2023-10-01 10:35:41 <OdaNobunaga[m]> Hi! How do I change the language under KDE plasma on Alpine? The region and language settings menu is greyed out. I've downloaded all the -lang packages I could find. I did work in the past.
2023-10-01 10:35:56 <OdaNobunaga[m]> s/I/It/
2023-10-01 10:40:57 <OdaNobunaga[m]> Interesting:
2023-10-01 10:40:57 <OdaNobunaga[m]> is supposed to fetch the language files of all installed packages. Maybe this will help.
2023-10-01 10:40:57 <OdaNobunaga[m]> doas apk add lang
2023-10-01 11:49:32 <singham> I want to do it over ethernet so once I add plugin pppoe.so eth0, it should not ask for a device
2023-10-01 11:49:58 <singham> I am configuring ppp-pppoe
2023-10-01 12:00:12 <singham> How does pppd take inputs of config file?
2023-10-01 14:46:59 <donoban> is the musl "problem" with dns already fixed or is probably the cause of https://gitlab.alpinelinux.org/alpine/aports/-/issues/15320 ?
2023-10-01 14:47:46 <Ermine> 64 IPs at once?????????
2023-10-01 14:48:03 <Ermine> well, 63, but still
2023-10-01 14:48:41 <minimal> when I saw that I guessed it might be due to a "sizing" limitation of the musl dns code
2023-10-01 15:03:25 <minimal> this is the discussion I had in mind: https://www.openwall.com/lists/musl/2023/07/04/3
2023-10-01 15:07:09 <donoban> I remember vaguely that glibc returns what it has if there is not enough space or something is wrong with the answer, while musl gives an error
2023-10-01 15:10:54 <minimal> donoban: why do you think it is already solved?
2023-10-01 15:10:58 <donoban> in 19b285aa13affa7e13db982b702c16b57674ade6 there are some commits from musl 
2023-10-01 15:11:09 <donoban> two of them related to dns
2023-10-01 15:11:18 <minimal> donoban: that was to do with adding TCP support for DNS
2023-10-01 15:11:34 <donoban> ah
2023-10-01 15:13:08 <minimal> musl 1.2.4 added TCP DNS support and Alpine 3.18.x (which they're using) contains musl 1.2.4
2023-10-01 15:13:51 <donoban> uhmm... https://git.musl-libc.org/cgit/musl/commit/?id=a4ecaf89a9b88df76e8bf9f28e1cc6cb89e4bfa8
2023-10-01 15:14:13 <donoban> buffer increase, is not related with this?
2023-10-01 15:15:04 <minimal> right, that seems to be a follow-on patch from the discussion I posted
2023-10-01 15:15:37 <donoban> well if the OP wants to retry the test with edge... is not a big deal :)
2023-10-01 15:16:00 <minimal> but that patch is ALSO present in Alpine 3.18.x's musl
2023-10-01 15:16:39 <minimal> it was committed on 27th July
2023-10-01 15:17:22 <donoban> ouch
2023-10-01 15:21:31 <minimal> editted my comment as that patch indicates there STILL is a 512byte limit for address answers (rather than CNAME chains)
2023-10-01 15:22:16 <minimal> I wonder why the heck someone would want more than 63 IP addresses to be returned for a name lookup
2023-10-01 15:22:51 <donoban> they feel like google :D
2023-10-01 15:23:33 <Ermine> feels more like gooooooooooooooooooooooooooooooooooooooogle
2023-10-01 15:24:19 <minimal> woah! just noticed, in their Dockerfile example they're using Alpine 3.18.3 base image but then using 3.14 repo for installing ca-certificates. Why the heck would they do that?
2023-10-01 15:24:49 <Ermine> Joins: Enterprise
2023-10-01 15:25:53 <Ermine> I think there's some weird certificate available in 3.14 ca-certificates but not available in new versions
2023-10-01 15:26:22 <minimal> oh, you mean like a CA removed due to them being a dodgy CA? lol
2023-10-01 15:27:17 <Ermine> Well, I have no better idea
2023-10-01 18:23:36 <singham> My pppd is unable to find chap-secrets file it seems
2023-10-01 18:24:11 <singham> Currently, I have edited init.d/pppd and added all the options directly there
2023-10-01 18:25:11 <Ermine> you may try running pppd under strace to see which files it tries to open
2023-10-01 18:25:11 <singham> does pppd find /etc/ppp/chap-secrets by default?
2023-10-01 18:38:41 <singham> require-chap is to be used for the server right?
2023-10-01 18:38:50 <singham> Not for client right?
2023-10-01 19:00:15 <quinq> singham, maybe this could help https://tldp.org/HOWTO/PPP-HOWTO/ (https://tldp.org/HOWTO/PPP-HOWTO/pap.html)
2023-10-01 19:03:10 <Ermine> Is there metapackage for all qt5 -dev packages?
2023-10-01 19:06:22 <ikke> unlikely as it wouldn't be that useful
2023-10-01 19:17:21 <singham> I made some progress. So now, it gets authentocated but stuck at IPCP
2023-10-01 19:28:43 <singham> peer refused to authenticate
2023-10-01 19:28:55 <singham> LCP terminated by peer
2023-10-01 19:43:20 <singham> ======
2023-10-01 19:43:39 <singham> peer from calling number MAC_SERVER authorized
2023-10-01 19:44:01 <singham> sent IPCP ConfReq
2023-10-01 19:44:52 <singham> rcvd [LCP TermReq id=0x4 "peer refused to authenticate"]
2023-10-01 19:44:58 <singham> ======
2023-10-01 19:45:39 <singham> Above is the error I'm stuck at.
2023-10-01 20:30:39 <m_ben> which line editor does ash use? it's certainly not readline
2023-10-01 20:30:54 <ikke> probably something bespoke
2023-10-01 20:31:34 <m_ben> is there a way to configure some keybindings in ash?
2023-10-01 20:38:05 <dwfreed> stop the self flagellation and just install a real shell :P
2023-10-01 20:41:14 <quinq> s/real/more interactive/
2023-10-01 20:44:04 <m_ben> well I'm curious to see what ash has to offer
2023-10-01 20:44:37 <ikke> ash is an okay shell, but it's not primarily an interactive shell
2023-10-01 20:52:09 <dwfreed> yeah, the only devices I have ash on are my openwrt boxes (I don't run alpine, to be fair), and the only reason I keep them that way is because I almost never have to do anything non-trivial in the shell on them
2023-10-01 20:52:40 <ikke> The only thing I really miss in ash is alt+dot
2023-10-01 20:53:50 <somercet1> TIL bash has alt-dot
2023-10-01 20:53:59 <dwfreed> oh, I use !$ usually, but alt-dot is definitely easier to type
2023-10-01 20:54:13 <ikke> yes, and you can still edit / append afterwards
2023-10-01 20:58:20 <m_ben> there doesn't seem to be a proper documentation on ash. for example, which backslash-escaped special characters are supported in PS1?
2023-10-01 20:59:03 <ikke> ash is mostly POSIX
2023-10-01 20:59:08 <ikke> look at the posix spec
2023-10-02 03:06:50 <zcrayfish> Also the source is insanely easy to read.
2023-10-02 03:06:58 <zcrayfish> if he comes back: https://gopher.zcrayfish.soy/1/phlog/20220914-busybox-ash-isn-t-very-customizable
2023-10-02 05:15:57 <ikke> zcrayfish: "did I miss something?" If ENV will set, ash will read the file it points to for non-login shells
2023-10-02 05:17:42 <zcrayfish> idk, I was probably spewing nonsense again :p
2023-10-02 05:18:26 <ikke> Was just pointing to something that was not mentioned on that page
2023-10-02 05:20:44 <zcrayfish> oh right, it wouldn't hurt for me to point out where the environment is sourced from
2023-10-02 05:26:39 <ikke> ~/.profile is only read for login shells, so people often ask how to get something similar to .bashrc
2023-10-02 05:27:58 <ikke> so if ENV is set and points to a file, it will source that for non-login shells as well
2023-10-02 06:40:22 <metallicus> Postfix needs "options edns0 trust-ad" in resolv.conf for DNSSEC and DANE support. It seems Musl recently got basic TCP support but I'm not sure issue 11455 is really fixed? Just disabling DANE for Postfix seems not quite right. https://gitlab.alpinelinux.org/alpine/aports/-/issues/11455
2023-10-02 06:44:16 <Saijin_Naib[m]> <ikke> "so if ENV is set and points to a..." <- What would that line look like exactly? 
2023-10-02 06:46:55 <ikke> export ENV=$HOME/.ashrc for example
2023-10-02 07:20:23 <Vishal_> Is alpine Linux as secure as some of the other distro like Ubuntu ? Does it have hardening like SElinux and stuff?
2023-10-02 07:23:02 <Vishal_> ??
2023-10-02 07:41:39 <i5yxt1550rd9gbko8o55[m]> <Vishal_> "Is alpine Linux as secure as..." <- nobody cares
2023-10-02 08:05:24 <humm> Linux has SELinux, you can use it if you want
2023-10-02 08:13:10 <winny> secure is pretty subjective
2023-10-02 08:13:25 <winny> it could mean all sorts of setups
2023-10-02 08:13:44 <winny> my application that runs as root in a VM appliance is secure :^)
2023-10-02 08:14:57 <aron> personally, I define systemd as a HUGE security issue :P
2023-10-02 08:16:30 <winny> see there you go, many perspectives to what is secure
2023-10-02 08:32:21 <lopid> "is x as secure as ubuntu?" heh
2023-10-02 08:46:26 <Ermine> Just 2 words: security model.
2023-10-02 09:16:17 <donoban> I wonder if a simple isolation model with uid+process-name would be easy to achieve with selinux, I think that Android does something like it but does it use a different uid for each process?
2023-10-02 10:06:48 <metallicus> Just one more attempt. Anybody here who knows more about this? Postfix needs "options edns0 trust-ad" in resolv.conf for DNSSEC and DANE support. It seems Musl recently got basic TCP support but I'm not sure issue 11455 is really fixed? Just disabling DANE for Postfix seems not quite right. https://gitlab.alpinelinux.org/alpine/aports/-/issues/11455
2023-10-02 11:09:56 <Ermine> donoban: in android each app is executed under its own uid
2023-10-02 11:15:13 <donoban> yep I believed it, do you know if selinux alone could provide a similar isolation only with 1 uid + process_context? something very simple like only the app that creates a file can reads/write/it, it would probably would need some kind of exception that ends breaking all the model :D
2023-10-02 11:18:08 <mercenary> guess that is why android has its 'virtual' filesystems, where the OS can presumably apply overrides
2023-10-02 11:19:32 <mercenary> proper SElinux policies (i.e. better than just running audit2allow and hoping for the best) are a lot of work
2023-10-02 11:22:24 <donoban> well the system needs more or less privileged power in any case, I was thinking in something like, ~/.bashrc can be only readed/modifed etc.. by bash, ok, now how to now edit it?
2023-10-02 11:23:08 <donoban> s/to now/to
2023-10-02 11:25:39 <mercenary> that'd have to be a separate context I guess, with a rule that only that context can write. But I haven't done selinux policies since the days of centos6
2023-10-02 11:31:17 <donoban> yep too weird in any case
2023-10-02 13:12:44 <Ermine> is there a reason why veracrypt is not packaged?
2023-10-02 13:28:27 <fluix> https://gitlab.alpinelinux.org/alpine/aports/-/commit/ce9644ed6d9226f5e81b04b122598c1fd5f1585a
2023-10-02 13:28:32 <fluix> Ermine: ^
2023-10-02 13:33:06 <fluix> now how does Arch package this
2023-10-02 13:36:22 <ikke> They still mention the Truecrypt license 
2023-10-02 13:37:07 <ikke> So they probably never checked the new license 
2023-10-02 13:37:17 <fluix> I don't think it's new?
2023-10-02 13:37:47 <fluix> I think the alpine commit means truecrypt because veracrypt source says "VeraCrypt is multi-licensed under Apache License 2.0 and the TrueCrypt License version 3.0"
2023-10-02 13:38:25 <fluix> and it contains the snippet psykose pasted
2023-10-02 13:39:36 <Ermine> Oic
2023-10-02 13:40:22 <Ermine> thank you
2023-10-02 13:40:54 <Ermine> fluix: Arch packages nonfree software
2023-10-02 13:41:15 <fluix> but what about the non-distribution part specifically?
2023-10-02 13:41:31 <fluix> anyways, talking in #archlinux
2023-10-02 13:42:41 <Ermine> I think this clause is ambiguous
2023-10-02 13:42:53 <fluix> > Foxboron  fluix: Hm, that license is problematic
2023-10-02 13:49:37 <minimal> metallicus: a quick look at the musl code for parsing resolv.conf shows that it does not support "edns0" or "trust-ad" as options: https://git.musl-libc.org/cgit/musl/tree/src/network/resolvconf.c#n41
2023-10-02 13:50:15 <minimal> metallicus: perhaps you should ask about this on the Libera #musl channel?
2023-10-02 14:00:33 <fluix> also I think it's beyond the clause, it's non-free
2023-10-02 14:43:25 <singham> I have 2 RJ45 ports. eth0 connects to ISP
2023-10-02 14:44:04 <singham> I want to connect another PC to eth1 and allow eth1 node to access internet through my PC.
2023-10-02 14:44:24 <singham> How do I do this? I don't want dhcp but static addressing
2023-10-02 14:45:12 <minimal> singham: does your ISP provide you a static IP address for eth0?
2023-10-02 14:45:41 <humm> What do you want? A LAN and masquerading? A subnet? Your PC to be entirely transparent?
2023-10-02 14:46:48 <singham> ? I just want the requests from eth1 to go to eth0 and return the packets to eth1
2023-10-02 14:47:28 <humm> Do you want to relay Ethernet frames between eth0 and eth1?
2023-10-02 14:47:39 <singham> Yes, thank you!
2023-10-02 14:48:09 <humm> May I ask why?
2023-10-02 14:49:49 <singham> My boss wants it man! I don't know what the hell is he fucking thinking. He says, do it like this!
2023-10-02 14:50:11 <ikke> Sounds like you want a bridge 
2023-10-02 14:50:35 <ikke> Create a bridge and then add the eth0 ports to that bridge 
2023-10-02 14:50:36 <aron> this is quite common task for a router or switch
2023-10-02 14:50:45 <ikke> A PC can be both 
2023-10-02 14:51:36 <ikke> I have a rv64 board where I have the 2 eth ports bridged so that I can hook another board to the 2nd port 
2023-10-02 14:52:20 <ikke> The bridge interface would have a static IP address
2023-10-02 14:52:39 <singham> So how do I set the bridge?
2023-10-02 14:53:07 <singham> And I connect internal IP with MAC address?
2023-10-02 14:54:01 <singham> Or the bridge software would automatically do it, fetch MAC, assign internal IP, and manage packets?
2023-10-02 14:54:55 <aron> you can connect them with openvswitch as well
2023-10-02 14:56:51 <Habbie> you rarely need openvswitch though
2023-10-02 15:00:01 <ikke> ip link add dev br0 type bridge
2023-10-02 15:07:54 <metallicus> minimal: thanks for your response :-)  
2023-10-02 15:44:30 <singham> ikke: Was offline, was that for me?
2023-10-02 15:44:44 <ikke> Yes
2023-10-02 15:44:54 <singham> ip link add dev eth1 type bridge 
2023-10-02 15:45:01 <singham> Do I do this?
2023-10-02 15:45:58 <ikke> That will create a bridge in runtime (ie, it's not permanent)
2023-10-02 15:46:39 <ikke> ip link set dev eth0 master br0; ip link set dev eth1 master br0;
2023-10-02 15:46:52 <ikke> ip addr add dev br0 <ip>/<subnet>
2023-10-02 15:46:56 <ikke> ip link set dev br0 up
2023-10-02 15:47:05 <ikke> ip route add default via <gateway>
2023-10-02 15:48:50 <singham> What is keyword dev?
2023-10-02 15:52:33 <singham> ikke: Thanks a lot!
2023-10-02 15:53:11 <quinq> dev stands for device
2023-10-02 15:58:41 <singham> But device is mapped to eth right?
2023-10-02 15:58:59 <singham> Say /dev/ttyS0 is connected to eth0
2023-10-02 15:59:13 <singham> Then why have keyword dev?
2023-10-02 16:32:27 <elly> what do you mean? it's labeling that argument as a device
2023-10-02 16:35:31 <singham> Aah ok
2023-10-02 16:35:47 <singham> understood
2023-10-02 17:31:09 <u0_a377> hey
2023-10-02 17:31:56 <u0_a377> i bought now usb flash 16G and after dd it not loads in liveiso
2023-10-02 17:33:42 <u0_a377> it writes: "Mounting boot media failed. initramfs emergency recovery shell launched. Type 'exit' to continue boot sh: cant access tty: job turned off
2023-10-02 17:35:58 <Ermine> what did you dd?
2023-10-02 17:37:19 <u0_a377> as usual iso
2023-10-02 17:37:31 <u0_a377> i now did same stepsb
2023-10-02 17:38:02 <u0_a377> i did same steps right now with other usb flash and it work
2023-10-02 17:39:03 <u0_a377> but fresh boughted flash dont work after same dd
2023-10-02 17:40:15 <u0_a377> mb this usb dont supports linux
2023-10-02 17:43:00 <u0_a377> mb its not possible
2023-10-02 17:43:13 <singham> ikke: What I want to do is, maintain a sheet of MAC addresses, and assign IPs to them
2023-10-02 17:43:39 <singham> like Aa:B0:...:6C 192.168.0.1
2023-10-02 17:44:29 <ikke> That sounds like dhcp
2023-10-02 17:44:46 <Habbie> i do wonder if singham wants bridging, or DHCP+NAT, or ...
2023-10-02 17:46:02 <singham> dhcp is when ips change right?
2023-10-02 17:46:11 <ikke> not necessarily
2023-10-02 17:46:22 <ikke> you can have static addresses (reserved addresses) with dhcp
2023-10-02 17:46:26 <singham> dynamically assign 192.168.0.[2-256]
2023-10-02 17:46:40 <singham> assign ip as connection is established
2023-10-02 17:46:55 <singham> oh ok
2023-10-02 17:47:36 <singham> In bridge what determines connection?
2023-10-02 17:48:31 <singham> Like in dhcp, there is internal ip address connected with MAC, so I see MAC as an inherent id of device and internal ip as an address assigned by server
2023-10-02 17:53:38 <u0_a377> still, why i have this error?
2023-10-02 17:58:24 <Habbie> singham, in a bridge, the networks on both sides become one network
2023-10-02 17:58:37 <Habbie> singham, just like with a switch
2023-10-02 17:59:58 <singham> That is dangerous. I don't want that. I merely want all requests to get resolved and returned. I'll have to read about this.
2023-10-02 18:00:22 <singham> Can you please recommend me some resource for learning basic networking in GNU/Linux ?
2023-10-02 18:01:34 <Habbie> then you do not want bridging
2023-10-02 18:01:50 <ikke> singham: you have 2 options: briding or routing
2023-10-02 18:01:52 <metallicus> minimal: got feedback on #musl from dalias; seems those options are not needed in resolv.conf for the musl resolver (the "broken" glibc resolver needs it though) as musl just uses EDNS with AD as per this commit: https://git.musl-libc.org/cgit/musl/commit/?id=fd7ec068efd590c0393a612599a4fab9bb0a8633
2023-10-02 18:02:21 <singham> Yes, I want routing
2023-10-02 18:02:30 <minimal> metallicus: I'd asked you via DM *where* you saw the mention of those resolv.conf options being needed...
2023-10-02 18:02:34 <singham> So switches bridge primarily?
2023-10-02 18:02:49 <ikke> yes
2023-10-02 18:02:52 <singham> The physical hardware devices?
2023-10-02 18:03:54 <Habbie> yes
2023-10-02 18:07:12 <u0_a377> help
2023-10-02 18:12:14 <u0_a377> help! after dd alpine on new usb i got this, when booting from usb Mounting boot media failed.
2023-10-02 18:12:17 <singham> Isn't there static host control protocol?
2023-10-02 18:12:17 <u0_a377> initramfs emergency recovery shell launched. Type 'exit' to continue boot
2023-10-02 18:14:41 <ikke> u0_a377: any messages before that?
2023-10-02 18:14:53 <u0_a377> no
2023-10-02 18:15:34 <u0_a377> i just downloaded alpine 3.18.4 and dd-ed on new usb
2023-10-02 18:15:44 <u0_a377> and i have thia error
2023-10-02 18:16:04 <u0_a377> but with old usb there is not this problem
2023-10-02 18:16:20 <u0_a377> same command with same iso
2023-10-02 18:16:42 <u0_a377> and after booting from old usb its as usual work
2023-10-02 18:16:44 <singham> So what I want is dhcp with a fixed mac, ip whitelist
2023-10-02 18:40:49 <ikke> u0_a377: You should verify if that other USB is working properly then
2023-10-02 18:40:55 <ikke> sometimes they lie about the sizes
2023-10-02 18:54:10 <u0_a377> i told yo it works
2023-10-02 18:55:25 <ikke> You said with some usb disks but not with others
2023-10-02 18:57:34 <u0_a377> ok i have 3 usb flash drives
2023-10-02 18:57:52 <u0_a377> 3 is new, i buy today
2023-10-02 18:57:58 <u0_a377> and on it this error
2023-10-02 19:02:16 <u0_a377> *i mean third is new
2023-10-02 19:09:07 <markand> can someone gives me its rc-update -av result as a paste please?
2023-10-02 19:09:53 <ikke> On what kind of system?
2023-10-02 19:10:01 <markand> desktop if possible
2023-10-02 19:10:42 <ikke> https://tpaste.us/vNOY
2023-10-02 19:11:54 <markand> muchas gracias 
2023-10-02 19:42:19 <u0_a377> there same text like my error https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device#Troubleshooting
2023-10-02 19:42:42 <u0_a377> but for me this error only on one of my flash drives
2023-10-02 19:42:56 <u0_a377> also idk how fix it still
2023-10-02 20:34:23 <metallicus> https://www.furorteutonicus.eu/2023-10-02-musl-alpine-dns
2023-10-02 21:52:26 <Mani_I6K> hi everyone 
2023-10-02 21:54:47 <Mani_I6K> i want to install alpine linux in dual boot with Mac OS , I have mounted the partition in alpine linux with bootable usb, I have followed the following commands apk add e2fsprogs mkfs.ext4 /dev/disk0s3 apk del e2fsprogs  mount -t ext4 /dev/disk0s3 /mnt  BOOTLOADER=none setup-disk /mnt 
2023-10-02 21:55:30 <Mani_I6K> i have the installation completed
2023-10-02 21:56:17 <Mani_I6K>  and : => initramfs: creating /boot/initramfs-its 
2023-10-02 21:58:54 <Mani_I6K> in the /mnt folder I have all the folders : bin; boot dev; etc;home;libmedia;mnt;opt;procroot;runsbin;srv;syst;tmp;usr;var
2023-10-02 21:59:45 <Mani_I6K> my main boot loader is handeld by Refind, but the boot partition of alpine Linux does not appear 
2023-10-02 22:00:00 <Mani_I6K> someone has an idea please ?
2023-10-02 23:04:45 <minimal> Mani_I6K: well you told setup-disk not to install a bootloader.....so there's nothing for Refind to see/run
2023-10-03 07:48:06 <username_> hey guys, need to cross-compile for aarch64 musl (alpine) on a x86 glibc system, is there any prebuilt toolchain for that? if not, how do I build one myself? thanks!
2023-10-03 07:50:32 <Habbie> username_, this is not the -best- answer but it is one that works for me - with docker + qemu you can run aarch64 alpine in full, and compile inside it
2023-10-03 07:50:35 <Habbie> it'll be slower though
2023-10-03 07:53:02 <username_> yeah, that's an obvious solution but it's very slow. I thought maybe there's some better option
2023-10-03 13:08:33 <elagost> Did we stop getting firefox+firefox-esr updates in stable releases?
2023-10-03 13:09:36 <kwadroke[m]> Edge appears to still get them
2023-10-03 13:11:26 <elagost> yeah, but not stable - I thought at least esr got updated in stable.
2023-10-03 13:11:57 <kwadroke[m]> yeah
2023-10-03 13:12:34 <kwadroke[m]> stable is still on 116.03
2023-10-03 13:13:47 <kwadroke[m]> there's an issue about this: https://gitlab.alpinelinux.org/alpine/aports/-/issues/15325
2023-10-03 17:03:48 <gvnic> hi guys/girls
2023-10-03 17:04:05 <gvnic> just seen that 3.18.4 is out, and this has been added: main/linux-lts: enable Google Virtual NIC (gVNIC) driver
2023-10-03 17:05:01 <gvnic> why not only to the virt release? i just always try to avoid google wherever possible and it makes me uncomfortable, but that's personal ofc.
2023-10-03 17:06:00 <minimal> gvnic: because the person who submitted the MR to add its support wanted it in linux-lts as well as linux-virt
2023-10-03 17:06:46 <minimal> I asked why would it be needed in linux-lts as I expected it to be unusual/unrealistic to use the linux-lits kernel on Google Cloud
2023-10-03 17:06:54 <kwadroke[m]> Isn't this mostly for running VMs on Chromebooks?
2023-10-03 17:07:09 <minimal> nope,m it is for running VMs on some Google Cloud instances
2023-10-03 17:07:23 <minimal> it is an alternative network interface to virtio-net
2023-10-03 17:07:39 <gvnic> yepp that's right @minimal
2023-10-03 17:07:44 <kwadroke[m]> ah ok. I have heard work on VMs for Chromebooks was coming down the line
2023-10-03 17:07:55 <gvnic> can't he just load a .ko if he (for who knows what reason) wants to run a non-virt kernel in GCE?
2023-10-03 17:07:59 <kwadroke[m]> so that's why I was confused
2023-10-03 17:08:41 <minimal> gvnic: you can see here where I asked why add it to linux-lts: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/51894
2023-10-03 17:08:43 <gvnic> i think it has no place in the lts kernel, and not because of my anti-google mentality :D
2023-10-03 17:10:36 <gvnic> yeah I just read that convo @minimal and I completely agree with you
2023-10-03 17:11:13 <gvnic> do you have a link to the sources (since it is in the kernel I can't easily look up a repo)
2023-10-03 17:11:28 <minimal> I just gave up making any more comments in that MR
2023-10-03 17:11:31 <gvnic> because this says it is kinda experimental :D https://github.com/GoogleCloudPlatform/compute-virtual-ethernet-linux
2023-10-03 17:12:54 <minimal> that doesn't mention recent kernel as supported...........as there is no need as it is already part of recent kernels
2023-10-03 17:14:48 <minimal> gvnic: https://github.com/torvalds/linux/tree/master/drivers/net/ethernet/google/gve
2023-10-03 17:15:47 <gvnic> makes sense. thanks, will dig into it a bit
2023-10-03 17:16:35 <minimal> dig into what exactly?
2023-10-03 17:16:47 <gvnic> yeah couple thousand lines as expected I'm not reviewing it :D
2023-10-03 17:16:55 <minimal> why would you?
2023-10-03 17:17:22 <gvnic> well I have systemd ptsd about google
2023-10-03 17:17:38 <minimal> it's a module - don't load it if it bothers you
2023-10-03 17:17:55 <gvnic> yeah I know it is in the modloop
2023-10-03 17:18:20 <minimal> so? if it is not actually loaded...
2023-10-03 17:18:21 <gvnic> is there any step-by-step tutorial how I could use abuild to build a custom kernel? preferably in docker
2023-10-03 17:19:01 <minimal> you want to build a custom kernel as you're worried that that module might magically get loaded some day?
