdiscards it smells like a bug
2020-07-28 14:47:36 <wener[m]> Seems not related to cryptdiscards, removed cryptdiscards, add root=, not it boot up
2020-07-28 14:47:39 <wener[m]> now it boot up
2020-07-28 14:55:22 <winny> wener[m]— here's a diff of my update-extlinux.conf https://0x0.st/ixPW.diff showing what was necessary to make my ext4 on lvm2 on luks on md root bootable (besides the edits needed to /etc/mkinitfs/mkinitfs.conf
2020-07-28 15:15:19 <wener[m]> winny:  My setup is almost the same, but I don't use lvm, the result is simpler
2020-07-28 15:15:57 <wener[m]> Here is my setup, 60 line of shell, https://github.com/wenerme/alpine-image/blob/master/scripts/loopdev-luks.sh
2020-07-28 15:21:57 <albr> ScottHamilton[d]: Hello, I have pkg-config problem with ruby bundler on a docker container. bundler tries to build eventmachine but it gets an error when checking openssl's pkg-config :
2020-07-28 15:21:58 <albr> ScottHamilton[d]: here is the relevant output of /usr/lib/ruby/gems/2.7.0/extensions/x86_64-linux-musl/2.7.0/eventmachine-1.2.7/mkmf.log :
2020-07-28 15:21:59 <albr> ScottHamilton[d]: "pkg-config --exists openssl"
2020-07-28 15:22:00 <albr> ScottHamilton[d]: package configuration for openssl is not found
2020-07-28 15:22:01 <albr> ScottHamilton[d]: 
2020-07-28 15:22:03 <albr> ScottHamilton[d]: What I don't understand it that pkg-config --exists openssl returns 0, I've added set PKG_CONFIG_PATH to /usr/lib/pkgconfig and ensured  /usr/lib/pkgconfig/openssl.pc exists. What's going wrong?
2020-07-28 16:08:34 <albr> zoinks[d]: Does anyone how I can download `php-libxml`
2020-07-28 16:13:23 <albr> zoinks[d]: I got it nvm
2020-07-28 16:27:56 <WeblordPepe> hey fellas. loving alpine so far. ran into an issue with cmake tho
2020-07-28 16:28:12 <WeblordPepe> it wont work with libarchive. it just buggers up saying it cant read libarchive
2020-07-28 16:28:33 <WeblordPepe> had to apk del cmake and then apk add icmake
2020-07-28 16:28:46 <WeblordPepe> from the thingie doodad repository
2020-07-28 16:29:04 <WeblordPepe> testing
2020-07-28 16:29:55 <WeblordPepe> altho that kinda doesnt work for this thing im tryiung to build heh
2020-07-28 16:45:53 <WeblordPepe> im trying to build cmake on my alpine
2020-07-28 16:46:18 <WeblordPepe> tho its running in a virtual machine that doesnt have vm acceleration and is already on an ancient cpu
2020-07-28 16:48:06 <ikke> WeblordPepe: Won't work with libarchive, how?
2020-07-28 16:48:45 <WeblordPepe> it gives errors saying that its the wrong executible format or wrong binary format. sorry its way up in my terminal now. will have a look
2020-07-28 16:49:04 <ikke> Sounds like wrong arch
2020-07-28 16:49:17 <WeblordPepe> something about libexpat not working right
2020-07-28 16:50:01 <WeblordPepe> yeah tho i havent specified any wired arches or nothin. just on x86_64
2020-07-28 16:52:42 <WeblordPepe> cmake libexpat.so.1: Exec format error (needed by /usr/lib/libarchive.so.13)
2020-07-28 16:53:00 <WeblordPepe> then it has a bunch of complaints about various XML functions not being found
2020-07-28 16:53:29 <WeblordPepe> such as
2020-07-28 16:53:31 <WeblordPepe> Error relocating /usr/lib/libarchive.so.13: XML_Parse: symbol not found libarchive
2020-07-28 16:54:37 <ikke> Is it built against musl?
2020-07-28 16:54:40 <ikke> or glibc?
2020-07-28 16:55:51 <WeblordPepe> the thing im building? or cmake? im super new to alpine so i couldnt tell ya how cmake was built here (except that its from the testing branch it looks like?) - the actual thing im trying to use cmake for is a doom port. but i type cmake anywhere and its the same complaints
2020-07-28 16:57:01 <WeblordPepe> /usr/bin/cmake: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib/ld-musl-x86_64.so.1, stripped
2020-07-28 16:58:34 <WeblordPepe> im really liking this OS tho. the package manager is so fast. its seems so ...i dunno tidy or something. straight forward. having some troubles here sure but the OS itself is so free of garbage
2020-07-28 16:59:10 <WeblordPepe> im running the whole thing at the moment inside a VM on an old amd athlon2 that doesnt have virtualization extensions so its all software emulation and its still fast
2020-07-28 16:59:33 <WeblordPepe> tho im compiling cmake now on it so that ... that is not so fast
2020-07-28 17:00:20 <ikke> can you check ldd libexpat.so.1
2020-07-28 17:00:35 <ikke> It seems like libexpat is the issue
2020-07-28 17:05:18 <WeblordPepe> sure thing. one sec
2020-07-28 17:06:13 <WeblordPepe> /lib/ld-musl-x86_64.so.1: /usr/lib/libexpat.so.1.6.11: Not a valid dynamic program
2020-07-28 17:06:22 <WeblordPepe> alpine:~/cmake-master# file /usr/lib/libexpat.so.1.6.11
2020-07-28 17:06:22 <WeblordPepe> /usr/lib/libexpat.so.1.6.11: data
2020-07-28 17:06:29 <WeblordPepe> that file response doesnt fill me with confidence
2020-07-28 17:07:37 <WeblordPepe> readelf: Error: Not an ELF file - it has the wrong magic bytes at the start
2020-07-28 17:07:58 <WeblordPepe> dun dun ddaaaAa
2020-07-28 17:08:55 <WeblordPepe> sorry if this is the wrong place to dump a bug report - im pretty new & couldnt figure anything online & not signed up to anyting or what not
2020-07-28 17:09:30 <WeblordPepe> i live on planet debian and this is my first trip off-world in a lot of years
2020-07-28 17:21:20 <ikke> libexpat is functional for me
2020-07-28 17:21:37 <ikke> ldd /usr/lib/libexpat.so.1.6.11
2020-07-28 17:21:42 <ikke> libc.musl-x86_64.so.1 => /lib/ld-musl-x86_64.so.1 (0x7f3ddd67a000)
2020-07-28 17:48:17 <WeblordPepe> hmm
2020-07-28 17:48:34 <WeblordPepe> wanna md5sum it?
2020-07-28 17:48:48 <WeblordPepe> a40d89eb685cbf20299f2f58ead1e9a9  libexpat.so.1
2020-07-28 17:48:51 <WeblordPepe> thats me
2020-07-28 17:49:08 <ikke> 0752cdb959724964cd0b7200aa773026
2020-07-28 17:49:25 <WeblordPepe> lddtree is pretty cool. i just found it
2020-07-28 17:49:36 <WeblordPepe> try lddtree /usr/bin/cmake
2020-07-28 17:49:41 <WeblordPepe> what a cool tool
2020-07-28 17:50:45 <ikke> nice
2020-07-28 17:51:27 <ikke> seems like your libexpat.so.1 library is corrupt somehow
2020-07-28 17:51:32 <telmich> I'm trying to bump graphviz to 2.44.1, however when building with `abuild -r` I get the error message `>>> ERROR: graphviz: builddeps failed` and no more details
2020-07-28 17:51:36 <telmich> How would I debug this?
2020-07-28 17:51:50 <WeblordPepe> ikke: can i have your one? :P
2020-07-28 17:52:06 <ikke> WeblordPepe: try apk fix expat
2020-07-28 17:52:16 <WeblordPepe> 
ACTION tries

2020-07-28 17:52:29 <ikke> telmich: try apk fix
2020-07-28 17:52:32 <WeblordPepe> oh shit it goes
2020-07-28 17:52:50 <WeblordPepe> heh thanks ikke
2020-07-28 17:53:00 <WeblordPepe> thanks man
2020-07-28 17:53:04 <ikke> no problem
2020-07-28 17:53:22 <telmich> ohh... tons of texlive errors like this: `ERROR: texlive-20200406-r0: trying to overwrite usr/share/texmf-dist/fonts/enc/dvips/base/dvips-all.enc owned by texmf-dist-2020.55416-r0.`
2020-07-28 17:53:37 <ikke> ah, that has been fixed in the mean time, I believe
2020-07-28 17:53:46 <ikke> at least, someone was looking at it
2020-07-28 17:54:01 <ikke> Cogitri:
2020-07-28 17:54:34 <Cogitri> Yes, I think there's a MR to fix it now
2020-07-28 17:54:42 <Cogitri> Gonna look into it once FF is done building
2020-07-28 17:54:48 <Cogitri> Sorry for the breakage
2020-07-28 17:55:31 <ikke> texlive suddenly started to provide the same files as texmf-dist
2020-07-28 17:59:38 <WeblordPepe> hmm how would one get openGL libraries for building shiz ? not sure which package
2020-07-28 18:00:01 <WeblordPepe> mesa right
2020-07-28 18:00:10 <Shiz> ?
2020-07-28 18:01:16 <mps> :)
2020-07-28 18:01:24 <WeblordPepe> haha i didnt see someone here was actually called shiz
2020-07-28 18:02:00 <WeblordPepe> hi :P pleased to meet you
2020-07-28 18:06:06 <telmich> what does $pkgdir usually refer to? A temporary installation / root directory?
2020-07-28 18:07:17 <telmich> ah, found it, it pkg/ inside the directory
2020-07-28 18:09:06 <wener[m]> <telmich "ah, found it, it pkg/ inside the"> usually refereed to  pkg/
2020-07-28 18:12:59 <mps> wener[m]: no need to quote on irc
2020-07-28 18:15:41 <telmich> It would be nice if the quote would stay in the matrix world, because here it is quite useful, but I can see it is not very natural in irc
2020-07-28 18:17:12 <mps> telmich: yes
2020-07-28 18:18:54 <telmich> I was also wondering whether it makes sense to have a "native" alpine linux matrix channel, but I see the problem of community splitting with that
2020-07-28 18:24:28 <mps> maybe this isn't bad idea :P
2020-07-28 18:37:18 <telmich> Wait...there is actually already #alpinelinux:matrix.org
2020-07-28 18:39:30 <telmich> Different question: for some reason bumping graphviz from 2.44.0 to 2.44.1 is not straight forward and even after adding missing groff dependency, the configure does not find Python.h - is this a common pitfall?
2020-07-28 19:51:48 <albr> Watynecc[d]: My GPU make sound i want to stop it how do that
2020-07-28 19:57:43 <albr> Watynecc[d]: thanks irl
2020-07-28 19:57:47 <albr> Watynecc[d]: irc*
2020-07-28 20:47:09 <gjabell> does alpine have any packages for h264/5 support in qtwebengine?
2020-07-28 20:48:54 <Cogitri> https://gitlab.alpinelinux.org/alpine/aports/-/issues/11331
2020-07-28 21:04:49 <gjabell> hmm
2020-07-28 21:04:54 <gjabell> guess I'm running into the same issue then
2020-07-29 01:06:30 <PhoenixMage> Anyone seen it where udhcpc attempts to get an address but the interface is down and it doesnt try to bring it up?
2020-07-29 01:06:57 <PhoenixMage> I essentiall have to "ip link set eth0 up" then restart networking
2020-07-29 01:17:50 <PhoenixMage> ncopa3: Thanks for the info on those patches, any idea if Alpine is going to go 5.7?
2020-07-29 03:47:41 <winny> after adding "blacklist kvm" to /etc/modprobe.d/blacklist.conf i noted kvm is still loaded on subsequent boots, is there a different way to ensure a module isn't loaded?
2020-07-29 04:14:25 <winny> adding stuff to blacklist==... does not appear to do what one might think either
2020-07-29 04:20:03 <winny> ah I think i want modules_blacklist=...
2020-07-29 04:21:07 <winny> anyway, interesting issue -- this machine i have hangs on "Requesting system reboot".
2020-07-29 04:54:52 <wener[m]> mps:  ok, I use riot, the quote is handy, haha
2020-07-29 04:57:08 <wener[m]> winny: maybe be cause by acpid not started ?
2020-07-29 05:25:12 <winny> huh i didn't think to check that
2020-07-29 05:26:10 <winny> i found after swithing the machine to uefi mode halt/rebooting works, but now i have to get this thumb drive bootable under grub and uefi. this particular machine is a bit picky about uefi stuff
2020-07-29 05:39:24 <PhoenixMage> winny: Make a grub.efi file and make that your default uefi boot?
2020-07-29 05:46:04 <wener[m]> Is syslinux efi working
2020-07-29 05:47:54 <winny> i'm not really sure what did the trick, i looked in efibootmgr, and i had around 100 entries for bogons (from an old fedora install many years ago), wiped those, and manually added one to boot grub
2020-07-29 05:48:57 <winny> good tips, i'll remember that for next time (it'll be sooner than i like, too!)
2020-07-29 05:49:29 <winny> i tried to get syslinux efi working on a another box with alpine, but was unsuccessful, would be very interested in using that too
2020-07-29 05:50:25 <wener[m]> I want to try uefi on qemu, but qemu with uefi is kind of hard
2020-07-29 05:53:46 <adu> wener[m] I didn't know syslinux was working on an efi version
2020-07-29 05:55:39 <adu> I thought syslinux was dead
2020-07-29 05:56:07 <adu> given that the most recent release was 6 years ago
2020-07-29 06:01:57 <adu> winny what is this syslinux efi thing you speak of?
2020-07-29 06:16:27 <winny> adu— https://wiki.gentoo.org/wiki/Syslinux#Making_Syslinux_known_to_EFI
2020-07-29 06:16:36 <winny> https://wiki.syslinux.org/wiki/index.php?title=Install#UEFI
2020-07-29 06:16:48 <winny> I did not try very hard, to be honest
2020-07-29 06:29:11 <mps> wener[m]: I believe quoting is good for matrix, but here it looks ugly
2020-07-29 06:37:16 <adu> winny, why not just rename it BOOTX64.efi
2020-07-29 06:38:49 <winny> that's also a good question adu
2020-07-29 06:39:01 <adu> then you don't need to add a boot entry
2020-07-29 06:39:51 <adu> https://wiki.syslinux.org/wiki/index.php?title=Install#UEFI
2020-07-29 06:39:59 <adu> that's what the official docs say
2020-07-29 06:40:27 <adu> honestly I didn't know there was existing code for this
2020-07-29 06:40:32 <winny> i had done than on my elitebook, actually, because grub-install never made itself bootable no matter what i did
2020-07-29 06:41:49 <winny> i wish i could get back the lost months due to dealing with various pcs and misconfigured efi
2020-07-29 06:44:45 <adu> I'm kind of a UEFI expert, and it's a mess
2020-07-29 06:45:12 <adu> pretty much the biggest bug in UEFI is the E
2020-07-29 06:46:34 <adu> It's just so huge, and there are so many variants and versions and vendors that it's almost impossible to get right
2020-07-29 06:50:23 <adu> my personal opinion is that there will probably be some kind of stretchy standard in the future, and BIOS and UEFI are going to be points on a spectrum
2020-07-29 06:53:54 <adu> my personal favorite bootloader is the UEFI shell with a shell script that calls linux with the options that you want, no syslinux, no grub
2020-07-29 07:01:31 <PhoenixMage> I boot grub efi from my various SBCs just for consistency
2020-07-29 08:02:06 <ncopa> PhoenixMage: we try to stick to LTS kernels
2020-07-29 08:03:05 <ikke> We have linux-edge in testing for linux latest stable kernels
2020-07-29 08:24:01 <komnpomat> mm... ---> Error relocating /usr/bin/links: __mktime64: symbol not found
2020-07-29 08:24:19 <ikke> komnpomat: what arch?
2020-07-29 08:24:20 <komnpomat> Error relocating /usr/bin/links: __utimes_time64: symbol not found
2020-07-29 08:24:26 <komnpomat> 32bit
2020-07-29 08:24:34 <komnpomat> alpine edge
2020-07-29 08:24:37 <komnpomat> oh well
2020-07-29 08:24:39 <ikke> what version of musl is installed?
2020-07-29 08:24:46 <ikke> You need to make sure you have musl-1.2
2020-07-29 08:24:51 <komnpomat> how to check
2020-07-29 08:24:58 <ikke> do you use a docker container?
2020-07-29 08:25:12 <komnpomat> no, full install
2020-07-29 08:25:14 <ikke> ok
2020-07-29 08:25:20 <ikke> just do a full system upgrade
2020-07-29 08:25:22 <ikke> apk upgrade -U
2020-07-29 08:25:58 <ikke> komnpomat: https://lists.alpinelinux.org/~alpine/devel/%3C20200721171650.48fa63a4%40ncopa-desktop.copa.dup.pw%3E
2020-07-29 08:27:21 <komnpomat> now upgrading
2020-07-29 08:27:37 <komnpomat> which means rebooting :(
2020-07-29 08:27:40 <ikke> you need to add -a (--available) as well
2020-07-29 08:28:55 <komnpomat> after upgrade?
2020-07-29 08:29:16 <ikke> No, just run that command again with -a
2020-07-29 08:29:18 <komnpomat> ERROR: sway-1.5-r0: BAD signature
2020-07-29 08:29:24 <komnpomat> lawl
2020-07-29 08:29:38 <komnpomat> ok
2020-07-29 08:29:39 <ikke> Yeah, that's mentioned in the e-mail as well
2020-07-29 08:30:25 <ikke> Can you try again?
2020-07-29 08:31:05 <komnpomat> I will, I am waiting to finish the earlier command
2020-07-29 08:31:10 <ikke> nod
2020-07-29 08:37:12 <komnpomat> now it's replacing
2020-07-29 08:39:34 <komnpomat> ikke: thanks a lot! your genious!
2020-07-29 10:14:34 <Hello71> winny: some older implementations have broken variables, so you need to specify grub-install --removable
2020-07-29 10:15:04 <Hello71> also regarding your reboot problem you should try adjusting reboot= kernel parameter
2020-07-29 10:26:52 <cvicpp> hello
2020-07-29 10:27:14 <cvicpp> I am using alpine linux in an open source side project of mine, called https://github.com/chrisvel/wreeto_official
2020-07-29 10:27:18 <cvicpp> There's an issue with rubygems and IPv6 which is open for some time and I can't find a solution https://github.com/chrisvel/wreeto_official/issues/5
2020-07-29 10:27:23 <cvicpp> unfortunately there is no gai.conf in alpine, so I have been searching for an alternative and couldn't find something useful. Any help will be appreciated
2020-07-29 10:47:14 <mps> isn't gai.conf glibc specific
2020-07-29 10:49:07 <cvicpp> I have no idea, what does it mean If it's glibc specific ?