2023-10-03 17:19:27 <strajder> on any system using Linux you can build it from source and cp the bzImage
2023-10-03 17:19:27 <gvnic> I just prefer it not to be loadable at all. maybe something there is designed that an udev event could trigger it getting loaded or whatever
2023-10-03 17:19:39 <minimal> then add a black list for it
2023-10-03 17:19:45 <gvnic> if it is not there it can not be loaded, much simpler for my paranoia to sleep better at night
2023-10-03 17:20:05 <gvnic> and I am also against unnecessary bloat as a principle
2023-10-03 17:20:20 <minimal> I don't believe paranoia is one of Alpine's packaging criteria
2023-10-03 17:20:26 <gvnic> I only use a few types of machines, so a custom kernel for my usecase is much justified since a long time
2023-10-03 17:21:10 <minimal> and then you'd rebuilt your custom kernel each time a new kernel version comes out or a patch is applied?
2023-10-03 17:21:17 <gvnic> no don't get me wrong my paranoia is not related to Alpine. at all! I am really grateful for what you are doing here, my favourite distro by far
2023-10-03 17:21:32 <gvnic> yepp I can set it up in CI easily
2023-10-03 17:21:48 <gvnic> I remember I tried to build a virt kernel with abuild quite some time ago
2023-10-03 17:22:04 <gvnic> and it ended up being 300M -ish so I didn't care to research further what I messed up
2023-10-03 17:22:18 <gvnic> if there's any good tutorial I'd appreciate it much
2023-10-03 17:22:28 <strajder> gvnic: compiling custom kernel used to be the norm, and one of the things done immediately after installation
2023-10-03 17:23:20 <gvnic> yeah it totally makes sense, dunno why it is not anymore. I mean look at me. I kinda know my way around linux (would not say I am an expert) and here I am struggling with a kernel build :D
2023-10-03 17:23:25 <strajder> esp. if you want to remove all the drivers you don't use, it can cut down on the size of kernel and provide a bit of optimization
2023-10-03 17:23:41 <gvnic> yepp for my hardware specifically. just beautiful!
2023-10-03 17:24:22 <minimal> have you doublechecks to make sure there are no other Google drivers lurking in your kernel?
2023-10-03 17:25:07 <gvnic> nah but I will now take time to check everything, not just Google. I've seen in your thread about the AWS drivers etc.
2023-10-03 17:25:24 <gvnic> basically I will experiment to get the bare minimum needed for my system to boot
2023-10-03 17:25:34 <gvnic> and then maybe add some additional stuff reasonable for my usecase
2023-10-03 17:25:49 <gvnic> not the other way around (stripping down the current lts)
2023-10-03 17:26:36 <gvnic_> sorry just had a conn issue
2023-10-03 17:28:18 <gvnic_> so yeah I think it is more sensible to experiment with the minimum requirements to boot
2023-10-03 17:28:30 <minimal> which may be a moving target
2023-10-03 17:28:32 <gvnic_> then maybe add a little extra (external NICs etc.)
2023-10-03 17:29:23 <gvnic_> right, but that means no bloat less risk of planted backdoors from state actors, which is a win for me
2023-10-03 17:29:52 <gvnic_> will recompile if I find myself starting to use new types of external NICs for example
2023-10-03 17:29:54 <strajder> if you need some information on how to build a custom kernel, there is a page on Alpine wiki: https://wiki.alpinelinux.org/wiki/Custom_Kernel but the one on Arch wiki is closer to how it was done back in the day https://wiki.archlinux.org/title/Kernel/Traditional_compilation
2023-10-03 17:30:04 <gvnic_> rebuild is not that painful for me since it can be done in my ci environment
2023-10-03 17:30:12 <minimal> you're expecting "planted backdoors from state actors"? rather than from bad actors
2023-10-03 17:30:46 <minimal> ...like Steven Seagal ;-)
2023-10-03 17:31:18 <gvnic_> of course. the best of the latter always end up working for the former after a single mistake they did :)
2023-10-03 17:33:08 <gvnic_> thanks @strajder !
2023-10-03 17:33:14 <strajder> np
2023-10-03 17:44:13 <singham> What is a gateway?
2023-10-03 17:45:03 <singham> Please explain 1 thing. Who assigns ip address? What differentiates internal and external ip addresses
2023-10-03 17:45:37 <minimal> those are general networking questions, not specific to Alpine
2023-10-03 17:45:47 <minimal> I think you need to read up on networking
2023-10-03 17:47:01 <singham> Alright
2023-10-03 17:47:27 <gvnic> hey my internet is not really stable. is there a way to check messages received since my last disconnect? i am using the oftc web client
2023-10-03 17:48:47 <singham> ikke: The issue is I am trying this
2023-10-03 17:49:01 <singham> https://wiki.alpinelinux.org/wiki/Bridge
2023-10-03 17:49:11 <singham> The latter part, configuration file
2023-10-03 17:49:34 <singham> And there is ppp0 which is achieved after negotiation
2023-10-03 17:49:51 <singham> This is not working when I replace ppp0 with eth0
2023-10-03 17:50:44 <heat_> gvnic, the channel is logged, see the topic
2023-10-03 18:01:33 <gvnic> @heat_ nice, thanks! 
2023-10-03 18:14:27 <WhyNotHugo> Are wlroot's protocol xml files shipped in any package?
2023-10-03 18:14:48 <WhyNotHugo> I.e.: the protocols from https://gitlab.freedesktop.org/wlroots/wlr-protocols
2023-10-03 19:05:32 <Saijin_Naib[m]> A local root vulnerability in glibc
2023-10-03 19:05:32 <Saijin_Naib[m]> https://lwn.net/Articles/946381/
2023-10-03 19:05:49 <Saijin_Naib[m]> Alpine and musl are exempted from this CVE
2023-10-03 19:14:55 <idkrn[m]> <Saijin_Naib[m]> "Alpine and musl are exempted..." <- Common Alpine/musl W
2023-10-03 19:15:21 <Saijin_Naib[m]> Right? 🤣 
2023-10-03 19:15:29 <Saijin_Naib[m]> Nice to see it acknowledged 
2023-10-03 19:16:05 <Saijin_Naib[m]> Seeing stuff like this would certainly have brought my attention to Alpine sooner 
2023-10-03 19:31:27 <quinq> Why didn't it?
2023-10-03 19:33:51 <Saijin_Naib[m]> I'm not sure if I understand your question 
2023-10-03 19:44:01 <quinq> My question is why didn't you bring your attention to Alpine sooner? :)
2023-10-03 19:44:27 <lopid> explain yourself
2023-10-03 19:47:02 <quinq> yes!
2023-10-03 19:49:21 <Saijin_Naib[m]> Oh, because all I saw about it was people saying it's awful as a Desktop OS and only appropriate for container or server usage 
2023-10-03 19:50:32 <Saijin_Naib[m]> Eventually, I got into it by using AlpineWSL under Windows 10, and then got curious enough to ditch my Debian and Solus installs 
2023-10-03 19:50:54 <Saijin_Naib[m]> Now it's my desktop OS for all my machines 🤷
2023-10-03 19:51:59 <Saijin_Naib[m]> I needed to run Linux at work on a managed machine with very limited resources, so needed to be via Store, so AlpineWSL fit the bill 
2023-10-03 19:52:27 <quinq> Thanks for explaining, Saijin_Naib[m] :)
2023-10-03 19:52:48 <lopid> judgement pending
2023-10-03 19:53:03 <quinq> Please tell us when it's ready!
2023-10-03 19:53:12 <Saijin_Naib[m]> Yeah, no problem. It's phenomenal as a Desktop OS, and by far the easiest to maintain, so I evangelize for it everywhere I can 
2023-10-03 19:53:58 <Saijin_Naib[m]> Biggest stumbling block I see for other users is the text mode installer. Everyone bounces off it, unless I set it up for them 
2023-10-03 19:54:31 <Saijin_Naib[m]> But once it's up, folks love it, and it's rock solid 
2023-10-03 19:56:00 <Saijin_Naib[m]> Having PIE and SSP and other nice musl/Alpinisms just makes it even better for general desktop usage as whole classes of vulnerabilities are largely not applicable, it seems, which is great when supporting non-technical users 
2023-10-03 19:56:51 <Saijin_Naib[m]> Once again , the most amazing aspects Alpine are its reduced cognitive load to manage and how robust it is 
2023-10-03 20:31:46 <Ermine> mglru is also google, be sure to turn it off!
2023-10-03 21:21:13 <Forza> Good thing Alpine uses musl: https://www.openwall.com/lists/oss-security/2023/10/03/2
2023-10-03 21:27:11 <quinq> ooold
2023-10-03 22:26:09 <stonley890> Hello! I'm trying to install apache2 with php but the guide (https://wiki.alpinelinux.org/wiki/Setting_Up_Apache_with_PHP) doesn't work. It looks like the php5-apache2 package isn't available in the pkg database. Is there any way I can work around this? Am I doing something wrong?
2023-10-03 22:31:13 <Ermine> try php8, as php 8.x is current version
2023-10-03 22:31:21 <Ermine> php8-apache2
2023-10-03 22:32:51 <stonley890> apk add apache2 php8-apache2 > ERROR: unable to select packages: php8-apache2 (no such package): required by: world[php8-apache2]
2023-10-03 22:34:56 <fluix> you need the community repository enabled
2023-10-03 22:35:28 <fluix> err sorry, you nedd php81 or php82 as the prefix instead
2023-10-03 22:35:44 <fluix> https://pkgs.alpinelinux.org/packages?name=php*apache*
2023-10-03 22:36:02 <stonley890> ah, that works perfectly. Thank you so much!
2023-10-04 02:34:23 <ahmadraniri[m]> Hello, I'm using dracut instead of the default initramfs generator. Everytime kernel upgrade dracut always generate the initramfs as "dracut-lts" instead of "initramfs-lts". That makes everytime we reboot, we will face a kernel panic, manually copy dracut-lts into initramfs-lts will fix that problem. My question is how to tell dracut to generate initramfs-lts by default instead of dracut-lts ? thanks.
2023-10-04 03:00:20 <zcrayfish> Dunno, but if you don't find a way to fix dracut... If your boot partition is a suitable type, a symlink could be another option 
2023-10-04 04:43:19 <jonno> is there a way to get selenium working with alpine? I keep getting: selenium.common.exceptions.WebDriverException: Message: Process unexpectedly closed with status signal
2023-10-04 10:53:43 <Forza> ahmadraniri[m]: is there an option in dracut.conf to change the output name? There is otherwise a command line flag for it. 
2023-10-04 10:54:50 <Forza> Is the guide https://wiki.alpinelinux.org/wiki/Setting_Up_Apache_with_PHP completely wrong nowadays. I saw the discussion before. I think we should just remove the complicated oho version test and just suggest looking at current packages instead? 
2023-10-04 10:56:34 <Forza> I mean, commands like this does not seem right in a simple setup guide : export phpverx=$(alpinever=$(cat /etc/alpine-release);[ ${alpinever//./} -ge 309 ] && echo  7|| echo 5)
2023-10-04 11:06:58 <Forza> Is it just for me, but I can't see the button to flag this out of date : https://pkgs.alpinelinux.org/package/v3.18/community/x86_64/caddy
2023-10-04 12:04:09 <minimal> ahmadraniri[m]: this has been discussed before for both dracut and booster and the suggested solution is to create a /boot/initramfs-lts symlink pointing to /boot/dracut-lfs, see: https://gitlab.alpinelinux.org/alpine/aports/-/issues/14164
2023-10-04 12:06:55 <Forza> Hm it is weird. I use Dracut on my Gentoo box and it doesnt create files names dracut-*
2023-10-04 12:08:39 <minimal> also this documented in the Alpine dracut package: https://git.alpinelinux.org/aports/tree/community/dracut/README.alpine#n40
2023-10-04 12:09:33 <minimal> based on past experience I think Alpine should rename all such files to DO-NOT-README.Alpine so that people do the opposite and actually read them lol
2023-10-04 12:11:52 <minimal> Forza: on Alpine that is controlled by the trigger, obviously on Gentoo it uses a different mechanism and names them differently
2023-10-04 12:15:04 <staceee> hey there. I reboot a machine that was sleeping for a long time. It now have expired ssl certs. How can I upgrade? o_O
2023-10-04 12:16:31 <Forza> minimal: yea, likely so. Which means we should be able to do the same here? 
2023-10-04 12:17:30 <minimal> Forza: do what, change the trigger contents in the dracut package? that would then prevent someone having both mkinitfs and dracut installed with mkinitfs still working as before
2023-10-04 12:17:43 <Forza> minimal: reading that readme, it seems a design choice for Alpine to have a separate Dracut initramfs 
2023-10-04 12:18:09 <minimal> yes, so that more than 1 initramfs tool can be installed at same time
2023-10-04 12:18:23 <Forza> I mean to chabge Dracut to create initramfs named the same as mkinitfs
2023-10-04 12:18:51 <Forza> Is it common to have more than one? 
2023-10-04 12:19:50 <minimal> Forza: that's what I'm referring to - if I have a system using mkinitfs and then I install dracut (whether just to look at it or I want to be able to boot using either mkinitramfs' or dracut's initramfs then there's a problem if dracut AND mkinitfsmfs want to create the initramfs-lts file
2023-10-04 12:19:51 <Forza> Tools can be the same, but is it really supported to use two tools to update the boot loader?
2023-10-04 12:20:35 <Forza> Ok there's a point. Can we do something with the bootloaders to include Dracut images? 
2023-10-04 12:21:07 <minimal> Forza: update the bootloader how exactly? the initramfs tools don't update the bootloader, they create a initramfs file (whether a bootloader referrences more than 1 initrqamfs in different boot menu options is another matter)
2023-10-04 12:21:51 <Forza> You know what I mean. 
2023-10-04 12:22:23 <Forza> To make sure grub and syalinux reference both/all initramfs files 
2023-10-04 12:22:50 <minimal> Forza: not easy, first there are multiple bootloaders that would require changes, also do you mean that the bootloaders should create multiple boot options when more than 1 initramfs file discovered? if so, how should they decide which is the default boot option? etc...
2023-10-04 12:23:17 <Forza> Another option is maybe renaming them initramfs-lts-dracut, I believe at least grub-mkconfig will find it
2023-10-04 12:23:44 <Forza> I have more than one grub entry on my Alpine server 
2023-10-04 12:24:08 <minimal> any "solution" needs to be consistent, not just "let's fix it for Grub and forget about other bootloaders"
2023-10-04 12:24:29 <minimal> what about Limine? u-boot? Raspberry Pi's bootloader? etc
2023-10-04 12:24:35 <Forza> It has both default initramfs-lts/kernel-lts and a custom built kernel/initramfs that we use 
2023-10-04 12:25:16 <Forza> Sure. So what loaders do we have in Alpine that is officially supported and make sure these work? 
2023-10-04 12:25:58 <Forza> The symlink option does not work on fat32, if boot is on the efi partition 
2023-10-04 12:26:07 <minimal> what about EFISTUB? which initramfs should it use? you can't have more than 1 per EFISTUB. What about UKI? (same issue)
2023-10-04 12:26:21 <Forza> Be creative and look for solutions instead of saying we can't fix it? 
2023-10-04 12:26:57 <minimal> I never said it cannot be fixed, indeed I said a consistent solution is needed, not just to "fix" your specific requirements
2023-10-04 12:27:57 <Forza> I just put proposed some options based on the current topic. But sure, I'm not going to be ignorant of other options 
2023-10-04 12:28:00 <minimal> as for FAT32, yeah that's a limitation but I never understood why /boot is placed on ESP partition rather than its' own partition (which is what I do myself)
2023-10-04 12:28:21 <Forza> Space usage? 
2023-10-04 12:29:09 <Forza> Especially if you use efi kernel boot thingy 
2023-10-04 12:29:37 <minimal> EFISTUB or UKI
2023-10-04 12:31:03 <minimal> if it was decided that you cannot have more than 1 initramfs creation tool installed then things would be easy to address - make mkinitfs/booster/dracut conflict with each other and have all 3 packages name their file initramfs-<flavour>
2023-10-04 12:31:31 <minimal> the present situation of permitting more than 1 to be installed leads to these complications
2023-10-04 12:33:54 <Forza> The issue is with how the boot loaders reference these, no? 
2023-10-04 12:35:15 <minimal> some bootloader don't provide a menu/don't let you define more than a single boot entry
2023-10-04 12:35:29 <Forza> And we can't exactly say we have to support all in all configurations either. Syslinux used to have an issue allowing for root/boot on btrfs subvols. It works in grub, but should we block grub or btrfs because of Syslinux? 
2023-10-04 12:36:07 <minimal> some bootloaders (i.e. Grub) do potentially support multiple entries using different initramfs' but even then the question of WHICH such entry comes first/is used by default comes up
2023-10-04 12:36:14 <Forza> Yea sure. No issue. And if you want a boot menu, use a loader that has one? 
2023-10-04 12:36:45 <Forza> That's the case. But it can be controlled I think
2023-10-04 12:37:27 <minimal> EFISTUB - there's only 1 initramfs used by a EFISTUB, so if more than 1 initramfs (booster/dracut/mkinitfs) is present do you create more than 1 EFISTUB?
2023-10-04 12:39:07 <minimal> Forza: with RPI there's no boot menu, it only boots 1 thing, changing the initramfs used requires manually modify 1 of its bootloader config files
2023-10-04 12:39:47 <Forza> What do you suggest? 
2023-10-04 12:41:55 <minimal> well I guess, someone should define the "requirements" in general, which includes how to handle cases such as where a particular bootloader can't handle more than 1 boot entry, how to order boot menus when more than 1 initramfs present
2023-10-04 12:42:20 <minimal> once the requirements are defined then the various iniramfs tools and bootloaders in Alpine can be modified to meet those requirements
2023-10-04 12:43:00 <minimal> and to note any limitations a particular bootloader may have with regard to the requirements
2023-10-04 12:45:30 <whatislinux> Hey, I am trying to learn about Linux. My goal is to have  a lean server thats used for running Node.js, Java  & MySQL applications. 
2023-10-04 12:45:42 <whatislinux> Is alpine-linux just as secure as Ubuntu?
2023-10-04 12:47:22 <minimal> as secure? or as insecure? ;-)
2023-10-04 12:48:34 <whatislinux> what do you mean?
2023-10-04 12:48:53 <whatislinux> how do i learn indepth of alpine-linux?
2023-10-04 12:50:45 <Ermine> by reading source code and documentation
2023-10-04 12:52:39 <whatislinux> ok thank you
2023-10-04 12:53:09 <lopid> not sure i would put much faith into the alpine docs. they can be quite sparse and out of date
2023-10-04 12:54:46 <whatislinux> did you guys read alpine source code? if so,  what was your use case?
2023-10-04 12:55:24 <Forza> minimal: maybe bring it up in a bug report on gitlab? 
2023-10-04 12:57:03 <Forza> whatislinux: security is so many different things. So, before knowing how secure a system is, you should find out what type of attack vectors you have or want to protect against. 
2023-10-04 12:58:30 <whatislinux> @forza my use case is a public server
2023-10-04 12:58:41 <minimal> Forza: I'd think it's more something for TSC, especially as it's "hard" to raise an Issue for multiple packages (i.e. booster & dracut & mkinitfs)
2023-10-04 13:05:08 <Forza> What's tsc? 
2023-10-04 13:05:51 <Forza> whatislinux: well. On public servers it depends a lot on the services exposed, how users interact and manage, etc etc. 
2023-10-04 13:05:51 <minimal> Technical Steering Committee
2023-10-04 13:06:03 <Forza> minimal: makes sense 
2023-10-04 13:06:14 <minimal> https://gitlab.alpinelinux.org/alpine/tsc
2023-10-04 13:06:20 <whatislinux> @forza, it will be an HTTPS server
2023-10-04 13:06:28 <Forza> whatislinux: maybe ask in #security on Libera.chat IRC network? 
2023-10-04 13:07:05 <whatislinux> @forza what do you want me to ask them?
2023-10-04 13:07:28 <Forza> whatislinux: sure. Then load an old version of WordPress or an node.js app and boom, you have fun visitors inside your root 
2023-10-04 13:09:09 <Forza> whatislinux: well, you have make an open ended question like you did, as there are many different answers and the most correct one is "it depends". 
2023-10-04 13:11:35 <whatislinux> @forza, what more info would you like from me?
2023-10-04 13:12:19 <Forza> I don't have definite answers anyway. I'm just a normal user with a bit of sys admin experience. 
2023-10-04 13:14:17 <Forza> IMO specific questions are easier to answer. Not broad ones. As an http server, alpine would do well. But my point being is that security depends more on the very specific application and setup you have. 
2023-10-04 13:14:59 <whatislinux> why is alpine linux not used in banks or any real workplace?
2023-10-04 13:15:06 <whatislinux> @forza
2023-10-04 13:15:40 <Forza> Maybe because of corporate things want enterprise support
2023-10-04 13:16:00 <Forza> And it is easy to go with red hat or suse or oracle 
2023-10-04 13:16:13 <whatislinux> Otherwise alpine linux is just as secure as Ubuntu, debian? Does it have security certifications like them?
2023-10-04 13:16:13 <Forza> (and pay millions for it... 🤦‍♂️) 
2023-10-04 13:16:26 <Forza> I do not know 
2023-10-04 13:16:30 <whatislinux> whats wrong with paying millions?
2023-10-04 13:16:43 <Forza> It may not be. 
2023-10-04 13:16:50 <Forza> Or it is a waste 
2023-10-04 13:17:03 <Forza> In the company I work for we have both 
2023-10-04 13:17:06 <whatislinux> so you are saying alpine linux is a waste?
2023-10-04 13:17:49 <whatislinux> How do you get in touch with official alpine-lix devs?
2023-10-04 13:17:53 <Forza> I have switched to Alpine from Ubuntu and RedHat because of my own interest and want an easier life administrating the stuff I am responsible for
2023-10-04 13:18:22 <Forza> Try #alpine-devel? Maybe some contact on their website 
2023-10-04 13:18:35 <whatislinux> @forza how do you know you made the right decision if you dont even know if its as secure as ubuntu
2023-10-04 13:18:38 <whatislinux> alpine-devel is dead
2023-10-04 13:19:11 <Forza> It's not? Maybe just not active at this moment. Remeber time zones... 
2023-10-04 13:20:59 <Forza> Security.. I have isolated environments, which means no Internet or unauthorised access. That's one level of security. Alpine is also small and easily understood, which decrease the risk of bad config. And it does only what I want it too. 
2023-10-04 13:21:51 <whatislinux> @forza I meant real security when something is connected to the internet
2023-10-04 13:22:58 <Forza> Maybe you want audit logging and all that then. I have not used that on Alpine yet 
2023-10-04 13:23:13 <Forza> You're not very specific 
2023-10-04 13:23:17 <Forza> :D
2023-10-04 13:23:36 <Forza> Why is air gap not real security? 
2023-10-04 13:24:22 <Forza> And stuff like jump environments, reverse proxies, application load balancers, and so on. It is a very complex topic 
2023-10-04 13:25:15 <whatislinux> Load balancers, reverse proxies are all conencted to the internet
2023-10-04 13:25:27 <whatislinux> because its paperweight, useless if no internet
2023-10-04 13:25:43 <Forza> They are. 
2023-10-04 13:26:51 <Forza> But the ip stack in Alpine is the same as on every other Linux. It comes down to how you set up things more than which exact os you use. 
2023-10-04 13:28:02 <durrendal> whatislinux: what does Ubuntu have insofar as security goes that you're not seeing in Alpine? Are you looking for something specific?
2023-10-04 13:28:35 <durrendal> security of your servers will come down to how you configure them. you can completely botch a "secure" RHEL install, even though it's "enterprise" and "secure"
2023-10-04 13:31:19 <whatislinux> durrendal. are you a linux-alpine dev?
2023-10-04 13:32:16 <mancomb_> Hopefully someone can help. I'm very new to Alpine benn running debian for years but I am trying to get to run some lxc containers on a fresh alpine install, but when I install lxc from apk it messes up my machine:
2023-10-04 13:32:16 <mancomb_> # apk add lxc
2023-10-04 13:32:16 <mancomb_> # mount
2023-10-04 13:32:16 <mancomb_> ...
2023-10-04 13:32:16 <mancomb_> Error relocating /bin/mount: mnt_context_enable_noautofs: symbol not found
2023-10-04 13:32:18 <mancomb_> Error relocating /bin/mount: mnt_context_enable_onlyonce: symbol not found
2023-10-04 13:32:20 <mancomb_> # apk del lxc
2023-10-04 13:32:22 <mancomb_> ...
2023-10-04 13:32:25 <mancomb_> # mount
2023-10-04 13:32:27 <mancomb_> -> works as expected
2023-10-04 13:33:33 <minimal> mancomb_: are you mixing Alpine versions perhaps? does your /etc/apk/repositories reference more than a single Alpine version?
2023-10-04 13:34:05 <mancomb_> I setup the repos with "setup-apkrepos -c" other than that no
2023-10-04 13:34:48 <mancomb_> But yes... it seems that there is both 3.18 and edge in there..  Will try to fix :/
2023-10-04 13:36:08 <mancomb_> minimal: Thx, not sure how I got both in there. But definetly the reason for this
2023-10-04 13:48:31 <whatislinux> @durrendal ?
2023-10-04 13:48:33 <whatislinux> there?
2023-10-04 13:53:38 <durrendal> whatislinux: not a core dev, just a contributor. But I've been running Alpine in production environments for about 6 years now, moved from RHEL/Debian
2023-10-04 13:55:10 <whatislinux> Can you teach me core of linux?
2023-10-04 13:55:20 <whatislinux> having a teacher makes it easy
2023-10-04 13:55:23 <whatislinux> are you on discord?
2023-10-04 13:58:20 <Kladky> What do you mean by "core of linux"?
2023-10-04 14:00:14 <whatislinux> how alpine linux works, what do the packages do. 
2023-10-04 14:02:00 <Forza> Best way is to simply start using it
2023-10-04 14:02:12 <Kladky> If you want to know how package management works check out the wiki page: https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper
2023-10-04 14:02:38 <Forza> And then ask if you find a promblem or issue doing something specific 
2023-10-04 14:07:22 <durrendal> I'll third that, the best way to learn is to do, homelabbing is a great way to learn by doing
2023-10-04 14:38:15 <Test111> Hi, I'm trying to install Alpine on a single drive where Windows is also installed. I followed the dual booting guide on the wiki but there's still something wrong with my installation.
2023-10-04 14:38:15 <Test111> 2) I am trying to install alpine in a partition, so the command I run is `setup-disk -m cryptsys /mnt`. Although the installation continues, I get 2 errors after the first step: `grub-install: error: failed to get canonical path of {2 symbols here}.` and `install: can't stat '/EFI/alpine/grubx64.efi': No such file or directory`.
2023-10-04 14:38:15 <Test111> 1) I managed to install refind, my device directly boots to it. Unfortunately, I used an unorthodox configuration to make it work. I don't have EFI/refind, I dumped refind into EFI/Boot.
2023-10-04 14:38:15 <Test111> I tried booting to Alpine through Refind and I get init issues, can't properly start up the system. I noticed `apk` couldn't download the repos and `doas` was nonexistent along with some other things. I'm not sure what I'm doing wrong.
2023-10-04 14:38:38 <Test111> Oops, it sent all the paragraphs as different messages, sorry.
2023-10-04 14:54:50 <Test111> Any ideas?
2023-10-04 15:00:54 <minimal> Test111: what "symbols"?
2023-10-04 15:01:09 <Test111> something like these `' 
2023-10-04 15:01:58 <Test111> I found the exact same issue here: https://gitlab.alpinelinux.org/alpine/alpine-conf/-/issues/10555
2023-10-04 15:02:21 <Test111> It's supposed to be fixed, and I used ext4 anyway
2023-10-04 15:14:27 <minimal> did you have the ESP partition mounted?
2023-10-04 15:15:29 <Test111> All I mounted was the partition I ran the setup-disk command on. `mount /dev/nvme0n1p5 /mnt`
2023-10-04 15:16:18 <minimal> you're using UEF~I so the ESP partition also needs to be mounted so setup-disk can write EFI file(s) etc to it
2023-10-04 15:16:25 <Test111> makes sense
2023-10-04 15:19:14 <Test111> Something shows up on refind: "boot\vmlinuz-lts from 90 GiB ext4 volume" which is /dev/nvme0n1p5. When I select it, I get mount errors: "mounting /dev/nvme0n1p5 on /sysroot failed: No such file or directory", "Mounting root: failed". Then "initramfs emergency recovery shell launched. Type 'exit' to continue boot"
2023-10-04 15:19:36 <Test111> Do you think I get these errors because I forgot to mount ESP during installation?
2023-10-04 15:21:00 <minimal> Test111: no idea. Is nvme module in the initramfs?
2023-10-04 15:24:35 <Test111> I'm not sure why I need to know that
2023-10-04 15:24:44 <Test111> I just followed the wiki, and this is where I am
2023-10-04 15:25:11 <Test111> Is there a module special for nvme?
2023-10-04 15:25:57 <minimal> yes nvme is handled via a loadable kernel module
2023-10-04 15:26:30 <minimal> if it is not present in the initramfs then obviously the initramfs cannot load it to access and mount the rootfs
2023-10-04 15:27:15 <minimal> I'm not saying the nvme module is definately not in the initramfs, I'm asking if it is not
2023-10-04 15:27:20 <Test111> I thought nvme was so common at this point it'd be built into the default kernel provided by alpine
2023-10-04 15:29:28 <j_v> does anyone know of a webpage or doc that might explain why armhf binaries are larger than armv7? i expected it to be the opposite, since armhf has the fpu so would have thought the soft float would bloat the armv7 binaries. been searching, but no joy so far
2023-10-04 15:29:44 <minimal> when you get to the emergency shell run "lsmod" and see if nvme is listed there
2023-10-04 15:30:12 <ikke> j_v: armhf is armv6 with hard float, it does not imply armv7 does not have it
2023-10-04 15:30:20 <ikke> j_v: there was an armv6 variant without hf
2023-10-04 15:32:44 <j_v> ikke: thanks for explanation.
2023-10-04 15:33:58 <Test111> minimal: Yeah nvme is there
2023-10-04 15:35:24 <minimal> Test111: well if you didn't have /boot and/or ESP partition mounted before you ran setup-disk then I doubt it installed things correctly
2023-10-04 15:36:22 <Test111> Okay, I'll try mounting ESP next time.
2023-10-04 15:38:15 <minimal> Test111: I assume you have either separate /boot or else a combined /boot and ESP partition
2023-10-05 04:18:07 <_sou-`_> https://www.youtube.com/watch?v=sqSA-SY5Hro
2023-10-05 04:18:08 <_sou-`_> These rich men north of Richmond
2023-10-05 04:18:08 <_sou-`_> Lord knows they all just wanna have total control
2023-10-05 04:18:09 <_sou-`_> Wanna know what you think, wanna know what you do
2023-10-05 04:18:09 <_sou-`_> And they don’t think you know, but I know that you do
2023-10-05 04:18:10 <_sou-`_> ‘Cause your dollar ain’t shit and it’s taxed to no end
2023-10-05 04:18:10 <_sou-`_> ‘Cause of rich men north of Richmond
2023-10-05 04:18:11 <_sou-`_> irc.supernets.org #superbowl
2023-10-05 04:35:13 <somercet1> Fed plant
2023-10-05 04:44:52 <dwfreed> please don't comment on the spam
2023-10-05 04:54:02 <somercet1> Can't we add some keywords to kill this? It's the same every time.
2023-10-05 04:58:42 <dwfreed> you will note they already were
2023-10-05 04:59:07 <dwfreed> not necessarily as quickly as possible, but they were handled
2023-10-05 05:13:52 <somercet1> I've seen it on Rizon and Libera.
2023-10-05 05:14:26 <Daanct12> have yet to see one on libera, maybe thats because im not in big channels or it didnt pick one in my list
2023-10-05 05:14:58 <somercet1> probably hit some threshold and got booted before it got to your chans
2023-10-05 05:15:32 <somercet1> I have a small personal room on Rizon and it ended up in there /me shrugs
2023-10-05 05:15:58 <Daanct12> ah
2023-10-05 05:16:26 <Daanct12> sounds like it'll hit any channel then
2023-10-05 05:17:25 <invoked> like dwfreed said, let staff handle it
2023-10-05 05:37:13 <dwfreed> the discussion regarding the spam has now exceeded the length of the spam, congrats
2023-10-05 05:48:01 <Forza> 
ACTION morning

2023-10-05 10:29:40 <Test111> I don't get how I'm supposed to mount ESP. I already mount the partition I want to install Alpine on at /mnt. Where can I mount ESP?
2023-10-05 10:31:29 <ikke> iirc, the esp is mounted at /mnt/boot then
2023-10-05 10:32:12 <Test111> Oh yeah I think that's right
2023-10-05 10:32:29 <Test111> You manually mount it at /mnt/boot afaik
2023-10-05 10:33:04 <Test111> but wait
2023-10-05 10:33:22 <Test111> `/mnt/boot` is the usual linux /boot.
2023-10-05 10:33:36 <Test111> I'm seeing it right now
2023-10-05 10:33:51 <ikke> https://wiki.archlinux.org/title/EFI_system_partition#Typical_mount_points
2023-10-05 10:33:52 <Test111> It's not my ESP where refind's located
2023-10-05 10:34:45 <strajder> ESP is usually mounted at /boot/efi in Alpine
2023-10-05 10:35:55 <strajder> if you are mounting for chroot at /mnt, then that will become /mnt/boot/efi
2023-10-05 10:39:52 <Test111> Okay I think I did it
2023-10-05 10:40:37 <Test111> Mounted the partition I want to install alpine on at /mnt and mounted ESP at /mnt/boot so that /mnt/boot/EFI is where refind is found
2023-10-05 10:46:08 <minimal> "it depends" - you can either have separate boot (i.e. /boot) and ESP (i.e. /boot/esp) partitions, or have a single ESP partition (i.e. /boot) that is also used for boot files
2023-10-05 10:46:35 <minimal> from memory setup-disk defaults to using a single/combined boot/ESP partition
2023-10-05 10:47:05 <minimal> doh, replace "/boot/esp" with "/boot/efi"
2023-10-05 10:49:40 <strajder> on two of my machines where I have Alpine installed (one full disk automatic partitioning, other manual partitioning with dual boot), ESP is assigned /boot/efi by setup
2023-10-05 11:02:03 <Test111> so I managed to install Alpine but now I can't boot into Refind
2023-10-05 11:04:06 <minimal> how did you install Refind?
2023-10-05 11:05:21 <Test111> I dumped refind's files into ESP's Boot directory.
2023-10-05 11:05:59 <Test111> Renamed refindx64.efi to bootx64.efi 
2023-10-05 11:06:14 <minimal> that's you problem then
2023-10-05 11:06:42 <minimal> setup-disk changed BOOT/BOOTX64.EFI to be the Grub EFI file...
2023-10-05 11:06:56 <Test111> Oh...
2023-10-05 11:07:04 <minimal> why are you using "fallback" booting?
2023-10-05 11:07:59 <minimal> "fallback" booting is only used when there are no UEFI boot entries/variables created
2023-10-05 11:08:02 <Test111> This pc was given to me and I don't know its bios admin password. I can't change the boot order or anything, and I think this is the reason why the pc just won't recognize any new boot entries
2023-10-05 11:08:17 <Test111> Refind or Alpine won't show up in the boot menu
2023-10-05 11:08:22 <minimal> you're using UEFI
2023-10-05 11:08:38 <minimal> normal practice is to create UEFI boot entries
2023-10-05 11:08:50 <Test111> I know but it wouldn't work
2023-10-05 11:08:53 <minimal> typically using efibootmgr
2023-10-05 11:08:59 <Test111> I created a normal refind entry
2023-10-05 11:09:08 <Test111> Still wouldn't show up in the boot menu
2023-10-05 11:09:11 <minimal> how?
2023-10-05 11:09:38 <minimal> also which "boot menu"? the UEFI's own menu?
2023-10-05 11:10:08 <Test111> Yeah that should be it minimal. I turn on the pc, hit F12 and a boot menu appears
2023-10-05 11:10:18 <Test111> I booted into a live distro, used refind's auto installer
2023-10-05 11:10:30 <minimal> but you never created a UEFI boot entry/variable for Refind?
2023-10-05 11:10:38 <Test111> I saw the refind entry in UEFI
2023-10-05 11:10:51 <Test111> I used a tool on Windows (unfortunately) to create it
2023-10-05 11:10:53 <Test111> still nothing
2023-10-05 11:11:28 <Test111> btw refind's guide didn't mention creating a uefi boot entry. I think it does that automatically
2023-10-05 11:12:01 <minimal> what does "uefibootmgr" show for "Boot000X" entries?
2023-10-05 11:12:40 <minimal> sorry, "efibootmgr"
2023-10-05 11:12:41 <Test111> let me boot into alpine
2023-10-05 11:14:45 <Test111> it shows nothing pertaining to refind
2023-10-05 11:14:48 <Test111> or alpine 
2023-10-05 11:15:00 <minimal> what does it show?
2023-10-05 11:15:58 <Test111> Is see Windows boot manager, setup fvfile, boot menu, diagnostic splash screen and a bunch of other vendor stuff. USB CD, USB FDD, nvme0, nvme1, PXE boot
2023-10-05 11:16:02 <Test111> I see*
2023-10-05 11:16:37 <Test111> some other stuff as well but nothing that has to do with linux
2023-10-05 11:19:01 <minimal> so then yes it sounds like no boot entries for either refind or alpine, which is why then the fallback is used
2023-10-05 11:19:27 <minimal> but, like Highlander, there can only be one fallback! ;-)
2023-10-05 11:20:56 <Test111> I booted to a refind session off a usb drive and reinstalled refind. there's now a refind directory in /boot/EFI
2023-10-05 11:21:15 <Test111> i'll try to add a uefi boot entry with efibootmgr
2023-10-05 11:22:31 <minimal> well as you want to use Refind for booting everything then you either add a boot entry for Refind, or else use Refind as the fallback file
2023-10-05 11:22:57 <Test111> is the file in EFI/boot the fallback?
2023-10-05 11:24:45 <Test111> since I can't change the boot order, I should make refind the fallback option, create a new efi entry for alpine, and add alpine's new boot entry to refind.conf
2023-10-05 11:24:59 <minimal> yes EFI/BOOT/BOOTX64.EFI is the fallback file
2023-10-05 11:25:42 <minimal> well you *can* change the boot order using "efibootmgr"
2023-10-05 11:26:51 <minimal> however many UEFI implementations don't show you a boot menu, they simply go through the boot order list and boot the first entry that works (unless you press the appropriate button when UEFI starts)
2023-10-05 11:35:14 <Test111> Okay it worked
2023-10-05 11:35:21 <Test111> Thank you minimal
2023-10-05 11:35:50 <Test111> You've been a great help
2023-10-05 13:27:59 <Test111> Any idea how I can specify the domain in iwd peap network config? I checked the iwd and arch wikis to no avail
2023-10-05 13:28:56 <Test111> or rather, i don't have a certificate but i'd like to connect to a network that uses peap regardless. 
2023-10-05 13:29:28 <Test111> I can do this just fine on Android.
2023-10-05 17:13:05 <vishal_> Does alpine linux get security updates as fast as Ubuntu or another popular distros? Will it be a good fit for prod?
2023-10-05 17:15:13 <strajder> Ubuntu, lol
2023-10-05 17:15:30 <vishal_> ?
2023-10-05 17:15:56 <Habbie> ubuntu is not fast with security updates for 90% of their packages
2023-10-05 17:15:59 <lopid> same guy
2023-10-05 17:16:00 <Habbie> so it's not a great comparison
2023-10-05 17:16:01 <strajder> the most recommended distro for beginners, and probably the worst
2023-10-05 17:17:36 <vishal_> yes, i am the same guy. 
2023-10-05 17:17:48 <vishal_> Did you think I was trying to hide?
2023-10-05 17:18:06 <vishal_> Ubuntu does not have in time security patches? You are the first person to say that.
2023-10-05 17:20:13 <Habbie> i don't want to trigger a whole off topic discussion here, but I want to leave you with this convenient table from Ubuntu that says so https://ubuntu.com/pro
2023-10-05 17:21:12 <vishal_> What are you pointing to in that table?
2023-10-05 17:21:23 <vishal_> Kernel livepatch?
2023-10-05 17:21:59 <Habbie> the numbers '2,300' and '23,000'
2023-10-05 17:22:37 <vishal_> Are you saying I should get Pro?
2023-10-05 17:22:46 <Habbie> no
2023-10-05 17:22:56 <Habbie> i can't imagine ever saying that to anybody :)
2023-10-05 17:23:00 <Habbie> anyway, none of this was your question
2023-10-05 17:23:22 <vishal_> Drop the elitist attitude man, I am just looking to understand.
2023-10-05 17:23:24 <Habbie> you wanted to know if alpine gets security updates fast
2023-10-05 17:23:49 <vishal_> yeah because its volunteers who maybe busy with their real jobs
2023-10-05 17:23:58 <vishal_> Where Ubuntu development is people's real job
2023-10-05 17:33:13 <vishal_> you all dead?
2023-10-05 17:33:13 <vishal_> fuck
2023-10-05 17:42:04 <fluix> nobody is dead, people just don't want to respond to your questions because they're very surface level and show a misunderstanding of different aspects of security
2023-10-05 17:42:50 <fluix> different packages get updated at different rates, look at the different support levels per repository on the alpinelinux website
2023-10-05 17:43:05 <vishal_> so you all are deep divers? You discuss individual packages like Libc and so on?
2023-10-05 17:43:31 <fluix> no, it's just that questions like "is Alpine Linux more secure than Ubuntu" is impossible to answer with "yes" or "no"
2023-10-05 17:44:20 <minimal> "that would be an ecumenical matter"
2023-10-05 17:44:28 <invoked> were you triggered by the glibc vuln or something? what's your deal
2023-10-05 17:45:45 <vishal_> If you say you cant answer it with a yes or no then you are definitely a very deep diver.
2023-10-05 17:45:47 <vishal_> very very deep
2023-10-05 17:46:15 <vishal_> isnt security just isolation? I assume all distros do it well. So are they all similarly secure?
2023-10-05 17:46:19 <invoked> so it was the glibc vuln then. gotcha.
2023-10-05 17:46:28 <vishal_> And alpine linux is just light weight?
2023-10-05 17:47:17 <vishal_> Or are you picking based on which distro has had least security issues?
2023-10-05 17:47:20 <fluix> no, security is not just isolation, security depends on a host of factors. for a simple example, if you have a completely offline server running in a physically secure location, everything is equally secure
2023-10-05 17:47:42 <vishal_> I am talking a real computer, not an isolated server in a blacksite in area 51
2023-10-05 17:47:46 <fluix> I pick distributions based on their package managers, the community around them, my understanding of them, and more
2023-10-05 17:47:48 <vishal_> I am talking real world, normal people usecase
2023-10-05 17:48:20 <minimal> how do you know the site is black? it could be beige
2023-10-05 17:49:11 <fluix> well just the fact that it's lightweight _might_ mean it's more secure in your case, say because you use musl and doas instead of glibc and sudo, but maybe you don't care about local exploits
2023-10-05 17:50:11 <vishal_> local exploits? as in rogue apps running on the same server? Ok, I see your point but lets focus a little more on production server use cases where we can control what apps run on the server
2023-10-05 17:50:27 <fluix> you went from "normal people usecase" to "production server use case"...
2023-10-05 17:51:00 <invoked> this person is wasting our time.
2023-10-05 17:51:06 <vishal_> Is hould have clarified it better in the beginning, i meant normal production server use case of a system that is connected to the public internet and runs public apps that normal people can access.
2023-10-05 17:51:23 <vishal_> invoked, no just looking to understand
2023-10-05 17:51:25 <vishal_> learn
2023-10-05 17:51:45 <vishal_> dont assume this much in your marriage or you ll end up divorced
2023-10-05 17:52:10 <vishal_> fluix, you said you care about package managers. Shouldnt you care about packages more than package managers?
2023-10-05 17:52:14 <fluix> I suggest you figure out what your threat model is an investigate which distribution is better based on that
2023-10-05 17:52:30 <fluix> s/an/and/
2023-10-05 17:53:43 <fluix> my goal is some optimal balance between security and convenience
2023-10-05 17:53:55 <vishal_> what do you think is the thread model in a public server with publicly accessabel apps? kernel exploits right?
2023-10-05 17:53:58 <vishal_> Thats about it
2023-10-05 17:54:23 <fluix> ...no
2023-10-05 17:54:51 <fluix> that's literally one of the last issues I have on my mind when running things on servers, right near the hosting provider getting pwn'd
2023-10-05 17:55:24 <vishal_> hosting provider getting pwnd? what do you mean?
2023-10-05 17:55:29 <vishal_> like getting hacked?
2023-10-05 17:55:36 <vishal_> getting sold?
2023-10-05 17:55:43 <vishal_> getting DDOSd?
2023-10-05 17:55:47 <vishal_> what do you mean by pwnd?
2023-10-05 18:00:22 <fluix> please put in some effort when asking questions. I'm going back to work now, gl
2023-10-05 18:00:57 <vishal_> i would call you a faggot if i didnt need your help
2023-10-05 18:01:05 <vishal_> useless elitist attitude
2023-10-05 18:03:03 <Habbie> thanks ikke
2023-10-05 18:03:24 <minimal> Alpine is Webscale ;-)
2023-10-05 18:42:32 <bl4ckb0ne> can setup-disk do encryption on multiple drives?
2023-10-05 18:42:57 <bl4ckb0ne> i have an nvme drive that id like to use for root, and a ssd for /home, both encrypted with a single key pass
2023-10-05 19:14:05 <invoked> you mean you want one password prompt for two mounts? i don't believe that is an option. but you can use a key file on / to automatically mount the 2nd drive after.
2023-10-05 19:14:45 <invoked> you can add a password also, as a fallback
2023-10-05 19:15:04 <invoked> setting up luks/lvm on another drive is pretty easy.
2023-10-05 19:17:24 <bl4ckb0ne> might be an idea, though about zfs as well
2023-10-05 19:17:38 <bl4ckb0ne> can lvm split the drives equally to still respect the physical drive?
2023-10-05 19:18:26 <invoked> once you have the volume created, you can partition that as you want
2023-10-05 19:21:00 <bl4ckb0ne> and setup-disk handles that?
2023-10-05 19:21:55 <invoked> not sure about setup-disk but just doing it with cryptsetup and etc is straightforward
2023-10-05 19:22:30 <bl4ckb0ne> i found a page on the wiki but its labeled as deprecated
2023-10-05 19:24:43 <bl4ckb0ne> https://wiki.alpinelinux.org/wiki/Full_disk_encryption_secure_boot 
2023-10-05 19:25:43 <invoked> i'm not sure about striping two luks devices with lvm, but the basic process of 1) creating a luks device on physical storage 2) creating a lvm volume group on that 3) adding whatever volumes you want, 4) formatting those volumes with the fs of your choice, 5) fstab edit ... that's not terribly difficult imo
2023-10-05 19:26:25 <bl4ckb0ne> i think i can reuse most of that
2023-10-05 19:26:30 <invoked> i always do it manually, anyhow
2023-10-05 19:27:10 <bl4ckb0ne> seems like ive reached the limit of setup-disk, ill do it manually as well
2023-10-05 19:27:40 <invoked> 1) cryptsetup 2) pvcreate 3) vgcreate 4) lvcreate ... 5) mkfs 6) test 7) fstab edit
2023-10-05 19:27:47 <invoked> that's the concise process
2023-10-05 19:28:07 <bl4ckb0ne> thanks
2023-10-05 19:28:40 <invoked> omitted: overwriting the headers, wiping the device post-luks before creating the lvm side
2023-10-05 19:30:47 <bl4ckb0ne> overwriting headers?
2023-10-05 19:34:21 <invoked> if you want to wipe the disk using luks
2023-10-05 19:38:05 <invoked> that probably seems confusing. what i mean is, you set up luks (crypsetup), then dd if=/dev/zero of=/dev/whatever_crypt, let that finish. then close the luks device. then dd over the first few mb of your physical device (/dev/whatever) with /dev/random, then set up the luks device fresh again.
2023-10-05 19:38:38 <invoked> this is if you have old data on that drive you need to make sure is not retrievable.
2023-10-05 19:38:52 <invoked> otherwise ignore
2023-10-05 19:39:40 <dwfreed> I mean, there are significantly faster and less resource intensive ways of doing that
2023-10-05 19:39:40 <invoked> but even if not, it's a good practice.  more wear for the drive, sure, but not a bad idea.
2023-10-05 19:39:42 <bl4ckb0ne> i think I got it with setup-disk
2023-10-05 19:39:42 <dwfreed> see 'shred'
2023-10-05 19:39:57 <bl4ckb0ne> did lvm and crypt and I now have both swap and drive crypted
2023-10-05 19:40:12 <bl4ckb0ne> for the ssd i think ill do zfs since i have another one laying around
2023-10-05 19:40:53 <invoked> shred is for files, i'm talking about the device
2023-10-05 19:41:38 <invoked> i'm not offering an opinion on whether it's overkill, though.
2023-10-05 19:41:59 <dwfreed> invoked: shred works for devices too
2023-10-05 19:42:31 <dwfreed> a snippet from the manpage: "it is common to operate on device files like /dev/hda"
2023-10-05 19:42:38 <invoked> ah, well, i stand corrected, but i don't know how shred works on devices or especially nvme
2023-10-05 19:43:00 <dwfreed> I mean, it's going to be the same as a regular device write would
2023-10-05 19:43:22 <invoked> fair.
2023-10-05 19:43:25 <dwfreed> you should probably follow up the shred with a blkdiscard, but other than that
2023-10-05 19:45:37 <invoked> anyhow, bl4ckb0ne it might be good to do it the manual way just to familiarize yourself with that in case your system becomes unbootable.
2023-10-05 19:45:53 <bl4ckb0ne> absolutely
2023-10-05 19:46:01 <bl4ckb0ne> thanks for your help
2023-10-05 19:46:07 <invoked> np
2023-10-05 20:01:41 <vkrishn> i wonder if one can ask the same question on other distro forum/irc, "hey guys is alpinelinux more secure than yours"
2023-10-05 20:03:59 <bl4ckb0ne> what would be the point
2023-10-05 20:04:11 <vkrishn> of slightly more fiery, "hey guys, i want to switch to al, tell me why i should not"
2023-10-05 20:04:47 <bl4ckb0ne> seems like a loss of time
2023-10-05 20:05:16 <vkrishn> instead of all these rucus, why not just install it and use for couple of days
2023-10-05 20:06:06 <bl4ckb0ne> me?
2023-10-05 20:09:26 <vkrishn> or slightly ^
2023-10-05 20:09:35 <bl4ckb0ne> 
ACTION shrugs

2023-10-05 20:23:31 <goran> I installed Alpine+Xen and changed my keyboard layout (once during setup and once later) and both times it wont survive a reboot. Hot to make it persistent?
2023-10-05 20:23:54 <goran> *How
2023-10-05 21:15:17 <lopid> setup-keymap?
2023-10-06 01:24:49 <skraitoo> hi all , wanna start some project [ 0day (xc) Our ] at http://worldhacker.org ... .
2023-10-06 05:06:55 <Test111> I tried getting pipewire, sway and a bunch of other things to work but everything got messed up. Then I installed gnome instead, only to face a black screen after boot. Not sure what's happening. Can't even do anything to check or fix it.
2023-10-06 05:13:10 <ikke> Test111: At what point after boot?
2023-10-06 05:13:26 <Test111> Starts chronyd, then black screen
2023-10-06 05:13:59 <ikke> Can you switch to a different tty with ctrl+alt+f1-6?
2023-10-06 05:14:05 <Test111> Nope
2023-10-06 05:15:51 <Test111> Just chrooted into it, let me know if there's anything I should check
2023-10-06 05:16:56 <ikke> what's enabled in /etc/runlevels/default?
2023-10-06 05:17:02 <ikke> (ie, what symlinks exist)
2023-10-06 05:19:20 <Test111> acpid, chronyd, cpufreqd, crond, dhcpcd, elogind, gdm, networkmanager, seatd, sshd, udev-postmount, wpa_supplicant
2023-10-06 05:38:07 <ikke> Test111: you could try to remove gdm and see if you get access again
2023-10-06 05:39:02 <Test111> I'll have a go at it
2023-10-06 05:48:36 <Test111> ikke: Thank you, it worked. Now I need to figure out a way to fix whatever's broken.
2023-10-06 06:21:49 <cyclisme24[m]> Test111 does X run successfully without gdm from the command line? 
2023-10-06 06:22:03 <Test111> I have wayland installed
2023-10-06 06:22:21 <Test111> Unless gdm depends on x, I doubt I have it installed
2023-10-06 09:27:38 <fabrice> Good morning, why isn't the translation available under firefox 118 on Alpine linux?
2023-10-06 09:28:57 <ptrc> fabrice: it works for me, but only on some websites
2023-10-06 09:29:24 <ptrc> i think the website itself needs to have a language different than your main one
2023-10-06 09:30:14 <ptrc> can you check if about:preferences#general shows section called "Translations"?
2023-10-06 09:33:41 <fabrice> I've installed all the languages in the preferences. In the firefox menu I don't have the "translate this page" option. Nor do I have the translation logo in the address bar.
2023-10-06 09:45:28 <fabrice> In the settings for the translation section, although I've installed the languages, I can't change the options in this section. Wouldn't there be a system package to install?
2023-10-06 09:57:45 <cheslavic> Hello, I've got question regarding installation of Alpine Linux on my Acer Chromebook with eMMC storage. How to prevent excessive wear of it? Use no swap, F2FS, logs in ram? Any other ideas?  
2023-10-06 10:00:44 <liske> cheslavic: use diskless mode installation - changes are written to tmpfs overlay, changes in protected paths like /etc can be made persistent with lbu command
2023-10-06 10:04:37 <cheslavic> liske: sure, but what if I run out of RAM? Diskless mode is RAM based, am I correct? All installed packages are kept there, or not?
2023-10-06 10:07:08 <liske> it depends what apps will be used and how much RAM is available
2023-10-06 10:09:57 <liske> I use it on some raspberry pi's running Xorg and a webkit based kiosk browser: about 1,5GB memory
2023-10-06 10:14:55 <zcrayfish> If the drive implements wear leveling, and it's got a lot of extra room for what you need, leave a bunch of unallocated space in your partition table; on many drives that will help the wear leveling algorithm work better.
2023-10-06 12:59:28 <synapse> hi there!
2023-10-06 13:00:02 <ncopa> hi
2023-10-06 13:00:37 <synapse> I'm trying to add 'docker-cli' and 'docker-compose-cli' to an aarch64 image, but I'm building them on an x86_64 machine. I've been avoiding installing the appropriate QEMU tools to make this happen in the way where 'apk add' works in the runtime image; rather, I've been trying to move the apk files from the builder, but this is turning out to be a mess.