2020-07-29 10:51:50 <ikke> cvicpp: seems like an issue on the host
2020-07-29 10:52:09 <ikke> THe Dockerfile seems to install fine for me on an ipv4 only host
2020-07-29 10:52:29 <ikke> Will test it also on a ipv6 capable host
2020-07-29 10:53:37 <mps>                  │(12:10:33) Я: rekao bih da jeste                                                                                 ▒│
2020-07-29 10:53:52 <mps> crap, sorry
2020-07-29 10:53:55 <ikke> :P
2020-07-29 10:54:03 <cvicpp> Thanks, it makes sense If it's related to binding of docker's & host's networks
2020-07-29 10:54:03 <mps> https://www.man7.org/linux/man-pages/man5/gai.conf.5.html
2020-07-29 10:55:58 <mps> cvicpp: I mean to say it is not defined by POSIX or I can't find
2020-07-29 10:56:11 <ikke> yeah, it's something glibc specific
2020-07-29 10:56:19 <ikke> but it would only be a workaround
2020-07-29 10:56:38 <mps> create empty file :)
2020-07-29 10:59:58 <ikke> cvicpp: yes, working on an ipv6 capable host as well
2020-07-29 11:00:19 <ikke> Issue is that for some reason you get an ipv6 dns record back while you cannot connect over ipv6
2020-07-29 11:00:56 <solar> that ipv6 address is @ fastly
2020-07-29 11:01:09 <solar> i just pinged an engineer over there to see if the issue is on their side :)
2020-07-29 11:02:15 <ikke> maybe
2020-07-29 11:02:37 <ikke> nice serendipity
2020-07-29 11:54:54 <cvicpp> solar: thanks
2020-07-29 13:54:16 <vejetary1> I'm getting http://ix.io/2sBu from `certbox --nginx` command on AWS. Anyone know how can I fix this?
2020-07-29 13:54:50 <vejetary1> /etc/alpine-release says `3.13.0_alpha20200626`.
2020-07-29 13:57:11 <ikke> vejetary1: hmm, cannot reproduce it on alpine:edge (docker)
2020-07-29 13:58:00 <vejetary1> I don't know why this happens either. It wasn't doing this when I first get Let's Encrypt certificate.
2020-07-29 13:58:11 <vejetary1> Maybe it got broken whe I did a system update.
2020-07-29 14:01:18 <vejetary1> ikke: I did another update & upgrade & re-install the package now, but still doesn't work. :'(
2020-07-29 14:05:59 <cvicpp> solar: sorry to interrupt, but do you have any updates ?
2020-07-29 14:06:18 <solar> no
2020-07-29 14:06:22 <solar> it's 7am
2020-07-29 15:52:28 <EF> Is it necessary to add root and non-OS in the cryptdm= in update-extlinux.conf? Cannot see to get the non-OS luks volume to decrypt automagically by using crypttab?
2020-07-29 15:54:06 <EF> My sentence is gibberish. Root and non-OS luks volumes and cryptdm in the default kernel options!
2020-07-29 15:56:01 <Cogitri> Still not sure if I understand what you mean
2020-07-29 15:57:21 <EF> I have two luks volumes. The root one gets decrypted automatically at boot. The second is a data volume and it doesn't get decrypted automatically
2020-07-29 15:57:44 <EF> Can make the volume decrypt on the cli.
2020-07-29 15:59:00 <EF> Apart from detailing the second luks volume in crypttab, is there anywhere else it is required, i.e. default_kernel_opts in update-extlinux.conf?
2020-07-29 16:04:34 <gjabell> EF you want the root system to decrypt a non-OS volume after system start is that right?
2020-07-29 16:05:32 <EF> Yes please
2020-07-29 16:06:31 <gjabell> if you are using a keyfile you should be able to add an entry to /etc/conf.d/dmcrypt and then do rc-update add dmcrypt boot
2020-07-29 16:06:55 <gjabell> I had the same question the other day :P
2020-07-29 16:07:14 <gjabell> apparently crypttab is ignored (at least for this case)
2020-07-29 16:08:56 <EF> Would that apply to swap which is an lvm partition on the root luks or just the second luks?
2020-07-29 16:09:39 <gjabell> hmm I think unless you have swap separately encrypted it should already be decrypted with the root luks
2020-07-29 16:09:54 <gjabell> so at that point you'd just need an fstab entry for swap
2020-07-29 16:11:09 <gjabell> the setup I used this with was: root drive encrypted with LUKS that requires entering a password to unlock, plus two data drives encrypted with a LUKS key that should get decrypted (and mounted) automatically (ie not having to enter passwords for them)
2020-07-29 16:11:13 <EF> Okay thanks. Swap is being ignored too but that's tomorrow's problem I guess. Thanks for the help. Simple when you know how
2020-07-29 16:11:38 <gjabell> yeah, couldn't find anything on the wiki for it unf, might should add a page
2020-07-29 16:12:15 <gjabell> for swap you should just be able to have something like `/dev/mapper/swap    swap    swap   defaults 0 0` in your /etc/fstab
2020-07-29 16:12:43 <gjabell> (replacing /dev/mapper/swap with whatever the path is from lvm)
2020-07-29 16:12:50 <EF> Mine is luks on raid10 for root, raid1 for /boot and a raid1 with luks for some config data. Aim is for root to be decrypted by a key on an sd card which they exposes the keyfile in /root/key for the second luks volume.
2020-07-29 16:15:14 <EF> Why make it simple! Haha. If syslinux booted raid10 non of this would be an issue
2020-07-29 16:16:12 <gjabell> ah, yeah :P
2020-07-29 16:16:41 <gjabell> I opted for a non-raid root drive with zfs mirror for the data drives; I figure as long as I keep good backups + write down the config steps then it's not a huge deal if the root drive dies
2020-07-29 16:16:58 <gjabell> it's for a server anyways so all the real stuff (ie VMs) is on the data drives
2020-07-29 16:28:01 <EF> mine is a synology replacement in a domestic servery-ish setup but nothing public or mission critical.
2020-07-29 16:38:44 <alot> o/
2020-07-29 16:38:53 <alot> is there a tool like archiso but for alpine?
2020-07-29 16:39:17 <Cogitri> Well, what does archiso do?
2020-07-29 16:39:53 <alot> Cogitri: it takes a package list that you craft (along with some configurations) and generates a liveusb image
2020-07-29 16:40:11 <alot> so I could, hypothetically, generate a custom liveusb for installation
2020-07-29 16:40:26 <alot> (in particular, so I can install alpine on a pinebookpro with LUKS with very little hassle)
2020-07-29 16:40:44 <alot> (a solid bit of documentation on how to accomplish that install would be a reasonable replacement)
2020-07-29 16:41:31 <ikke> alot: The documentation is the missing part, but there is a scripts directory in aports which contains the scripts used to build the isos
2020-07-29 16:41:55 <alot> ikke: interesting
2020-07-29 16:42:17 <alot> ikke: and the isos are valid for liveusb as well as spinning media?
2020-07-29 16:42:21 <ikke> yes
2020-07-29 16:42:26 <alot> excellent
2020-07-29 16:42:40 <alot> that seems like an option
2020-07-29 16:57:32 <winny> Hello71: thanks. I had used the --removable flag which might had done the trick. TIL about the reboot parameter
2020-07-29 17:02:47 <afontain_> alot: technically, postmarketOS already allows installing almost-Alpine on a pinebook pro with very little hassle
2020-07-29 17:04:11 <afontain_> note that it's a bit less minimal than Alpine, for example, you'll have chronie (ntp client) installed by default
2020-07-29 17:07:19 <ggg> Hi!I need a little help. I have an ODROID H2 + SBC and would like to use it with Alpine. However, I could not find a solution for my network card. It contains an RTL8125, which in turn recognizes it as r8169. There is also a description of how to set it up, but I would like to use it from USB and I would like to ask for your help. Thank you very
2020-07-29 17:07:19 <ggg> much in advance and sorry for my bad English!
2020-07-29 17:09:11 <vejetary1> Anyone using Searx with Morty on Alpine?
2020-07-29 17:15:06 <mps> ggg: is this driver supported in mainline kernels
2020-07-29 17:17:44 <vejetary1> Is it normal for ash to not read ~/.profile when I switch to that user with `su`?
2020-07-29 17:17:58 <ikke> vejetary1: 'su -;
2020-07-29 17:18:01 <ikke> vejetary1: 'su -'
2020-07-29 17:18:07 <vejetary1> What do you mean?
2020-07-29 17:18:26 <mps> I use rtl8152 but can't see rtl8125 in kernel config
2020-07-29 17:18:34 <vejetary1> ikke: What do you mean?
2020-07-29 17:18:40 <TBB> vejetary1: man su
2020-07-29 17:18:54 <ikke> vejetary1: you need to pass '-' to su to get a login shell
2020-07-29 17:18:56 <vejetary1> oh, login shell...
2020-07-29 17:18:57 <ggg> @mps thanks your answer, but i think no, not supported :(  .
2020-07-29 17:19:02 <vejetary1> dang it
2020-07-29 17:19:28 <mps> ggg: then we don't support it
2020-07-29 17:19:33 <ggg> mps .These are the links that contain the description for Ubuntu:https://wiki.odroid.com/odroid-h2/hardware/install_ethernet_driver_on_h2plushttps://wiki.odroid.com/odroid-h2/hardware/install_r8125_dkms#tab__using_a_installation_script
2020-07-29 17:20:29 <mps> uh, what is real url
2020-07-29 17:20:30 <TBB> hopefully one day Realtek will start properly supporting Linux
2020-07-29 17:20:32 <ggg> mps And isn't there a chance that it will work somehow in diskless-mode?
2020-07-29 17:20:38 <ggg> Oh
2020-07-29 17:20:53 <ggg> DKMS: https://wiki.odroid.com/odroid-h2/hardware/install_r8125_dkms#tab__using_a_installation_script
2020-07-29 17:21:04 <ggg> And the driver: https://wiki.odroid.com/odroid-h2/hardware/install_ethernet_driver_on_h2plus
2020-07-29 17:21:37 <ggg> Sorry, but i am not a linux guru :(
2020-07-29 17:22:41 <mps> alpine doesn't have dkms but looks like this driver could be built 'manually'
2020-07-29 17:23:43 <ggg> mps Can you possibly help me make it work
2020-07-29 17:23:54 <ikke> See rtl8821ce-lts for an example, although ncopa doesn't prefer to add out-of-tree drivers
2020-07-29 17:25:58 <mps> ikke: also I don't understand why you introduced it
2020-07-29 17:26:20 <ikke> because we don't have dkms :)
2020-07-29 17:26:41 <mps> :D
2020-07-29 17:26:45 <ikke> So end-users have a hard time getting alpine to work with their hardware
2020-07-29 17:27:11 <mps> ggg: so I can follow ikke reasoning :)
2020-07-29 17:28:12 <ggg> @mps If I understand correctly, shouldn't I even hope for it?
2020-07-29 17:28:26 <mps> joke aside, you need to install linux-lts or linux-edge, and linux-lts-dev or linux-edge-dev, depends which kernel you use
2020-07-29 17:29:23 <mps> follow instructions on wiki to unpack driver
2020-07-29 17:29:24 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/rtl8821ce-lts/APKBUILD
2020-07-29 17:29:58 <mps> then tweak Makefile or autorun.sh script
2020-07-29 17:30:11 <mps> looks like it is not complicated
2020-07-29 17:30:36 <mps> right now I don't have much free time to test this
2020-07-29 17:31:17 <ikke> that APKBUILD is for a similar driver, you can try to adopt it for the driver that you need
2020-07-29 17:31:44 <mps> yes, I think ikke is right
2020-07-29 17:32:19 <mps> but that will require setting complete alpine devel system
2020-07-29 17:32:29 <ikke> yup
2020-07-29 17:35:04 <wener[m]> why nlplug-findfs hang event if the dev is showed up
2020-07-29 17:36:24 <wener[m]> vm is ok, but hang on machine
2020-07-29 17:36:36 <wener[m]> nlplug-findfs -p /sbin/mdev UUID
2020-07-29 17:37:08 <wener[m]> 
ACTION uploaded an image: image.png (2424KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/rZbOZBlmtKRBApNtPsgkMhjh >

2020-07-29 17:37:33 <wener[m]>         blkid
2020-07-29 17:37:33 <wener[m]>         echo Need $UUID
2020-07-29 17:37:33 <wener[m]>         nlplug-findfs -p /sbin/mdev $UUID
2020-07-29 17:45:07 <mps> never looked 'in' dkms inner working, but iiuc they have base dkms pkg and for each driver separate dkms-some-driver pkg?
2020-07-29 17:46:07 <ikke> mps: this is the repo that arch uses for the rtl8821ce driver: https://github.com/tomaspinho/rtl8821ce/
2020-07-29 17:46:08 <ggg> Wow, it doesn't seem easy ... And do this again every time when i update? Do I understand correctly?
2020-07-29 17:46:21 <ikke> Yes, for every kernel upgrade
2020-07-29 17:46:38 <ikke> But once you have the package, that should be relatively easy
2020-07-29 17:47:43 <mps> well, whole idea with dkms looks strange, you must have running/installed kernel for which you have to build driver
2020-07-29 17:48:05 <ikke> The idea is that kernel modules are automatically rebuilt when you upgrade the kernel
2020-07-29 17:48:24 <ikke> so you need to have a dev environment installed
2020-07-29 17:48:59 <mps> would it work on alpine, because we remove old kernel on upgrade
2020-07-29 17:49:00 <ikke> So it would use an apk trigger to start it
2020-07-29 17:49:04 <ikke> yes
2020-07-29 17:49:07 <ikke> that doesn't matter
2020-07-29 17:50:02 <telmich> Are you sure, ikke? What if you upgrade the kernel and then install a dkms module that you want to load into the current kernel?
2020-07-29 17:50:07 <mps> I have some scripts which I used to build some out-of-tree modules but I always first reboot and then run these scripts
2020-07-29 17:51:41 <mps> let me try this what ggg need
2020-07-29 17:51:44 <ikke> telmich: the same as if you want to load in-tree modules after upgrading
2020-07-29 17:53:51 <telmich> True. This was/is actually my biggest complaint about arch linux and one of the great things about Debian, that loading kernel modules always works. Is there any plan/notion/general support for fixing this in Alpine mid term?
2020-07-29 17:54:38 <mps> no, I will not, they require me to enter mail :(
2020-07-29 17:54:40 <ikke> telmich: I haven't heard any concrete plans for it
2020-07-29 17:55:21 <mps> this is non-free driver, sorry ggg
2020-07-29 17:55:55 <ikke> shouldn't it be GPL if it's based on the kernel?
2020-07-29 17:57:10 <ikke> GPL compliance even includes just sending it per snail-mail
2020-07-29 17:57:21 <mps> I didn't looked further if I have to enter my mail, or any data, to download it
2020-07-29 17:58:23 <mps> ikke: is this same driver as ggg asked
2020-07-29 17:58:46 <mps> ikke: sorry
2020-07-29 17:58:54 <mps> I missed something
2020-07-29 17:59:26 <ikke> mps: https://github.com/kac89/r8125
2020-07-29 18:01:38 <ikke> sorry, that does not include the source itself
2020-07-29 18:01:45 <mps> hmm, this looks better
2020-07-29 18:03:02 <mps> this look familiar :) https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=r8125
2020-07-29 18:03:25 <ikke> yes, that's the same file
2020-07-29 18:06:27 <mps> running autorun.sh gives 'find: /lib/modules/5.4.34-0-lts/kernel/drivers/net/ethernet: No such file or directory' which is good sign
2020-07-29 18:18:05 <mps> hmm, ggg left and I found solution to her/his problem :)
2020-07-29 18:18:21 <mps> it is not complicated ;)
2020-07-29 18:19:55 <ikke> perhaps they will come back
2020-07-29 18:20:15 <ikke> I've built the rtl8821ce driver someone who was here on-and-off all the time
2020-07-29 18:20:48 <u0jQx9gPyrYg> hah, i was looking for the same driver :)
2020-07-29 18:20:56 <arrde> hei
2020-07-29 18:20:59 <u0jQx9gPyrYg> foudn this one: https://github.com/TallGuy74/r8125
2020-07-29 18:21:08 <arrde> lawl
2020-07-29 18:21:17 <arrde> xorg failing hard with latest upgrade
2020-07-29 18:21:48 <arrde> as a result am in sway now
2020-07-29 18:21:53 <arrde> weirdest thing ever
2020-07-29 18:23:01 <arrde> woah! alacritty smoothiest!
2020-07-29 18:24:20 <ikke> u0jQx9gPyrYg: did you manage to build it?
2020-07-29 18:25:16 <u0jQx9gPyrYg> not yet, but i'll have to. but first i have to setup a system for building it...
2020-07-29 18:25:37 <ikke> right
2020-07-29 18:25:45 <u0jQx9gPyrYg> at least this one seems to be a local guy judging from the name ;)
2020-07-29 18:25:55 <mps> I've got ' error: implicit declaration of function 'pci_set_mwi' [-Werror=implicit-function-declaration]' errors
2020-07-29 18:26:08 <ikke> u0jQx9gPyrYg: heh
2020-07-29 18:27:08 <fyr> guys, i'm lost to why my wifi chip on a c720 (ath9k) isn't loading the firmware, ip a only show loopback, I've installed linux-firmware-ath9k_htc using another machine
2020-07-29 18:27:20 <fyr> the modules show
2020-07-29 18:27:39 <fyr> in /lib/modules
2020-07-29 18:27:50 <ikke> did you try to modprobe it?
2020-07-29 18:33:20 <ggg> Hi again. Im so sorry, but my net was down :(
2020-07-29 18:33:57 <ikke> :)
2020-07-29 18:35:21 <ggg> The last time I saw this was https://github.com/kac89/r8125
2020-07-29 18:35:36 <ikke> ggg: mps was looking it
2020-07-29 18:37:24 <mps> ggg: https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=r8125
2020-07-29 18:40:05 <ggg> Sorry for my stupid question, but should I be able to make a package of this?
2020-07-29 18:40:17 <mps> yes
2020-07-29 18:40:34 <arrde> any other users of sway?
2020-07-29 18:40:48 <mps> you need to install linux-lts-dev and alpine-sdk
2020-07-29 18:41:32 <arrde> by the way, latest upgrade killed /etc/default/grub
2020-07-29 18:41:33 <mps> download this driver, unpack it, tweak Makefile and build it
2020-07-29 18:41:42 <mps> that's all
2020-07-29 18:41:48 <arrde> so now I can't have custom kernel command
2020-07-29 18:42:59 <ggg> And then can it be installed smoothly from a package? Can I do the package in docker too?
2020-07-29 18:43:13 <mps> no, no package
2020-07-29 18:43:22 <mps> by 'hand'
2020-07-29 18:43:44 <mps> to make package will need some more time, which I don't have now
2020-07-29 18:44:23 <arrde> any sway experts around? cursor is upside down
2020-07-29 18:44:26 <mps> maybe u0jQx9gPyrYg will make it
2020-07-29 18:45:12 <ikke> well, it should be easy enough to create a package out of it
2020-07-29 18:46:00 <ggg> Seriously? And will it get into the aports? If so, how much time does that usually be?
2020-07-29 18:46:17 <ikke> oh, gone
2020-07-29 18:46:32 <u0jQx9gPyrYg> a bit impatient :)
2020-07-29 18:46:51 <ikke> ggg: That's what we discussed earlier
2020-07-29 18:47:07 <mps> people who need drivers are usually impatient :)
2020-07-29 18:47:17 <ikke> These drivers should ultimately be included in Linux upstream
2020-07-29 18:48:38 <ggg> Don’t get mad, I’m a very beginner on a topic like this :(
2020-07-29 18:48:58 <ikke> ggg: no worry
2020-07-29 18:49:06 <u0jQx9gPyrYg> afaik this driver should be in 5.4 but 5.7 latest...
2020-07-29 18:50:03 <ggg> So is it the fastest to make for myself? However, another question will come soon :). How can I use it in diskless-mode?
2020-07-29 18:50:19 <u0jQx9gPyrYg> if you want to run it on a odroid h2+ then some of the device ids are off in 5.4 i think.
2020-07-29 18:50:48 <ikke> ggg: if you make it as a package and include it in the package cache, it can just be installed like other packages on boot
2020-07-29 18:51:06 <ikke> another option is to commit the kernel module with lbu
2020-07-29 18:53:44 <mps> u0jQx9gPyrYg: I don't see it in 5.7
2020-07-29 18:55:53 <mps> hmm, some refs in drivers/net/ethernet/realtek/r8169_main.c
2020-07-29 18:56:04 <u0jQx9gPyrYg> yeah.
2020-07-29 18:56:16 <ikke> but that looks like a different driver, not?
2020-07-29 18:56:37 <u0jQx9gPyrYg> but i heard theres an A and a B revision. and those are not compatible. so maybe i'm wrong about support in 5.4 for the B revision
2020-07-29 18:57:16 <c705> for mkimage.sh (aports/scripts), how can I add modules to the initramfs? Is it enough to add /etc/modules and /etc/modules-conf.d/ to the apkovl I have alongside my mkimg script?
2020-07-29 18:58:42 <ikke> /etc/mkinitfs/mkinitfs.conf and /etc/mkinitfs/features.d/
2020-07-29 18:58:43 <u0jQx9gPyrYg> btw there's a howto and exra info on the odroid wiki for this device: https://wiki.odroid.com/odroid-h2/hardware/install_ethernet_driver_on_h2plus#how_to_install_ethernet_driver_on_h2_plus
2020-07-29 18:59:35 <c705> ikke: I'm using the scripts at the directory I mentioned which ios creating all this in a chroot I believe
2020-07-29 19:00:28 <c705> Are you saying it's enough to have mkinitfs.conf on my host? Will these scripts read the entries in the file on my host even if they're building in a chroot?
2020-07-29 19:00:42 <mps> no mentioning of B version in drivers/net/ethernet/realtek/r8169_main.c, only A
2020-07-29 19:00:57 <ikke> c705: not sure if it uses the one on the host
2020-07-29 19:00:59 <ikke> I don't think so
2020-07-29 19:01:25 <fyr> ikke, sry modprobe complains that the module can't be found despite it existing in the path
2020-07-29 19:02:54 <ikke> c705: in the profiles it defines what features are incldued
2020-07-29 19:03:01 <ikke> but I guess they come from the package
2020-07-29 19:03:03 <c705> ikke: i didn't think it did which is why I asked if having the same contents in my apkovl would be enough
2020-07-29 19:03:49 <c705> ikke: yes, correct, but I specifically need a particular module with a particular config in the initramfs. I am hoping it's enough to include it in the apkovl file (i'll try tonight)
2020-07-29 19:04:23 <c705> high level question i guess is: is the apkovl file applied inside the chroot before the initramfs is built 9i'm hoping yes)
2020-07-29 19:05:09 <ikke> I don't know enough about this to be able to confirm
2020-07-29 19:05:51 <ikke> I assume you mean the initramfs that's part of the iso?
2020-07-29 19:06:13 <c705> ikke: is there some other way I can more easily make my own isos? Eventually I was going to make my own script for this (on the list), but if there's some other better documented way, I'm all ears
2020-07-29 19:06:19 <c705> ikke: yes, correct
2020-07-29 19:06:47 <ikke> c705: I'm only aware of that mkimage.sh script
2020-07-29 19:06:56 <ggg> Jesus guys, push very well! But I didn't understand a word of it anymore :)
2020-07-29 19:07:45 <c705> ikke: yeah, me too. I wish it was better documented. One day, i'll work on acghieving that (also on the list ;-) )
2020-07-29 19:08:01 <c705> maybe i'll learn to type properly as well..let's see
2020-07-29 19:09:22 <ikke> That's something we all could probably do :P
2020-07-29 19:14:47 <mps> ggg: if no one make it remind me on weekend, I will try to find some time for it
2020-07-29 19:15:09 <ikke> mps: did you solve that build error?
2020-07-29 19:15:17 <ikke> sounds like some header is missing
2020-07-29 19:16:46 <mps> no, I didn't and I stopped there
2020-07-29 19:17:07 <mps> should pass -Wno-error to gcc
2020-07-29 19:17:48 <ikke> ok
2020-07-29 19:18:05 <mps> it is implicit-function-declaration warning
2020-07-29 19:18:16 <ikke> yes, which means the function prototype is missing, right?
2020-07-29 19:18:28 <mps> I think so
2020-07-29 19:19:50 <mps> I added pciutils-dev and linux-headers but didn't helped
2020-07-29 19:19:57 <ikke> https://www.kernel.org/doc/html/latest/driver-api/pci/pci.html
2020-07-29 19:20:41 <ggg> I don't even know what to say :). Thank you very much, though I was very afraid it would be much harder or slower. Thank you very much and very grateful.
2020-07-29 19:20:43 <mps> hmm, so source doesn't include proper headers?
2020-07-29 19:22:45 <ikke> mps: where did you get this error?
2020-07-29 19:22:59 <mps> with 'make'
2020-07-29 19:23:06 <ikke> I mean, where in the source
2020-07-29 19:23:08 <mps> in src subdir
2020-07-29 19:25:07 <mps> ikke: https://tpaste.us/DDdd
2020-07-29 19:38:23 <xopowo> dayum!!
2020-07-29 19:38:31 <xopowo> alactritty fast as hell!!
2020-07-29 19:40:14 <ikke> mps: include/linux/pci.h:int __must_check pci_set_mwi(struct pci_dev *dev);
2020-07-29 19:40:36 <ikke> that's for linux 5.4, so it should be defined
2020-07-29 19:41:41 <xopowo> my /etc/defaults/grub.conf disappeared with latesnt edge upgrade
2020-07-29 19:41:46 <xopowo> any thoughts?
2020-07-29 19:42:53 <ikke> xopowo: I don't see any package that included that file
2020-07-29 19:43:34 <ikke> xopowo: you mean /etc/default/grub?
2020-07-29 19:43:55 <alot> afontain_: yeah, the reason I'm interested in alpine rather than arch is specifically for the minimality
2020-07-29 19:44:04 <alot> namely, I want to only install the things that I specifically want
2020-07-29 19:44:12 <alot> it's going to be… a very minimal system
2020-07-29 19:46:09 <xopowo> ikke: exactly, I did had modyfied it but with the latest upgrade it has vanished
2020-07-29 19:46:32 <ikke> xopowo: is grub still installed?
2020-07-29 19:46:46 <ikke> And I find it unlikely that the upgrade removed it, especially if you modified it
2020-07-29 19:46:48 <xopowo> dunno, isn't the default bootloader?
2020-07-29 19:47:16 <ikke> xopowo: depends, syslinux is the default in most cases. grub is mostly used when you have uefi
2020-07-29 19:48:03 <ikke> xopowo: https://pkgs.alpinelinux.org/contents?file=&path=%2Fetc%2Fdefault&name=grub&branch=edge&repo=main&arch=x86
2020-07-29 19:48:06 <ikke> it still exists
2020-07-29 19:48:38 <xopowo> ikke: can I add kernel command option to syslinux?
2020-07-29 19:52:39 <ikke> xopowo: /etc/update-extlinux.conf
2020-07-29 19:53:52 <mps> ikke: our linux-headers missing PCIE_LINK_STATE_* declarations
2020-07-29 19:54:42 <mps> at least on armv7
2020-07-29 19:55:08 <xopowo> ikke: thanks, apparently that's what's installed, since update-extlinux.conf is there
2020-07-29 19:55:24 <mps> хорошо :)
2020-07-29 19:55:43 <ikke> xopowo: also check /boot
2020-07-29 19:55:56 <xopowo> for what?
2020-07-29 19:56:16 <ikke> mentions of grub or ext/syslinux
2020-07-29 20:18:44 <ikke> mps: seems like the Makefile does not support MODULES_DIR / KVER
2020-07-29 20:19:31 <ikke> hmm, BASEDIR
2020-07-29 20:19:59 <mps> yes, needs changes or passing through make
2020-07-29 20:20:30 <mps>  BASEDIR := /lib/modules/5.4.54-0-lts
2020-07-29 20:21:35 <mps> #include "/usr/src/linux-headers-5.4.54-0-lts/include/linux/pci.h"
2020-07-29 20:21:40 <ikke> it just builds for me
2020-07-29 20:22:02 <mps> nice :)
2020-07-29 20:22:10 <mps> I will stop then
2020-07-29 20:22:11 <ikke> Did you use the one from TallGuy74?
2020-07-29 20:22:36 <mps> no, from arch linux source in PKGBUILD
2020-07-29 20:23:03 <ikke> ah ok
2020-07-29 20:23:13 <ikke> I used that one
2020-07-29 20:23:34 <mps> good, u0jQx9gPyrYg was right about it
2020-07-29 20:24:45 <ikke> But it tries to install it to /
2020-07-29 20:25:06 <ikke> but I'll just manuall install it
2020-07-29 20:25:12 <ikke> just make modules
2020-07-29 20:28:03 <ikke> mps: it uses kernels modules_install, do you happen to know how to override where it installs?
2020-07-29 20:29:29 <mps> yes, look at linux-lts or linux-edge, I can't tell from the head
2020-07-29 20:30:06 <mps> make INSTALL_MOD_PATH="${pkgdir}" modules_install
2020-07-29 20:30:18 <ikke> right
2020-07-29 20:30:37 <ikke> Wonder if that works from an outer Makefile invocation
2020-07-29 20:31:09 <mps> it should, I use it from some scripts to build custom kernels
2020-07-29 20:31:16 <ikke> ok
2020-07-29 20:32:15 <mps> but iirc $pkgdir should be absolute path
2020-07-29 20:32:53 <ikke> it is
2020-07-29 20:33:32 <mps> hmm, not needed, I have this 'pkgdir=".."' in one script
2020-07-29 20:33:57 <ikke> nice, it works
2020-07-29 20:34:51 <ikke> I have an APK now :)
2020-07-29 20:34:55 <mps> you can look at my armv7 and aarch64 lxc's under ~/kernel dir
2020-07-29 20:34:59 <mps> :)
2020-07-29 20:35:54 <ikke> ggg: https://tpaste.us/BjZb
2020-07-29 20:36:07 <mps> huh, ncopa will start to harass you :P
2020-07-29 20:36:19 <ikke> I don't say I submit it
2020-07-29 20:36:47 <mps> someone probably will
2020-07-29 20:37:05 <ikke> We need something like dkms
2020-07-29 20:37:14 <mps> I agree
2020-07-29 20:38:07 <mps> but I'm too tired to think anything, going afk
2020-07-29 20:38:18 <ikke> o/
2020-07-29 20:53:39 <xopowo> sway pros around?
2020-07-29 20:53:47 <xopowo> is there a way to change resolution?
2020-07-29 20:54:06 <ikke> never used sway before
2020-07-29 20:54:15 <xopowo> it's kinda neat
2020-07-29 20:54:41 <xopowo> ikke: Xorg crashing hard with latest edge upgrade, something about timer or timing
2020-07-29 20:54:45 <xopowo> hold on
2020-07-29 20:55:14 <xopowo> (EE) client bug: timer event2 debounce short: scheduled expiry is in the past (-3
2020-07-29 20:55:17 <xopowo> ms), your system is too slow
2020-07-29 20:55:24 <ikke> I don't think sway (the window manager) manages the resolution of the screen
2020-07-29 20:55:28 <xopowo> (EE) client bug: timer event2 debounce short: scheduled expiry is in the past (-3
2020-07-29 20:55:38 <xopowo> ikke: not, its unrelated
2020-07-29 20:55:49 <xopowo> resolution issue is for sway.
2020-07-29 20:56:05 <xopowo> sway working fine, latest upgrade broke hard xorg
2020-07-29 20:56:18 <xopowo> in desperation I got sway up and running.
2020-07-29 20:56:43 <xopowo> some discoveries such as alactritty butter smooth rendering
2020-07-29 20:59:10 <vejetary1> xopowo: Little bit Searxing let me to this: https://www.reddit.com/r/swaywm/comments/fskmd6/how_to_change_screen_resolution/ and from there, to this: https://github.com/swaywm/sway/wiki#display-configuration
2020-07-29 20:59:35 <vejetary1> I hope it is enough for you as I can't test it right now. :D
2020-07-29 21:00:08 <xopowo> vejetary1: I forgot to mention that I already got that on sway conf
2020-07-29 21:00:20 <ikke> u0jQx9gPyrYg: https://tpaste.us/BjZb this is the APKBUILD for that driver btw
2020-07-29 21:01:06 <vejetary1> u0jQx9gPyrYg: Did you accidentally set your password as username or something? :D
2020-07-29 21:01:21 <ikke> vejetary1: some people like random usernames
2020-07-29 21:01:38 <vejetary1> ikke: I was just kidding, lol.
2020-07-29 21:01:41 <xopowo> vejetary1: actually thanks because I had the output with the wrong caps
2020-07-29 21:01:57 <vejetary1> xopowo: You're more than welcome, my good sir.
2020-07-29 21:02:08 <xopowo> it was autogenerated by firefox
2020-07-29 21:02:33 <xopowo> let's see if tmux survives killing sway
2020-07-29 21:02:42 <ikke> it should
2020-07-29 21:03:17 <vejetary1> IT will.
2020-07-29 21:03:25 <vejetary1> But don't kill the server or the session.
2020-07-29 21:03:53 <vejetary1> Just close your terminal through its "Close icon" or close shortcut/whatever, you get the idea.
2020-07-29 21:04:07 <vejetary1> Or just do `C-b d` to detach (default keys).
2020-07-29 21:04:27 <bougyman> it will on alpine, probably.
2020-07-29 21:04:39 <bougyman> it wouldn't on systemd without modifications in how its run.
2020-07-29 21:04:58 <vejetary1> I didn't know that.
2020-07-29 21:05:17 <vejetary1> Well, I don't use systemd on any of my machines. :D
2020-07-29 21:05:27 <bougyman> I have to run ubuntu on my $dayjob laptop.
2020-07-29 21:05:32 <bougyman> so I have systemd-run --user --remain-after-exit tmux -S /tmp/0 new-session -d zsh stuff all over.
2020-07-29 21:06:02 <ikke> bougyman: I run tmux in archlinux, and I have never run into that
2020-07-29 21:06:30 <vejetary1> ikke: Maybe Arch has some saner defaults for systemd.
2020-07-29 21:06:38 <bougyman> ikke: you probably have KillUserProcesses=no in logind.conf
2020-07-29 21:06:49 <bougyman> There's a few ways to do it. None of them are "normal" defaults for systemd.
2020-07-29 21:06:55 <vejetary1> Like I heard that Arch does not use Google's NTD servers by default (which systemd does).
2020-07-29 21:07:02 <ikke> bougyman: right, it's commented out (so default)
2020-07-29 21:07:10 <bougyman> ikke: then that shouldn't work.
2020-07-29 21:07:24 <bougyman> If you start a tmux in an x session and kill that x session, the tmux should die.
2020-07-29 21:07:31 <bougyman> It has to, systemd kills the whole process group
2020-07-29 21:07:47 <ikke> Yes, I'm aware of that, but iirc, tmux has always survived for me
2020-07-29 21:07:52 <bougyman> wild
2020-07-29 21:08:27 <bougyman> I've seen this exact problem crop up in #archlinux (and #ubuntu, and #systemd) many times.
2020-07-29 21:08:33 <ikke> hmm
2020-07-29 21:08:58 <bougyman> which is all offtopic because alpine doesn't use systemd. Sorry I'm ranting.
2020-07-29 21:08:58 <xopowo> bougyman: LAWL!!!
2020-07-29 21:09:19 <xopowo> it shouldn't kill it
2020-07-29 21:09:23 <vejetary1> bougyman: Good to know, though. Thanks.
2020-07-29 21:09:26 <bougyman> xopowo: I thought it should?
2020-07-29 21:12:17 <ikke> basically this: https://unix.stackexchange.com/questions/490267/prevent-logoff-from-killing-tmux-session, but I have no idea why it's working for me
2020-07-29 21:13:14 <xopowo> so sway didn't kill it
2020-07-29 21:13:20 <xopowo> bougyman: mm.. are you young?
2020-07-29 21:14:03 <xopowo> the idea is not to kill your work and being able  resume under any other circumstances
2020-07-29 21:14:17 <xopowo> bougyman: extremely important, systemd is doing it wrong again
2020-07-29 21:14:42 <ikke> also relevant: https://github.com/tmux/tmux/issues/428
2020-07-29 21:15:07 <bougyman> xopowo: no, not young.
2020-07-29 21:15:57 <bougyman> I might be the oldest person in this channel.
2020-07-29 21:16:47 <u0jQx9gPyrYg> ikke:thanks! very cool
2020-07-29 21:16:49 <xopowo> bougyman: ok, it wasn't meant to demean or anything. But I thought a young person would not know the purpose of screen/tmux/dtach
2020-07-29 21:17:10 <bougyman> Of course I do. That's why the systemd behavior was so odd.
2020-07-29 21:17:20 <bougyman> I don't normally use systemd, so when it happened to me I was surprised.
2020-07-29 21:17:33 <ikke> It's a different philosophy
2020-07-29 21:17:36 <bougyman> I follow #systemd, cause I have to know what's up with it, but I generally don't use it.
2020-07-29 21:18:12 <bougyman> I use voidlinux on most things. But my dayjob mandates ubuntu and amazon linux 2, and then we use alpine containers.
2020-07-29 21:18:32 <ikke> This user finds it for example completely reasonable that everything gets killed: https://github.com/tmux/tmux/issues/428#issuecomment-222353206
2020-07-29 21:19:02 <u0jQx9gPyrYg> ikke: are you sure it's the same i was looking at? i'm into the rtl8125b - this looks like something else
2020-07-29 21:19:38 <ikke> u0jQx9gPyrYg: it is at least based on the repo that you linked to
2020-07-29 21:27:19 <u0jQx9gPyrYg> i linked to this: https://github.com/TallGuy74/r8125 - the apk is based on the other repo linked by someone else: https://github.com/kac89/r8125
2020-07-29 21:28:36 <ikke> u0jQx9gPyrYg: "source="$pkgname-$_commit.tar.gz::https://github.com/TallGuy74/r8125/archive/$_commit.tar.gz""
2020-07-29 21:56:16 <u0jQx9gPyrYg> ooooh, ikke i only parsed until this line: url="https://github.com/tomaspinho/rtl8821ce"
2020-07-29 22:00:40 <ikke> yeah, I accidentally left that from the package I copied
2020-07-29 22:04:13 <hbug__> is there already a fix in the work to make libvirt not hang? > https://gitlab.alpinelinux.org/alpine/aports/-/issues/11602
2020-07-29 22:04:42 <hbug__> i found the workarounds, but is there a discussion about a serious fix? upstream seems not to care at all.