2023-10-06 13:02:06 <synapse> doing 'apk add --repository=...edge/community' still makes it fetch the x86_64 index because that architecture can be inferred from the builder image, so the 'docker-cli:aarch64' packages are naturally not found, even though they're in the aarch64 index. seems like I'm going against the current here.
2023-10-06 13:03:03 <ncopa> you could do apk add --arch=aarch64 ...?
2023-10-06 13:03:39 <synapse> I... could! I'll just try that.
2023-10-06 13:04:39 <minimal> synapse: you're building aarch64 packages? or building a aarch64 disk image?
2023-10-06 13:04:49 <ikke> Problem is going to be pre/post instal scripts 
2023-10-06 13:06:38 <synapse> yeah, 'RUN apk add --arch=aarch64 --repository=https://dl-cdn.alpinelinux.org/alpine/edge/community/ docker-cli:aarch64 docker-cli-compose:aarch64' still fails after fetching three different '.../x86_64/APKINDEX.tar.gz' URLs.
2023-10-06 13:07:06 <synapse> so I think wanting to install an aarch64 binary on an x86_64 (builder) image is nonsensical enough that the CLI won't let me.
2023-10-06 13:07:42 <synapse> I'm really just trying to avoid adding various QEMU dependencies in the host environment, but I think that's where I'm at now.
2023-10-06 13:08:09 <ncopa> usually there is only needed for a single qemu-user binary for the target architecture
2023-10-06 13:08:13 <synapse> I avoided this once with 'apk install tzdata' and just did that on the host and copied the data files over, no binaries involved.
2023-10-06 13:08:17 <ncopa> and configure binfmt
2023-10-06 13:08:56 <ncopa> apk files are also just tarballs so you can also : curl .../file.apk | tar -zx
2023-10-06 13:10:01 <synapse> yeah, and there's only three files in two APKs. I did consider that.
2023-10-06 13:13:51 <minimal> synapse: if you're trying to create a aarch64 disk image on a x86_64 machine then you really need to setup qemu-user + binfmt and then use chroot, that's what I do
2023-10-06 13:14:34 <synapse> minimal, it's the right way, I know.
2023-10-06 13:14:44 <synapse> I feel like this gets more and more hacky if I don't.
2023-10-06 13:15:08 <minimal> as ikke indicated things like the pore/post install scripts for packages won't work right otherwise
2023-10-06 13:16:17 <minimal> synapse: I have it all scripted if you want to have a look
2023-10-06 13:17:29 <synapse> minimal, sure!
2023-10-06 13:20:21 <synapse> I installed the Arch packages qemu-user-static qemu-user-static-binfmt and it just worked.
2023-10-06 15:43:02 <sewn> is anyone  currently facing a 'I/O error' with grub 2.06r14 trigger?
2023-10-06 16:44:45 <Saijin_Naib[m]> I'm trying to package Spectator, by treagod, a GTK/GUI REST API workbench, and I'm having trouble with the ```granite``` depends. It is declared in the makedepends, but neither pkgconfig nor cmake are able to find it for whatever reason.
2023-10-06 23:38:25 <jrcresawn> I need help configuring openconnect to start as an OpenRC service. I created the symbolic link in /etc/init.d, but what I struggle with is where to put the openconnect.conf file. I also don't understand the variables named server, vpnopts, and password defined in the start() function. Could someone point me to documentation that explains how this works? Many thanks for any help.
2023-10-06 23:39:55 <dwfreed> openconnect.conf would go in /etc/openconnect/<whatever you put after the dot in your symlink>
2023-10-06 23:40:13 <dwfreed> so if your symlink is openconnect.corp, you'd put openconnect.conf in /etc/openconnect/corp/ (make that dir)
2023-10-06 23:41:38 <dwfreed> for the vars, in /etc/conf.d/openconnect *or* /etc/conf.d/openconnect.corp , you'd put server_corp=192.0.2.1; vpnopts_corp=whatever; password_corp=passwordgoeshere
2023-10-06 23:42:37 <jrcresawn> Thank you dwfreed. I will test that now.
2023-10-06 23:44:08 <jrcresawn> So server_corp should be set to an IP address and not a URL?
2023-10-06 23:44:37 <jrcresawn> I use server=https://etc/ in openconnect.conf.
2023-10-06 23:45:32 <dwfreed> idk, whatever openconnect accepts when running
2023-10-06 23:46:00 <dwfreed> I don't use it
2023-10-06 23:46:04 <dwfreed> I just can read openrc
2023-10-06 23:55:19 <jrcresawn> I'm still struggling to get this to work. openconnect.err contains this output "Failed to open HTTPS connection to 
2023-10-06 23:55:42 <jrcresawn> Failed to complete authentication * start-stop-daemon: failed to start `/usr/bin/openconnect'"
2023-10-06 23:56:00 <jrcresawn> Sorry for the poor formatting.
2023-10-07 02:35:26 <TomawsCock> Anybody here into n i g g e r s
2023-10-07 02:35:28 <TomawsCock> ?
2023-10-07 04:09:28 <jrcresawn> dwfreed, thanks for you suggestions earlier. I have it working.
2023-10-07 04:09:38 <dwfreed> \o/
2023-10-07 04:10:52 <jrcresawn> I was completely unaware of /etc/conf.d/openconnect.
2023-10-07 04:15:27 <dwfreed> every init.d script has a corresponding conf.d script; init.d of the form foo.bar get both the foo.bar and the foo conf.d script sourced
2023-10-07 04:15:41 <dwfreed> some conf.d scripts are more useful than others
2023-10-07 04:18:36 <winny> dang, i guess they weren't interested in learning how to fish
2023-10-07 04:19:08 <dwfreed> completely missed they quit, lol
2023-10-07 05:39:16 <sewn> has the latest grub update broken anyone's grub?
2023-10-07 05:40:40 <sewn> I updated my system and grub had suddenly broke
2023-10-07 05:44:42 <dwfreed> did you remember to update the bootloader on disk after updating the package?
2023-10-07 05:45:35 <sewn> alpine hooks are supposed to do that
2023-10-07 05:46:04 <ikke> sewn: do you get an error messages?
2023-10-07 05:46:41 <sewn> sh: write error: I/O error
2023-10-07 05:46:48 <sewn> on grub hook
2023-10-07 05:47:15 <sewn> currently reinstalling grub, probably will switch to uefistub
2023-10-07 08:45:46 <linfan> hello :-) I am trying out alpine 3.18 on vbox and find it well working. I have gnome desktop 44.3 installed. All but sound works satisfactory. I try to figure out how to get pipewire to work without cracking.
2023-10-07 08:49:10 <linfan> When executing the pipewire-launcher I get errors from rt.c etc., which I have not been able to interpret. What could I further try?
2023-10-07 08:53:19 <bangbang> I kept hearing about alpine linux , then i found out that there is also archbang
2023-10-07 08:53:26 <bangbang> what is the difference between these tiny distros?
2023-10-07 08:53:33 <bangbang> There are a ton more: https://www.makeuseof.com/tag/linux-distro-space/
2023-10-07 09:25:48 <quinq> “requires just 256MB of memory” *just* xD
2023-10-07 09:26:00 <quinq> “It isn't the smallest Linux distro in terms of actual download or installation size, rounding up to about 2GB” haha
2023-10-07 09:27:59 <bangbang> glibc is more performant
2023-10-07 09:28:09 <bangbang> are you reading archlinux?
2023-10-07 09:28:14 <bangbang> archbang*
2023-10-07 09:28:22 <quinq> Is that a book?
2023-10-07 09:28:38 <bangbang> The first one in the blog is archbang https://www.makeuseof.com/tag/linux-distro-space/
2023-10-07 09:28:53 <bangbang> The blog is not mean for your confirmation bias. Its a discussion that provides many options
2023-10-07 09:29:52 <quinq> If you asked if I read the entry about archbang, then yes as I quoted part of it
2023-10-07 09:30:00 <LLVMclang[m]> tinycore uses glibc and way lighter than alpine 
2023-10-07 09:30:29 <quinq> Sure, it provides option, what I'm questionning is how much relevant they are, according to the title of the page that is about “The 9 Smallest Linux Distros That Are Minimal and Super Lightweight”
2023-10-07 09:30:47 <bangbang> archbang said that it needs 2gb? Are you blind?
2023-10-07 09:31:15 <quinq> It doesn't, please read again the web page :)
2023-10-07 09:34:16 <bangbang> 700mb, thats what archbang needs
2023-10-07 09:34:20 <bangbang> typical alpine user
2023-10-07 09:34:26 <bangbang> small os, small everything
2023-10-07 09:34:30 <bangbang> i mean brain
2023-10-07 09:47:26 <i5yxt1550rd9gbko8o55[m]> <bangbang> "700mb, thats what archbang needs" <- boring
2023-10-07 09:47:33 <i5yxt1550rd9gbko8o55[m]> i need more flexable
2023-10-07 11:34:58 <asdas> I just installed ubuntu server and it takes 200mb ram.
2023-10-07 11:35:07 <asdas> What is even the point of alpine linux if it requires 150mb
2023-10-07 11:36:55 <lopid> why are you here?
2023-10-07 11:37:30 <asdas> to learn
2023-10-07 11:37:57 <asdas> what can i do to make ubuntu 128mb like alpine
2023-10-07 11:38:26 <lopid> best ask in an ubuntu forum
2023-10-07 11:42:09 <asdas> are you guys not experts?
2023-10-07 11:43:55 <asdas> what is the difference between you and me then?
2023-10-07 11:51:41 <asdas> ??
2023-10-07 13:01:45 <openlinux21[m]> ![Alpine Linux Logo](https://upload.wikimedia.org/wikipedia/commons/thumb/e/e6/Alpine_Linux.svg/2560px-Alpine_Linux.svg.png)
2023-10-07 13:58:06 <strajder> I'm getting the "illegal file descriptor name" error with `apk add vlc-qt`: http://ix.io/4Iog
2023-10-07 13:58:54 <strajder> in vlc-libs-3.0.18-r8.trigger
2023-10-07 13:59:33 <strajder> apparently something redirects to ">&/dev/null" ?!
2023-10-07 14:30:10 <elly> strajder: hm! I thought that syntax wasn't allowed but it seems to work in my busybox sh
2023-10-07 14:31:55 <elly> what shell are you using?
2023-10-07 14:33:00 <strajder> lksh
2023-10-07 14:35:07 <strajder> that's mksh compiled with -L and -DMKSH_BINSHPOSIX
2023-10-07 14:36:50 <strajder> well, to be more precise I have /bin/sh pointing to /bin/lksh, and I use /bin/mksh as interactive shell
2023-10-07 14:43:40 <elly> gotcha - I think the apk scripts generally assume /bin/sh is busybox
2023-10-07 14:55:36 <vanveldt> 6
2023-10-07 15:04:19 <donoban> anyone using sway with autotiling? I'm testing it, looks nice but there are some windows that I can't move to left/rigth, up/down works
2023-10-07 15:04:40 <donoban> when I try to move some of this windows, autoting window says: Debug: Switched to splitv
2023-10-07 15:43:46 <minimal> sewn: no, grub trigger only re-runs grub-mkconfig, it doesn't *currently* update the files in /boot/grub/grub-* and also the MBR and other sectors (for BIOS) or the EFI file(s) for UEFI
2023-10-07 15:44:04 <minimal> sewn: I'm working on a solution for this that I hope to have ready in time for Alpine 3.19 release
2023-10-07 18:24:42 <JellyWasCoding> Hello!
2023-10-07 19:00:03 <wravoc> Hello good people! Just joined. I've got several repos of BSD security software on GH and Wallpapers to go with them. I've gotten permission from the BSD orgs to use them. Is there anything I need to do to use the Alpine Logo in a Wallpaper? I looked online and it was ambiguous. Where would I get that info if you could direct me please.
2023-10-07 19:07:54 <wravoc> For instance on Wikimedia commons it says the SVG logo is not original enough to be copyrighted, but the PNG version is under CC BY-SA 4.0 DEED
2023-10-07 19:08:38 <wravoc> The website copyright page on home does not like image copyright, only OS, as far as I could tell.
2023-10-07 19:33:39 <bearick> hi how do i request a package to be added to alpine linux
2023-10-07 19:35:55 <minimal> bearnick: create an issue here: joe ~/f
2023-10-07 19:36:03 <minimal> oops
2023-10-07 19:36:04 <fluix> https://gitlab.alpinelinux.org/alpine/aports/-/issues
2023-10-07 19:36:14 <fluix> close enough :p
2023-10-07 19:36:24 <minimal> https://gitlab.alpinelinux.org/alpine/aports/-/issues
2023-10-07 19:36:33 <bearick> wow which one should i click
2023-10-07 19:36:45 <fluix> this one, actually: https://gitlab.alpinelinux.org/alpine/aports/-/issues/new
2023-10-07 19:37:42 <bearick> nice virus link
2023-10-07 19:37:46 <bearick> haha jk thank you both
2023-10-07 19:38:55 <minimal> virus link?
2023-10-07 19:44:36 <bearick> dont ask about my $1,000,000,000,000,000 html zero day rce exploit(s) .... *tips black fedora*
2023-10-07 19:45:51 <bearick> https://gitlab.alpinelinux.org/alpine/aports/-/issues/15337 please bump it guys please
2023-10-07 19:56:26 <minimal> bearnick: handbrake is *already* packaged
2023-10-07 19:57:21 <minimal> bearick: ^^^
2023-10-07 20:02:17 <fluix> ._. can someone change that to a request to move it from testing to community
2023-10-07 20:03:28 <fluix> nvm, writing a comment
2023-10-07 20:09:58 <fluix> their profile is... something
2023-10-07 20:13:49 <elly> no accounting for taste
2023-10-07 20:38:37 <Habbie> elly, you mean, they didn't account for taste? :)
2023-10-07 20:42:45 <elly> :)
2023-10-08 02:58:06 <i5yxt1550rd9gbko8o55[m]> what happend on aport
2023-10-08 03:31:05 <fluix> ?
2023-10-08 04:06:04 <i5yxt1550rd9gbko8o55[m]> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/zfs
2023-10-08 04:06:26 <dwfreed> alpine doesn't package RCs for ZFS
2023-10-08 04:07:11 <i5yxt1550rd9gbko8o55[m]> dwfreed: but 2.1.13 doesn't update too
2023-10-08 04:07:58 <i5yxt1550rd9gbko8o55[m]> it seems something wrong with that apkbuild
2023-10-08 04:08:15 <dwfreed> I mean, if ZFS is no longer fixing issues in 2.1 even though they haven't even released 2.2 yet, that's their problem
2023-10-08 04:08:24 <i5yxt1550rd9gbko8o55[m]> it stopped at 2.1.11
2023-10-08 04:11:23 <IRCNOWea`y> BR⁠E⁠A⁠D OF GOD CO⁠MP​EL⁠S YOU TO JO​IN OUR IRC N⁠ET​W⁠ORK
2023-10-08 04:11:24 <IRCNOWea`y> IRC.BR​E⁠A​DOF​GOD​.ORG​ 
2023-10-08 04:11:25 <IRCNOWea`y> JRMU THE PREACHER⁠ A​ND GOD COM⁠PELS YOU
2023-10-08 04:11:26 <IRCNOWea`y> IRC​.IR⁠CNOW.O​RG​ F​OR FREE SHELLS⁠ AND BO​TS
2023-10-08 04:11:27 <IRCNOWea`y> IRCNOWea`y zcrayfish IsambardPrince Forza smcavoy8 crazymax somercet1 Misthios isak rootbeerdan Piraty uuidNuniq JellyWasCoding zbychuk redgloboli mbx caskd tomaw g0shfuckingdarn1t gabes4 solaarae mpmc vanveldt rz Naib liske toe vkrishn nicbk iXNyNe661440 callmepk rvalue kitzman corysanin jn kattjevfel floof58 balrog an3223_ L1Cafe82 Apachez lehm QwertzAM loneplanet leahcl elly sewn hanez avoidr 
2023-10-08 04:11:28 <IRCNOWea`y> cg theobjectivedad km Amber_Harmonia Hello71 terrorjack5483 lazywalk1 wener Kabaya qsx afofu jacksonchen666 nickersonm jangho mbrndtgn GNUmoon2 synapse Terminus miostreams garo odrling wasutton3 p1fk6ohrq9 rapmoc defec7ed cols shinzonremus coppsilgold[m] Sean[m]12 ezntek[m] mikeoxbig[m] marek_racko[m] herkulessi[m] mjh[m] thirdBeedellfirstRokesecondJul Invarianz[m] RicardoDuarte[m] a3e3a[m] 
2023-10-08 04:11:29 <IRCNOWea`y> Sean[m]1234 rnonce[m] cyclisme24[m] boombim[m] qaqland[m] sdrag0[m] Samson[m] NepNepdmsalwaysopen[m] clius[m] antholop[m] Kazu[m] AsutoshVariar[m] thenightterror[m] thu[m] Nulo[m] baalajimaestro Alxhr0[m] boscovn[m] t21222878[m] appendage[m] msrd0 tgamblin [][l0][][m] cache_return[m] nu[m]12 Heliopolis[m] Kadar[m] dryya[m] emlen[m] robertobot[m] kc2bez w[m]12 lnl flabi[m] userdocs[m] s
2023-10-08 04:11:30 <IRCNOWea`y> irotecer[m] Inga[m] ash[m]12345 fabienabc[m] niv fancsali[m] pnael[m] blackhelicoptersdotnet[m] palarran[m] dani-g5x[m] nikolashachka[m] nu[m]1 idkrn[m] bahamat holy-blindness[m] IvnMaxiPereyra[m] maksim[m] temporary_username[m] pivotwizard[m] rAajusS[m] mparisi futune capsized[m] AverageOSRSenjoyer[m] luks2[m] infernal_dynasty[m] GeorgeVassilev[m] M4s3ti[m] AnyCreatureMorph[m] nfbyte Logos[m] 
2023-10-08 04:11:32 <IRCNOWea`y> sifoavlinuxmxedition[m] magnolia_mayhem[m] openlinux21[m] Zeroine blt[m] ash[m]123 g90[m] KlingonBoyKlingon[m] duckl1ng[m] fossmen[m] tomjb[m] caskd[m] inthewalls[m] ash[m]1234 Sertonix[m] iztokf[m] Duck[m] ribalgz[m] danqo[m] Khalidibnal-Walid[m] malmi70[m] DylanVanAssche stdedos[m] PappGborSndor[m] pj halontka89[m] v1r9l|0[m] moltylollipop[m] hexagram23 kwadroke[m] imk4711[m] DieYoung[m] 
2023-10-08 04:11:32 <IRCNOWea`y> erminea[m] minux[m] extratone[m] emil2011[m] scorpion2185[m] hypergenesis[m] mm[m]1 the_very_sokol[m] Virtual[m] nm17[m] CodeIter[m] vahvist_ce_maraamat_on[m] alexhaydock gebbie[m] iota[m]1 ahmadraniri[m] tit4n[m] iamak[m] noyau[m] DanielBrenot[m] jimnextdoor[m] chris[m]12345678910111213 MattWiseman[m] Tom[m]12345678 alikateshethey[m] acrsoft[m] cleatis[m] fny[m] Janik[m] Armote[m] ruanime[m] 
2023-10-08 04:11:34 <IRCNOWea`y> Pavel[m]1 SmthSweethinher[m] poukman148 jomat[m] teiresias wisemansfear[m] vincentrabah[m] Saijin_Naib[m] Someone[m] dmitz monoid[m] DavidBlue[m] Androw bitmuncher[m] Someone[m]12 SUPERB[m] pitch[m] z3ntu telmich chillbruh[m] jarvis-owl[m] cockliuser[m] indigaz38833048167822834571522 Newbyte saad987654321[m] komrade-g[m] Guest2240 linex[m] alexw-578[m] KaliOmegaRogueshethey[m] IntelRide[m] 
2023-10-08 04:11:34 <IRCNOWea`y> rsuinux[m] Greg[m]123 LLVMclang[m] maakari[m] ayakael lindsay j-k[m] emersion bl4ckb0ne AppAraat[m] JoshuaHarris[m] Guest2224 chris218 marco[m]123 eric[m]123 IgnacyKuchciskiignapk[m] decke[m] roebuck[m] RoyalTiger[m] slime[m]1 feuerteufel i5yxt1550rd9gbko8o55[m] shubhisroking[m] Asuras[m] nazar-pc[m] ttjaguar[m] nulltemp[m] chexo4[m] Ruka[m] helmen[m] RyanPaulLBrulinski[m] Adahehim[m] eloitor[m] 
2023-10-08 04:11:36 <IRCNOWea`y> blt[m]1 ribal[m] v7a9a[m] grattykraat900[m] tiddlypom[m] krzno[m] BroPro40[m] scottmax[m] aureldsn[m] the4oo4 luma Case_Of OdaNobunaga[m] JohannesJ[m] IZEASG crapStone[m] sertonix[m]1 pkgbuild[m] unrznbl[m] PureTryOut sethmhurst[m] virtualuser[m] Mari[m] Andre[m]1 tyler[m]123 Enissay hariselldon[m] tskalyo xerz Cubixusin[m] blakes vhns hanetzer1 dpirate lxjv silasrech jbg rann markpash donoban 
2023-10-08 04:11:36 <IRCNOWea`y> Metroid adhawkins dari PopLamina bdju linuxgemini0 maria ashie[m] teapot9 invoked mid brettgilio Noisytoot_ markand vlp dionisis tugrik manwich jjuran pinkieval latex veecue sauce tetsumaki torresjrjr dallemon artok cultpony g_u_e_s_t PocketRat timlegge moxie rcf terminaldweller xordspar0 coc0nut ePirat veremitz xfbs asriel shdw_ tru_tru bri- nepeat azahi matthias11 Xe Guest1579 jeffrom indy 
2023-10-08 04:11:38 <IRCNOWea`y> leah2 deckard programmerq bunny mosinut sentriz haise01 jwindelborg ghormoon midasi nemo| blaboon af4 chili-b KindOne realtime-neil JakeSays1 gabrielgio earboxer rycwo exprez135 ajhalili2006 sooc marderbot_ sm2n euclaise akspecs micahrl truita sliwkr WhyNotHugo dekedro humm gjabell lemontree ashpool shinobi57474858 staceee stonks3141 tr1nity jay funspectre psnszsn z812 raiaq tobtobxx Guest1248 
2023-10-08 04:11:48 <qaqland[m]> 🤔
2023-10-08 04:12:04 <dwfreed> ignore it
2023-10-08 04:12:15 <elly> i5yxt1550rd9gbko8o55[m]: are you on 3.18 instead of edge?
2023-10-08 04:12:23 <elly> the latest version in 3.18 is 2.1.11
2023-10-08 04:13:01 <elly> https://gitlab.alpinelinux.org/alpine/aports/-/blob/3.18-stable/main/zfs/APKBUILD
2023-10-08 04:15:07 <fluix> shouldn't 3.18 also be updated?
2023-10-08 04:15:31 <fluix> given "support level: bug fixes"
2023-10-08 04:18:02 <elly> seems like an ncopa question
2023-10-08 04:18:15 <elly> (the maintainer)
2023-10-08 04:27:38 <i5yxt1550rd9gbko8o55[m]> <elly> "楊嘉晨: are you on 3.18 instead..." <- also on edge
2023-10-08 04:28:08 <dwfreed> edge is on 2.1.13
2023-10-08 04:29:41 <i5yxt1550rd9gbko8o55[m]> dwfreed: zfs version
2023-10-08 04:30:23 <dwfreed> https://github.com/alpinelinux/aports/blob/master/main/zfs/APKBUILD#L106
2023-10-08 04:30:29 <dwfreed> zfs-2.1.13
2023-10-08 04:34:00 <i5yxt1550rd9gbko8o55[m]> dwfreed: zfs-lts is correct
2023-10-08 04:34:16 <i5yxt1550rd9gbko8o55[m]> but zfs is wrong since 2.1.11
2023-10-08 04:34:31 <i5yxt1550rd9gbko8o55[m]> strat alpine zfs version
2023-10-08 04:34:31 <i5yxt1550rd9gbko8o55[m]> zfs-2.1.11-1
2023-10-08 04:34:31 <i5yxt1550rd9gbko8o55[m]> zfs-kmod-2.1.13-1
2023-10-08 04:34:46 <i5yxt1550rd9gbko8o55[m]> i5yxt1550rd9gbko8o55[m]: > <@i5yxt1550rd9gbko8o55:matrix.org> strat alpine zfs version... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/kivBRXjaEoOwmsqgHopseXrD>)
2023-10-08 04:35:03 <i5yxt1550rd9gbko8o55[m]> what's wrong with userland
2023-10-08 04:39:43 <dwfreed> I don't know what you're on, but I see zfs-2.1.13-r0.apk on the download server
2023-10-08 04:39:55 <dwfreed> https://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/zfs-2.1.13-r0.apk
2023-10-08 04:51:09 <i5yxt1550rd9gbko8o55[m]> <dwfreed> "I don't know what you're on, but..." <- you have to install to proof it
2023-10-08 04:51:45 <dwfreed> I don't run alpine
2023-10-08 04:52:02 <dwfreed> and I'm not bothering to spin one up to prove you wrong
2023-10-08 04:58:56 <fluix> podman run -ti alpine:edge; apk update; apk info zfs; zfs-2.1.13-r0 description...
2023-10-08 04:59:30 <fluix> make sure you're _really_ on edge
2023-10-08 05:57:40 <i5yxt1550rd9gbko8o55[m]> <fluix> "podman run -ti alpine:edge..." <- lmao,you have to modprobe zfs from host
2023-10-08 06:05:55 <dwfreed> not to look at package info to prove that 2.1.13 is available
2023-10-08 06:06:43 <quinq> $ cat /sys/module/zfs/version
2023-10-08 06:06:43 <quinq> 2.1.13-1
2023-10-08 06:21:13 <i5yxt1550rd9gbko8o55[m]> <quinq> "$ cat /sys/module/zfs/version" <- this is kernel space zfs-kmod / zfs-lts version
2023-10-08 06:22:42 <quinq> $ zfs -V
2023-10-08 06:22:42 <quinq> zfs-2.1.13-1
2023-10-08 06:22:42 <quinq> zfs-kmod-2.1.13-1
2023-10-08 06:33:27 <i5yxt1550rd9gbko8o55[m]> i see,maybe non-x86_64 is wrong
2023-10-08 06:34:57 <quinq> Yes, here it's on x86_64 indeed
2023-10-08 06:35:54 <dwfreed> I see 2.1.13 packages on every arch that alpine has on their download server
2023-10-08 06:41:54 <i5yxt1550rd9gbko8o55[m]> zfs-libs?
2023-10-08 06:42:44 <i5yxt1550rd9gbko8o55[m]> i see
2023-10-08 07:42:49 <JellyWasCoding> Hello, Where is setup-alpine script repo, I have good idea for new feature.
2023-10-08 07:44:21 <donoban> JellyWasCoding: https://gitlab.alpinelinux.org/alpine/alpine-conf
2023-10-08 11:51:12 <RoyalTiger[m]> <qaqland[m]> "🤔" <- Even God is interested in Alpine Linux. 😄
2023-10-08 13:37:23 <armin> [src/libout123/libout123.c:out123_open():439] error: Found no driver out of [alsa,oss] working with device <default>.
2023-10-08 13:37:27 <armin> any hint how to fix this?
2023-10-08 13:38:53 <armin> that's the internal soundcard (0), the external one i got connected via usb (1) plays fine.