2020-07-29 22:11:02 <afontain_> xopowo: a way to change resolution is wdisplay, packaged. There is also the command line wrender
2020-07-29 23:54:56 <c705> hbug__: correct, upstream doesn't care to write to POSIX standards
2020-07-29 23:55:15 <c705> though I have not noticed the bug myself in a few days for whatever reason
2020-07-30 03:47:01 <PhoenixMage> mtneug: Are you using cloud-init to install any packages (like ansible for the next level config)?
2020-07-30 06:57:44 <mtneug> PhoenixMage: I'm not entirely sure if I ever did this with Alpine.
2020-07-30 06:58:23 <mtneug> I use Packer to build Alpine VM images and install all packages in that way
2020-07-30 07:00:20 <mtneug> in regular intervals, the VM is then replaced with a new image (i.e. immutable infrastructure)
2020-07-30 07:01:37 <maldridge> mtneug: what is the hueristic you drive machine replacement with?
2020-07-30 07:04:28 <TBB> hbug___: libvirt has been discussed some days back, and while fixing libvirt and maintaining it would be possible, it takes effort and it's a problem with limited dev resources... would be better to send patches upstream, but like you said, upstream doesn't care
2020-07-30 07:05:20 <TBB> one of the devs had a good understanding of what kind of care libvirt needs tho
2020-07-30 07:29:28 <mtneug> maldridge: this approach is relatively new for me. Currently, I build a new VM image in CI every week and replace the VM using Terraform.
2020-07-30 07:31:09 <mtneug> The Alpine VM is more or less a test for this approach. It is not running critical user-facing services.
2020-07-30 07:37:03 <PhoenixMage> mtneug: All good I found the packages module on cloud-init, worked a treat
2020-07-30 08:07:55 <maldridge> ah, a bit harsh for replacement, but does work
2020-07-30 08:35:13 <mtneug> PhoenixMage: you can always use runcmd for custom actions not covered declaratively
2020-07-30 08:38:39 <mtneug> maldridge: would you recommend something else? I think Terraform supports rolling updates and a properly configured load balancer should lead to zero downtime.
2020-07-30 08:53:29 <maldridge> I tend to work with fleets that are large enough where an autoscaller and an orchestrator can cycle machines out over a period of time with no downtime
2020-07-30 08:53:35 <maldridge> but those solutions aren't practical for all use cases
2020-07-30 08:53:38 <telmich> Something is funky with pdflatex since the upgrade today
2020-07-30 08:53:42 <telmich> It's running at 100% cpu and does not return
2020-07-30 08:54:43 <telmich> And it creates a lot of .vrb files
2020-07-30 09:01:07 <telmich> nvm, was a broken org mode file
2020-07-30 09:07:45 <xopowo> afontain_: weird --> ERROR: wdisplays-1.0-r1: BAD signature
2020-07-30 09:08:33 <xopowo> and 'wrender' doesn't exist
2020-07-30 09:09:59 <afontain_> notice that I only said wdisplays was packaged
2020-07-30 09:10:30 <afontain_> not that wrender was too
2020-07-30 09:11:06 <afontain_> The bad signature error is because your mirror is syncing
2020-07-30 09:11:25 <afontain_> try again, it'll work after some tile
2020-07-30 09:11:30 <afontain_> *time
2020-07-30 09:11:45 <xopowo> http://ftp.halifax.rwth-aachen.de/alpine/edge/main
2020-07-30 09:12:43 <xopowo> perhaps I should change mirrors
2020-07-30 09:15:56 <xopowo> afontain_: you seem to be knowledgeable with sway
2020-07-30 09:16:07 <xopowo> is there a way to turn off mouse focus?
2020-07-30 09:21:34 <hbug___> xopowo: what do you mean with mouse focus?
2020-07-30 09:22:57 <afontain_> in `man 5 sway`, see focus_follows_mouse
2020-07-30 09:38:12 <xopowo> hbug___: what afontain just suggested
2020-07-30 09:38:18 <xopowo> I wonder if it's true
2020-07-30 09:53:20 <xopowo> another question does sway support shading?
2020-07-30 09:53:50 <EF> if swap is not automounted with fstab line but works with swapon -a, any suggestions on how do I start to debug it?
2020-07-30 09:54:32 <xopowo> I just realized I don't have a application bar
2020-07-30 09:55:20 <vejetary1> What is the purpose of the Virtual install?
2020-07-30 09:55:27 <vejetary1> installation image*
2020-07-30 09:55:41 <TBB> virtual machines
2020-07-30 09:55:50 <vejetary1> I get it, but like why?
2020-07-30 09:55:54 <xopowo> vejetary1: perhaps optimise, with the modules and all for virtual machines
2020-07-30 09:56:09 <xopowo> vejetary1: give it a try, it's free and quick
2020-07-30 09:56:12 <hbug___> maybe qemu-guest stuff?
2020-07-30 09:56:13 <TBB> the kernel, mostly, it has all the virtual i/o stuff in it
2020-07-30 09:56:13 <vejetary1> Is it something like for VPS, or VirtualBox testing?
2020-07-30 09:56:47 <vejetary1> By VirtualBox, I mean also QEMU and all other programs that I forgot their name.
2020-07-30 09:57:00 <xopowo> vejetary1: there are a bunch of kernel modules that help the vm images perform better.
2020-07-30 09:57:44 <vejetary1> Does VMs do something different that requires specific kernel modules in order to get better performance?
2020-07-30 09:58:10 <vejetary1> And does that cover say VPSes as they are also Virtual Machines?
2020-07-30 09:58:12 <xopowo> vejetary1: yes, such as vhost-net, vhost-socket, etc.
2020-07-30 09:59:02 <xopowo> virtio drivers, qxl display driver, virtio-vga, virtio-net, m$$10.dll own performance options, xen options, virtualbox modules, etc
2020-07-30 09:59:30 <vejetary1> Do you have any possible anwser to my secon question, too? xopowo
2020-07-30 10:00:12 <xopowo> VPS that's a cloud root hosting?
2020-07-30 10:00:30 <vejetary1> Yes.
2020-07-30 10:01:07 <xopowo> vejetary1: then most definately, inform yourself which virtualization is the hosting using and apply those.
2020-07-30 10:01:38 <xopowo> however, most generix images tend to work albeit not in their optimal mode perhaps.
2020-07-30 10:01:49 <xopowo> many distros now offer cloud images.
2020-07-30 10:02:30 <vejetary1> Then is Virtual installation image for cloud computing?
2020-07-30 10:02:50 <xopowo> vejetary1: either inforem which technology does the hosting use, or boot a liveISO and do some  hardware probing.
2020-07-30 10:03:05 <vejetary1> Let me see.
2020-07-30 10:03:53 <xopowo> vejetary1: that's relative and depends on your goals and case scenario.
2020-07-30 10:04:39 <xopowo> vejetary1: since qemu can emulate whole physical hardware, so you want cloud computing using real hardware emulation for whatever reason you can do it.
2020-07-30 10:05:13 <xopowo> vejetary1: boot a liveCD, then some 'dmesg', 'lspci' and the like.
2020-07-30 10:08:39 <vejetary1> It says "KVM emulation". xopowo
2020-07-30 10:11:07 <xopowo> vejetary1: excellent, then the cloud image is worth a shot
2020-07-30 10:11:45 <vejetary1> Which one is the could image one? Virtual one?
2020-07-30 10:13:23 <vejetary1> xopowo: ^
2020-07-30 10:13:53 <xopowo> vejetary1: yeah
2020-07-30 10:14:10 <vejetary1> Thanks a lot. ^^
2020-07-30 10:46:26 <Ashpool> What are some up-to-date tips or guides on how to install Alpine Linux 3.12 as Xen dom0 with UEFI and GRUB? I'm stuck at the point where XEN boots but instead of running Linux it just relinquishes the console and reboots. I can't see anything bad in the log so I don't know where to begin troubleshooting
2020-07-30 10:46:46 <Ashpool> Thing is, the `alpine-xen` iso boots just fine.
2020-07-30 10:54:25 <xopowo> Ashpool: so where's the problem again?
2020-07-30 10:58:43 <Ashpool> After GRUB boots a Xen entry, Xen outputs some log messages on the screen, which to me look like a totally normal startup sequence. After that, I would expect the installed OS to launch. Instead, the screen goes blank and the server reboots, leaving me with no idea on how to debug my configuration.
2020-07-30 10:59:39 <Ashpool> I can show my bootloader config. I can also upload a Xen log, but it's going to be in form of an .avi video, since the whole thing is happening on a remote server to which I have IPMI access only.
2020-07-30 11:18:58 <EF> Okay, so swap needs rc-update add swap!
2020-07-30 12:29:09 <sshine> hello. I'm trying to create a Dockerfile based on the alpine image that installs a specific version of GHC, but doing 'RUN apk add --no-cache ... ghc=8.6.5-r0', I get ERROR: unsatisfiable constraints: ghc-8.8.3-r0: breaks: world[ghc=8.6.5-r0]. but I didn't ask for 8.8.3-r0; this is the newest one in alpine, but not the one I'm looking for. :)
2020-07-30 12:31:01 <sshine> I know there's a Docker image called 'haskell:8.6.5', but this uses Cabal 2.4.0.0 and I specifically need 3.x. I thought I'd go the minimal way and only add stuff, rather than subtract.
2020-07-30 12:42:18 <ikke> We don't provide all versions of ghc
2020-07-30 12:42:52 <ikke> 3.6.5 is only avaible in alpine:3.11
2020-07-30 12:43:13 <ikke> We don't have multiple versions of packages in the same repo
2020-07-30 13:01:56 <sshine> ikke, ahh, so I'd have to use a specific version of alpine to reach a specific version of GHC. good to know.
2020-07-30 13:08:35 <sshine> ikke, thanks. :)
2020-07-30 13:09:06 <ikke> no pronlem
2020-07-30 14:05:24 <ggg> Hey guys! I'm back, I'm the Realtek 8125 NIC dude: P.
2020-07-30 14:18:44 <wener[m]> alpine has mips64, is mips64le or mips64be ?
2020-07-30 14:25:09 <ikke> wener[m]: afaik, it's BE
2020-07-30 14:27:30 <ikke> yes, i'ts BE
2020-07-30 14:41:35 <wener[m]> thanks
2020-07-30 14:53:57 <mps> ggg: did you build it using APKBUILD which ikke prepared
2020-07-30 15:37:31 <ggg> mps I'm sorry, I missed that. Although I’ve never done this on Alpine before, I’m happy to try. Where can I find info and files?
2020-07-30 15:38:37 <mps> ask ikke, he probably have it
2020-07-30 15:39:37 <ikke> https://tpaste.us/BjZb
2020-07-30 16:51:42 <markand> hey it looks like there is something wrong in 3.12
2020-07-30 16:51:45 <markand> apk add gnome ends with:
2020-07-30 16:51:46 <markand> ERROR: unsatisfiable constraints:
2020-07-30 16:51:46 <markand>   so:libvncserver.so.1 (missing):
2020-07-30 16:51:47 <markand>     required by: gnome-remote-desktop-0.1.8-r0[so:libvncserver.so.1]
2020-07-30 16:52:51 <markand> or maybe it's the mirror, let's try another
2020-07-30 17:18:56 <ikke> Looks like a build / package issue with libvncserver
2020-07-30 17:19:11 <ikke> https://pkgs.alpinelinux.org/contents?file=libvncserver.so.1&path=&name=&branch=v3.12
2020-07-30 17:19:39 <ikke> It's installed under lib64
2020-07-30 17:29:20 <ikke> markand: I think this should fix it: https://tpaste.us/WRXY
2020-07-30 17:30:35 <ikke> strange that edge does not have this issue
2020-07-30 17:31:09 <ikke> aha
2020-07-30 17:31:16 <ikke> 10f8ef02b1e254a838c16ce53a39abd9ba341418
2020-07-30 17:33:54 <ikke> markand: I backported that fix
2020-07-30 17:33:59 <ikke> should be fixed soon
2020-07-30 20:53:42 <oliv3r[m]> Hey Guys, I know I asked this before :) but what was the 'bad signature' reason on the main alpine cdn again? I see the package was rebuilt on the 22nd (memtester in this case) on edge/testing and while it's fine on a diff mirror, the cdn still shows it as bad. Is there anything we can do to help 'fix' this bad signature errors on the CDN?
2020-07-30 21:06:20 <Cogitri> https://lists.alpinelinux.org/~alpine/devel/%3C20200721171650.48fa63a4%40ncopa-desktop.copa.dup.pw%3E
2020-07-30 21:18:29 <oliv3r[m]> ah right, the checksum cache ... how long is that cache going to be used for. I reccon 7 days is quite long already?
2020-07-30 21:19:03 <oliv3r[m]> so nothing we can do to help; but having a mirror be 'wrong' for a week is quite annoying :)
2020-07-30 21:31:49 <Cogitri> Ah well, it wasn't that way for a week, but as we rebuild packages it keeps getting invalidatrd for the new packages
2020-07-30 21:32:03 <Cogitri> You could use a non CDN mirror for now, then it should work
2020-07-30 21:32:07 <ggg> ikke: Hi Ikke! I tried to compile based on APKBUILD, but unfortunately it ran into an error. What have I done wrong?
2020-07-30 21:32:19 <ggg> https://tpaste.us/gkDm
2020-07-30 21:32:44 <Cogitri> hopefully we'll work out something with fastly soon to so this doesn't happen again, sorry for the inconvenience, oliv3r[m]
2020-07-30 21:35:39 <mps> ggg: you run it as root?
2020-07-30 21:38:06 <ggg> mps: Yes, why? Is that wrong?
2020-07-30 21:38:44 <mps> yes, it is wrong
2020-07-30 21:39:06 <mps> first clone git.alpinelinux.org/aports
2020-07-30 21:39:17 <mps> cd aports/testing
2020-07-30 21:39:52 <mps> mkdir rtl8125-lts
2020-07-30 21:40:05 <mps> cd rtl8125-lts
2020-07-30 21:40:19 <mps> copy ikkes APKBUILD there
2020-07-30 21:40:28 <mps> then run 'abuild -r'
2020-07-30 21:40:50 <mps> you will probably need to fix _KVER in APKBUILD
2020-07-30 21:42:57 <TBB> I have said this before, but that a repository can end up in that state sounds like a bad method used to sync
2020-07-30 21:43:43 <TBB> and I've had to deal with that problem in a previous project of mine, basically had to come up with another sync method for package distribution
2020-07-30 21:46:50 <markand> ikke, thanks :)
2020-07-30 21:46:56 <mps> ggg: you want this module for armv7 ?
2020-07-30 21:47:36 <mps> or x86_64?
2020-07-30 21:48:20 <ggg> mps: x86_64
2020-07-30 21:48:40 <mps> aha, ok. it still fails on armv7
2020-07-30 21:49:17 <ggg> mps: What is the correct _kvar value? uname -r is ok?
2020-07-30 21:51:16 <mps> no
2020-07-30 21:52:19 <mps> if you run abuild -r it will tell you what version you should set
2020-07-30 21:53:22 <mps> for current aports, _kver=5.4.54, _krel=1
2020-07-30 21:54:05 <ggg> ok, i understand. i rewrite _krel from 0 to 1, thanks
2020-07-30 21:55:31 <mps> and you are building on alpine edge not 3.12-stable?
2020-07-30 21:56:21 <ggg> grep -i version /etc/os-release
2020-07-30 21:56:27 <ggg> VERSION_ID=3.12.0
2020-07-30 21:56:58 <ggg> cat /etc/apk/repositories #/media/sda/apkshttp://dl-cdn.alpinelinux.org/alpine/v3.12/main
2020-07-30 21:57:09 <mps> then the _kver and _krel will be something else
2020-07-30 21:58:23 <mps> uhm, that will require more work, 'git checkout 3.12-stable' in aports iirc
2020-07-30 21:59:15 <ggg> Sorry, i told you, i'am a very noob :(
2020-07-30 22:00:02 <mps> someone more familiar with git than me could help
2020-07-30 22:01:27 <mps> it is not important if you are inexperienced till you are ready to try and learn
2020-07-30 22:01:45 <mps> we all were novice one day :)
2020-07-30 22:02:25 <ggg> https://tpaste.us/evoa
2020-07-30 22:03:48 <ggg> mps: The value of _kver had to be rewritten to 5.4.43 and then it ended :)  . What do you think?
2020-07-30 22:04:07 <mps> :)
2020-07-30 22:04:21 <mps> success, congrats
2020-07-30 22:06:04 <ggg> Thanks :)  . And now?
2020-07-30 22:06:11 <mps> look under ~/testing/x86_64 dir
2020-07-30 22:07:21 <mps> sudo apk add --allow-untrusted ~/path_to_apk_filename
2020-07-30 22:10:25 <ggg> sudo apk add --allow-untrusted packages/testing/x86_64/rtl8125-lts-5.4.43-r1.apk [sudo] password for ggg: (1/1) Installing rtl8125-lts (5.4.43-r1)Executing busybox-1.31.1-r19.triggerExecuting kmod-27-r0.triggerOK: 1051 MiB in 180 packages
2020-07-30 22:10:53 <ggg> It looks like it went well too :)
2020-07-30 22:17:24 <c705> hi. couple more questions about mkinitfs. Am I correct in assuming that /etc/modules.d/* is copied to the intramfs via the 'base" feature in mkinitfs/features.d? Also is the order of the features in mkinitfs.conf preserved (ie: the first feature of the list is run first)?
2020-07-30 22:18:12 <mps> c705: I think only modules listed in mkinitfs.conf feature
2020-07-30 22:22:09 <ggg> mps: It also works :). And now what?
2020-07-30 22:22:41 <mps> ggg: sudo depmod
2020-07-30 22:23:01 <mps> sudo modprobe what_is_module_name
2020-07-30 22:23:10 <mps> and good night :)
2020-07-30 22:29:27 <ggg> mps: Thank you very much! And should i do that before each update?
2020-07-31 03:04:14 <amosbird> Hello, how can I use "apk add fio" to get a static binary?
2020-07-31 03:08:21 <adu> some packages have {name}-static
2020-07-31 03:09:12 <amosbird> / # apk add fio-static
2020-07-31 03:09:15 <amosbird> ERROR: unsatisfiable constraints:
2020-07-31 03:09:17 <amosbird>   fio-static (missing):
2020-07-31 03:09:19 <amosbird>     required by: world[fio-static]
2020-07-31 03:09:28 <amosbird> Hmm, is there a docker image that provide static world?
2020-07-31 03:09:56 <amosbird> I thought alpine always provides a way to build static binary...
2020-07-31 03:10:03 <adu> maybe you could compile it yourself
2020-07-31 03:10:43 <adu> amosbird: you might be thinking of golang
2020-07-31 03:11:20 <amosbird> oh
2020-07-31 03:11:25 <adu> almost every single package links with musl dynamically, so I'm not sure where you got that idea
2020-07-31 03:11:34 <amosbird> ok
2020-07-31 03:11:57 <amosbird> could you help me with this fio-static?
2020-07-31 03:12:02 <adu> sure
2020-07-31 03:12:10 <amosbird> I don't think I'm capable of doing that
2020-07-31 03:12:21 <amosbird> https://github.com/axboe/fio
2020-07-31 03:13:39 <adu> You're looking at the wrong thing
2020-07-31 03:13:40 <adu> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/fio/APKBUILD
2020-07-31 03:13:55 <adu> That's the alpine build file
2020-07-31 03:14:07 <adu> What I would do is modify that to build statically
2020-07-31 03:15:02 <amosbird> oh!
2020-07-31 03:15:27 <amosbird> i'm completely new to alpine. Could you help me modify that build?
2020-07-31 03:15:48 <amosbird> the way to build fio staticly is    EXTFLAGS="-static" make
2020-07-31 03:17:25 <adu> what platform are you on?
2020-07-31 03:17:49 <adu> alpine?
2020-07-31 03:18:28 <amosbird> yes
2020-07-31 03:18:37 <amosbird> archlinux  running alpine docker
2020-07-31 03:18:50 <amosbird> I'd like to use alpine to build fio static to run on centos 7
2020-07-31 03:19:09 <adu> So I use dabuild
2020-07-31 03:19:23 <adu> but abuild works from within docker
2020-07-31 03:19:27 <amosbird> hmm, could you clarify?
2020-07-31 03:19:52 <adu> git clone https://gitlab.alpinelinux.org/alpine/aports.git
2020-07-31 03:20:11 <adu> cd aports/community/fio
2020-07-31 03:20:16 <adu> vi APKBUILD
2020-07-31 03:20:22 <adu> abuild
2020-07-31 03:20:35 <adu> wait no, there are deps, so abuild -r
2020-07-31 03:20:44 <amosbird> does it run on archlinux or alpine?
2020-07-31 03:21:22 <adu> alpine
2020-07-31 03:21:38 <adu> if you are doing it from archlinux, try docker-abuild (or dabuild, that's what I use on mac)
2020-07-31 03:22:00 <amosbird> interesting
2020-07-31 03:22:27 <adu> https://github.com/alpinelinux/docker-abuild
2020-07-31 03:26:35 <adu> amosbird: so you might just need to modify the build() function
2020-07-31 03:27:02 <adu> but you may also have to change all it's deps to build statically too
2020-07-31 03:27:42 <amosbird> adu: yeah,...
2020-07-31 03:27:50 <amosbird> but thoes deps...
2020-07-31 03:27:54 <amosbird> I don't know ...
2020-07-31 03:28:40 <adu> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/libaio/APKBUILD
2020-07-31 03:29:31 <amosbird> adu: hmm
2020-07-31 03:29:33 <adu> oh zlib-static exists
2020-07-31 03:29:39 <adu> you wouldn't need to rebuild that
2020-07-31 03:29:55 <amosbird> ok...
2020-07-31 03:30:00 <amosbird> Receiving objects:   0% (580/615522), 140.01 KiB | 6.00 KiB/s
2020-07-31 03:30:05 <amosbird> ..... bad network
2020-07-31 03:31:02 <uplime> oof
2020-07-31 03:33:09 <adu> amosbird: also have you tried https://dl.fedoraproject.org/pub/epel/7/x86_64/Packages/f/fio-3.1-1.el7.x86_64.rpm
2020-07-31 03:37:41 <amosbird> adu: I don't have root ...
2020-07-31 03:37:45 <amosbird> on that centos machine
2020-07-31 04:15:23 <PhoenixMage> Irony at its finest, went to add apk suppot to ansible package_facts, turns out its already added just in a newer version then Alpine packages
2020-07-31 04:22:57 <PhoenixMage> And that would be because someone had the same idea 2 weeks ago :/
2020-07-31 04:25:24 <glats> hi! i installed alpine 3.12 and i couldn't find man command. why isn't shipped with this version? should i use v3.11?
2020-07-31 04:26:04 <PhoenixMage> Its not shipped by default
2020-07-31 04:26:12 <PhoenixMage> apk add man
2020-07-31 04:26:19 <PhoenixMage> Perhaps
2020-07-31 04:26:23 <glats> already did it
2020-07-31 04:26:27 <glats> says missing
2020-07-31 04:26:32 <glats> https://pkgs.alpinelinux.org/packages?name=man&branch=v3.12
2020-07-31 04:26:57 <glats> i couldn't find it there neither
2020-07-31 04:27:39 <PhoenixMage> apk add mandoc
2020-07-31 04:28:11 <glats> oh
2020-07-31 04:28:25 <glats> so then the name was changed
2020-07-31 04:28:27 <PhoenixMage> Changes are that the man page for the command you are trying to man isnt installed either
2020-07-31 04:28:57 <PhoenixMage> tbh I usually google 'man <command>' saves hassle
2020-07-31 04:30:31 <PhoenixMage> glats: If you are after a missing command try 'apk search <command>' or https://pkgs.alpinelinux.org/contents
2020-07-31 04:30:58 <glats> ok. thankj you
2020-07-31 04:31:07 <PhoenixMage> np
2020-07-31 07:08:53 <EF> glats: apk search cmd:man
2020-07-31 08:10:17 <oliv3r[m]> Cogitri no pressure; I know using a different mirror works around it ;) but if there's something technical we can do to help for now, and something technical for the future? e.g. if there's a breaking change, 'for pkg in APKBUILD sed pkgver=n+1; git add/push; done' for example :)
2020-07-31 08:10:55 <Cogitri> Well, we usually always increase pkgrel, so this should be a non-issue normally
2020-07-31 08:11:35 <Cogitri> But we didn't do this for this ports-wide rebuild of packages for 32-bit since that'd have meant that everyone would have had to reinstall their system :)
2020-07-31 08:11:44 <Cogitri> Since we'd have had to bump pkgrel on every package
2020-07-31 09:05:38 <vegankumpir> Is there a length or character limit of a password of a user in Alpine?
2020-07-31 10:01:40 <xopowo> vegankumpir: once on irssi somebody told me that his password was several gigabyets.
2020-07-31 10:02:49 <xopowo> to put it into perspective, how many characters takes to fill 1GB?
2020-07-31 10:03:16 <vegankumpir> GB or GiB?
2020-07-31 10:04:42 <xopowo> regular internet 1GB, like 1GB mkv movie.
2020-07-31 10:04:44 <afontain_> "GB or GiB": answering the real question :D
2020-07-31 10:04:53 <ikke> xopowo: depends on the encoding, but in ascii, each character is 1 byte
2020-07-31 10:05:14 <xopowo> no, he mentioned he used different utf8 codepoints
2020-07-31 10:05:48 <xopowo> so let's say 8kb per unicode character
2020-07-31 10:06:13 <xopowo> sorry 8b
2020-07-31 10:06:56 <afontain_> a bound is "somewhere between 250 000 000 characters and than 1 100 000 000"
2020-07-31 10:07:02 <afontain_> s/than /
2020-07-31 10:07:22 <xopowo> damn you were faster than I
2020-07-31 10:08:12 <afontain_> well, I rounded up because I didn't want to pull out a calculator
2020-07-31 10:08:15 <xopowo> vegankumpir: 250 000 000 characters is enough for you? and that's not the hard limit.
2020-07-31 10:08:47 <ikke> A character could take up multiple unicode codepoints though
2020-07-31 10:08:48 <afontain_> that's also ridiculously big
2020-07-31 10:08:49 <vegankumpir> xopowo: Fairly enough, my good sir. ^^
2020-07-31 10:09:19 <afontain_> to the point it doesn't make it more secure
2020-07-31 10:12:19 <xopowo> afontain_: why not? now it could even be more convenient with NVME, store the /etc/passwd in a separate nvme device.
2020-07-31 10:12:31 <afontain_> (the actual upper bound is 1 073 741 824 chars)
2020-07-31 10:12:43 <xopowo> login 250 000 000 characters effortlessly
2020-07-31 10:12:59 <afontain_> have you heard of hashing before?
2020-07-31 10:13:14 <afontain_> your password is not actually stored on disk
2020-07-31 10:13:17 <afontain_> and that's good for security
2020-07-31 10:13:27 <vegankumpir> afontain_: You mean as plain text?
2020-07-31 10:13:51 <afontain_> or even encrypted
2020-07-31 10:14:06 <vegankumpir> Well, it is stored as hash.
2020-07-31 10:15:04 <xopowo> vegankumpir: so is not that you're short of options
2020-07-31 10:15:13 <afontain_> here's what's stored if your password is "hi" and the system is setup to use a sha512 $6$rATLSwTzGObI9Rnr$gXWC4eyn9sn8bNnT8H9xfyY1dorZt6tRy9Ayv2BZ3aFI0M.lD7WPbux0NIGPUWN99.yhd5ZkGLwmL8d.aEnKR1
2020-07-31 10:15:16 <vegankumpir> I wanted to know.
2020-07-31 10:15:30 <xopowo> ha ha ha!!
2020-07-31 10:16:24 <ikke> due to the salt, it's different for everyone though
2020-07-31 10:16:43 <afontain_> good point
2020-07-31 10:17:06 <afontain_> here the salt is rATLSwTzGObI9Rnr, if I'm not mistaken
2020-07-31 10:22:16 <afontain_> xopowo: why not it's not more secure? because since the hash is 512bits, and you'll have to cram 1GB into 64B, your many 1GB passwords will actually match
2020-07-31 10:23:29 <afontain_> (well, still only one in 2⁵¹² = 13407807929942597099574024998205846127479365820592393377723561443721764030073546976801874298166903427690031858186486050853753882811946569946433649006084096 chance for each)
2020-07-31 10:23:51 <TBB> just type 2^512, no need to be posh about it ;)
2020-07-31 10:25:37 <afontain_> s/your //
2020-07-31 10:25:38 <afontain_> and you don't even force the attacker to try 1GB passwords (which are slower to check), because it *might be* that "hi" produces the same hash for this salt
2020-07-31 10:26:45 <vegankumpir> TBB: lol
2020-07-31 10:31:27 <afontain_> wow, hashing 1GB of data is is actually very fast
2020-07-31 10:31:46 <afontain_> dd if=/dev/urandom bs=1GB count=1 | tr '\n' '\t' | busybox time openssl passwd -6 -noverify -stdin
2020-07-31 10:31:59 <afontain_> sys 0m 0.00s
2020-07-31 10:59:12 <xopowo> afontain_: is it hashing each character or 1GB as continous data?
2020-07-31 11:01:54 <ikke> xopowo: it works with blocks of 512 bits
2020-07-31 11:02:35 <ikke> or 256 for sha256
2020-07-31 11:04:56 <ikke> sorry, for sha256 the block size is 512, for sha512 it's 1024
2020-07-31 12:10:34 <wener[m]> https://github.com/alpinelinux/mkinitfs/blob/master/mkinitfs.in#L197 I think the feature dir should set after the getops, otherwise, this always use the none chroot feature.d
2020-07-31 12:12:01 <wener[m]> sorry, ignore me.
2020-07-31 12:12:41 <wener[m]> need to set -P to override that
2020-07-31 12:27:41 <EF> Is aconf a cli based admin tool for Alpine?
2020-07-31 12:55:06 <azarus_> Anybody else have firefox freeze often?
2020-07-31 12:55:35 <Ganwell> EF: https://pkgs.alpinelinux.org/package/edge/main/x86_64/aconf
2020-07-31 12:58:49 <MY-R> azarus_, is freezing usually after reboot when running first time and website got some audio/video?
2020-07-31 12:59:56 <azarus_> MY-R: indeed
2020-07-31 13:00:43 <MY-R> azarus_, check your "dmesg" if dont have something with IRQ bdl_pos_adj blablabla
2020-07-31 13:01:02 <amosbird> hello
2020-07-31 13:01:19 <Ganwell> EF: I wanted to see what it is, but it has not been moved to gitlab.alpinelinux.org
2020-07-31 13:02:09 <azarus_> MY-R: I'll try to recreate it and check my dmesg.
2020-07-31 13:02:40 <MY-R> azarus_, just grep it and be sure you played onece or twice something on youtube for example
2020-07-31 13:03:49 <Ganwell> EF: it seems to be something web-based: https://git.io/JJaJJ
2020-07-31 13:04:16 <azarus_> MY-R: what would be the resolution, if any?
2020-07-31 13:04:32 <MY-R> azarus_, got something like "hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj"  ?
2020-07-31 13:04:52 <azarus_> MY-R: not yet.
2020-07-31 13:05:14 <azarus_> Ah, so a bigger bdl_pos_adj would fix it?
2020-07-31 13:05:38 <MY-R> problem is that first time when it doing that auto adjust then firefox ye freezing, when close it then still in background sitting some firefox audio process so that is why cant run it without killall firefox
2020-07-31 13:06:26 <Ganwell> @EF: alpine-conf can setup various things, it is used in most how-tos... https://gitlab.alpinelinux.org/alpine/alpine-conf
2020-07-31 13:06:32 <MY-R> azarus_, and yes, need increase it, I was checking some values like 2.2 1.4 etc etc but still same and value 64 done the job, no more freezes and that message in logs
2020-07-31 13:07:28 <azarus_> MY-R: Okay, I'll try that. Thanks!
2020-07-31 13:07:28 <amosbird> how can I use dabuild to fio ?
2020-07-31 13:08:24 <MY-R> azarus_, just create for example /etc/modprobe.d/zzz-alsa.conf and put there "options snd-hda-intel index=0 bdl_pos_adj=64" dont need put index=0 if your card order is just fine
2020-07-31 13:11:41 <amosbird> how can I build https://github.com/axboe/fio statically
2020-07-31 13:12:35 <wsinatra> you'd need a $pkgname-static or $pkgname-dev subpackage in your APKBUILD
2020-07-31 13:12:50 <mps> wsinatra: no
2020-07-31 13:12:50 <wsinatra> but just a rough look, the default hooks for that might pick that up
2020-07-31 13:12:59 <mps> -static is for libs
2020-07-31 13:13:16 <wsinatra> I couldn't remember if it was -static or -dev, thanks for clearing that up mps :)
2020-07-31 13:17:26 <amosbird> wsinatra: please could you help me?
2020-07-31 13:18:40 <mps> wsinatra: this is what I want to change in alpine
2020-07-31 13:20:02 <mps> -static should be static binaries, and static libs should go to -dev
2020-07-31 13:22:33 <wsinatra> I mean, that makes sense if you ask me, it's easier to remember
2020-07-31 13:22:44 <wsinatra> amosbird: do you know if there's an Arch Linux AUR package for it?
2020-07-31 13:22:58 <wsinatra> (I know that seems like an odd question, but if there is it's useful)
2020-07-31 13:23:33 <mps> wsinatra: I can count on your vote when I raise this on ML :)
2020-07-31 13:24:01 <wsinatra> hahaha I'm all for logical changes :)
2020-07-31 13:24:31 <afontain_> IMO, it would be better too. There isn't much sense in having a development tool split (you have the header, but you can't use it? why?)
2020-07-31 13:25:06 <mps> historical mistake, which no one want to start fixing
2020-07-31 13:25:12 <wsinatra> I think it's a call back to the lightweight mindset
2020-07-31 13:25:23 <wsinatra> Always has been that way kind of thing
2020-07-31 13:25:55 <mps> I intentionally don't add -static for libs I packaged
2020-07-31 13:27:04 <mps> though fixing that will not be fast and easy, a lot of pkgs have $libname-static in makedepends
2020-07-31 13:28:33 <wsinatra> Yeah no joke, I'm working on a jemalloc package right now which has it
2020-07-31 13:29:12 <mps> iirc, nothing depends on jemalloc so it is easy
2020-07-31 13:29:38 <mps> wsinatra: look at MRs, someone already working on it
2020-07-31 13:33:36 <wsinatra> Oh really?
2020-07-31 13:34:18 <wsinatra> huh I don't see anything when I search through the MRs for jemalloc
2020-07-31 13:34:46 <wsinatra> got the # for it? Because I already have a working package for it, I just couldn't push anything until my Nagios changes went through
2020-07-31 13:35:02 <wsinatra> (because I'm dumb and don't branch my aports fork before I make MRs..)
2020-07-31 13:35:08 <mps> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/9020
2020-07-31 13:35:57 <mps> but, imo jemmaloc is not good for general use, maybe some specialized apps
2020-07-31 13:36:29 <wsinatra> Oh it's in unmaintained, that makes sense
2020-07-31 13:36:50 <wsinatra> I was trying to port Mastodon specifically. Wanted to build an Alpine Mastodon container
2020-07-31 13:37:00 <mps> especially when dalias releases new musl with malloc-ng
2020-07-31 13:37:24 <wsinatra> but I hate compiling source in docker containers when I can just push packages to aports for everyone. Seems like backwards thinking to not build the package
2020-07-31 13:39:13 <wsinatra> I honestly don't have a usecase for jemalloc outside it being a dep for mastodon haha
2020-07-31 13:41:47 <mps> I don't know how Mastodon works but I doubt it need jemalloc
2020-07-31 13:43:07 <wsinatra> That might be true, I'm mostly basing that guess work off of their official Ubuntu based container
2020-07-31 13:43:30 <wsinatra> The package itself will build without it, so it could just be something inside the container that needs it. I just haven't gotten that fair figuring it out
2020-07-31 13:43:42 <amosbird> wsinatra: what do you mean?
2020-07-31 13:43:52 <amosbird> I have dabuild at hand
2020-07-31 13:44:03 <amosbird> but I don't know how to build a static fio package
2020-07-31 13:45:30 <wsinatra> I was going to suggest using toAPK to convert the AUR package to APKBUILD to start with, it likely would just need some tweaks after that.
2020-07-31 13:46:18 <wsinatra> if you have a subpackages="$pkgname-dev" in your APKBUILD the default_dev() function should pick it up
2020-07-31 13:46:32 <wsinatra> in most cases that is, assuming it needs no tweaks. Are are you not asking about packaging?
2020-07-31 13:46:57 <wsinatra> I forget not everyone in here is actually working to package things, since it's more of a help channel
2020-07-31 13:47:56 <amosbird> .....
2020-07-31 13:48:07 <amosbird> I don't know anything of your sayings ...
2020-07-31 13:48:11 <amosbird> I just need a static fio
2020-07-31 13:48:48 <amosbird> if alpine linux cannot help me build static fio, I will try another distro...
2020-07-31 13:49:12 <wener[m]> I submit fio to alpine
2020-07-31 13:49:40 <amosbird> wener[m]: awesome
2020-07-31 13:49:47 <amosbird> could you help me to build a static fio
2020-07-31 13:49:57 <wener[m]> I think fio only depends on musl
2020-07-31 13:50:04 <ikke> wsinatra: fyi, I also prefer to be able to install packages rathat than compile from source in containers :)
2020-07-31 13:50:08 <wener[m]> Not only musl
2020-07-31 13:50:09 <amosbird> it has a lot deps
2020-07-31 13:50:23 <ikke> s/rathat/rather/
2020-07-31 13:50:34 <wener[m]> you can build anything full static on alpine, but not easy
2020-07-31 13:50:44 <amosbird> wener[m]: ok....