2023-10-08 13:43:57 <armin> not much of a biggie as i'm using the usb one anyways, but that's weird
2023-10-08 15:50:51 <AnyCreatureMorph[m]> Why most people prefer gnome on alpine?
2023-10-08 15:53:09 <ikke> citation needed
2023-10-08 15:53:15 <ikke> No idea though
2023-10-08 15:53:40 <ikke> maybe because of wayland
2023-10-08 15:57:45 <Ermine> Maybe you should ask those people?
2023-10-08 17:59:52 <john3dc> hi. i use alpine-linux mainly in virtualbox and so far, everything works as except. but i have a problem with virtalbox's shared clipboard. i've installed the two apk packages and also started the two service. shared folders and screen work without any problems. is this a known issue or did i forget something?
2023-10-08 18:01:26 <minimal> john3dc: I believe this relates to there being no Virtualbox Addons ISO available for Alpine (specifically for musl)
2023-10-08 18:02:03 <Ermine> minimal: there's vbox-guest-additions package
2023-10-08 18:02:45 <minimal> Ermine: yeah, I just realised I was talking nonsense, lol, when I looked at a VB Alpine VM here lol
2023-10-08 18:03:44 <Ermine> john3dc: do you use any windowing system? Is it X11 or Wayland?
2023-10-08 18:04:38 <minimal> john3dc: for the particular VM have you looked in the Virtualbox settings in General -> Advanced and enabled "Shared Clipboard" ?
2023-10-08 18:04:57 <Ermine> john3dc: also, what version of alpine do you use? And what version of virtualbox?
2023-10-08 18:05:55 <minimal> it appears to be disabled by default
2023-10-08 18:07:09 <Ermine> Is pkgs.a.o up?
2023-10-08 18:07:32 <john3dc> clipboard is enabled in virtualbox and i'm using the 3.18.4 x86_64 standard
2023-10-08 18:07:58 <john3dc> and i'm using gnome on wayland
2023-10-08 18:09:28 <john3dc> Ermine: what do you mean wrh pkgs.a.o.up?
2023-10-08 18:10:33 <ikke> It is up (it's the webapp that shows what packages there are)
2023-10-08 18:10:35 <Ermine> I wanted to check versions at pkgs.alpinelinux.org, but it loaded slowly for me. It was not a question for you, but rather broadcast
2023-10-08 18:11:20 <Ermine> john3dc: my first guess is that vbox doesn't support wayland clipboard. May I ask you to try with x11?
2023-10-08 18:13:47 <john3dc> okay. i give it a try. i have to change the irc-client.
2023-10-08 18:15:54 <john3dc_> so now in windows again
2023-10-08 18:17:40 <minimal> never actualyl used the VB clipboard. Is it supposed to work with a text console?
2023-10-08 18:18:10 <ikke> I use it together with X
2023-10-08 18:18:13 <ikke> X11
2023-10-08 18:18:23 <Ermine> minimal: does text console have any sort of clipboard?
2023-10-08 18:18:23 <ikke> it does require an agent to be running
2023-10-08 18:18:36 <minimal> not running either X or Wayland on any of my Virtualbox VMs lol
2023-10-08 18:19:59 <minimal> s/text console/console/ as console on UEFI systems is graphics based ;-)
2023-10-08 18:20:32 <Ermine> i guess it is so on bios systems as well
2023-10-08 18:21:09 <Ermine> I'd not support vesa text modes if I was programming an OS. Probably
2023-10-08 18:21:14 <minimal> Ermine: not if you're using "VGA text output" by specifying "nomodeset" on cmdline
2023-10-08 18:22:56 <rtyler> minimal: o/ :) 
2023-10-08 18:23:18 <minimal> hey
2023-10-08 18:25:48 <Ermine> john3dc_: so, how things are going?
2023-10-08 18:25:51 <rtyler> my next exercise is to figure out how to convince udhcpc to take the hostname from the DHCP server, rather than to try to tell the DHCP server the hostname
2023-10-08 18:29:27 <minimal> rtyler: isn't that controlled by an entry it its config file?
2023-10-08 18:31:31 <rtyler> not that I've found, somewhere there's an `-x hostname:localhost` getting pushed onto the command line and I'm still digging for it
2023-10-08 18:34:37 <minimal> rtyler: you need the default.script file to look at the "hostname" environment variable and set the hostname using that
2023-10-08 18:35:27 <minimal> "udhcp client scripts" section here: https://udhcp.busybox.net/README.udhcpc
2023-10-08 18:36:58 <minimal> the "-x" is setting the hostname in the DHCP *request*
2023-10-08 18:38:46 <minimal> it is coming from ifupdown-ng
2023-10-08 18:39:12 <minimal> https://github.com/ifupdown-ng/ifupdown-ng/blob/main/executor-scripts/linux/dhcp#L33
2023-10-08 18:42:42 <minimal> rtyler: you need to edit /usr/share/udhcpc/default.script accordingly - current version doesn't use "hostname"
2023-10-08 18:44:03 <rtyler> yes, I'm giving that a try now. 
2023-10-08 18:45:53 <Ermine> shared clipboard doesn't work for me with gnome wayland 
2023-10-08 18:46:13 <Ermine> and gnome on x11 doesn't work at all
2023-10-08 18:47:39 <john3dc_> Ermine: Thx for testing. i never had any problems on arch.
2023-10-08 18:48:13 <rtyler> minimal: I was hoping to specify a different script, but it looks like specifying a new -s path/to/script for udhcpc_opts in the interfaces file doesn't do the trick
2023-10-08 18:50:28 <minimal> rtyler: seems to be controlled in ifupdown-ng by IF_DHCP_SCRIPT variable, so try setting that?
2023-10-08 18:54:23 <rtyler> where would I end up setting that variable? /etc/network/interfaces
2023-10-08 18:58:26 <Ermine> john3dc_: I think I've found a culprit
2023-10-08 18:59:06 <Ermine> john3dc_: as a workaround, please run VBoxClient --clipboard as root
2023-10-08 19:02:32 <minimal> rtyler: not entirely sure, I've never tried setting such stuff. You'd have to look at the ifupdown-ng docs
2023-10-08 19:03:49 <john3dc_> Ermine: I've already tried that.but not on x11
2023-10-08 19:03:56 <rtyler> fwiw overwriting the default.script in my overlay is sufficient, albeit a bit janky
2023-10-08 19:04:56 <Ermine> john3dc_: it worked for me on wayland
2023-10-08 19:07:21 <Ermine> Can you please send me output of this command (through some pastebin service) ?
2023-10-08 19:21:04 <john3dc_> Ermine: Hmmm. So can you help out with the installation, please?  maybe i missed something. are you on discord?
2023-10-08 19:25:10 <u0_a377> hey
2023-10-08 19:25:21 <u0_a377> how i can install gnome?
2023-10-08 19:25:39 <u0_a377> as i understand wiki method for xorg
2023-10-08 19:26:06 <u0_a377> `setup-desktop gnome`
2023-10-08 19:26:20 <u0_a377> but if i want wayland?
2023-10-08 19:38:51 <Ermine> john3dc_: you can DM me on IRC as well, no need for discord
2023-10-08 19:39:20 <Ermine> u0_a377: setup-desktop gnome installs wayland session as well
2023-10-08 20:12:32 <rtyler> muharharharhar, my frankenstein experiment works. minimal I now can boot N number of these diskless machines I have lying about and they're auto-registering in nomad and consul, ready to take on jobs ^_^
2023-10-08 20:56:13 <idkrn[m]> I was wondering why I got pinged lol
2023-10-08 20:56:29 <Habbie> turns out you're not special today ;)
2023-10-08 20:56:49 <Habbie> but i am because i did not make the cut ;)
2023-10-09 07:52:15 <frag> why #alpine-linux instead of #alpinelinux? anyway.. :>
2023-10-09 07:52:39 <frag> anyone experiencing that qutebrowser dont work everywhere anymore? eg finn.no is hard to use..
2023-10-09 07:54:05 <frag> page is mostly white
2023-10-09 07:54:41 <ncopa> frag: we have #alpine- prefix, so we have also #alpine-devel, #alpine-infra, #alpine-security etc
2023-10-09 07:55:07 <quinq> #alpine-apple
2023-10-09 07:55:25 <ncopa> im also experiencing issues with chromium at the moment, with lounge irc app
2023-10-09 07:56:08 <frag> ncopa: aha & aha :]
2023-10-09 07:56:34 <ncopa> finn.no appears to work ok in chromium though
2023-10-09 07:57:01 <ncopa> reminds me that i should sell my old motherboard there today
2023-10-09 07:59:31 <pukko> finn.no works fine for me in qutebrowser, v3.0.0
2023-10-09 08:00:00 <frag> hmm
2023-10-09 08:00:17 <pukko> ...with qt6 backend
2023-10-09 08:21:07 <synapse> frag, why #alpine-linux instead of #alpinelinux? we wouldn't want to confuse users of #al-pinelinux ("al" being "towards" in Esperanto)! :-D
2023-10-09 08:28:32 <Daanct12> pine linux
2023-10-09 09:49:54 <Ermine> towards pine linux?
2023-10-09 10:07:00 <vkrishn_> coz, then #alpinedevel would not sound cool ;)
2023-10-09 10:17:41 <jothe> hey. another day another question. how can i change the kb-layout in for gdm?
2023-10-09 10:18:15 <jothe> the ez way please ;)
2023-10-09 10:22:21 <LLVMclang[m]> windows key + space
2023-10-09 12:12:36 <staceee> hey there o/ I'm trying to run element-desktop as native Wayland client. How to do so? --ozone-platform-hint=auto seems useless
2023-10-09 12:14:24 <staceee> mhh it looks blurry but maybe it is actually already Wayland
2023-10-09 12:16:21 <somercet1> staceee, try clicking on it with xprop
2023-10-09 12:18:29 <staceee> cant
2023-10-09 12:18:36 <staceee> also xeyes don't look at it
2023-10-09 12:18:41 <staceee> but still, it is blurry o_O
2023-10-09 12:18:44 <somercet1> mmm
2023-10-09 12:18:47 <staceee> like a xwayland client
2023-10-09 12:18:52 <somercet1> yah
2023-10-09 12:19:05 <somercet1> Does XWayland come on when you start it?
2023-10-09 12:19:26 <staceee> I don't have any xwayland process running atm
2023-10-09 12:19:40 <staceee> with the program running
2023-10-09 12:20:33 <somercet1> Any messages when you start it from the command line?
2023-10-09 12:20:55 <staceee> nothing interesting
2023-10-09 12:21:56 <somercet1> what about `unset DISPLAY` and then running it? Just for laughs
2023-10-09 12:22:46 <staceee> same thing
2023-10-09 12:23:00 <staceee> I guess it is just blurry because element-desktop sucks? o_O
2023-10-09 12:23:06 <somercet1> sorry, out of ideas
2023-10-09 12:23:36 <staceee> somercet1: can it be related to fractional scaling problem?
2023-10-09 12:23:53 <somercet1> dunno, never used that
2023-10-09 12:24:12 <staceee> okay thanks still :)
2023-10-09 12:49:36 <xMPLR> Anybody here using cloudflared with Alpine? I have a weird condition when using `service cloudflared stop` that it kills the process, but the *.pid stays behind in /run -- this causes an issue when using `service cloudflared start` as it thinks the process is still running =x
2023-10-09 12:54:23 <xMPLR> I think the init.d script might not be compatible with ash as it uses: for i in {1..10} now that I look at it
2023-10-09 12:54:44 <xMPLR> 
ACTION tries with bash

2023-10-09 13:08:28 <minimal> xMPLR: use shellcheck to verify?
2023-10-09 13:12:43 <dalias> xmplr, should probably send a bug report and patch for the script
2023-10-09 13:12:54 <dalias> ideally it should be fixed not to use bash stuff
2023-10-09 13:13:06 <dalias> otherwise, edited to be #!/bin/bash not /bin/sh
2023-10-09 14:05:39 <Saijin_Naib[m]> Is anyone good with Busybox date module? I'm trying to figure out how to use it to convert ISO 8601 Timestamps into Unix Epoch in milliseconds 
2023-10-09 14:07:16 <ikke> Does coreutils even support that?
2023-10-09 14:08:14 <dalias> just do the math..?
2023-10-09 14:08:16 <Saijin_Naib[m]> I believe the full one does, or at least according to a StackExchange answer 🤣
2023-10-09 14:09:00 <quinq> worksforme, printf '%d000\n' "$(date +%s -d '2023-10-09T13:53:32Z')"
2023-10-09 14:09:10 <Saijin_Naib[m]> I want to use it as a XFCE Clipman action, so it will need to be a one-line for the terminal, ideally without pulling in stuff not in base Alpine 
2023-10-09 14:09:30 <dalias> quinq, invalid date
2023-10-09 14:10:09 <quinq> I meant with the GNU implementation
2023-10-09 14:10:10 <Saijin_Naib[m]> quinq thanks, I'll try that 
2023-10-09 14:10:18 <quinq> I don't think that busybox supports ISO-8601
2023-10-09 14:10:41 <Saijin_Naib[m]> dalias accounting for leaps isn't trivial if I'm just doing the math, no? 
2023-10-09 14:10:43 <dalias> try date +%s -d '2023-10-09 13:53:32'
2023-10-09 14:10:44 <quinq> But you can remove the T and the Z and it'd work
2023-10-09 14:10:59 <dalias> yes
2023-10-09 14:11:10 <ikke> dalias: that's in seconds, not?
2023-10-09 14:11:13 <dalias> saijin_naib[m], the posix formula handles that
2023-10-09 14:11:30 <dalias> ikke, yes, add the ms separately *shrug*
2023-10-09 14:11:31 <quinq> ikke, we can let the user the part converting seconds to milliseconds ;p
2023-10-09 14:11:50 <dalias> the only thing missing in the posix formula is converting the mm-dd to julian
2023-10-09 14:11:55 <Saijin_Naib[m]> expr does that kind of math, right? 
2023-10-09 14:12:06 <dalias> https://www.etalabs.net/sh_tricks.html
2023-10-09 14:12:08 <quinq> You don't need maths here
2023-10-09 14:12:13 <quinq> Just add 000 zeroes :p
2023-10-09 14:12:25 <dalias> but that's assuming you don't have a working %s
2023-10-09 14:12:29 <dalias> if you have working %s, just use it
2023-10-09 14:12:49 <ikke> Bb date supports +%s
2023-10-09 14:12:59 <Saijin_Naib[m]> quinq: I need milliseconds to get accurate Timestamps, unfortunately. I gather that isn't really posix 
2023-10-09 14:13:44 <quinq> Saijin_Naib[m], you asked for a convertion
2023-10-09 14:13:54 <quinq> Does your input provide milliseconds?
2023-10-09 14:13:56 <ikke> Saijin_Naib[m]: the man pages for coreutils date do not mention miliseconds 
2023-10-09 14:14:07 <dalias> well an ISO 8601 timestamp has ms==0
2023-10-09 14:14:13 <Saijin_Naib[m]> Right now I'm doing it all in LibreOffice Calc
2023-10-09 14:14:15 <dalias> but you can add the ms separately from wherever you get it
2023-10-09 14:14:33 <Saijin_Naib[m]> quinq: Yes, it's ISO 8601 with millisecond
2023-10-09 14:14:44 <quinq> I doubt that's ISO 8601 ten
2023-10-09 14:14:47 <quinq> s/ten/then/
2023-10-09 14:15:00 <quinq> But as dalias suggested, yes just add the ms part
2023-10-09 14:15:15 <Saijin_Naib[m]> Non-conformant? I was told it is ISO 8601
2023-10-09 14:15:35 <dalias> oh maybe ISO 8601 allows fractional seconds with . ?
2023-10-09 14:15:42 <dalias> in any case just separate out the ms
2023-10-09 14:15:45 <dalias> and add them back at the end
2023-10-09 14:15:54 <dalias> there's no nontrivial computation involved in that
2023-10-09 14:16:02 <dalias> it's just string manipulation
2023-10-09 14:16:05 <elly> ISO 8601 does allow milliseconds, yeah
2023-10-09 14:16:39 <elly> it's RFC 2822 that doesn't
2023-10-09 14:19:13 <dalias> ahh nice
2023-10-09 14:20:44 <quinq> Not according to https://www.iso.org/obp/ui/#iso:std:iso:8601:-1:ed-1:v1:en
2023-10-09 14:24:13 <quinq> Sorry I take that back, I misread
2023-10-09 14:24:14 <quinq> In this International Standard these marks, which are collectively referred to as time of day, are represented by the number of hours elapsed after the beginning of the day, the number of minutes elapsed after the last full hour, the number of seconds elapsed after the last full minute, with decimal parts of a second if necessary.
2023-10-09 14:26:20 <quinq> Nice, actually any part (hh, mm, ss) accepts a decimal part :D
2023-10-09 14:26:30 <dalias> uhg wtf
2023-10-09 14:26:36 <dalias> that sounds rather horrible
2023-10-09 14:26:37 <quinq> So you can have [hhmm,mm]
2023-10-09 14:26:41 <quinq> Enjoy implementing your parser :D
2023-10-09 14:26:56 <dalias> and ambiguous in the presence of leap seconds, perhaps?
2023-10-09 14:28:11 <quinq> I guess so ^^
2023-10-09 14:33:26 <elly> surely if there is a decimal present on the hours you can't have a number of minutes as well?
2023-10-09 14:33:29 <elly> that would be deranged
2023-10-09 14:33:37 <elly> on the other hand this *is* an ISO standard
2023-10-09 14:34:47 <quinq> Seems like that yeah, elly 
2023-10-09 14:56:33 <Saijin_Naib[m]> 😭 This is not what I thought it would be, haha
2023-10-09 14:56:42 <Saijin_Naib[m]> Thanks all for the help, I'll see what I come up with 
2023-10-09 15:00:32 <Saijin_Naib[m]> If it works, I'll post it and see if I can get it included in Clipman, so thank you all! 
2023-10-09 17:19:34 <jothe> hi. anybode can explain how i can change the kb-layout for for gdm?
2023-10-09 17:43:22 <ikke> anyone knows if / where udhcpc stores previously obtained leases? I'm trying to force it to get another ip, but it keeps getting the same
2023-10-09 18:06:41 <Habbie> ikke, any chance it's your router remembering?
2023-10-09 18:07:02 <Habbie> try '-x hostname:blabla123'
2023-10-09 18:07:19 <Habbie> or, from the manpage, -x 0x3d:0100BEEFC0FFEE - option 61 (client id)
2023-10-09 18:07:28 <ikke> It's dnsmasq that is giving the leases. I remove it from the dnsmasq lease file and restart dnsmasq, but every time the client gets the same IP
2023-10-09 18:12:56 <ikke> It's strange, if I do a tcpdump, I see the request (but can't detect it request the ip), the next packet is immediately to the ip address that the client will receive
2023-10-09 18:13:28 <Habbie> where do you dump?
2023-10-09 18:14:03 <ikke> on the host (the clients are lxc containers)
2023-10-09 18:14:19 <Habbie> can i see the tcpdump line?
2023-10-09 18:14:55 <ikke> https://tpaste.us/d1Yl 
2023-10-09 18:15:52 <Habbie> can you add -e to your tcpdump line?
2023-10-09 18:17:51 <Habbie> this looks useful: https://networkengineering.stackexchange.com/questions/64463/why-is-dhcp-client-able-to-accept-dhcp-offer-packet
2023-10-09 18:18:00 <Habbie> have to say i missed that detail and i tcpdumped a lot of dhcp this month
2023-10-09 18:18:20 <ikke> https://tpaste.us/4a4Z 
2023-10-09 18:19:13 <Habbie> ok, i think this exchange looks normal, once we explain how dnsmasq is managing to remember that IP
2023-10-09 18:19:22 <ikke> yeah
2023-10-09 18:19:27 <Habbie> is it perhaps the lowest free one?
2023-10-09 18:19:56 <ikke> let me confirm, I thought I had raised it, but that's a good question
2023-10-09 18:20:21 <Habbie> btw, you can do a 'free' request with udhcpc -q -s /bin/true
2023-10-09 18:20:36 <Habbie> this gets a lease, then does -not- run the local script that tells your interface, then exits
2023-10-09 18:20:43 <ikke> ok
2023-10-09 18:20:53 <Habbie> if you add one of the -x from above, you might be able to grab a few IPs
2023-10-09 18:20:59 <Habbie> .. or exhaust the entire range in 30 seconds
2023-10-09 18:21:18 <ikke> from the process list: --dhcp-range 172.16.27.5,172.16.27.240,255.255.255.0
2023-10-09 18:21:29 <ikke> so .5 should be the lowest available
2023-10-09 18:24:20 <ikke> Habbie: That command returned a high ip address
2023-10-09 18:24:22 <ikke> 237
2023-10-09 18:25:12 <Habbie> ok
2023-10-09 18:25:14 <Habbie> and if you run it again?
2023-10-09 18:25:20 <ikke> same ip
2023-10-09 18:25:25 <Habbie> ok, seeing the same here
2023-10-09 18:25:29 <Lord> 🧌
2023-10-09 18:25:32 <Habbie> wonder if dnsmasq is doing something clever
2023-10-09 18:25:37 <ikke> yeah
2023-10-09 18:25:49 <ikke> I have noticed this more often, but never got to the bottom of it
2023-10-09 18:26:33 <minimal> ikke, from dnsmasq manpage: "Dnsmasq is designed to choose IP addresses for DHCP clients using a hash of the client's MAC address."
2023-10-09 18:26:39 <Habbie> ah
2023-10-09 18:26:40 <ikke> Habbie: if I run that command with the hostname (-x hostname:..) I get the same ip address again
2023-10-09 18:26:42 <Habbie> was looking for a text like that
2023-10-09 18:27:08 <Habbie> ikke, yes, i just remembered, i used -x hostname because my testing was focused on making hostname resolution work
2023-10-09 18:27:09 <minimal> you can use "--dhcp-sequential-ip" to change this behaviour
2023-10-09 18:27:10 <ikke> right, was also suspecting something like that
2023-10-09 18:27:22 <Habbie> in fact i'm not managing to get udhcpc to assign separate IPs now from my dhcpd (which is not dnsmasq)
2023-10-09 18:27:28 <ikke> But the question is, why does dnsmasq return an ip outside the configured range
2023-10-09 18:27:36 <Habbie> 237 < 240
2023-10-09 18:27:40 <ikke> I mean .3
2023-10-09 18:27:47 <Habbie> oh
2023-10-09 18:27:55 <ikke> https://tpaste.us/MRy1 
2023-10-09 18:27:57 <Habbie> static assignment?
2023-10-09 18:28:11 <Habbie> so dhcp-host
2023-10-09 18:28:18 <minimal> yeah static would make sense
2023-10-09 18:28:40 <ikke> 🤦
2023-10-09 18:28:48 <Habbie> alright
2023-10-09 18:29:57 <ikke> somehow seeing the address in the lease file made me assume it was a dynamic lease 
2023-10-09 18:30:14 <Habbie> right
2023-10-09 18:30:14 <ikke> thanks for the reality check
2023-10-09 18:30:17 <Habbie> no problem!
2023-10-09 18:31:02 <minimal> I tend to use /etc/ethers for static addresses definitions rather than --dhcp-host
2023-10-09 18:31:25 <ikke> problem solved
2023-10-09 18:31:31 <Habbie> yay
2023-10-09 18:31:54 <ikke> it doesn't help where the opposite has happened as well
2023-10-09 18:32:55 <ikke> The issue was that .3 was clashing with another address, so I needed to move it
2023-10-09 18:37:22 <Habbie> right
2023-10-09 18:43:16 <minimal> ikke: doesn't dnsmasq ping an address before allocating it?
2023-10-09 18:43:24 <ikke> apparently not
2023-10-09 18:43:38 <Habbie> you sure? it blocks the dhcp 'thread' for a few seconds while processing a request
2023-10-09 18:43:43 <Habbie> but perhaps that's configurable
2023-10-09 18:44:03 <ikke> --no-ping, so it should do by default
2023-10-09 18:44:07 <Habbie> ah right
2023-10-09 18:44:17 <ikke> but I guess the static assignment was the issue
2023-10-09 18:44:24 <minimal> "y default, the DHCP server will attempt to ensure that an address is not in use before allocating it to a host. It does this by sending an ICMP echo request (aka "ping") to the address in question. If it gets a reply, then the address must already be in use, and another is tried. This flag disables this check. Use with caution."
2023-10-09 19:11:45 <quinq> minimal, I don't know about dnsmasq specifically, but in the DHCP protocol, it's the responsibility of the client to verify that another host doesn't have the proposed address on the same link
2023-10-09 19:11:52 <quinq> (not the server)
2023-10-09 19:12:23 <ikke> It's a feature of dhcp servers nonetheless
2023-10-09 19:12:29 <quinq> *some
2023-10-09 19:12:39 <ikke> yes
2023-10-09 19:12:48 <quinq> As an extention, but one can't rely on that
2023-10-09 19:13:11 <quinq> (it also sounds a bit superfluous, as the client will do that anyway)
2023-10-09 19:13:32 <ikke> well, apparently not :P
2023-10-09 19:14:10 <minimal> quinq: I was referring to dnsmasq specifically
2023-10-09 19:15:28 <quinq> ikke, my bad, I didn't follow the start of the conversation, I'll scroll up to see which client we have to fix! :D
2023-10-09 19:15:49 <quinq> minimal, I know, that's why I said “I don't know about dnsmasq specifically”
2023-10-09 19:15:50 <ikke> udhcpc
2023-10-09 19:16:15 <quinq> udhcpc ok :>
2023-10-09 19:16:57 <minimal> for udhcpc that's controlled by the FEATURE_UDHCPC_ARPING build setting
2023-10-09 19:17:02 <Habbie> right
2023-10-09 19:17:06 <Habbie> arping is better than ping anyway
2023-10-09 19:17:46 <minimal> which is set to "y" in Alpine
2023-10-09 19:21:50 <quinq> Ah, the problem isn't about ensuring the address is available
2023-10-09 19:22:06 <quinq> It was about dnsmasq serving an IP out of the configured range, ok.
2023-10-09 19:22:30 <Habbie> yes
2023-10-09 19:22:48 <Habbie> which, it turned out, ikke told dnsmasq to do
2023-10-09 19:23:30 <ikke> 
ACTION puts on a cone of shame

2023-10-09 19:23:57 <Habbie> no worries :)
2023-10-09 19:24:05 <Habbie> the cone just makes sure you won't spot it next time either
2023-10-09 19:24:06 <Habbie> take it off ;)
2023-10-09 19:24:06 <minimal> oh no, not the cone of shame!
2023-10-09 19:24:29 <quinq> jaja yeah, please give it back to minimal 
2023-10-09 19:25:04 <minimal> I only like cones when they contain ice cream ;-)
2023-10-09 19:25:21 <minimal> "the cone of calories"
2023-10-09 19:27:09 <quinq> The bests :>
2023-10-09 19:30:13 <Habbie> :)
2023-10-09 19:30:32 <dani> hi
2023-10-09 19:31:46 <ikke> bye
2023-10-09 19:39:13 <quinq> ikke, it seems we failed to answer dani's implicit question in less than a minute
2023-10-09 19:39:54 <ikke> Yes, it's a disappointment
2023-10-09 19:49:43 <invoked> maybe one could persuade opers to tarpit webirc sessions for a minimum of 2m, just for funsies
2023-10-09 19:50:56 <invoked> not that it would block someone closing a browser tab
2023-10-09 19:53:39 <Ermine> you failed your KPIs!
2023-10-09 20:29:38 <jothe> hi. anybode can explain how i can change the kb-layout for for gdm?
2023-10-09 20:34:38 <ikke> I have no idea about gdm, but someone suggested windows key + space
2023-10-09 20:35:46 <jothe> ikke: thx but didnt work
2023-10-09 20:37:26 <jothe> the gnome-shell end terminal was no problem...but gdm is a hard nut
2023-10-09 20:53:21 <goran> Why does alpine-xen-...-x86_64.iso exist? Can't it be handled with the ordinary x86_64 version?
2023-10-09 20:56:49 <dwfreed> it's intended for dom0, so probably pre-includes the hypervisor
2023-10-09 20:57:15 <dwfreed> Xen is a Type 1 hypervisor, you have to boot it before you boot Linux
2023-10-09 20:57:48 <dwfreed> (KVM and other Linux hypervisors are Type 2, meaning they're set up after Linux boots)
2023-10-09 20:59:23 <goran> Yes, but after an installation of Xen within an standard x86_64 installation, it should be good to go too or am I missing something?
2023-10-09 20:59:45 <dwfreed> how do you install and reboot into something if you have no persistent storage
2023-10-09 21:01:33 <goran> Why no persistent storage? After GRUB gets a new entry it should be able to boot Xen too.
2023-10-09 21:01:47 <dwfreed> assume you're using the ISO as a live media
2023-10-09 21:02:08 <dwfreed> your VM images and config could be on a disk somewhere, but otherwise everything runs off the ISO
2023-10-09 21:02:09 <ikke> (which is not uncommon to do with alpinelinux)
2023-10-09 21:02:48 <goran> Folks are using alpine-xen-...-x86_64.iso to adhoc run some VMs?
2023-10-09 21:03:28 <dwfreed> I mean, I'm sure somebody is, it also makes a great recovery media for your VM hosts, if you need to get the VMs back up fast while fixing other things
2023-10-09 21:03:49 <ikke> not ad-hoc
2023-10-09 21:05:14 <goran> I see. I never used Xen without a rw-filesystem. Ordinary installations. Thats why I was wondering.
2023-10-09 21:06:24 <goran> I played around with some distributions and most do have problems to boot Xen. An installation via alpine-xen-...-x86_64.iso did not.
2023-10-09 23:08:01 <ahmadraniri[m]> Hello, is it just me or adb tool on stable alpine seems not work ? I'm getting "could not install *smartsocket* listener" problem when I run it. I tried to kill it using "adb kill-server" and restart it using "adb start-server" but the problem still exist.
2023-10-10 01:11:40 <bleb> top says I am using almost all of my memory
2023-10-10 01:11:59 <bleb> but when I sort by memory usage there's like nothing... most memory hungry process is using 0.6 %MEM
2023-10-10 01:12:08 <bleb> could it be zfs?
2023-10-10 01:16:44 <dwfreed> what does 'free -m' say
2023-10-10 01:23:56 <bleb> reports almost no usage
2023-10-10 01:24:01 <bleb> I guess top is just wrong?
2023-10-10 01:24:33 <dwfreed> more likely top is just doing a poor job of reflecting the difference between actually used and used for buffers and cache that can be freed if needed
2023-10-10 01:24:57 <dwfreed> this is why 'free' grew an 'available' column a few years ago
2023-10-10 01:25:47 <dwfreed> available is how much RAM you could use if you needed to, and kernel would drop buffers and caches to make room
2023-10-10 01:25:47 <bleb> hmm interesting
2023-10-10 01:26:24 <dwfreed> the 'free' column in 'free' the command is how much RAM is not actually used for anything at the moment
2023-10-10 01:30:18 <winny> ^ excess free can indicate too much resources on a host (save $$$) , or bursty workflows
2023-10-10 01:30:57 <dwfreed> wat
2023-10-10 01:34:09 <winny> Not sure what the question is 
2023-10-10 01:34:42 <dwfreed> a lot of free could also just mean you just turned the system on and it hasn't done anything yet
2023-10-10 01:35:08 <dwfreed> I wouldn't really put any value in that as a metric
2023-10-10 01:36:41 <winny> fair point :)  I usually turn down boxes after monitoring free usage over time, no need to throw money away
2023-10-10 02:57:46 <Saijin_Naib[m]> quinq, yeah the coreutils date does the conversion without issue, even keeping the timezone and milliseconds... I can't figure out how to get BusyBox date to do it yet
2023-10-10 05:43:02 <quinq> Saijin_Naib[m], your use-case still hasn't been explained correctly
2023-10-10 05:43:16 <quinq> Do you want to convert any ISO date, or just one fixed format
2023-10-10 05:43:43 <quinq> Do you want date to actually convert it, or are you just looking for any way of doing this
2023-10-10 05:43:51 <quinq> etc.
2023-10-10 05:50:22 <Saijin_Naib[m]> quinq, sorry. So, I have logs that give me timestamps in what I understand to be ISO 8601, with milliseconds, and varying timezones. I need to convert them, as-is, to Unix Epoch.
2023-10-10 05:50:35 <Saijin_Naib[m]> Here's an example of one:  2023-09-07T16:13:35.503985
2023-10-10 05:51:02 <Saijin_Naib[m]> I am looking to make it "portable" or dependency-less on Alpine, so hopefully just using BusyBox built-in tools like date
2023-10-10 05:51:34 <Saijin_Naib[m]> I will be calling whatever the command is from XFCE's ClipMan tool, so that when I copy a date in ISO 8601 from my logs, it is converted at time of paste directly to Unix Epoch
2023-10-10 05:53:16 <Saijin_Naib[m]> Right now, I'm doing this with a combination of time.lol and a LibreOffice Calc spreadsheet... It is not ideal
2023-10-10 05:53:51 <Saijin_Naib[m]> Why do I want Unix Epochs? Way easier to do maths on to get how many seconds a stage of a given Task ran by subtracting one Unix Epoch from the other
2023-10-10 06:19:59 <quinq> varying timezones, so you do have dates in the format +/-hh?
2023-10-10 06:24:18 <quinq> (also .503985 are us, not ms)
2023-10-10 06:24:31 <Saijin_Naib[m]> Ah, sorry
2023-10-10 06:27:02 <Saijin_Naib[m]> Sorry, no, my date-times are in UTC already I guess, given the T notation in the date/time stamp above
2023-10-10 06:27:13 <Saijin_Naib[m]> So, I guess timezones have already been accounted for by the logging agent
2023-10-10 06:27:29 <quinq> UTC is given by a trailing Z
2023-10-10 06:27:43 <quinq> But if you don't actually have timezoned dates, easier :)
2023-10-10 06:27:55 <quinq> You could try something like:
2023-10-10 06:28:27 <quinq> /bin/sh -c 'd="2023-09-07T16:13:35.503985"; _d="${d%%T*} ${d##*T}"; echo $(busybox date -d "${_d%%.*}" +%s)${_d##*.}'
2023-10-10 06:29:21 <quinq> Otherwise a sed call instead of the variable manipulation if you don't mind
2023-10-10 06:29:54 <Saijin_Naib[m]> Oh, that's awesome, thank you!
2023-10-10 06:30:18 <Saijin_Naib[m]> Either the timezones have been worked out, or they are recorded Unqualified which is... Ugh, haha
2023-10-10 06:30:30 <quinq> Well, that's another problem ;)
2023-10-10 06:30:34 <Saijin_Naib[m]> :P
2023-10-10 06:31:07 <Saijin_Naib[m]> It shouldn't matter unless I'm comparing data across timezones, so... A problem for another night when I can bang my head against the code to see where and what logs and why
2023-10-10 06:31:26 <Saijin_Naib[m]> I'll give that code a go, thank you so much!
2023-10-10 06:31:43 <Saijin_Naib[m]> Want to be credited when I post it up for a PR to Clipman?
2023-10-10 06:36:40 <quinq> Good luck!
2023-10-10 06:36:49 <quinq> Na, it's fine, but thanks for asking that's appreciated
2023-10-10 06:38:12 <Saijin_Naib[m]> Thanks so much for the help!
2023-10-10 07:18:46 <Sofia> Hello world. I've just installed alpine in qemu, trialing before probably migrating back to alpine for the first time in many years. I have `-net user` and `nslookup` works within the virtual machine, yet `apk` errors with "DNS lookup error". Any suggestions or theories?
2023-10-10 07:20:25 <Habbie> what is '-net user'?
2023-10-10 07:21:37 <dwfreed> Habbie: basically qemu performs your network operations as a regular application instead of doing something like tap
2023-10-10 07:21:48 <Habbie> ah on the qemu side, ok
2023-10-10 07:22:22 <herkulessi[m]> Sofia: Pretty sure nslookup ignores some DNS Caching, may try ping?
2023-10-10 07:22:32 <dwfreed> ping doesn't work with -net user
2023-10-10 07:22:45 <dwfreed> I mean, I guess the resolution part would, the rest won't
2023-10-10 07:23:18 <Sofia> herkulessi[m]: "ping: bad address 'example.com'"
2023-10-10 07:23:55 <dwfreed> I mean, you should use a real DNS name, like dl-cdn.alpinelinux.org
2023-10-10 07:24:38 <herkulessi[m]> well example.com does exist as a webpage, and now we know it is DNS
2023-10-10 07:24:43 <herkulessi[m]> It ALWAYS is DNS
2023-10-10 07:24:55 <Sofia> Wait no. ping works on these domains
2023-10-10 07:25:02 <Sofia> I did another domain and no idea why that doesn't work.
2023-10-10 07:25:10 <Sofia> So ping gets IP addresses fine.
2023-10-10 07:25:42 <Sofia> o.o
2023-10-10 07:25:46 <Sofia> Well, apk is downloading now
2023-10-10 07:25:48 <Sofia> 
ACTION shrugs

2023-10-10 07:25:51 <Sofia> Mystery.
2023-10-10 07:28:08 <Sofia> dwfreed: Ping won't work on my system anyway for other reasons. :)
2023-10-10 07:28:23 <dwfreed> well, and -net user can't do ICMP
2023-10-10 07:28:40 <Sofia> I mean neither can the rest of my system.
2023-10-10 07:30:16 <herkulessi[m]> Well user networking seems to break DNS sometimes, next time adding 8.8.8.8 as dns server in your resolv.conf should fix it from what I googled in the last 5 minutes
2023-10-10 07:31:27 <Sofia> Does alpine support 9p mounts? Tried and get "Invalid argument"
2023-10-10 07:32:09 <Sofia> herkulessi[m]: resolv.conf correctly says the IP of my host, where I have tor setup for my system-wide DNS.
2023-10-10 07:32:55 <Sofia> Works, typo'd my command.
2023-10-10 07:37:01 <herkulessi[m]> maybe it timed out inside the vm, but the upstream resolver cached the result and at some point it started working?
2023-10-10 07:37:18 <herkulessi[m]> Well, DNS fails in mysterious ways, we might never know
2023-10-10 07:38:45 <Sofia> It is always DNS...
2023-10-10 07:39:39 <Sofia> 
ACTION discovers her /etc/apk/repositories don't include community. <_<

2023-10-10 07:39:59 <Sofia> 
ACTION installs a bunch of things and hopes for the best.

2023-10-10 07:46:38 <Sofia> Getting DNS errors again hmm
2023-10-10 07:47:09 <Sofia> nslookup works but ping and pk fail (again)
2023-10-10 07:47:36 <Sofia> pkill -1 -x tor
2023-10-10 07:47:41 <Sofia> On host, fixed it.
2023-10-10 08:08:54 <Sofia> BTW I'm switching to migrate back to alpine because it really was my favorite system, but I hit glibc et al frictions and just needed some packages which didn't work on alpine years ago. Thus trialing gcompat, distrobox + podman to fill in the gaps.
2023-10-10 08:10:29 <Sofia> Has anyone here got bcachefs working on alpine yet? Will probably wait til it gets mainlined. I want it. Better auth-enc than luks' un-auth enc. Has snapshots. Should perform better too.
2023-10-10 08:27:27 <Sofia> switching/trialing
2023-10-10 09:34:55 <invoked> when bcachefs gets scrub and ideally some fsck, that'd be when i'd take a look, but not until.
2023-10-10 09:45:14 <Sofia> invoked: Are you sure it doesn't have fsck?
2023-10-10 09:45:35 <Sofia> Pretty sure it has both?
2023-10-10 09:46:09 <Sofia> https://bcachefs.org/Fsck/
2023-10-10 09:47:03 <invoked> i think that's offline fsck
2023-10-10 09:47:29 <invoked> my info might be dated, i dunno.
2023-10-10 09:48:03 <Sofia> https://bcachefs.org/Roadmap <- missing scrub at least.
2023-10-10 09:48:10 <Sofia> And online fsck.
2023-10-10 09:48:27 <invoked> yeah.
2023-10-10 09:50:50 <Sofia> `distrobox enter arch`, with archlinux:latest, is this expected to fail with "applet not found" ?
2023-10-10 09:51:03 <invoked> i don't know why i'm nitpicking, i ran btrfs early and that still doesn't have a proper fsck. that said btrfs hasn't failed me in the limited sense i've used it.
2023-10-10 09:51:35 <Sofia> adding `-- sh` works. Have sudo within the environment to just become root, despite being rootless/totally running as my user. So this is good.
2023-10-10 09:52:35 <Sofia> When you don't know what package to install, try to vim /etc/pacman.conf, and find you don't have vim. Well, install that. :)
2023-10-10 10:00:52 <witcher> am i the only one that can never get lvm on luks with grub working and always has to debug it for different reasons? asking because, if that's not the case, maybe the docs for installing alpine like that could use a few more sentences for clarity or something
2023-10-10 10:03:22 <john3dc> hey. another day...but the same question. :) i still have the problem to change to keymap language for gdm on wayland. can i use any script or make a change in config-files?
2023-10-10 10:06:22 <invoked> witcher: i assume you're trying to do this with setup-disk
2023-10-10 10:06:37 <invoked> (this bug report sucks)
2023-10-10 10:07:57 <witcher> invoked: nope, by hand. this wiki article is what i'm following: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Grub_with_UEFI
2023-10-10 10:08:35 <invoked> i never look at the wiki, but i'll look.
2023-10-10 10:08:38 <witcher> this is intentionally vague and not a bug report as I'm trying to see if there's an issue with me or otherwise
2023-10-10 10:11:40 <Sofia> invoked: I tried btrfs early; it destroyed all my data.
2023-10-10 10:13:38 <invoked> witcher: at a glance that looks ok, but i'm not going to walk through all the steps there to 100% verify.  but this is one of these things where everything has to be done exactly right or you don't boot.
2023-10-10 10:14:40 <invoked> i think where some people get messed up is on encrypted /boot, if they want that.
2023-10-10 10:14:44 <witcher> invoked: no that's all I wanted to hear, thank you. I guess I'll have to redo the steps to make sure after all
2023-10-10 10:16:34 <invoked> godspeed :)
2023-10-10 10:20:42 <invoked> Sofia: some devices will eat any filesystem though.  devices can lie (and storage devices especially do) and there are a lot of quirks that the fs and os can't (and imo should not) compensate for. that's not a defense of btrfs, though.
2023-10-10 10:20:50 <invoked> i've only use btrfs for casual stuff.
2023-10-10 10:25:22 <Sofia> It was almost certainly btrfs which ate my data.
2023-10-10 10:25:58 <Sofia> A different problem, while on an ext4 disk however, it was the Intel Optane-SSD hybrid which disconnected at the NVMe interface level, causing data corruption and system halts.
2023-10-10 10:26:11 <Sofia> Had to immediately backup what I could, a few times over, reading distinct data each time.
2023-10-10 10:26:18 <Sofia> Best effort recovery.
2023-10-10 10:26:31 <Sofia> Now I use an older, faster, bigger Samsung NVMe SSD.
2023-10-10 10:26:34 <Sofia> No problems.
2023-10-10 10:26:46 <invoked> samsung nvmes have had some weird shit too.
2023-10-10 10:26:52 <invoked> anyway, it's impossible to know
2023-10-10 10:27:12 <invoked> you can only see what you can see via software
2023-10-10 10:27:36 <Sofia> Indeed
2023-10-10 10:28:12 <Sofia> At least disconnecting at the NVMe interface level seems pretty evident it was the Intel hybrid Optane-SSD.
2023-10-10 10:28:22 <Sofia> Replacing that fixed it. Reseating didn't help.
2023-10-10 10:29:24 <invoked> i had a great experience with sandisk, they were the real mass-market innovators, but now they're owned by wd -- so all prior assumptions have been invalidated
2023-10-10 10:29:48 <invoked> there's always micron, i suppose.
2023-10-10 10:33:17 <Sofia> dd is always the scariest command
2023-10-10 10:43:08 <invoked> looking at the luks/lvm wiki page again, it should probably be refactored(? for lack of a better term) into several pages instead of one page with toc.  mixing bios/mbr and uefi together along with optional stuff like encrypted /boot, is probably very confusing to people who have never done it before.
2023-10-10 10:44:04 <invoked> there are some concepts there that probably deserve their own page so that people understand what's going on.
2023-10-10 10:44:12 <invoked> then a simplified list of steps at the end.
2023-10-10 10:45:59 <Sofia> Installing on a laptop to a 32 GiB USB. Cryptsys doesn't write to the whole disk, right?
2023-10-10 10:46:53 <invoked> usually you would use the whole disk.
2023-10-10 10:47:31 <invoked> but you can partition it however you want
2023-10-10 10:48:18 <Sofia> Yes, not quite what I meant. I'm asking if the install attempts to write zeros over the full encrypted volume to securely erase what was previously on the disk and to hide the size of what does get written. Is an optional hardening step. Seems it doesn't do it.
2023-10-10 10:49:05 <Sofia> Seems stuck on initramfs... not a step I expect to take any real time?
2023-10-10 10:49:38 <invoked> i don't know what setup-disk does. but it probably does not wipe the disk for you.
2023-10-10 10:50:06 <Sofia> 
ACTION is borrowing this laptop and just for safety, she disassembled to remove the SSD. :)

2023-10-10 10:50:24 <Sofia> Already scary enough dd'n to the right disk.
2023-10-10 10:50:57 <invoked> do backups first then destructive operations second :)
2023-10-10 10:51:31 <Sofia> Lovely. htop -> status: Dead, grub-probe.
2023-10-10 10:52:35 <Sofia> How to do backup: mkfs on the wrong disk
2023-10-10 10:52:36 <Sofia> Uh oh!
2023-10-10 10:53:24 <Sofia> Had to reboot. Install take #2.
2023-10-10 10:58:44 <Sofia> Now grub-install is dead :/
2023-10-10 10:58:57 <Sofia> Didn't even reach the same progress...
2023-10-10 10:59:39 <Sofia> Or not. It managed to alive itself. o.o
2023-10-10 11:00:30 <Sofia> Oh my bad
2023-10-10 11:00:34 <Sofia> D isn't dead, it is "disk sleep"
2023-10-10 11:01:16 <Sofia> Just slow USB I guess...
2023-10-10 11:09:03 <Sofia> Wow.... slowest system ever. What is this.
2023-10-10 11:09:14 <Sofia> Booted from it. Takes forever to do anything on this USB....
2023-10-10 11:21:20 <Sofia> Installed sway; in sway everything is responsive. Curiously the slow commands, according to the `time` program, rightfully took no time. Yet visually took a minute to make a directory...
2023-10-10 11:21:44 <Sofia> Can switch between TTYs while waiting for the mkdir in /tmp to return.
2023-10-10 11:22:45 <Sofia> htop shows `addgroup -S flatpak` as state "disk wait"
2023-10-10 11:35:46 <Sofia> Never had a system this responsive, yet not, at the same time.
2023-10-10 11:36:48 <ikke> Sounds like your storage has issues
2023-10-10 11:37:03 <ikke> disk wait means that the process waits for an IO operation to reutrn
2023-10-10 11:39:10 <Sofia> Right. But... how is a 32 GiB USB this slow..?
2023-10-10 11:39:30 <Sofia> Is USB2 but ... I've never had this issue and I used to run alpine from a microsd card ... a decade ago.
2023-10-10 11:42:37 <ikke> Sofia: Maybe that specific drive has issues?
2023-10-10 11:42:59 <ikke> These kinds of problems often indicate some problem
2023-10-10 11:43:56 <Sofia> Certainly plausible. Going to reboot and see if I can get any better from a different port at least.
2023-10-10 11:44:33 <Sofia> At least I got sway, firefox, foot, working. Flatpak isn't even half way through installing steam. Distrobox ... has at least made some progress with an arch container.
2023-10-10 11:50:39 <Sofia> Didn't help.
2023-10-10 11:57:10 <Sofia> At least I'm confident in alpine, and everything I care about seems to just work. Just this disk is insanely slow.
2023-10-10 13:20:08 <minimal> Sofia: no setup-disk does not "securely wipe" the storage device before creating LUKS on it. Indeed writing zeros to a whole USB flash drive might not be a good idea in terms of drive health
2023-10-10 13:22:32 <ikke> Wiring zero's is also not recomended as it makes actual data stand out 
2023-10-10 13:23:41 <minimal> witcher: what problems exactly are you having with LVM-on-LUKS & Grub? Are you trying to set up LUKSv2 including for encrypted /boot ?
2023-10-10 13:24:07 <Sofia> ikke: You write the zeros to the encrypted disk, so the full disk isn't zeros on the outside, it is random junk.
2023-10-10 13:24:37 <Sofia> minimal: A global write may not be great but it is ultimately wear leveled so really not an issue.
2023-10-10 13:26:10 <minimal> Sofia: also remember that flash-based devices has extra "hidden" space (to replace some faulty cells) and, and depending on the wear levelling algo on the device so existing "blocks" with data on them may get switched out rather than written to
2023-10-10 13:26:27 <minimal> it is hard ot ensure data is completely overwritten in that way
2023-10-10 13:27:43 <Sofia> minimal: Yes, this is why you never write to an unencrypted disk.
2023-10-10 13:28:27 <minimal> witcher: if you were trying to setup LUKSv2 with encrypted boot the answer is.....it doesn't (currently) work as it's broken in (upstream) Grub 2.06 - they've fixed it AFAIK in Grub git (which will become Grub 2.12 some time this decade hopefully lol)
2023-10-10 13:28:28 <Sofia> If encrypted, it suffices to just rekey and re-encrypt the active content if you want to be assured even hidden encrypted blocks remain useless.
2023-10-10 13:28:46 <Sofia> But that is really pushing defense in depth. A bigger problem is the lack of authentication with luks.
2023-10-10 13:29:15 <Sofia> luke2 has verify things though? Not caught up on those details to properly discuss that though.
2023-10-10 13:29:37 <witcher> minimal: Not on purpose at least. I just want an encrypted rootfs. AFAIK GRUB doesn't support encrypted boot with LUKS2 (yet)
2023-10-10 13:30:28 <minimal> witcher: in theory Grub was supposed to support LUKSv2 encrypted boot - but in reality it didn't work
2023-10-10 13:30:35 <witcher> the setup I was trying this time is /boot partition and encrypted ESP on /boot/efi. Is this already considered to be encrypted boot partition?
2023-10-10 13:31:22 <minimal> witcher: so you haven't been clear on exactly what/how you've set things up. I have LVM-over-LUKS with Grub working fine here
2023-10-10 13:31:36 <minimal> encrypted ESP? nope,, ESP can't be encrypted
2023-10-10 13:31:53 <minimal> UEFI doesn't understand encryption
2023-10-10 13:32:34 <minimal> encryption is not part of the UEFI specification/standard
2023-10-10 13:33:03 <witcher> hmmm I might be mixing something up majorly
2023-10-10 13:34:24 <Sofia> witcher: My plan for my laptop when I bought it (before I learned the hard way the UEFI lacks the drivers for the microsd card reader) was to luks-encrypt the full NVMe disk, then use a /boot disk with the luks headers on it, such that I could remove the microsd card and the laptop is a brick. To avoid corruption, the /boot kernel is signed with my own keys registed into the
2023-10-10 13:34:26 <Sofia> UEFI for secure boot.
2023-10-10 13:34:51 <witcher> I've had LVM over LUKS set up fine in the past, too (i. e. 2 days ago on the same machine) but have had trouble setting that up as well. this seems to be ignorance on my end more than anything so I'll read up on bootloaders a little bit before asking anything else
2023-10-10 13:35:38 <witcher> Sofia: sounds like a nice idea. bummer it didn't work out
2023-10-10 13:35:50 <minimal> witcher: I don't believe that Wiki article actually mentioned encrypting the ESP partition
2023-10-10 13:36:05 <Sofia> (UEFI lacking that driver forced me to just put my /boot on the NVMe, though I could have similarly stored the headers or a keyfile on the microsd, I didn't bother in the end.)
2023-10-10 13:37:19 <Sofia> The only valid tampering with the signed kernels _should be_ just downgrading. You can harden it further if you wanted using the TPM to certify the latest (thus expected) kernel+initramfs. 
2023-10-10 13:38:07 <Sofia> If you're using the TPM, you can have it display a TOTP, which _you_ verify by checking against your other device. If the system were corrupted, the laptop would be unable to prove to you it wasn't tampered with.
2023-10-10 13:38:17 <Sofia> Before you type your passphrase to unlock it.
2023-10-10 13:45:49 <invoked> there is anti-evil-maid type stuff that could be done like hashing the efistub, but i am not sure anyone has done that kind of thing in alpine yet.
2023-10-10 13:47:44 <invoked> the gentoo/sakaki setup (which was i guess controversial in its time) had something like that.
2023-10-10 13:48:35 <minimal> invoked: if you're using UEFI Secure Boot and things are signed then wouldn't that cover the Evil Maid scenario?
2023-10-10 13:50:12 <invoked> not an expert, i think most evil maid solutions involved disabling secure boot.
2023-10-10 13:51:12 <invoked> qubes does it with tpm
2023-10-10 13:52:46 <Sofia> invoked: My current kernel (custom, not Alpine) is a signed efi kernel+initrd standalone file.
2023-10-10 13:53:34 <Sofia> sakaki sounds familiar, perhaps influenced my process. Ended up with my own script using dracut. Don't use Gentoo though, Exherbo, somewhat close enough.
2023-10-10 13:54:33 <minimal> Sofia: it's a UKI in other words (Unified Kernel Image)
2023-10-10 13:55:22 <Sofia> Yes.
2023-10-10 13:58:00 <invoked> sakaki had a guide for full system encryption involving (among other things) booting off usb keys, and there was a bit of conflict betweeen her and gentoo install orthodoxy. ultimately she got too busy with $dayjob, i think.
2023-10-10 13:58:47 <invoked> oh, it's still there, i guess.  https://wiki.gentoo.org/wiki/User:Sakaki/Sakaki%27s_EFI_Install_Guide
2023-10-10 14:00:07 <invoked> i dunno, as a matter of opinion, i just don't trust bios that much.
2023-10-10 14:00:30 <Sofia> I don't trust x86 or arm.
2023-10-10 14:01:32 <Sofia> Waiting on decent RISC-V workstations. Then want seL4 as hypervisor whitelisting select hardware to an alpine virtual machine, with nested qemu-x86_64-user as necessary to run software which just doesn't have RISC-V builds.
2023-10-10 14:01:49 <invoked> you'll probably be waiting a while. (years)
2023-10-10 14:02:05 <Sofia> Probably less than you think but yeah.
2023-10-10 14:02:12 <invoked> riscv amongst other problems, has no patent pool
2023-10-10 14:02:32 <Sofia> ?
2023-10-10 14:03:20 <invoked> in other words, it'll be a litigous space until they do.
2023-10-10 14:03:34 <invoked> there's not much you can count on there
2023-10-10 14:03:45 <invoked> at least for general purpose computing
2023-10-10 14:03:46 <Sofia> You want patents?