2020-07-31 13:50:57 <amosbird> what's distro is better for this kind of job?
2020-07-31 13:51:02 <wener[m]> last I build a full static core miner, need to change the cmake, make the musl static link
2020-07-31 13:51:28 <wsinatra> @ikke: at least I'm not the only one. That was the justification for the Nagios updates, I didn't realize we were using the 2013 version until I started to dockerize our instance at work
2020-07-31 13:51:36 <wener[m]> alpine is the best choose, because you can static link musl, other disto's libs is not easy to static link
2020-07-31 13:51:45 <amosbird> ok...
2020-07-31 13:51:56 <amosbird> so what's the way of building static fio
2020-07-31 13:52:02 <amosbird> could anyone help me out?
2020-07-31 13:52:22 <amosbird> gentoo has a flag of saying  static-libs, then emerge magically works everything out
2020-07-31 13:52:31 <amosbird> I wish alpine does that too but for executable
2020-07-31 13:52:49 <ikke> amosbird: you can pass --build-static to configure
2020-07-31 13:52:52 <wener[m]> start from https://git.alpinelinux.org/aports/tree/community/fio/APKBUILD?id=cf191c9b5b350effa46c027bc7cd2a348c772fd6 , tweak the flag
2020-07-31 13:52:53 <ikke> Not sure if that's enough though
2020-07-31 13:53:11 <amosbird> ......
2020-07-31 13:53:22 <amosbird> I don't know how to begin
2020-07-31 13:53:31 <ikke> https://github.com/axboe/fio/blob/master/configure#L178
2020-07-31 13:53:37 <amosbird> makedepends="linux-headers zlib-dev coreutils libaio-dev"   doesn't look like static
2020-07-31 13:53:45 <amosbird> and I don't know which to change
2020-07-31 13:54:27 <wener[m]> some -static  will not static link libc, just be carefull
2020-07-31 13:54:34 <wener[m]> google, gcc static link
2020-07-31 13:55:30 <amosbird> I guess this is out of everyone's pay grade T_T. I give up
2020-07-31 13:56:30 <wener[m]> If you don't know how to build a package, then start from that, next step is static build.
2020-07-31 14:02:20 <mps> fully static bins can be built on musl even on other glibc based distros
2020-07-31 14:04:09 <mps> on alpine adding '-static' to CFLAGS is usually enough
2020-07-31 14:36:23 <TBB> there's some work ahead any time you learn to do new things; I found Alpine to be quite packaging friendly
2020-07-31 14:44:18 <wener[m]> I found alpine is the only disto that I can find out and understand almost anything, lol
2020-07-31 15:15:36 <albr> ScottHamilton[d]: Hello I'm trying to use the isso self hosted commenting server in a Docker image. But when I run the container, I get : ```
2020-07-31 15:15:36 <albr> ScottHamilton[d]: ImportError: cannot import name 'SharedDataMiddleware' from 'werkzeug.wsgi' (/usr/lib/python3.8/site-packages/werkzeug/wsgi.py)
2020-07-31 15:15:37 <albr> ScottHamilton[d]: ```. Do you knwo what could go wrong? (the same works on ubuntu image)
2020-07-31 16:54:07 <albr> ScottHamilton[d]: So seems like this SharedDataMiddleware class location changed since release 0.15.6 of werkzeug whereas isso requires werkzeug>=1.0.0
2020-07-31 17:31:39 <WeblordPepe> hey fellas im just sorta looking at LXC and realized its probably more simpler than docker or at least more straight forward
2020-07-31 17:33:38 <WeblordPepe> am i missing something with lxc-create? like im sitting on a debian system and wanting to fire up an alpine container
2020-07-31 17:34:07 <WeblordPepe> i see theres like template names. is this a debian thing? or is lxc-create sniffing for template names from some repo
2020-07-31 17:34:10 <WeblordPepe> i dunno how this works
2020-07-31 17:40:24 <WeblordPepe> oh wow cool so my distro (debian) just has a bunch of templates for lxc-create. fascinating
2020-07-31 17:40:41 <WeblordPepe> zomg it has an alpine one. wonderful
2020-07-31 17:42:10 <WeblordPepe> yikes what a huge script. surely this was created with some tool
2020-07-31 17:44:03 <WeblordPepe> 
ACTION discovers distrobuilder

2020-07-31 17:46:02 <TBB> lxc is nice. I built my own build system on top of lxc once.
2020-07-31 17:49:47 <WeblordPepe> neat
2020-07-31 17:50:27 <WeblordPepe> im wanting to setup a doom server in a lxc for funsies and learning. sorta fell off the docker boat cos it seems people just use it for shipping messy programs
2020-07-31 17:50:48 <WeblordPepe> 'it works on my computer' so they ship their whole computer, instead of making lean portable applications
2020-07-31 17:50:54 <WeblordPepe> thats how i see it anyway
2020-07-31 17:55:17 <matthewcroughan> Hey
2020-07-31 17:55:31 <matthewcroughan> So what's the recommended way of overlaying changes on top of alpine, if I wanted something to run baremetal?
2020-07-31 17:56:10 <matthewcroughan> Or if I wanted to provide something to install on top of Alpine, a collection of packages, that will remain working forever.
2020-07-31 17:56:48 <matthewcroughan> I could provide them a non-specific bash script that does `apk add thing thing2 thing3` but that seems like it would be a mistake
2020-07-31 17:57:23 <matthewcroughan> Is the right answer to use Docker and provide them with a dockerfile?
2020-07-31 17:58:16 <matthewcroughan> Not only a Dockerfile, sorry. I mean a Dockerfile source PLUS an image they can download from the hub, and then it doesn't even matter what distro I'm on.
2020-07-31 18:06:20 <wsinatra> mattewcroughan: I think the multiarch/qemu package does that
2020-07-31 18:06:26 <wsinatra> err docker container, not package
2020-07-31 18:07:32 <matthewcroughan> wsinatra Well it's kind of pointless what distribution you choose then though, isn't it :P
2020-07-31 18:08:34 <matthewcroughan> The reason being is that I want to make a video "Making a VST Workstation with Alpine Linux + Docker on an Atomic Pi"
2020-07-31 18:09:36 <matthewcroughan> If I made an Alpine based container that did all of this, they could probably run it just fine on any distribution right? Kinda sad that distros don't matter on the host anymore, in a way :D
2020-07-31 18:10:02 <wsinatra> Oh I wasn't suggesting that you use the qemu package, but it produces binfmt mounts on the host and populates them with emulators, which sounds ISH like what you wanted
2020-07-31 18:10:15 <matthewcroughan> Hmm, could you explain that a bit more?
2020-07-31 18:10:16 <matthewcroughan> binfmt?
2020-07-31 18:10:22 <wsinatra> or I interpreted it that way, since you said you wanted to overlay packages on top of alpine
2020-07-31 18:10:28 <wsinatra> I guess I assumed host
2020-07-31 18:10:36 <matthewcroughan> I want to do stuff on Alpine that is going to work in 2025
2020-07-31 18:10:49 <matthewcroughan> I want to be able to abandon the project and just have it work so that people can pick it up
2020-07-31 18:10:58 <wsinatra> https://hub.docker.com/r/multiarch/qemu-user-static/
2020-07-31 18:11:06 <wsinatra> this is the container I was referencing
2020-07-31 18:11:52 <wsinatra> Since docker containers are typically pulled from dockerhub, assuming docker doesn't stop existing, you should be okay to build and push. Though anything 5 years old with no updates would be pretty security poor haha
2020-07-31 18:12:09 <matthewcroughan> Yeah this isn't about security, this is about making some things work
2020-07-31 18:12:18 <wsinatra> Can you give me more details on the project? Maybe if I understand more I can make better suggestions :)
2020-07-31 18:12:36 <matthewcroughan> I have a board, x86. I want to run Alpine on it, apk add a bunch of packages, then show people how easy that was
2020-07-31 18:12:47 <matthewcroughan> But I want this to work well into the future without providing them an ISO
2020-07-31 18:13:19 <matthewcroughan> I don't wanna have to update things, although I will.
2020-07-31 18:13:28 <matthewcroughan> This is why I'm saying, docker images are fine, right?
2020-07-31 18:13:47 <WeblordPepe> hey cool im looking at VSTs in linux too
2020-07-31 18:13:53 <matthewcroughan> if the hub goes away, sure, screwed. But for as long as that's around, my static build will be too?
2020-07-31 18:14:18 <matthewcroughan> but then if Github goes away, so does my project, so yknow, we can only get so redundany
2020-07-31 18:14:24 <matthewcroughan> bigger problems arise if either go
2020-07-31 18:15:00 <matthewcroughan> I mean, is it possible to use docker to produce an ISO someone could flash to a USB stick?
2020-07-31 18:15:02 <WeblordPepe> i dont use github. got my own server
2020-07-31 18:15:25 <matthewcroughan> ya but with git it doesn't matter, since you can use both github and your own server as remotes and just sync to both
2020-07-31 18:15:32 <WeblordPepe> yeah
2020-07-31 18:15:40 <matthewcroughan> more the better with git
2020-07-31 18:15:48 <WeblordPepe> i like your thinking though i also want stuff to last for decades
2020-07-31 18:16:05 <WeblordPepe> nothing has enough longevity anymore
2020-07-31 18:16:17 <matthewcroughan> It's hard. The only way to make that possible is to have an artifact like an ISO that does not rely upon services or registries like the docker hub
2020-07-31 18:16:51 <WeblordPepe> when i make programs that need libraries i statically link in the libraries -often i copy the source code of the library itself into my program
2020-07-31 18:16:59 <matthewcroughan> This way, if the hub or github go down, you've still got that build artifact, hopefully stored somewhere with the project sources
2020-07-31 18:17:17 <matthewcroughan>  Yes, exactly, it's just like that. I want a static build of my distribution that I'm building on top of Alpine.
2020-07-31 18:17:46 <wsinatra> matthewcroughan: if you look in gitlab.alpinelinux.org under the aports repo there's a directory labelled scripts which contains all of the scripts used to build the ISO images for alpine
2020-07-31 18:18:10 <WeblordPepe> i feel that all this docker stuff is gonna bite people in the backside. you wanted to ship a program ya made, but only made it available as some image that requires a balls load of other things to even go
2020-07-31 18:18:20 <wsinatra> so yes, what you're wanting is probably feasible. The exact logistics I'm not certain of it, but the tooling to at least build custom Alpine ISOs is there
2020-07-31 18:18:36 <matthewcroughan> Yes, and that's not easy
2020-07-31 18:18:56 <WeblordPepe> every OS i install expect 10 year out of it
2020-07-31 18:18:59 <matthewcroughan> Whereas if I have a docker image, it's easy. I just need to figure out if it's possible to store and host those artifacts/images myself.
2020-07-31 18:19:12 <matthewcroughan> I can host my own docker hub, but what do the resultant builds look like? Tarballs?
2020-07-31 18:19:35 <wsinatra> Well if you were hosting your own dockerhub, but the end goal was shipping an ISO to people
2020-07-31 18:19:39 <wsinatra> why not just host the ISO too?
2020-07-31 18:19:48 <wsinatra> and use the docker containers as CI/CD builders
2020-07-31 18:19:56 <wsinatra> that's more the intention of the system anyways
2020-07-31 18:19:57 <matthewcroughan> It's all too complex, I do not plan on hosting my own dockerhub, I'm stating that I *can*
2020-07-31 18:20:22 <matthewcroughan> I'm saying that if I can just provide the image in the case of the hub going down, and my Dockerfiles don't work anymore since alpine has updated, then things will be fine
2020-07-31 18:20:54 <matthewcroughan> 10 years from now, my Dockerfile that says `FROM alpine:3.1.2 RUN apk add ardour6` will cease to function, right?
2020-07-31 18:21:07 <matthewcroughan> But the image that was built when that worked should still be something I can give to people, just like the iso
2020-07-31 18:21:26 <ikke> depends if docker hub still exists :)
2020-07-31 18:21:35 <matthewcroughan> Does it? Why?
2020-07-31 18:21:36 <wsinatra> If you have a built container that people can pull down, that would work. Attempting to rebuild it wouldn't work if you couldn't get the host alpine:3.12 to build on
2020-07-31 18:21:53 <matthewcroughan> Docker builds OCI container images, right? Can't I just distribute that OCI image for usage with whatever tool? E.g Podman, Buldah
2020-07-31 18:21:55 <ikke> if docker hub doesn't exist, where would you get alpine:3.1.2 from?
2020-07-31 18:21:57 <matthewcroughan> Buildah*
2020-07-31 18:22:17 <matthewcroughan> That's not what I am saying. I am saying that I should be able to distribute the OCI image I build 10 years ago.
2020-07-31 18:22:19 <ikke> We still have http://ancient.alpinelinux.org/alpine/
2020-07-31 18:22:25 <ikke> sure
2020-07-31 18:22:31 <ikke> I think so
2020-07-31 18:22:33 <matthewcroughan> I just need to understand what the caveats of that are
2020-07-31 18:22:41 <matthewcroughan> Or how to even do it. Is it a tarball?
2020-07-31 18:22:51 <ikke> standards might have shifted
2020-07-31 18:22:56 <ikke> kernel support
2020-07-31 18:23:08 <matthewcroughan> Heh. So the ISO image is just the best method.
2020-07-31 18:23:16 <matthewcroughan> A lot more work, but will work forever.
2020-07-31 18:23:21 <ikke> though, the docker image is just a tar
2020-07-31 18:23:25 <ikke> with layers
2020-07-31 18:23:44 <wsinatra> work forever is arbitrary
2020-07-31 18:23:48 <matthewcroughan> Maybe I should just keep things up to date. Doesn't require much maintenance.
2020-07-31 18:23:57 <afontain_> matthewcroughan : it's much harder to update, too
2020-07-31 18:23:58 <wsinatra> nothing will work forever, drivers in your ISO might not be compatible with hardware down the line.
2020-07-31 18:24:13 <matthewcroughan> Yes, that is not what I am saying.
2020-07-31 18:24:30 <matthewcroughan> I want to take a snapshot, and have it work forever on all x86 cpus today.
2020-07-31 18:24:31 <matthewcroughan> Not in the future.
2020-07-31 18:25:05 <matthewcroughan> Future proofing isn't a thing, obviously lol.
2020-07-31 18:26:17 <matthewcroughan> There's two approaches here. One sees me generating an ISO, at which point I am making my own distro instead of building on top of Alpine.
2020-07-31 18:26:33 <matthewcroughan> The other approach sees me telling people instructions that hopefully work in the future.
2020-07-31 18:26:59 <wsinatra> provide a metapackage?
2020-07-31 18:27:05 <matthewcroughan> Explain?
2020-07-31 18:27:08 <wsinatra> What is it that you're implementing on said system?
2020-07-31 18:27:34 <matthewcroughan> It's more of a "Here's some cool apps that go together, here's how easy alpine makes it"
2020-07-31 18:27:34 <wsinatra> If it's a collection of packages, that produce X looking system, or Y consumable product, just package the software
2020-07-31 18:27:38 <afontain_> wsinatra : oh, nice idea
2020-07-31 18:27:52 <wsinatra> metapackage is your solution my friend, build on Alpine edge and you won't have to fuss much
2020-07-31 18:28:00 <wsinatra> you could even contribute the metapackage to aports
2020-07-31 18:28:14 <wsinatra> https://wiki.alpinelinux.org/wiki/APKBUILD_meta:Meta
2020-07-31 18:28:21 <matthewcroughan> I don't understand quite though, how would this be better than a Dockerfile?
2020-07-31 18:28:28 <wsinatra> a metapackage is just an APKBUILD that installs a series of packages
2020-07-31 18:28:33 <matthewcroughan> It would be worse, since this means it will not work on any distribution of Linux.
2020-07-31 18:28:42 <matthewcroughan> The metapackage is only useful for Alpine.
2020-07-31 18:29:16 <afontain_> well, technically, if you base on some Alpine stable, it's lower maintainance
2020-07-31 18:29:22 <matthewcroughan> AHhh
2020-07-31 18:29:27 <wsinatra> https://gitlab.com/postmarketOS/pmaports/-/blob/master/device/community/device-nokia-n900/APKBUILD
2020-07-31 18:29:33 <wsinatra> more in depth metapackaging ^
2020-07-31 18:29:39 <matthewcroughan> Okay. So, can I make a metapackage for personal usage, and not contribute it to APORTS?
2020-07-31 18:29:45 <wsinatra> sure easily
2020-07-31 18:30:05 <matthewcroughan> Alright, sounds awesome. I've been exposed to this a bit in Arch which does the same thing in some places
2020-07-31 18:30:23 <matthewcroughan> In 10 years, my metapackage will fail right?
2020-07-31 18:30:41 <matthewcroughan> As will all 3.1.2 isos, since they will call out to servers which will fail because of the fact that their mirrors are way too new
2020-07-31 18:30:45 <afontain_> postmarketos-ui-* are also metapackages
2020-07-31 18:30:52 <matthewcroughan> In general, there is no way to make things last that long other than maintenance
2020-07-31 18:31:00 <matthewcroughan> I think my goal is impossible, and a fools errand
2020-07-31 18:31:12 <matthewcroughan> Maintenance is the only way to guarantee any sort of longevity.
2020-07-31 18:31:34 <matthewcroughan> Yeah I'm convinced of that.
2020-07-31 18:31:47 <matthewcroughan> afontain_ how do I get `postmarket-ui` packages on base Alpine?
2020-07-31 18:32:32 <wsinatra> you can grab the postmarketOS repos, but I would suggest just looking at their pmaports repos for examples
2020-07-31 18:32:48 <wsinatra> it's a fork of alpine built to deploy on phones, which might not meet your use case
2020-07-31 18:32:53 <matthewcroughan> Let's say I want to get the APKBUILD for `device-nokia-n900` from the repo wsinatra linked
2020-07-31 18:33:02 <matthewcroughan> How do I add that to my repos?
2020-07-31 18:33:07 <matthewcroughan> do I put it in /etc/apk/repositories?
2020-07-31 18:33:25 <wsinatra> https://engineering.fundingcircle.com/blog/2015/04/28/create-alpine-linux-repository/
2020-07-31 18:33:34 <wsinatra> this is a pretty good write up on making your own APK repo
2020-07-31 18:34:00 <wsinatra> you can copy the APKBUILDS from Alpine's aports, and how them on your own HTTP server, and direct distros to that if you wanted
2020-07-31 18:34:10 <wsinatra> I do precisely this at work for our Alpine workstations
2020-07-31 18:34:12 <afontain_> Well, you can also "just" create the package
2020-07-31 18:34:41 <matthewcroughan> I would most definitely not want to do that.
2020-07-31 18:34:46 <wsinatra> haha yeah abuild -r the APKBUILD and then just install whatever_package.apk --allow-untrusted with apk
2020-07-31 18:34:53 <matthewcroughan> though it's amazing to see how easy it is to set that infrastructure up
2020-07-31 18:35:08 <wsinatra> but I think hosting makes sense based on the "distribute it" desire
2020-07-31 18:35:19 <matthewcroughan> I only want that to be true *if* things are down
2020-07-31 18:35:28 <matthewcroughan> if things aren't down, no maintenance for me
2020-07-31 18:35:32 <ikke> abuild already creates a repo
2020-07-31 18:35:36 <matthewcroughan> but then this means that I may as well just stay up to date, since it requires work *either way*
2020-07-31 18:35:41 <ikke> just point a webserver to it, and you can install it from everywhere
2020-07-31 18:35:49 <ikke> distribute the public key, and it's even trusted
2020-07-31 18:36:00 <matthewcroughan> that stands for nothing unless it implements TuF
2020-07-31 18:36:09 <matthewcroughan> https://github.com/theupdateframework/specification/blob/master/tuf-spec.md#1-introduction
2020-07-31 18:36:36 <matthewcroughan> I don't want to bother with key rotation, creating more trust is no good, I'd rather trust the alpine packages.
2020-07-31 18:36:46 <matthewcroughan> All too complex for what I want to teach people :P
2020-07-31 18:36:52 <ikke> heh
2020-07-31 18:36:54 <matthewcroughan> I just wanna mainly show how cool Alpine is
2020-07-31 18:37:36 <matthewcroughan> The way I see it. If I make a repo, distribute packages, etc. That is *more* work than just keeping my damn instructions up to date.
2020-07-31 18:37:42 <wsinatra> upstream whatever metapackage into Alpine's testing repo
2020-07-31 18:38:09 <matthewcroughan> wsinatra I did not notice metapackages existing. How do I browse all the metapackages in a web browser?
2020-07-31 18:38:22 <matthewcroughan> https://pkgs.alpinelinux.org/packages
2020-07-31 18:38:54 <wsinatra> I don't believe they're labelled differently, but I can suggest a few for reference. Such as nagios-plugins-all
2020-07-31 18:39:17 <matthewcroughan> If I summon an alpine container now, can I use `apk` to search only metapackages?
2020-07-31 18:39:43 <wsinatra> they're just kind of there, they exist as packages, it's the fact that they just install a collection of packages, instead of having a build,package, etc function in them that makes them different
2020-07-31 18:40:12 <matthewcroughan> That's just asking for breakage haha.
2020-07-31 18:40:28 <matthewcroughan> When stuff breaks, how do you know?
2020-07-31 18:40:41 <ggg> Hello guys! I came back again :D. I did a diskless USB today and then tried to put up the builded package but failed unfortunately :( . https://tpaste.us/ObPb
2020-07-31 18:40:41 <matthewcroughan> It seems like a massive risk to take, increasing the complexity. Is there a way you guys kjnow?
2020-07-31 18:40:43 <matthewcroughan> know*
2020-07-31 18:43:47 <wsinatra> what are you trying to bundle together that is so fragile?
2020-07-31 18:44:05 <wsinatra> If your instructions work, but a metapackage doesn't, then isn't it broken by design?
2020-07-31 18:44:47 <wsinatra> either you're maintaining the instructions, or a list of packages, but if a dependency breaks either you get the same result. This is a methodology for simplified distribution, not resilience
2020-07-31 18:46:03 <matthewcroughan> Well I'm just pre-emptively scared of all the dependencies going missing.
2020-07-31 18:46:56 <matthewcroughan> The only thing that conquers this fear atm is NixOS or Gentoo, right? :D
2020-07-31 18:47:08 <wsinatra> If a package maintainer in Alpine stops doing their job, the package is moved to unmaintained, and anyone can take over
2020-07-31 18:47:15 <matthewcroughan> All it takes is one rogue maintainer to nuke someone's metapackage. Am I mistaken?
2020-07-31 18:47:28 <wsinatra> So if a dependency is lost, you can take ownership of it. Or ask someone to fix it in the community.
2020-07-31 18:47:30 <matthewcroughan> And then if that meta package is important, we're soooooooo screwed.
2020-07-31 18:47:36 <matthewcroughan> It's like satellites in space.
2020-07-31 18:48:01 <wsinatra> If you're packaging this for something production grade, keep your own repo, build your own infra. Don't rely on something that's not in your control
2020-07-31 18:48:16 <wsinatra> but the "easy" solution is relying on others good will
2020-07-31 18:48:28 <matthewcroughan> Too true.
2020-07-31 18:48:37 <matthewcroughan> It has gotten us to 2020 alive and well.
2020-07-31 18:49:13 <matthewcroughan> We can rely on continental drift, the sun rising and setting, AND on people's good will.
2020-07-31 18:50:17 <wsinatra> There's no less risk in using a different distro. Even if it's Debian, or Ubuntu
2020-07-31 18:50:27 <wsinatra> the company/foundations could be bought or stopped.
2020-07-31 18:50:53 <wsinatra> Linux torvalds could die tomorrow in a freak accident. The world has a funny way of still going on, and you realistically can't plan for every scenario
2020-07-31 18:51:53 <matthewcroughan> Linus Tech Tips
2020-07-31 18:52:18 <matthewcroughan> Linus Sebastian could die tomorrow and Linux would still go on :'(
2020-07-31 18:53:24 <afontain_> The advantage of distributions is actually the men and women behind it. That's what drives them updgraded, working and well
2020-07-31 18:53:47 <wsinatra> Yep, and your metapackage dependencies will break, and either someone will fix it for you, or you can get into the nitty gritty of FOSS and contribute a fix yourself
2020-07-31 18:54:01 <matthewcroughan> Distros are pretty smart. About as smart and useful as a compiler, for getting stuff done
2020-07-31 18:54:17 <wsinatra> Unlike with vendor owned stuff, where your fancy smart glasses no longer work because google bought the company that made them and bricked them
2020-07-31 18:54:23 <matthewcroughan> Except its cogs are flesh and blood :P
2020-07-31 18:54:57 <matthewcroughan> Y'all ain't got the Hikari window manager packaged.
2020-07-31 18:55:13 <ikke> matthewcroughan: patches welcome :P
2020-07-31 18:55:20 <matthewcroughan> That's what I'm thinking
2020-07-31 18:55:32 <matthewcroughan> The more I observe, the easier it seems to look to do that.
2020-07-31 18:56:07 <matthewcroughan> So how do I actually provide a package to you guys? Is it like Arch with its PKGBuilds?
2020-07-31 18:56:10 <afontain_> I can speak for postmarketOS, at least. We do use quite some metapackages. We do add packages on top of Alpine and keep using Alpine's repos
2020-07-31 18:56:16 <ikke> matthewcroughan: yes, very similar
2020-07-31 18:56:33 <matthewcroughan> On the high level, since I don't have time to do it rn, do I just provide you build instructions?
2020-07-31 18:57:01 <ikke> matthewcroughan: fastest way to get it in alpine is to provide a merge request against aports
2020-07-31 18:57:36 <afontain_> pmOS has reached the size where if the lead developer was to die, or the two lead developers were to die, the project would live on
2020-07-31 18:59:14 <wsinatra> afontain_ that's good feels right there :)
2020-07-31 18:59:36 <wsinatra> matthewcroughan: if an AUR package exists for Hikari you can use toAPK to convert it to an APKBUILD
2020-07-31 19:00:05 <matthewcroughan> Hmm, how though? Does it not matter whether it's a binary vs source based aur package?
2020-07-31 19:00:07 <wsinatra> warning though, it will need tweaking. Not all the build functions carry over, and there will likely be formatting issues (only 4 character tabs allowed, but toAPK produces spaces)
2020-07-31 19:00:16 <ikke> and policy differences
2020-07-31 19:00:35 <wsinatra> the APKBUILDs are flat files, they're build instructions
2020-07-31 19:00:40 <matthewcroughan> Alpine is a lot like Arch, in a few ways. I dig it.
2020-07-31 19:01:11 <ikke> wsinatra: I think matthewcroughan is aware of that, as it's very similar to Arch' PKGBUILDs
2020-07-31 19:01:25 <ikke> But in Alpine it's a lot easier to contribute packages for others
2020-07-31 19:01:27 <wsinatra> fair point
2020-07-31 19:01:28 <matthewcroughan> wsinatra AUR packages come in -bin and -git variety. So it's hikari-git or hikari-darcs
2020-07-31 19:01:34 <matthewcroughan> or hikari-bin if it's just a binary file
2020-07-31 19:01:43 <matthewcroughan> the other problematic thing is that yes, hikari doesn't use git, it uses darcs.
2020-07-31 19:01:45 <afontain_> I'd advise pushing as much as possible to Alpine, as it's much less of a maintainance burden that way
2020-07-31 19:02:03 <wsinatra> second what afontain_ said, much much easier
2020-07-31 19:02:03 <ikke> It does mean you need to maintain it though :)
2020-07-31 19:02:18 <matthewcroughan> CI
2020-07-31 19:02:45 <ikke> matthewcroughan: we prefer to use releases as much as possible
2020-07-31 19:03:26 <afontain_> ikke: true
2020-07-31 19:03:35 <ikke> matthewcroughan: as a matter of fact, abuild does not support directly cloning from git
2020-07-31 19:04:03 <matthewcroughan> I'm free to do CI though right? It should require basically no maintenance if I just have a CI script that does all the procedure for me?
2020-07-31 19:04:20 <ikke> matthewcroughan: Not sure what you mean wiht a CI script
2020-07-31 19:04:41 <matthewcroughan> When a github release is pushed, trigger a script to run which ultimately deploys to Alpine a new release of the package.
2020-07-31 19:05:20 <wsinatra> Assuming nothing breaks, it's typically a simple matter of changing the package version and pushing a MR to aports
2020-07-31 19:05:26 <ikke> New versions are deployed by pushing new versions of APKBUILds to aports
2020-07-31 19:05:36 <ikke> We do have our own CI to test build the packages
2020-07-31 19:05:36 <wsinatra> If something changes in the package release, and it breaks, you'll have to fix that before the MR will be accepted
2020-07-31 19:05:57 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests
2020-07-31 19:06:58 <matthewcroughan> ikke yes, but let's say I have my own CI system. Is anything going to stop me using it?
2020-07-31 19:07:08 <matthewcroughan> intentional barriers to automation?
2020-07-31 19:07:35 <ikke> We do not directly accept packages
2020-07-31 19:07:45 <matthewcroughan> I don't know what that means.
2020-07-31 19:08:02 <matthewcroughan> Can I have a script run whilst I'm asleep that triggered due to something someone else did on github, and have that merge without my intervention?
2020-07-31 19:08:03 <matthewcroughan> Yes or no?
2020-07-31 19:08:36 <afontain_> That means that you can't push packages to Alpine's repos without Alpine's approval
2020-07-31 19:08:36 <ikke> matthewcroughan: if that script would create a merge request against aports, and someone who has permission to merge, merges it, then yes
2020-07-31 19:08:47 <matthewcroughan> Great, that answers it.
2020-07-31 19:08:53 <matthewcroughan> All I need is git right?
2020-07-31 19:09:19 <matthewcroughan> APKBUILD + Git, just like the AUR?
2020-07-31 19:09:37 <ikke> yes
2020-07-31 19:09:56 <afontain_> Well, technically I'm not even sure you need git
2020-07-31 19:10:07 <matthewcroughan> Do you guys do binary packages? Or must we include build instructions that actually build from source?
2020-07-31 19:10:19 <ikke> matthewcroughan: build from source preferably
2020-07-31 19:10:33 <afontain_> depends if you upstream to Alpine
2020-07-31 19:10:40 <ikke> ie, unless there are good reasons not to, build from source
2020-07-31 19:10:42 <matthewcroughan> But in the case of a GoLang application, where a static binary is built by *our* procedure, what then?
2020-07-31 19:11:26 <afontain_> You can always do whatever you want in your own repos
2020-07-31 19:11:33 <matthewcroughan> In a way it's more secure, since that binary is verifiable to *us* the creators of the package, rather than alpine itself
2020-07-31 19:11:35 <ikke> We want users to be able to build things from source
2020-07-31 19:11:45 <matthewcroughan> creators of the binary*
2020-07-31 19:12:31 <matthewcroughan> Yeah okay that makes sense, I'm just thinking about the fact that you guys could could easily modify gpg this way and we'd never know
2020-07-31 19:12:46 <matthewcroughan> you distribute the binaries that you claim you build from the same source that's available on aports
2020-07-31 19:12:58 <matthewcroughan> but there is no trustless way of verifying that indeed you have built from the correct sources
2020-07-31 19:12:59 <ikke> matthewcroughan: we are working on supporting reproducable builds
2020-07-31 19:13:13 <afontain_> well, reproducible builds would help
2020-07-31 19:13:16 <ikke> which means that anyone can build from the same source, and verify that they would get the same binary
2020-07-31 19:13:25 <matthewcroughan> ah yeah that's cool
2020-07-31 19:13:36 <matthewcroughan> Actually yeah that's absolutely the best of both worlds
2020-07-31 19:13:49 <ikke> matthewcroughan: What we want to avoid is that upstreams get lazy and make it hard for others to build from source because they can just use the binary distributions anwya
2020-07-31 19:13:52 <ikke> anyway
2020-07-31 19:13:55 <matthewcroughan> So do you guys do TuF yet?
2020-07-31 19:14:04 <ikke> matthewcroughan: first time I've heard of it
2020-07-31 19:14:23 <ikke> matthewcroughan: chef works that way for example
2020-07-31 19:14:30 <matthewcroughan> I can link you to a talk that explains it superbly well. Docker images are signed by a service called Notary which implements TuF for docker images.
2020-07-31 19:14:31 <ikke> very difficult for others to build
2020-07-31 19:14:50 <matthewcroughan> https://www.youtube.com/watch?v=JvjdfQC8jxM
2020-07-31 19:15:10 <ikke> heh, +1 for the youtube title
2020-07-31 19:15:26 <matthewcroughan> The TL;DR is that this allows mitigation of compromise.
2020-07-31 19:15:52 <matthewcroughan> You can revoke keys by rotating them, you get an offline and and online key that are needed for complete compromise, and replay attacks are prevented.
2020-07-31 19:16:25 <matthewcroughan> The signing key for key rotation is offline, so your entire service cannot be compromised unless both are present.
2020-07-31 19:16:49 <matthewcroughan> And of course, there's a concept of freshness. Keys expire and require rotation.
2020-07-31 19:19:31 <matthewcroughan> The docker hub and images and docker cli handle TuF. He gives a live demo of compromising a docker image,
2020-07-31 19:19:39 <ikke> Watching it now
2020-07-31 19:20:00 <matthewcroughan> :D
2020-07-31 19:20:25 <ikke> One concept he's mentioning is about replay attacks, which on it's own, is valid
2020-07-31 19:21:52 <ikke> But it does require you to have an infrastructure that keeps these signatures alive
2020-07-31 19:22:14 <matthewcroughan> He talks about this later.
2020-07-31 19:22:26 <ikke> ok
2020-07-31 19:22:27 <matthewcroughan> Towards the end, hold on for it. It's smart, you have a server doing that.
2020-07-31 19:22:42 <matthewcroughan> because of the offline key, compromise to that server is mitigatable anyway.
2020-07-31 19:23:12 <ikke> yeah, just worried about forced depracation of older versions
2020-07-31 19:27:29 <ikke> offline - online keys sounds a lot like what GPG already has, the main key + subkeys (there is ofcourse more to it, but the concept sounds similar)
2020-07-31 19:27:49 <ikke> or CA root keys + intermediate keys
2020-07-31 19:33:16 <matthewcroughan> he uses alpine in this talk actually, and compromises the alpine base image as an example
2020-07-31 19:33:47 <matthewcroughan> oh nvm, thought he would based on his slide
2020-07-31 19:37:58 <afontain_> Well, I don't think Alpine is vulnerable to the "mixing version in an unexpected way" attack
2020-07-31 19:39:17 <wsinatra> doesn't pmOS explicitly do this to ensure certain packages are sourced from pmOS and not alpine?
2020-07-31 19:39:34 <wsinatra> I remember qemu was packages with pkgver=9999 a ways back
2020-07-31 19:41:38 <afontain_> Well, it's a different repository you trust
2020-07-31 19:44:18 <ikke> matthewcroughan: certainly an interesting model
2020-07-31 19:44:42 <ikke> matthewcroughan: ncopa mentioned he wanted to update the way we do package signatures
2020-07-31 19:45:57 <wsinatra> afontain_: that's true, no complaints of course, just thought it was interesting
2020-07-31 20:04:52 <matthewcroughan>  ikke wsinatra do you know of any ways to preseed an alpine image?
2020-07-31 20:05:04 <matthewcroughan> Basically, I don't want to bother installing, ever again, using alpine-setup.
2020-07-31 20:05:17 <matthewcroughan> I want to flash the media, go in there, enable ssh, disable ram mode.
2020-07-31 20:11:20 <wsinatra> setup-alpine -c will create an answer file for you
2020-07-31 20:11:32 <matthewcroughan> I don't really want to boot into the system.
2020-07-31 20:11:37 <matthewcroughan> I want to do this before the system boots.
2020-07-31 20:11:59 <wsinatra> an APKVOL might be a good way to go about that then
2020-07-31 20:12:06 <matthewcroughan> You know how Raspbian works, right?
2020-07-31 20:12:34 <matthewcroughan> Most of what you do is modifying the sd card, then plugging it  into the pi, power it on. Alpine should be like this as much as possible, IMO
2020-07-31 20:12:39 <matthewcroughan> APKVOL?
2020-07-31 20:14:08 <wsinatra> yeah the lbu program lets you create a sort of snapshot called an APKVOL and restore from it. Though it's meant to be used with run from ram builds
2020-07-31 20:14:27 <matthewcroughan> Yeah, I just want to flash alpine and that's it. Lol.
2020-07-31 20:14:30 <wsinatra> I've build build systems for SBCL using that and Qemu, and it works decently
2020-07-31 20:14:38 <matthewcroughan> Raspbian works like this. OpenWRT works like this.
2020-07-31 20:15:13 <wsinatra> Right that's kind of what this does. the APKVOL lives on an SD card and is bootstrapped into the Alpine image on boot. It contains a localized APK repo and other persisted data (such as system configurations)
2020-07-31 20:16:31 <matthewcroughan> I've just flashed `alpine-std 3.12.0 x86`
2020-07-31 20:16:47 <matthewcroughan> I want now, before I boot the media up, to set its hostname, tell it to use dhcp, enable ssh
2020-07-31 20:17:14 <matthewcroughan> I see one partition. I've mounted it, there's 3 folders inside it. `apks` `efi` `boot`.
2020-07-31 20:18:02 <matthewcroughan> wsinatra is there a doc on APKVOL?
2020-07-31 20:18:22 <matthewcroughan> https://wiki.alpinelinux.org/wiki/Alpine_local_backup ah
2020-07-31 20:18:41 <matthewcroughan> you mean APKOVL?
2020-07-31 20:19:04 <ikke> but apkovl is for run-from-ram
2020-07-31 20:19:07 <matthewcroughan> This is still no good. This seems not to have any impact on first-boot behavior.
2020-07-31 20:19:17 <matthewcroughan> And like ikke just stated, it only relates to run from ram.
2020-07-31 20:19:50 <wsinatra> sorry dyslexia, but yeah the APKOVL is what I meant
2020-07-31 20:20:26 <matthewcroughan> What I want, is a way to install to a disk, without having to launch alpine from boot media.