2023-10-10 14:04:01 <invoked> no, but in hardware there's no getting around it.
2023-10-10 14:04:34 <Sofia> Closest interpretation I get is like licensing for proprietary periferials?
2023-10-10 14:04:41 <invoked> in riscv people don't really care about the cpu that much actually. it's about what they package around the cpu.
2023-10-10 14:04:55 <invoked> that's where it gets into murky legal shit
2023-10-10 14:05:26 <Sofia> AFAIK there is sufficient fully open implementations of specifications which may need licensing but not royalties or patents?
2023-10-10 14:05:45 <invoked> that's just for the riscv core
2023-10-10 14:05:48 <Sofia> Iffiest to come to mind is probably radios, especially GSM.
2023-10-10 14:05:55 <Sofia> I'm talking about outside of it.
2023-10-10 14:06:18 <Sofia> PCIe, now attach an AMD GPU or wifi card, works perfectly fine today.
2023-10-10 14:06:30 <Sofia> Likewise USB et al.
2023-10-10 14:06:38 <Sofia> What are you claiming is missing?
2023-10-10 14:07:21 <invoked> much of that stuff is patent encumbered and/or trademarked.
2023-10-10 14:13:26 <nixgoat> hello everyone! i'm having some issues with tailscale and networkmanager. i've read some docs on tailscale and it seems like they take advantage of systemd-resolved for DNS resolution, which of course alpine doesn't have. therefore for systems which don't use it, tailscale rewrites resolv.conf, which networkmanager also rewrites. so what workaround can i use to have both operating? is there an 
2023-10-10 14:13:32 <nixgoat> equivalent to systemd-resolved on alpine?
2023-10-10 14:15:56 <nixgoat> errata: i meant overwrite instead of rewrite
2023-10-10 14:16:44 <minimal> witcher: have a look at this as an example of how to setup LVM-on-LUKS1 with Grub for UEFI in Alpine: https://github.com/dermotbradley/alpine-misc-notes/blob/main/phy-x86_64-uefi-grub-encrypt-lvm.sh#L415
2023-10-10 14:45:06 <chereskata> Hi, after trying to pipe find . -type f | xargs grep -e "cmake-modules-extra" 2>/dev/null into grep, it only greps in /testing of aports. running just find . -type f lists all files correctly. When i limit the find to ./community for example, everything works correct. I have a suspicion, that grep does not like .git or other hidden dirs
2023-10-10 14:45:13 <chereskata> Can anyone confirm?
2023-10-10 14:45:23 <chereskata> the folder in question is aports git repo clone
2023-10-10 16:05:48 <witcher> minimal: thanks for the help! i managed to get it working, i seem to have messed up something about the boot partition previously
2023-10-10 16:49:26 <singham> I enabled udhcpd
2023-10-10 16:49:39 <singham> An ip was assigned to client
2023-10-10 16:49:53 <singham> But from client I am unable to access server
2023-10-10 16:50:03 <singham> on server, I executed
2023-10-10 16:50:24 <singham> wget -qO - http://192.168.34.1
2023-10-10 16:50:41 <singham> and I can see the content on port 80
2023-10-10 16:50:57 <singham> But not from client computer
2023-10-10 16:51:09 <quinq> Hi singham, is there some context from a previous conversation here?
2023-10-10 16:51:18 <singham> Nope
2023-10-10 16:51:42 <quinq> So I guess your question is “what's happening?”
2023-10-10 16:51:54 <quinq> What does the HTTP client tell you?
2023-10-10 16:52:18 <singham> Network unreachable
2023-10-10 16:52:33 <singham> I'm first doing ping
2023-10-10 16:53:11 <singham> ping 192.168.34.1
2023-10-10 16:53:20 <singham> Network unreachable
2023-10-10 16:53:31 <quinq> So it's not an HTTP problem
2023-10-10 16:53:41 <quinq> You're missing a route it seems
2023-10-10 16:54:12 <quinq> On the client: ip -4 route
2023-10-10 16:54:24 <kwadroke[m]> try pinging the DHCP server. If that doesn't work, you've got DHCP issues
2023-10-10 16:55:27 <singham> Yes, 192.168.34.1 is the dhcp server
2023-10-10 16:55:28 <quinq> Most likely they're the same, IIUC
2023-10-10 16:56:18 <singham> Also, whatI feel strange is udhcpd first asks to up interface, assign ip to self and only then works
2023-10-10 16:56:34 <quinq> No, that's expected
2023-10-10 16:56:52 <singham> udhcpd should do that by default right
2023-10-10 16:57:05 <quinq> It shouldn't even do that
2023-10-10 16:57:14 <quinq> That should be the responsibility of the admin to have this setup before hand
2023-10-10 16:57:23 <quinq> Not of a DHCP daemon
2023-10-10 16:57:52 <singham> ok
2023-10-10 16:58:17 <singham> why am I getting network unreachable though
2023-10-10 16:58:21 <singham> server and client are connected with a cat6
2023-10-10 16:59:20 <quinq> 18:53:31  quinq$ So it's not an HTTP problem
2023-10-10 16:59:21 <quinq> 18:53:41  quinq$ You're missing a route it seems
2023-10-10 16:59:21 <quinq> 18:54:11  quinq$ On the client: ip -4 route
2023-10-10 17:01:00 <singham> Yes
2023-10-10 17:01:05 <singham> There is a line
2023-10-10 17:01:21 <quinq> great
2023-10-10 17:01:24 <quinq> what does it say
2023-10-10 17:01:26 <singham> 192.168.0.0/24
2023-10-10 17:01:37 <mrHeavenli> Hi, small question here. I have a laptop that I want to connect to wifi, but I dont have the packages I need to do this. And no wifi to install them over the network. My idea was to put the packages on a usb stick and install them that way. I wanted to ask how I could get the APK files for that
2023-10-10 17:01:47 <quinq> Then that's incorrect for the network you're trying to use
2023-10-10 17:01:49 <minimal> singham: do you mean udhcpc rather than udhcpd?
2023-10-10 17:01:54 <singham> dev eth0 proto dhcp ... 192.168.0.172 metric 1002
2023-10-10 17:02:12 <quinq> wait
2023-10-10 17:02:13 <singham> No, udhcpd on server
2023-10-10 17:02:31 <quinq> So you're giving an IP to a different network from your HTTP server
2023-10-10 17:02:32 <singham> dhcpcd on client
2023-10-10 17:03:01 <quinq> You're getting 192.168.0.172/24 but your server is on 192.168.34.1
2023-10-10 17:03:02 <singham> There's just 1 network
2023-10-10 17:03:10 <quinq> Nope
2023-10-10 17:03:11 <singham> eth0 interface
2023-10-10 17:03:15 <quinq> Those are two distinct networks
2023-10-10 17:03:24 <quinq> You might have one cable
2023-10-10 17:03:25 <singham> 172 is the client
2023-10-10 17:03:29 <quinq> BuNo
2023-10-10 17:03:41 <quinq> 0.172/24 is the client
2023-10-10 17:03:52 <quinq> The server is 34.1/24 (/24 most likely)
2023-10-10 17:04:09 <quinq> Meaning the client is on network 0, while the server is on network 34
2023-10-10 17:04:10 <singham> I simply don't get what /24 is
2023-10-10 17:04:18 <quinq> It's the netmask
2023-10-10 17:04:23 <quinq> 255.255.255.0
2023-10-10 17:04:28 <singham> Server is 34.1 yes
2023-10-10 17:04:45 <singham> Yes, I read on it, very confusing!
2023-10-10 17:04:52 <quinq> Meaning the network is 192.168.0 and the host is 172
2023-10-10 17:04:56 <quinq> Not really
2023-10-10 17:06:01 <quinq> If you want to use two distinct networks on one line, you need to configure your machines routing so
2023-10-10 17:06:41 <quinq> Telling that 192.168.31.0/24 is reachable on your eth0 link, and on the server that 192.168.0.0/24 same
2023-10-10 17:06:42 <singham> When the ips are 34.1 and 0.172, it should ping directly right?
2023-10-10 17:06:48 <quinq> But really just use one network instead of two
2023-10-10 17:06:55 <quinq> No it shouldn't
2023-10-10 17:07:16 <quinq> Because, I'm starting to feel like I'm repeating, those are two distinct networks, and you don't have a route for them
2023-10-10 17:07:21 <quinq> s/them/the other one/
2023-10-10 17:07:32 <singham> So, I set ips in udhcpd from 34.2-34.256?
2023-10-10 17:07:43 <quinq> I advise revising a bit IPv4 knowledge :)
2023-10-10 17:07:58 <quinq> Yeah, that would be an easier setup
2023-10-10 17:08:10 <quinq> Also 256 is out of range, but you got the idea
2023-10-10 17:15:48 <singham> Yes
2023-10-10 17:15:51 <singham> :)
2023-10-10 17:16:01 <singham> quinq: Long live you
2023-10-10 17:16:11 <singham> Long live doorknob
2023-10-10 17:16:49 <singham> I'll see you fine folks later!
2023-10-10 17:21:29 <invoked> long live doorknob?
2023-10-10 17:22:02 <quinq> no sé
2023-10-10 17:23:09 <mrHeavenli> someone able to help me out? long live doorknob.
2023-10-10 17:24:53 <quinq> Damn, you found out the secret password, now we're bound by oath to help you
2023-10-10 17:25:18 <mrHeavenli> :0 damn
2023-10-10 17:26:43 <quinq> mrHeavenli, I actually have no idea what I'm talking about, but
2023-10-10 17:27:10 <quinq> It seems you can just create a directory with packages on your usb stick, and run apk index on it to make it a repository
2023-10-10 17:28:00 <quinq> Then with apk, use -X to specify this (mounted) repository path
2023-10-10 17:28:48 <mrHeavenli> afaik i can just install from the file "apk add --allow-untrusted /path/to/file.apk"
2023-10-10 17:29:04 <mrHeavenli> problem is i dont know how to get the package file to put on the stick
2023-10-10 17:29:31 <quinq> ah
2023-10-10 17:29:35 <quinq> apk-fetch maybe
2023-10-10 17:29:45 <quinq> apk fetch - download packages from global repositories to a local directory
2023-10-10 17:29:55 <quinq> apk fetch [<options>...] packages...
2023-10-10 17:30:16 <quinq> What might help too, -R, --recursive Fetch packages and all of their dependencies.
2023-10-10 17:30:22 <quinq> But you can also read a man-page ;)
2023-10-10 17:30:46 <mrHeavenli> yes, thanks for the help
2023-10-10 17:31:50 <quinq> Good luck!
2023-10-10 17:39:32 <fluix> and for an even more "basic" solution, get them manually from the http(s) mirror(s): http://dl-cdn.alpinelinux.org/alpine/
2023-10-10 17:42:25 <mrHeavenli> fluix: thanks, this was just what i was looking for :D im on a different distro so apk didnt work when i tried just now
2023-10-10 17:57:29 <quinq> :D
2023-10-10 19:57:37 <frag> links is pretty cool (also in graphical mode), but you cant yank urls and/or open them in other application, or change binds?
2023-10-10 19:57:55 <frag> i thought this was linux! jk
2023-10-10 19:59:52 <lopid> it is linux
2023-10-10 20:05:44 <frag> hmm, seems you can copy url with mouse ..
2023-10-10 21:28:26 <pbutler> I'm new to trying to run alpine on SOC boards. I was successful in getting it running on a nanopi neo (armv7), but now can't get rk3328-roc-cc board to boot correctly. Is there a channel specific to embedded SOC board bootup?
2023-10-10 21:46:40 <ikke> No, we only have a channel for riscv64
2023-10-10 21:51:02 <pbutler> Thanks. Do you know of anyone that works on embedded boards that could help?
2023-10-10 22:10:04 <john3d> hey.  I've tried everything to change the keyboard layout in GDM, but nothing has worked so far. Is this a known issue?
2023-10-10 22:12:33 <john3d> In SDDM, I had the same issue, but I was able to solve it by using "setxkbmap" in the "/usr/share/sddm/scripts/Xsetup" file.
2023-10-11 01:09:46 <Sofia> Re alpine in qemu. I have sway installed and have the mesa-virtio drivers. Sway segfaults. Yet an archlinux distrobox with sway, does run, but has its own issue of not launching the terminal and doesn't give any errors or debug warnings.
2023-10-11 01:11:38 <Sofia> Huh, actually, new day, new boot, no segfault, but I still get errors.
2023-10-11 01:13:27 <Sofia> still/do get errors this time*
2023-10-11 01:18:03 <Sofia> http://ix.io/4IIk
2023-10-11 01:28:52 <Sofia> Meanwhile following the same instructions, adjusting for intel video drivers instead of virtio, on the bare metal laptop did work. (But that had insanely slow I/O operations. Unresolved, probably faulty USB.)
2023-10-11 02:56:02 <Sofia> #sway on libera conclusion. Errors are in deeper things. Undetermined, unpatched, but workaround found. WLR_RENDERER=pixman
2023-10-11 02:56:43 <Sofia> Latest log file (with hardware cursors disabled; prior pixman success). http://ix.io/4IIw
2023-10-11 04:24:10 <fluix> hm, I think I managed to get an arch host with some kind of virt gpu drivers in qemu running sway in alpine working
2023-10-11 04:55:55 <Sofia> fluix: I had sway running in an arch distrobox container before I got it working on the host alpine.
2023-10-11 05:31:39 <fluix> I disn't use the pixman renderer thougg
2023-10-11 05:32:35 <fluix> mobile typing hard, gn o/
2023-10-11 06:33:46 <voidz0r> moin guise hru? I just installed using "Bootstrapping Alpine Linux" on the wiki except I already got grub and a kernel so I skipped linux-lts. For some reason whenever I boot into it, openrc kinda skips itself as if running a minimal container. It goes straight into login prompt even though inittab is almost identical to the iso, only difference is getty on ttyS0. I already regenerated initramfs using mkinitfs but it works the same
2023-10-11 10:07:21 <singham> Why does lan port need a ethernet controller chip?
2023-10-11 10:07:40 <singham> Why not connect the 4 wires straight to gpio?
2023-10-11 10:07:42 <WhyNotHugo> Is docs.alpinelinux.org still work in progress? I feel that it overlaps with the wiki a lot; any thoughts on folding the two into one?
2023-10-11 10:10:17 <quinq> singham, because that would be highly inefficient (also missing electrical isolation)
2023-10-11 10:10:35 <singham> How?
2023-10-11 10:10:41 <singham> What is EI?
2023-10-11 10:10:55 <quinq> I don't know, what is EI?
2023-10-11 10:11:07 <singham> electrical isolation
2023-10-11 10:11:07 <quinq> How what?
2023-10-11 10:11:28 <quinq> https://en.wikipedia.org/wiki/Electrical_isolation
2023-10-11 10:11:33 <singham> How would that be inefficient? OS would handle it right?
2023-10-11 10:11:42 <quinq> right
2023-10-11 10:11:48 <quinq> All would be done by CPU
2023-10-11 10:13:05 <singham> So nic is like a separate CPU?
2023-10-11 10:13:26 <singham> A proper reader of packets and which separates them to right addresses?
2023-10-11 10:13:40 <quinq> It's a specialized IC
2023-10-11 10:13:51 <ikke> singham: It handles things like supplying the right voltage at the right time
2023-10-11 10:14:23 <singham> So it merely stabilizes signal?
2023-10-11 10:14:30 <quinq> (that'd be the phy)
2023-10-11 10:14:40 <singham> And corrects the 0 and 1s properly?
2023-10-11 10:14:54 <singham> What is RTL8305S?
2023-10-11 10:15:10 <quinq> The ethernet controller rather deals with proto handling, verifying frame format, filtering, reassembly, etc.
2023-10-11 10:16:57 <ikke> Wouldn't putting wires on a GPIO bypass the PHY as well?
2023-10-11 10:17:58 <WhyNotHugo> I'm setting up a media centre for my living room display. Unsure if I should go with diskless mode, data disk or system mode.
2023-10-11 10:18:14 <WhyNotHugo> Data disk sounds like more work to set up but easier to later keep in an immutable state.
2023-10-11 10:18:25 <WhyNotHugo> System sounds easier to set up.
2023-10-11 10:18:31 <WhyNotHugo> Any advise on this?
2023-10-11 10:18:55 <singham> OS can handle that!
2023-10-11 10:19:12 <singham> in theory ofc
2023-10-11 10:19:14 <ikke> singham: the problem is that it can't.
2023-10-11 10:19:22 <ikke> I mean, the problem is not that it cant*
2023-10-11 10:19:35 <ikke> Everything you can offload from the OS is a win
2023-10-11 10:19:51 <ikke> Because the CPU can use those cycles for things that matter
2023-10-11 10:20:19 <WhyNotHugo> singham: A dedicatd controller is a lot more power efficient. Like, the CPU to almost completely wind down while the ethernet controller stays awake waiting for new packages using extremely little power.
2023-10-11 10:20:20 <ikke> Same reason things like DMA exist
2023-10-11 10:20:28 <singham> User controls OS whereas these things are baked into hardware
2023-10-11 10:20:36 <WhyNotHugo> And yes, also freeing up CPU cycles is important, especially given that ethernet nics are comparatively cheap.
2023-10-11 10:21:00 <singham> One can always have 1 more core within processor for such processing
2023-10-11 10:21:10 <ikke> and still perform poorly
2023-10-11 10:21:21 <quinq> haha yeah
2023-10-11 10:21:27 <singham> Like dedicated network core which runs processes 1-10 say
2023-10-11 10:21:32 <quinq> Good luck trying to get 1Gb
2023-10-11 10:22:03 <WhyNotHugo> singham: The ethernet controller IS a separate core, and one highly specialised at handling ethernet packets more efficiently than your main CPU's core.
2023-10-11 10:22:21 <singham> Aah yes
2023-10-11 10:22:35 <quinq> Sure, if your target is to have *1* core dedicated to *1* ethernet interface (up to 100Mbps), why not
2023-10-11 10:22:38 <singham> true. back to the current system
2023-10-11 10:22:40 <quinq> But that's an awful waste
2023-10-11 10:22:48 <WhyNotHugo> It's also physically wired in a way to make the whole thing as fast as possible -- if you did that for the CPU, you'd be compromising on some other aspect of it.
2023-10-11 10:23:27 <singham> But still, ethernet controller should be configurable but processor
2023-10-11 10:23:36 <singham> fully configurable
2023-10-11 10:23:45 <singham> maybe an FPGA would be nice there
2023-10-11 10:25:12 <WhyNotHugo> singham: If you need a programmable ethernet controller, you buy specialised harware. The non-programmable one is a lot cheaper and a good-enough fit for over 99% of the cases.
2023-10-11 10:25:33 <quinq> An FPGA would also be a huge waste
2023-10-11 10:25:37 <quinq> But yes, it would work
2023-10-11 10:25:54 <WhyNotHugo> Like, imagine the price jump of an rpi if you replace the ethernet controller with an FPGA.
2023-10-11 10:25:57 <ikke> It's not that you have a lot of options to implement ethernet properly
2023-10-11 10:25:59 <quinq> If your concern is about open hardware, that's a different topic
2023-10-11 10:26:07 <quinq> It's not an unmeaningful one!
2023-10-11 10:26:23 <WhyNotHugo> 
ACTION agrees that the hardware shuold be open, even if single-purpose and not configurable.

2023-10-11 10:26:28 <quinq> But I don't think that trying to put things in CPUs (that are not more open than your ethernet controller) isn't the way to go
2023-10-11 10:26:50 <singham> WhyNotHugo: You are an intelligent man!
2023-10-11 10:27:27 <singham> Joanna Rutkowska says, distrust peripherals
2023-10-11 10:28:15 <quinq> Like a GPIO controller?
2023-10-11 10:29:00 <singham> A CPU should distrust all peripherals and trust just keyboard connection and mouse
2023-10-11 10:29:18 <singham> And ofc, the core of kernel
2023-10-11 10:29:21 <singham> TCB
2023-10-11 10:33:11 <WhyNotHugo> singham: What if my keyboard is connected via a USB hub at the back of my monitor?
2023-10-11 10:33:44 <singham> PS2s existed for a reason!
2023-10-11 10:33:57 <singham> USBs are so bad
2023-10-11 10:34:20 <singham> Your pc can easily assume a usb flash drive to be a keyboard
2023-10-11 10:34:39 <singham> Our old desktops were designed so beautifully and carefully!
2023-10-11 10:34:48 <singham> where GPU was untrusted
2023-10-11 10:35:39 <singham> processor had a nice cooling metal plate which acted as something which prevents emw leak
2023-10-11 10:36:12 <singham> swappable processors with standard pin distancing etc
2023-10-11 10:36:45 <singham> PCIE for having more untrusted hardware! So beautiful!
2023-10-11 10:37:24 <singham> And today, put everything on a small chip, throw away everything and buy new one on problems
2023-10-11 10:37:48 <singham> Surrender yourself completely to the manufacturer! 
2023-10-11 10:44:11 <singham> Later folks
2023-10-11 10:44:17 <singham> ====
2023-10-11 10:44:24 <singham> Thanks for the help
2023-10-11 10:44:27 <singham> ======
2023-10-11 11:03:58 <linfan> Is there any up to date wiki page?
2023-10-11 11:08:11 <Sofia> linfan: What feels out of date?
2023-10-11 11:08:20 <Sofia> Everything I've followed seems up to date.
2023-10-11 11:08:24 <linfan> everything, lol
2023-10-11 11:08:35 <Sofia> Example?
2023-10-11 11:09:12 <linfan> install desktop only gnome works and then pipewire does not work properly
2023-10-11 11:09:14 <Sofia> Sway page mentions pipewire. Granted I've had audio issues, but at least that indicates somewhat modern practices.
2023-10-11 11:09:26 <Sofia> Ah so I'm not the only having audio issues :p
2023-10-11 11:09:56 <Sofia> I seem to have issues without pipewire, with just alsa though. Working with qemu, learning its -manyarguments.
2023-10-11 11:10:07 <linfan> i tried setup desktop xfce and it complained there is no lighdm service running
2023-10-11 11:10:42 <Sofia> I don't use X/xfce, haven't seen those docs
2023-10-11 11:10:45 <linfan> i am setting it up in virtualbox
2023-10-11 11:11:13 <Sofia> https://wiki.alpinelinux.org/wiki/Xfce
2023-10-11 11:11:21 <Sofia> Docs seem to indicate you're to start lightdm
2023-10-11 11:11:26 <linfan> gnome works as expected but there are the pipewire, wireplumber, pipewire-pulse oddities
2023-10-11 11:12:58 <linfan> sometimes you are told to chose setup-desktop and pick one, gnome, plasma or xfce. Only gnome installs completely
2023-10-11 11:13:39 <linfan> On the other hand wikis tell you to install the desktop files separately after install...
2023-10-11 11:15:36 <linfan> there is so much in alpine I love, but the wikis suck and the package-handler does not show me what I have already installed
2023-10-11 11:16:07 <linfan> it only informs the number of installed packages
2023-10-11 11:17:58 <linfan> but for now i will install gnome desktop for the n:th time and wait some months, maybe some day pipewire will work
2023-10-11 11:19:10 <Sofia> Can list what is installed, and installed which are dependencies. No list for non-deps?
2023-10-11 11:21:17 <linfan> other package manager, e.g. pacman, gives you a list when you search whick pkg is installed and which are not, also if you install something already installed it asks you if you want to reinstall it
2023-10-11 11:21:51 <Sofia> linfan: cat /etc/apk/world
2023-10-11 11:22:18 <linfan> have to set it up again for the 20th + time
2023-10-11 11:23:44 <Sofia> I haven't used alpine properly in about a decade, was always my favorite system. Very simple package manager with small, unbloated, hardened, binary packages.
2023-10-11 11:24:14 <Sofia> Stopped using it because of glibc friction.
2023-10-11 11:24:29 <Sofia> Since seemingly resolved thus trialing in qemu.
2023-10-11 11:24:46 <linfan> yeah, understandable
2023-10-11 11:25:39 <Sofia> Resolved with gcompat, and flatpak + distrobox to use slimmed down alternate systems (which are many times larger than alpine despite slimming efforts. :p)
2023-10-11 11:27:12 <linfan> i have been true to arch linux since start mostly, but also enjoy slackware - all debian clones are no-no and also sceptical to rpi-based stuff
2023-10-11 11:31:22 <Ermine> linfan: what problems do you have with pipewire?
2023-10-11 11:31:26 <Sofia> Arch was certainly a notable distro for me. Unsure if I've been using exherbo longer now. I've hopped between: debian, arch, freebsd, alpine, gentoo, nixos, and exherbo. Exherbo is pretty decent but I've really been wasting too much time with source based distros. Feel like it has only caused me more troubles than its benefits offered. Really missed Alpine. Soon time to migrate
2023-10-11 11:31:29 <Sofia> back to it as my host.
2023-10-11 11:33:08 <linfan> Ermine: i run it on wayland gnome and system sounds do not work at all and the sound is, how should I put it, crackling maybe
2023-10-11 11:33:46 <Sofia> 
ACTION hasn't managed to get pipewire working on her exherbo system; part of why she just wants to switch to alpine. And then has issues with just alsa in qemu. :p

2023-10-11 11:34:11 <Ermine> linfan: is it bare metal or VM or something?
2023-10-11 11:34:27 <Sofia> They mentioned VirtualBox.
2023-10-11 11:34:29 <linfan> bare metal?
2023-10-11 11:34:45 <Sofia> Bare metal as in, no virtual machines or containers.
2023-10-11 11:35:57 <linfan> i have and presently for the 20th+ time am setting up gnome desktop and I guess there are some conflicts with the pulse, which, as I understan, is a gnomish thing
2023-10-11 11:36:27 <Ermine> neither pulseaudio or pipewire are gnomish
2023-10-11 11:36:54 <linfan> now but in order to get sound out of gnome i need to install pipewire-pulse
2023-10-11 11:36:56 <Sofia> 
ACTION creates a new qemu image because she mistakenly made it too small. Reinstall time too. :p

2023-10-11 11:37:25 <Ermine> Seems like gnome wants to do system sounds through pulseaudio
2023-10-11 11:37:57 <Ermine> So with PW you need conpatibility layer, pipewire-pulse. This one is documented
2023-10-11 11:39:05 <Ermine> So, with pulseaudio uninstalled, and pipewire and pipewire-pulse installed, do you get crackling sounds?
2023-10-11 11:39:21 <linfan> but for some reason I can get a flicker out of sound on first start, but then it stops responding
2023-10-11 11:39:41 <linfan> yes i do get crackling sounds that way
2023-10-11 11:40:15 <linfan> like listening to youtube music it is quite irritating
2023-10-11 11:42:56 <linfan> ok gnome desktop set up, now i will reboot, without wireplumber and will have no sound
2023-10-11 11:43:24 <Ermine> without wireplumber you'll have no sound, that's for sure
2023-10-11 11:43:45 <linfan> i mean w/o any sound server
2023-10-11 11:44:05 <linfan> except what gnome eventually has installed
2023-10-11 11:44:06 <Ermine> As for crackling, I'll need to get home first. I'll contact you then
2023-10-11 11:44:23 <linfan> ok, nice
2023-10-11 11:44:40 <linfan> on alpine which irc client to use?