2020-07-31 20:20:54 <matthewcroughan> I could do that on a container on my laptop, I guess :/
2020-07-31 20:21:10 <matthewcroughan> You guys should provide such media for modification.
2020-07-31 20:22:01 <matthewcroughan> When I flash OpenWRT to any disk, that's it. It's done. There is no installation procedure, that was the installation procedure :D
2020-07-31 20:22:12 <matthewcroughan> If I want to change things, I just modify the FS.
2020-07-31 20:22:18 <mps> matthewcroughan: I started to work on it unofficially but didn't finished yet
2020-07-31 20:22:46 <matthewcroughan> Nice!
2020-07-31 20:22:52 <mps> you know how much different SBCs are around
2020-07-31 20:23:13 <matthewcroughan> How does that impact things?
2020-07-31 20:23:48 <mps> for example for each SBC different u-boot is required
2020-07-31 20:24:14 <matthewcroughan> I don't get how a disked rather than diskless would make this problem worse?
2020-07-31 20:24:22 <matthewcroughan> The diskless media won't boot on that either, right?
2020-07-31 20:24:30 <mps> different boot params, drivers, dtbs
2020-07-31 20:24:36 <matthewcroughan> yes, how does this matter?
2020-07-31 20:24:43 <mps> time
2020-07-31 20:24:51 <matthewcroughan> the current boot media, already does not work on those systems?
2020-07-31 20:25:01 <matthewcroughan> How would providing a disked version make this worse?
2020-07-31 20:25:54 <mps> whoever have time is free to do that
2020-07-31 20:27:50 <adu> matthewcroughan: something like minirootfs but gpt+ext4 instead of tar+gz?
2020-07-31 20:28:28 <matthewcroughan> I have no idea what you mean, since I'm not seeing the issue :P
2020-07-31 20:28:43 <mps> adu: miniroot is for docker/lxc
2020-07-31 20:28:51 <adu> mps: I understand that
2020-07-31 20:29:10 <adu> but it's the closest thing to a preinstalled version in the downloadable packages
2020-07-31 20:29:50 <mps> adu: then 'normal' tarballs are better
2020-07-31 20:30:15 <adu> you can't boot from tarballs
2020-07-31 20:30:41 <mps> are you sure :)
2020-07-31 20:31:15 <mps> I use tarballs to test releases
2020-07-31 20:31:51 <mps> made some scripts to unpack and 'flash' to sdcards, put in box and boot
2020-07-31 20:32:26 <adu> mps: so If matthewcroughan dd's a tarball to disk then the installation process is complete?
2020-07-31 20:32:43 <mps> heh, funny
2020-07-31 20:32:59 <mps> you are kidding with me
2020-07-31 20:33:19 <matthewcroughan> Are you Russian? I want to know
2020-07-31 20:33:33 <matthewcroughan> That sentence sounds like something a Russian speaking English would say :D
2020-07-31 20:33:43 <mps> no, we don't have images ready to be 'dd'-ed to media
2020-07-31 20:33:57 <mps> matthewcroughan: me?
2020-07-31 20:34:36 <afontain_> well, there are bootable ISOs
2020-07-31 20:35:31 <matthewcroughan> Yes mps
2020-07-31 20:35:42 <matthewcroughan> afontain_ the bootable iso is not enough, that's not what I want.
2020-07-31 20:35:52 <mps> no, not Russian though I speak Russian
2020-07-31 20:35:56 <adu> I honestly don't understand why we still use a format associated with a media that no longer exists
2020-07-31 20:36:05 <matthewcroughan> Like I said. When I dd openwrt isos to disk, they are done. The installation is 100% complete, I can now modify the filesystem of the disk and make my changes, *then* boot it
2020-07-31 20:36:08 <mps> matthewcroughan: I'm Serbian
2020-07-31 20:36:56 <mps> but most of aren't native English speakers
2020-07-31 20:37:04 <mps> most of us*
2020-07-31 20:39:31 <mps> alpine doesn't have much developers for arm arches
2020-07-31 20:39:59 <mps> afaik only I started to work on images which could be dd-ed
2020-07-31 20:40:17 <mps> and I have limited number of SBCs to test
2020-07-31 20:40:53 <matthewcroughan> My board is the atomic pi.
2020-07-31 20:41:13 <matthewcroughan> I just really don't want to tell people "Flash alpine to a usb, then get a usb hub, then boot it and run alpine-setup, then reboot"
2020-07-31 20:41:39 <adu> matthewcroughan: are you a compsci teacher?
2020-07-31 20:41:43 <matthewcroughan> Why not just "flash alpine to your usb disk, modify the filesystem here, then boot your alpine install"
2020-07-31 20:41:58 <matthewcroughan> adu No, I just want to make a youtube video that utilises alpine for running ardour6 and using VSTs in it.
2020-07-31 20:42:47 <matthewcroughan> which should just be as simple as flashing alpine to a disk, setting the hostname, enabling ssh, then booting it.
2020-07-31 20:43:01 <matthewcroughan> Just like Raspbian, PostmarketOS, OpenWRT.
2020-07-31 20:44:52 <RootWyrm_> The amount of pain involved in installing Alpine to RPis really has crossed firmly into the territory of ridiculous.
2020-07-31 20:45:24 <matthewcroughan> Really? Is it that bad? :D
2020-07-31 20:46:02 <RootWyrm_> https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi
2020-07-31 20:46:54 <matthewcroughan> But why though :(
2020-07-31 20:46:59 <mps> RPis are based on broadcom, so what could be expected
2020-07-31 20:47:07 <RootWyrm_> And half of that is wrong.
2020-07-31 20:47:52 <RootWyrm_> mps: yeah, no, that's not an excuse when Broadcom has made it clear they only accept Linux. And when FreeBSD is easier to install on a RPi4 than Alpine.
2020-07-31 20:47:55 <matthewcroughan> But why do we have to boot the device to configure stuff at all? I'm not getting it.
2020-07-31 20:48:20 <RootWyrm_> matthewcroughan: because there is no actual installed image for rpi, only a live image.
2020-07-31 20:48:32 <matthewcroughan> I like setup-alpine, I think it's useful and convenient. But when I'm doing this over and over, I just wanna flash the rootfs and modify some stuff.
2020-07-31 20:48:52 <matthewcroughan> RootWyrm_ ? I can't even figure this out for x86
2020-07-31 20:48:57 <matthewcroughan> How do I do it? There is no page on it afaik
2020-07-31 20:49:22 <RootWyrm_> The x86 stuff will install to disk, been long enough that I've forgotten the steps because I just clone a VM.
2020-07-31 20:49:22 <afontain_> pmOS installs as sys on raspberry pi
2020-07-31 20:49:42 <matthewcroughan> RootWyrm_ Oh my aching head :D
2020-07-31 20:49:59 <matthewcroughan> Where is the download page for the rootfs so I can just flash it to disk?
2020-07-31 20:50:15 <mps> :)
2020-07-31 20:50:19 <RootWyrm_> IIRC that won't work
2020-07-31 20:50:41 <matthewcroughan> But why gdi
2020-07-31 20:50:47 <RootWyrm_> Can't remember if it was because the rootfs was wrong or broken or what.
2020-07-31 20:50:52 <matthewcroughan> https://wiki.alpinelinux.org/wiki/Installation#2._Dump.2C_burn_or_flash_the_image
2020-07-31 20:51:01 <matthewcroughan> We need to burn step 2.3
2020-07-31 20:51:09 <adu> rootfs is a tarball, what you need is a gpt image
2020-07-31 20:51:30 <RootWyrm_> adu: uh, no, definitely not correct.
2020-07-31 20:51:32 <matthewcroughan> Step 2.3 should be definable, in a file in /boot/
2020-07-31 20:51:47 <RootWyrm_> the SoC firmware literally cannot support GPT
2020-07-31 20:52:04 <adu> but also minirootfs doesn't have linux kernel
2020-07-31 20:52:10 <RootWyrm_> (unless you want to get into UEFI which you don't.)
2020-07-31 20:52:22 <matthewcroughan> I should be able to provide an answer file, pre-emptively, to a disk that is not yet booted. Then boot it, and have the installation procedure ran through on the disk I used to boot.
2020-07-31 20:52:31 <adu> RootWyrm_: why don't you want UEFI?
2020-07-31 20:52:33 <RootWyrm_> Which is also moot because the kernel doesn't live on the rootfs _anyway_
2020-07-31 20:53:03 <RootWyrm_> adu: https://github.com/pftf/RPi4 - have fun
2020-07-31 20:53:16 <RootWyrm_> adu: p.s. it can and will brick hardware.
2020-07-31 20:53:51 <adu> RootWyrm_: what firmware does the Rpi come with?
2020-07-31 20:54:27 <RootWyrm_> obnoxious u-boot blobs.
2020-07-31 20:56:42 <adu> u-boot at least has partial UEFI support
2020-07-31 21:02:13 <adu> RootWyrm_: u-boot does support GPT, what do you mean by "the SoC firmware literally cannot support GPT"?
2020-07-31 21:04:27 <RootWyrm_> I would suggest you go actually read the technical documentation on the RPi boot process.
2020-07-31 21:13:23 <matthewcroughan> Is pmbootstrap portable to many different places?
2020-07-31 21:13:31 <matthewcroughan> whoops wrong channel
2020-07-31 21:16:05 <matthewcroughan> Are there any distributions that do what I'm asking? I can't use alpine if it requires user interaction to setup like this.
2020-07-31 21:16:23 <matthewcroughan> It's so funny that OpenWRT looks like my only option hahaha
2020-07-31 21:16:36 <matthewcroughan> Raspbian is hardware specific, and OpenWRT is for damned routing.
2020-07-31 21:17:20 <matthewcroughan> It's not a big ask. I just want to flash alpine-3.1.2.iso to my disk. Then boot it. Then ssh into it. Then do my stuff. No installation.
2020-07-31 21:17:41 <matthewcroughan> All within a literal 240 second timespan. The whole thing.
2020-07-31 21:18:13 <matthewcroughan> This setup-alpine step is a waste of time, in my opinion.
2020-07-31 21:18:50 <matthewcroughan> pmbootstrap makes things like this, but it requires pmbootstrap. Why do we need this dependency? It seems like this could be worked around.
2020-07-31 21:20:39 <afontain_> well, you need something to do your setup
2020-07-31 21:20:51 <afontain_> though you can build an image with pmbootstrap beforehand
2020-07-31 21:20:54 <matthewcroughan> Yes, but why should it involve me, the user?
2020-07-31 21:21:11 <afontain_> http://images.postmarketos.org/
2020-07-31 21:21:16 <matthewcroughan> E.G when I flash raspbian, everything is done. If I want to enable ssh I `touch /boot/ssh` **BEFORE** booting the machine.
2020-07-31 21:21:55 <afontain_> well, that's what pmbootstrap does
2020-07-31 21:21:58 <matthewcroughan> I do not want to have to pull a keyboard out to ssh into a machine.
2020-07-31 21:22:06 <matthewcroughan> Yes, but then I need pmbootstrap.
2020-07-31 21:22:17 <matthewcroughan> The reason I need pmbootstrap is because Alpine isn't standard enough not to need abstraction.
2020-07-31 21:22:51 <matthewcroughan> FreeBSD is about as basic too. If I want to turn FreeBSD's boot media into a *real freeBSD*. I can do it. All I need to do it, before even booting the disk.
2020-07-31 21:23:00 <matthewcroughan> I don't need freebsd-bootstrap
2020-07-31 21:23:09 <matthewcroughan> All I need to do is remove rc.conf
2020-07-31 21:24:08 <matthewcroughan> How many steps are required to install alpine? Far too many, in my opinion. We can do better.
2020-07-31 21:24:35 <matthewcroughan> No interactive installation, I don't want it. How do we get rid of it?
2020-07-31 21:25:08 <matthewcroughan> Obviously setup-alpine is nice and convenient. But I do not want to use it for the 20th time. I want simple static setups. Even Arch for Arm is like this!
2020-07-31 21:25:46 <matthewcroughan> Check it out https://archlinuxarm.org/platforms/armv6/raspberry-pi
2020-07-31 21:29:09 <matthewcroughan> I don't want another dependency (pmbootstrap) to do this. I want to be able to mess with this with simple posix utilities.
2020-07-31 21:29:36 <afontain_> define this
2020-07-31 21:30:06 <afontain_> as I said, you can build an image with pmbootstrap beforehand
2020-07-31 21:30:15 <matthewcroughan> There are no docs, that I can find, that tell me how I could replace pmbootstrap with mv, cp, mount, tar. Yet it should be possible.
2020-07-31 21:30:29 <matthewcroughan> I do not want to use pmbootstrap.
2020-07-31 21:30:39 <matthewcroughan> Pmbootstrap is a dependency. It messes up my host system.
2020-07-31 21:30:52 <matthewcroughan> I wand an image, that I tar, move some files, nothing more. No chroots, no nonsense.
2020-07-31 21:31:02 <afontain_> well, no it, it doesn't
2020-07-31 21:31:16 <afontain_> it will only touch ~/.local/var/pmbootstrap
2020-07-31 21:31:23 <matthewcroughan> That's too much lol.
2020-07-31 21:31:31 <matthewcroughan> I don't want yet another piece of software on my host.
2020-07-31 21:31:42 <afontain_> (well, and ~/.config/pmbootstrap.cfg)
2020-07-31 21:31:47 <matthewcroughan> Yes, too much for me.
2020-07-31 21:32:00 <matthewcroughan> No dotfiles, no python. No chroots. Just posix please.
2020-07-31 21:32:13 <afontain_> well, it can touch another directory if you prefer
2020-07-31 21:32:21 <matthewcroughan> No. Why do I need this utility?
2020-07-31 21:32:24 <afontain_> python and chroot are required though
2020-07-31 21:32:32 <matthewcroughan> Then I'm not using it.
2020-07-31 21:32:45 <matthewcroughan> I would rather use Docker and `FROM alpine:3.1.2` and do the install in there.
2020-07-31 21:32:59 <afontain_> hey, you choose if you want to use it, I've never said you need it
2020-07-31 21:33:00 <matthewcroughan> At least that way, there are no disgusting crumbs left behind.
2020-07-31 21:33:27 <matthewcroughan> I'm asking for an explanation of how pmbootstrap works, so I don't have to use pmbootstrap.
2020-07-31 21:34:09 <afontain_> pmbootstrap doesn't install itself on the target system
2020-07-31 21:34:19 <afontain_> so no crumbs left
2020-07-31 21:34:55 <afontain_> at some point your requirement are quite exigents, though: 10 years without any maintainance, TUF, no dependencies, doesn't write any file…
2020-07-31 21:35:00 <matthewcroughan> pmbootstrap is required. Pmbootstrap is now on my laptop. I don't want it on my laptop. Why did I need it in the first place?
2020-07-31 21:35:21 <afontain_> at some point, it's not actually possible
2020-07-31 21:35:27 <matthewcroughan> Why can't I do it the same as every other compliant and standard distribution?
2020-07-31 21:35:44 <matthewcroughan> I bet you I can. You just don't know how, because it's not clear.
2020-07-31 21:36:02 <matthewcroughan> It's just not in the docs because nobody but me wants it lol
2020-07-31 21:36:42 <afontain_> well, pmbootstrap allows you to use any distribution to build the image. You can also dd the image from any distribution. That was one of your requirements
2020-07-31 21:36:55 <matthewcroughan> ?
2020-07-31 21:37:00 <matthewcroughan> Have you ever used a pi?
2020-07-31 21:37:30 <matthewcroughan> Have you ever had the experience of just making *all* of your assumed changes to the disk you just flashed, before even booting it, and not even having to ssh into the machine?
2020-07-31 21:37:43 <matthewcroughan> Better yet, have you experienced not having to install anything, just flashing it to disk and booting it?
2020-07-31 21:37:53 <afontain_> that's not what postmarketOS uses
2020-07-31 21:38:06 <afontain_> also, you were talking about x86 two hours ago
2020-07-31 21:38:30 <matthewcroughan> There is no difference between the experience on ARM vs X86 in this same exact regard, with OpenWRT.
2020-07-31 21:38:38 <matthewcroughan> Architecture literally does not matter here.
2020-07-31 21:38:39 <matthewcroughan> At all.
2020-07-31 21:38:57 <afontain_> well, it's the case on postmarketOS too
2020-07-31 21:39:31 <afontain_> the basics of pmbootstrap, is that it grabs a static version of apk (for example from http://dl-2.alpinelinux.org/alpine/edge/main/x86_64/apk-tools-static-2.10.5-r1.apk), install the packages the user wants, and flash that rootfs, makes an image out of it, whatever you want
2020-07-31 21:39:40 <matthewcroughan> https://openwrt.org/docs/techref/targets/start
2020-07-31 21:39:46 <matthewcroughan> If I download any of these images. All I have to do is flash it.
2020-07-31 21:39:54 <matthewcroughan> That is the installation procedure. There is no setup-alpine.
2020-07-31 21:40:11 <matthewcroughan> If I want, I can make modifications to /etc just by mounting it and playing with files.
2020-07-31 21:40:16 <mps> I think this discussion is OT here
2020-07-31 21:40:19 <matthewcroughan> Then, I plug the disk in and boot. *done*.
2020-07-31 21:40:27 <matthewcroughan> I want this same experience with alpine.
2020-07-31 21:40:51 <afontain_> > /etc just by mounting it and playing with files. You can do that with any distro
2020-07-31 21:40:59 <matthewcroughan> Yes. But not before you install it.
2020-07-31 21:41:05 <matthewcroughan> In this case, there is no installation procedure.
2020-07-31 21:41:09 <afontain_> yeah, let's move to #alpine-offtopic
2020-07-31 21:41:13 <afontain_> or anything else
2020-07-31 21:42:13 <mps> pmOS? :)
2020-07-31 21:43:02 <afontain_> well, it's not really about pmOS either
2020-07-31 21:43:39 <mps> afontain_: right, but you understand I think
2020-07-31 22:11:46 <Hello71> >disgusting crumbs
2020-07-31 22:11:48 <Hello71> >docker
2020-07-31 22:46:46 <RootWyrm_> matthewcroughan: you can use an answer-file for the alpine-setup crap, feed it to an openrc firstboot script, but still doesn't solve for lack of actual rpi images
2020-07-31 22:48:36 <RootWyrm_> I may actually get time to poke at that this weekend.
2020-07-31 23:44:44 <matthewcroughan> RootWyrm_ It's not for the RPI
2020-07-31 23:44:51 <matthewcroughan> I do not want this for the RPI. I want this for all platforms.
2020-07-31 23:45:12 <matthewcroughan> I want to be able to flash it with no setup procedure, then the first boot *is* the installed system. It's just a better way of working.
2020-07-31 23:45:35 <matthewcroughan> This way, I have an installed system within 240 seconds, if that.
2020-07-31 23:45:52 <matthewcroughan> No monitor, keyboard or serial required. And I can SSH in after first boot.
2020-07-31 23:46:23 <matthewcroughan> pmbootstrap allows me to do it. It's a fine tool. I would just prefer that this be possible to do without it.