2023-10-11 11:45:48 <donoban> linfan: probably you can use your favourite :)
2023-10-11 11:45:49 <Ermine> any client you like
2023-10-11 11:46:00 <linfan> ok, hexchat
2023-10-11 11:46:22 <linfan> well, with neither sound service just what gnome puts up sound sounds awesome
2023-10-11 11:47:04 <linfan> cyl
2023-10-11 11:47:06 <Ermine> You may want to try changing virtualbox audio parameters
2023-10-11 11:47:36 <linfan> that is done thousands of times, only works with ac97
2023-10-11 11:48:16 <linfan> will now customize gnome reboot and check if sound persists
2023-10-11 12:01:16 <Sofia> I get crackling sounds with just alsa with qemu.
2023-10-11 12:01:38 <Sofia> When the sound even decided to work at all that is.
2023-10-11 12:04:25 <invoked> i haven't tried sndio in linux yet but it'd probably be good if everything just went to that.
2023-10-11 12:06:00 <Sofia> sndio?
2023-10-11 12:06:20 <invoked> sndio.org
2023-10-11 12:07:12 <Sofia> I hear "pro audio software devs" approve of Jack as "the right interface", and won't support pipewire directly because there exists the Jack support with all the benefits and matches low latency.
2023-10-11 12:08:17 <Sofia> Sounds good at least.
2023-10-11 12:08:19 <invoked> maybe, for people who are in audio production
2023-10-11 12:08:25 <Sofia> Can't say anything more than that. Not my area. :)
2023-10-11 12:08:41 <invoked> but outside of that i'm not sure. sndio would be better default generally (opinions being what they are)
2023-10-11 12:09:22 <invoked> i think pipewire's okay, though
2023-10-11 12:09:31 <invoked> but mainly in comparison to pulseaudio
2023-10-11 12:57:57 <sewn> how can I debug a kernel panic when the panic backtrace is stuck in a low resolution and I can't see many details?
2023-10-11 12:58:34 <lopid> squint
2023-10-11 13:00:15 <sewn> the backtrace is in a resolution of which the details, modules n shit are hidden
2023-10-11 13:00:36 <sewn> during the panic before the kms initializes it was in a bios resolution 
2023-10-11 13:01:11 <sewn> I'll try disabling that actually
2023-10-11 13:03:20 <sewn> disabling kms fixes the kernel panic
2023-10-11 14:04:34 <linfan> hey, i am back. phone calls and stuff. yeah, I was wondering about sound issues in gnome in vbox earlier. Don't remember who I was talking with
2023-10-11 16:58:04 <Ermine> linfan: did you install rtkit?
2023-10-11 16:58:19 <linfan> hi Ermine, yes i did
2023-10-11 16:59:19 <linfan> I wonder if it is some codecs missing
2023-10-11 16:59:43 <linfan> or ffmpeg that alppine uses
2023-10-11 17:01:00 <Ermine> if you were missing codecs, you wouldn't be able to play the file at all
2023-10-11 17:01:29 <Ermine> ffmpeg is used not by Alpine, but by programs which needs its services
2023-10-11 17:01:39 <linfan> k, i have installed wireplumber and pipewire and pipewire-pulse
2023-10-11 17:01:58 <Ermine> and everything works for me 
2023-10-11 17:02:07 <Ermine> no crackling at all
2023-10-11 17:02:09 <linfan> and it purged pulseaudio and alsa
2023-10-11 17:02:44 <linfan> how could i start debugging the sound
2023-10-11 17:05:24 <linfan> have run wpctl inspect but don't know what i am looking for really
2023-10-11 17:06:54 <Ermine> That said, I changed device to Intel HD Audio in virtualbox settings
2023-10-11 17:07:01 <linfan> when running /usr/libexec/pipewire-launcher volume control disappears
2023-10-11 17:07:29 <Ermine> You don't need that in gnome
2023-10-11 17:07:33 <linfan> ok, will check, although i have done that before
2023-10-11 17:08:20 <Ermine> Well, no sound with ICH AC97 controller
2023-10-11 17:08:32 <Ermine> did you really try Intel HD Audio?
2023-10-11 17:19:29 <linfan> sound only possible with ac97, laptop runs realtek so no direct sound
2023-10-12 02:14:18 <Saijin_Naib[m]> Anyone using Proton Mail under Alpine without the bridge? If so, just the webapp, right? I'm not finding a client packaged, nor an official flatpak... I might need to pressure my organization to reconsider Proton
2023-10-12 02:40:10 <idkrn[m]> There is no official desktop client
2023-10-12 02:40:31 <idkrn[m]> That's why the bridge exists
2023-10-12 04:52:33 <quinq> Doesn't the mail in proton mail implies that it supports standard smtp/imap?
2023-10-12 05:14:44 <Saijin_Naib[m]> It does, if you pay for the bridge, otherwise you can't use anything but their clients and the webui
2023-10-12 05:14:56 <Saijin_Naib[m]> There's an unofficial electron wrapper in Flathub, so that does the trick for now
2023-10-12 05:15:21 <Saijin_Naib[m]> I'm going to push for moving to NextCloud for mail, alongside literally everything else we use
2023-10-12 05:15:31 <Saijin_Naib[m]> <idkrn[m]> "There is no official desktop..." <- Lovely of them -_-
2023-10-12 07:17:05 <frojnd> I got a t14 3rd gen that I am in a process of installing alpine linux on it. Problem is this laptop is just a temporary piece of hw since I'll get a new t14 in a 2, 3 months... so how can I somehow save my setup (I am gonna do now) so I can apply some sort of automation in 2,3 months on a new t14? Any suggestions? I do now I'm gonna have zfs full encryption so that part can be
2023-10-12 07:17:06 <frojnd> done manually
2023-10-12 07:46:47 <quinq> frojnd, just move the disk from setup 1 to setup 2
2023-10-12 07:46:59 <quinq> s/disk/storage media/
2023-10-12 08:06:21 <Sofia> Alpine with sway, flatpak, distrobox, mpv and a few other things installed: less than 600 MiB
2023-10-12 08:06:23 <Sofia> Archlinux distrobox container with mpv: 3.6 GiB
2023-10-12 08:06:25 <Sofia> Flatpak dedicated for mpv: 1.7 GiB
2023-10-12 08:06:27 <Sofia> Enough said? :p
2023-10-12 08:08:38 <frojnd> Hm. Maybe I missed somewhere but...
2023-10-12 08:09:35 <frojnd> Let's say I'm total newbie (which I am) I enter live env and trying to install alpinelinux. I know about alpine-* but is there any help like a file that I can read that says something about manual pages and such :)
2023-10-12 08:09:57 <frojnd> s/salpine-*/setup-*/
2023-10-12 08:10:49 <Sofia> frojnd: https://wiki.alpinelinux.org/wiki/Alpine_Linux:FAQ#Why_don't_I_have_man_pages_or_where_is_the_'man'_command?
2023-10-12 08:11:16 <frojnd> I mean is there any help already inside live env while booting live usb?
2023-10-12 08:12:08 <Sofia> setup-alpine --help ;)
2023-10-12 08:15:51 <frojnd> Hm interesting. dd --help shows no progress flag
2023-10-12 08:15:59 <Sofia> status=progress
2023-10-12 08:16:32 <Sofia> Hmm
2023-10-12 08:16:37 <Sofia> Oh, this is probably a busybox dd.
2023-10-12 08:16:42 <Sofia> Yeah it is
2023-10-12 08:16:45 <frojnd> dd: invalid argument 'progress' to s'status'
2023-10-12 08:17:07 <frojnd> It's preinstalled one
2023-10-12 08:17:20 <frojnd> So I have to install non busybox dd
2023-10-12 08:17:33 <frojnd> In order to see how many hours/days will urandom take for 256GB hdd
2023-10-12 08:18:02 <Sofia> Trying to secure-erase it?
2023-10-12 08:18:52 <frojnd> Yes
2023-10-12 08:18:59 <frojnd> I always do on a new hwe
2023-10-12 08:20:09 <Sofia> Oh new disk, not secure erase, just global random mask to hide real data usage footprint?
2023-10-12 08:20:50 <Sofia> "HDD" as you said or really SSD?
2023-10-12 08:20:54 <frojnd> Windows preinstalled in domain name
2023-10-12 08:21:12 <frojnd> "new computer" Someone already did their work using Windows on it 
2023-10-12 08:21:38 <frojnd> Sofia: but yeah, just to hide real data
2023-10-12 08:22:01 <Sofia> Okay. Then yeah urandom that. :)
2023-10-12 08:22:14 <frojnd> Was hoping for a progress ;P
2023-10-12 08:22:24 <Sofia> https://pkgs.alpinelinux.org/contents?file=dd&path=&name=&branch=edge
2023-10-12 08:22:39 <Sofia> Pick a package. :)
2023-10-12 08:23:13 <Sofia> Probably after coreutils.
2023-10-12 08:23:30 <frojnd> Sofia: yes, coreutils.
2023-10-12 08:23:51 <frojnd> Now I Need to learn how to use apk in order to show info like on pkgs web page.
2023-10-12 08:24:30 <frojnd> At least I have docs now :)
2023-10-12 08:24:37 <Sofia> :)
2023-10-12 08:24:48 <Sofia> The site or installed man pages and the docs metapackage?
2023-10-12 08:25:14 <frojnd> Sofia latter. Btw, anyone using s6 supervision tools?
2023-10-12 08:25:58 <frojnd> Sofia: installed man pages and the docs metapackages
2023-10-12 08:26:18 <Sofia> frojnd: HDD or HDD?
2023-10-12 08:26:24 <Sofia> SSD or HDD*
2023-10-12 08:26:29 <frojnd> It's nvme
2023-10-12 08:26:35 <Sofia> So SSD?
2023-10-12 08:26:47 <Sofia> https://wiki.archlinux.org/title/Solid_state_drive/Memory_cell_clearing#NVMe_drive
2023-10-12 08:27:31 <frojnd> Oh thank you!
2023-10-12 08:27:40 <frojnd> Not sure if dd urandom is needed now
2023-10-12 08:27:46 <frojnd> I guess it is.
2023-10-12 08:28:03 <Sofia> urandom won't be if you can instruct the disk to destroy the data securely for you.
2023-10-12 08:28:25 <Sofia> Which may even be deeper than urandom could try.
2023-10-12 08:29:09 <frojnd> Gonna do both methods. urandom is "quite" fast: 144MB/s I have total of 256GB of storage on that laptop
2023-10-12 08:29:34 <frojnd> So you also use Alpine linux for desktop usage?
2023-10-12 08:30:11 <frojnd> I think I will need to use flatpak since Android Studio is not working out of the box for musl systems.
2023-10-12 08:30:21 <frojnd> Probably a few more "dedicated" apps from flatpak
2023-10-12 08:30:44 <Sofia> I haven't used alpine properly for about a decade. Just installed in qemu recently to verify a few things. As far as I can tell it will work perfectly for everything I need, and any holes can be filled with flatpak, distrobox, or chroots.
2023-10-12 08:31:53 <Sofia> Returning to alpine's minimalism from a source-based distro where its benefits really haven't been enough to outweigh the complexity over alpine.
2023-10-12 08:33:32 <Sofia> My issues with alpine were all related to glibc and any software which relied upon it. gcompat should fill a decent chunk of this. Have the other fallbacks if necessary.
2023-10-12 08:34:51 <frojnd> I was using Arch (still am untill I backup sway and my scripts) but it became quite complex and bulky over the years. So I'm gonna try Alpine with zfs full encrpytion (and hopefully secure boot enabled) with maybe virtual envs like qemu or don't know to have separated my main system from tools I need on a job. Not sure yet how to do that.
2023-10-12 08:37:03 <frojnd> I do love Arc's wiki though. I mean who doesn't.
2023-10-12 08:38:00 <Sofia> The Archwiki will remain useful while on Alpine. :)
2023-10-12 08:38:46 <Sofia> I'll potentially try zfs, though really I'll probably use btrfs short-term while waiting for bcachefs to replace it.
2023-10-12 08:40:26 <frojnd> I assume it's gonna be more light than zfs (zfs is amazing for data integrity)
2023-10-12 08:42:22 <Sofia> For now, just ext4 + lvm + luks with the default setup-alpine process. Not custom.
2023-10-12 08:42:57 <Sofia> Was surprised to find the non-lvm cryptsys option leaves an unencrypted swap. That is a very big no-no. :/
2023-10-12 08:43:37 <Sofia> Unless I misinterpreted it. Certainly looked like a plain swap, not even an ephemeral encrypted swap.
2023-10-12 08:51:13 <Sofia> Re flatpak I'm very disappointed to have _only_ a "dedicated" mpv, weighs 1.7 GiB.
2023-10-12 08:52:35 <Sofia> Suspect plenty of that blame goes to the packager's choices, but at the very least, if the platform/locale data is that big, surely it can be slimmed down with optional subsets.
2023-10-12 09:17:35 <Sofia> Re gcompat. Should I expect tor-browser to work? It _segfaults_ after a mutex error message.
2023-10-12 09:18:37 <Ermine> What a week
2023-10-12 09:18:50 <linfan> uh?
2023-10-12 09:19:21 <Ermine> linfan: I tried to run Alpine with gnome in vbox on laptop with realtek audio controller
2023-10-12 09:19:34 <Ermine> intel hd audio + directsound works gere
2023-10-12 09:20:10 <linfan> here only ac97 direct sound works, whit a crackling sound
2023-10-12 09:20:31 <Sofia> direct sound?
2023-10-12 09:20:43 <linfan> win 11 has realtec sound codec and for mic intel
2023-10-12 09:21:03 <linfan> in vobox window directsound/ac97
2023-10-12 09:21:24 <linfan> Default/ac97 works too
2023-10-12 09:21:42 <Ermine> Sofia: 1) Why don't you run tor browser without gcompat; 2) Can you please get core dump, or backtrace of tor when it crashes (and probably make an issue on alpine gitlab)?
2023-10-12 09:21:51 <linfan> in vbox audio settings
2023-10-12 09:22:24 <Ermine> Heck, on my machines it doesn't even detect ac97
2023-10-12 09:23:21 <Ermine> what vbox version do you have btw?
2023-10-12 09:23:29 <Sofia> Ermine: How do you suggest (1) ? Already trying in an arch distrobox, installing gtk3 currently.
2023-10-12 09:24:47 <linfan> well, virtualbox 6.0.10 lists default, windows direct sound, medea session and subsection ac97 (which is default), hda-intel, sb16 and null. 
2023-10-12 09:25:00 <Sofia> (2) potentially but it took a bit to figure out how for sway. Turns out `ulimit -c unlimited` wasn't sufficient to enable core dumps or I couldn't find where they went. Had to use ... `gdb --batch sway -ex 'run --debug' -ex 'thread apply all bt full'` to get a core with everything; after installing a bunch of -dbg for symbols.
2023-10-12 09:25:28 <Ermine> linfan: you mean 7.0.10 ?
2023-10-12 09:26:14 <linfan> yeah, sorry 7.0.10
2023-10-12 09:27:25 <linfan> is it a bit complex, coz both arch linux and slackware plays fine
2023-10-12 09:28:11 <linfan> archlinux plays with direct sound hd-intel and slack with direct sound ac97, lol
2023-10-12 09:28:43 <linfan> but have not tried alpine with sb16 yet, maybe worth a try
2023-10-12 09:30:48 <Sofia> Ermine: Works under archlinux after installing gtk3 and dbus-glib.
2023-10-12 09:30:54 <linfan> some compination of these vbox audio combos render hich pitch music
2023-10-12 09:31:44 <linfan> well arch and slack as i understand runs glibc, while alpine at least mine runs musl
2023-10-12 09:32:33 <linfan> e.g. libc (w/o the g)
2023-10-12 09:33:54 <linfan> but if gnome is ported to use musl, wouldn't the sound modules also do the same?
2023-10-12 09:33:57 <quinq> Yes, GNU libc vs. musl libc
2023-10-12 09:34:14 <quinq> There is no need to port something to musl libc
2023-10-12 09:34:29 <quinq> There might be a need to remove GNU libc extensions though
2023-10-12 09:34:55 <Sofia> Ermine: Correction. Needed gdb --batch because I needed sway to terminate to close its handles. Guess I don't need the core for the backtrace.
2023-10-12 09:35:18 <Sofia> Or can generate-core-file from within gdb.
2023-10-12 09:35:23 <linfan> it was specifically the GNU bloat that attracted my attention in alpine plus the absence of systemd
2023-10-12 09:37:41 <linfan> but i find it a bit hard to navigate the alpine system still. Would be nice to get a pkg description and information if a package is already installed, when doing apk search
2023-10-12 09:38:18 <Sofia> Ermine: core acquired. No debug symbols and no indication which *-dbg might help, if any.
2023-10-12 09:38:27 <linfan> gettin information about how many packages i now have installed is quite nonsensical
2023-10-12 09:38:58 <Sofia> https://share.riseup.net/#5bEIolz3lIa86QGT6xHIRQ
2023-10-12 09:40:35 <quinq> linfan, I agree about information
2023-10-12 09:40:44 <quinq> There's apk list though
2023-10-12 09:43:51 <linfan> hmm, how come I have this glib-2.76.4-r0 x86_64 {glib} (LGPL-2.1-or-later) [installed]
2023-10-12 09:44:10 <Sofia> The backtrace is especially unhelpful. http://ix.io/4IOy 
2023-10-12 09:45:07 <linfan> guess it is gnome dragging in those ugly ones :-)
2023-10-12 09:46:52 <Ermine> linfan: sb16 is ancient, I don't think it's a suitable option
2023-10-12 09:47:35 <linfan> i presume gnome is to blame for the glib steamroller
2023-10-12 09:47:36 <Ermine> Maybe something is wrong about kernel configuration
2023-10-12 09:48:15 <Ermine> it has nothing to do with gliv
2023-10-12 09:48:18 <Ermine> glib*
2023-10-12 09:48:47 <quinq> linfan, apk info -r glib
2023-10-12 09:50:05 <Ermine> Sofia: in Arch linux you don't need gcompat, since Arch is glibc-based
2023-10-12 09:50:23 <linfan> required by gnome but also wireplumber, so no clean musl alpine, it is a mishmash, eh...?
2023-10-12 09:50:40 <Sofia> frojnd: Re s6, last I heard people were trying to replace openrc in alpine.
2023-10-12 09:50:48 <quinq> What do you mean by clean?
2023-10-12 09:50:59 <quinq> If you don't want gnome, don't install gnome
2023-10-12 09:51:11 <linfan> a linux w/o glib*
2023-10-12 09:51:15 <quinq> If you install gnome, you'll have gnome libs
2023-10-12 09:51:19 <Sofia> Ermine: Right but I don't want to use the arch container by default. I'm okay to have it as a fallback when I need it.
2023-10-12 09:51:59 <quinq> linfan, maybe you're mixing glib (Gnome lib) and glibc (GNU libc)
2023-10-12 09:52:14 <Sofia> The segfault is on alpine (with gcompat). The arch distrobox runs tor-browser totally fine.
2023-10-12 09:52:18 <linfan> ah, i see, yes, ty
2023-10-12 09:52:51 <quinq> (well, I'm not sure the g actually stands for gnome, maybe rather GIMP like originally for GTK, dunno)
2023-10-12 09:53:17 <linfan> well as opposed to qt, maybe gtk
2023-10-12 09:53:33 <quinq> gtk isn't the opposite of qt
2023-10-12 09:53:41 <quinq> Those are just two graphical toolkits
2023-10-12 09:53:46 <quinq> There are others
2023-10-12 09:54:15 <linfan> but i understand that kde (plasma) uses qt, while gnome not
2023-10-12 09:54:55 <quinq> Sure, kde isn't gnome
2023-10-12 09:55:01 <Ermine> Sofia: so, on Alpine, how do you run Tor browser? Is it from Alpine repos, or have you built it yourself
2023-10-12 09:55:43 <Ermine> linfan: which kernel do you use? linux-virt? linux-lts? linux-edge?
2023-10-12 09:56:12 <linfan> lts, also tried edge, no difference as to sound
2023-10-12 09:56:45 <linfan> i have 2.18 (standard)
2023-10-12 09:56:53 <Sofia> Ermine: There is no tor-browser package for alpine. This is their official, reproducible build and you generally don't _want_ to alter that build setup because you want it to be consistent for anonymity set reasons.
2023-10-12 09:56:55 <linfan> 3.18*
2023-10-12 09:57:13 <Ermine> there's virt edition for virtual machines
2023-10-12 09:57:18 <Ermine> try linux-virt kernel
2023-10-12 09:57:40 <Ermine> Maybe it describes why intel hd audio works for me and ac97 doesn't
2023-10-12 09:57:45 <Sofia> Thus tor browser is an instance of a tar-extracted almost standalone archive with few-ish dependencies on the host. Annoyingly it isn't fully self contained, thus dependency issues.
2023-10-12 09:57:51 <linfan> do i need to reinstall for that? the kernel changing is quite unclear to me on alpine
2023-10-12 09:58:14 <Sofia> At the very least: glibc, gtk3, and dbus-glib. Probably more.
2023-10-12 09:58:42 <Ermine> linfan: apk add linux-virt ; apk del linux-lts and then reboot
2023-10-12 09:58:50 <linfan> so is it that simple as to do doas apk linux-vir or something, will it replace the kernel
2023-10-12 09:59:16 <Ermine> First command will install virt kernel, second one will remove lts one
2023-10-12 09:59:21 <linfan> will do that, ty
2023-10-12 09:59:41 <Sofia> Ermine: That runs sufficient grub* hooks? 
2023-10-12 09:59:50 <Ermine> Sofia: it should
2023-10-12 09:59:54 <Sofia> Neat
2023-10-12 10:00:07 <Ermine> or syslinux hooks if syslinux is installed
2023-10-12 10:01:44 <linfan> done, lets see what happens, maybe also change to hd intel, before firing up it again
2023-10-12 10:02:43 <Sofia> And they were never seen here again...
2023-10-12 10:04:13 <frojnd> Sofia: trying as in default alpine image?
2023-10-12 10:05:43 <Ermine> Sofia: I'll get back to your issue on the weekend. We'll see if there are any improvements that need to be made on gcompat side
2023-10-12 10:08:11 <Ermine> the fact linfan isn't coming back worries me
2023-10-12 10:12:33 <Sofia> Ermine: Thanks, though you may want to save that coredump from my link, it will be deleted in "12 hours" aka I think 24? (Configuration vs. claimed label is a bit off.)
2023-10-12 10:14:24 <Sofia> frojnd: Default alpine image re which exactly?
2023-10-12 10:14:24 <Sofia> I'm running the virtual image installed in qemu (and have my own audio issues with crackling); with sway setup with pipewire verified to "work" with mpv on alpine, in arch, and in flatpak.
2023-10-12 10:14:39 <Ermine> Yeah, did this
2023-10-12 10:15:02 <Sofia> :)
2023-10-12 10:15:18 <frojnd> Sofia: I mean if people are trying to make s6 deafult when installing alpine for the first time?
2023-10-12 10:15:50 <Ermine> To make s6 default you need to write all service dirs yourself
2023-10-12 10:15:53 <Sofia> Ah that. I believe so but I didn't hear it directly from alpine people, I think from the people working on s6.
2023-10-12 10:16:07 <Sofia> Welcome back linfan, we were starting to worry. <3
2023-10-12 10:16:46 <Sofia> Was a femdom talk.
2023-10-12 10:16:49 <Sofia> Not again... fosdem*
2023-10-12 10:16:52 <linfan> I tried all kinds of combos, but now all sound is gone, but ac97 started dummy card, though
2023-10-12 10:16:57 <Sofia> https://yt.drgnz.club/watch?v=gZqIEstv5lM
2023-10-12 10:17:36 <linfan> now have direct sound and hd intel, card identified but gives no sound
2023-10-12 10:18:01 <Ermine> At least you're back!
2023-10-12 10:18:13 <Sofia> =^,^=
2023-10-12 10:18:55 <linfan>  └─ Default Configured Node Names:
2023-10-12 10:18:55 <linfan> wpctl status reports ettings
2023-10-12 10:18:55 <linfan>          0. Audio/Sink    auto_null
2023-10-12 10:19:00 <Ermine> So, at this point you have gnome, pipewire and wireplumber, right?
2023-10-12 10:19:10 <linfan> sure
2023-10-12 10:19:10 <Ermine> and intel hd audio
2023-10-12 10:19:22 <linfan> yes, as direct sound
2023-10-12 10:19:32 <linfan> not as default or media session
2023-10-12 10:20:02 <linfan> i tried all those combinations, therefore the delay
2023-10-12 10:21:03 <Sofia> frojnd: Did you try the nvme-sanitize tool?
2023-10-12 10:21:21 <Ermine> dumb guess, do you have audio unmuted on both host and guest and audio output is on in virtualbox?
2023-10-12 10:22:10 <linfan> nope
2023-10-12 10:22:30 <linfan> but i wonder if it thinks my audio comes from monitor speakers
2023-10-12 10:23:49 <linfan> is there any latest kernel or is it only visible by enabling edge?
2023-10-12 10:24:09 <ikke> community has linux-edge
2023-10-12 10:24:13 <Ermine> Virtual machine should see only one output
2023-10-12 10:24:22 <ikke> Not sure if it's the very latest though
2023-10-12 10:24:44 <Ermine> And virtualbox should output sound in the same place as everyone else
2023-10-12 10:24:49 <Sofia> Re my own crackling audio. Outside the vm, I have only alsa, with qemu as the only alsa client active or trying to do any audio. qemu is running with `-vga virtio -audiodev sdl,id=snd0 -device intel-hda -device hda-duplex,audiodev=snd0`. Alpine has sway, pipewire{,-pulse,-alsa,-jack}, and wireplumber; where sway is configured to launch pipewire with the Alpine-specific libexec
2023-10-12 10:24:51 <linfan> Linux alpis 6.1.57-0-virt #1-Alpine SMP PREEMPT_DYNAMIC Wed, 11 Oct 2023 07:01:12 +0000 x86_64 Linux
2023-10-12 10:24:52 <Sofia> helper. The crackling is consistent for all mpv test instances; on the host alpine, guest distrobox arch, and flatpak.
2023-10-12 10:26:07 <Ermine> Ah, linfan, did you install pipewire-alsa?
2023-10-12 10:26:08 <linfan> when i get it to work it is also crackling
2023-10-12 10:27:18 <Sofia> IIUC the -alsa et al redirect the respective clients, it isn't necessary to use the system-alsa under the hood itself. Or whatever pipewire uses to interact with the sound cards itself.
2023-10-12 10:27:23 <linfan> yes I did i installd all pipewire-*:s i could find
2023-10-12 10:28:05 <linfan> alsamixer maybe and alsactl store like in olden days
2023-10-12 10:28:54 <Sofia> frojnd: Re s6. https://ariadne.space/2021/03/25/lets-build-a-new-service-manager-for-alpine/
2023-10-12 10:29:03 <Ermine> So, in gnome settings there's audio section where you can test your sound, it doesn't work too?
2023-10-12 10:29:34 <linfan> it is inaudible as if deaf :-)
2023-10-12 10:29:44 <linfan> dumb*
2023-10-12 10:29:54 <Sofia> frojnd: https://skarnet.com/projects/service-manager.html
2023-10-12 10:29:57 <Ermine> Now this is getting mysterious
2023-10-12 10:31:29 <frojnd> Hm I've encountered an issue while trying to install zfs on nvme (first install). I'm following this guide: https://wiki.alpinelinux.org/wiki/Root_on_ZFS_with_native_encryption after I mount mount -t ext4 /dev/nvme0n1p1 /mnt/boot/ and rc-update add zfs-import sysinit; rc-update add zfs-mount sysinit
2023-10-12 10:31:35 <linfan> will rm the wireplumber and pipewire files from /etc
2023-10-12 10:32:38 <linfan> und run /usr/libexec/wireplumber-launch
2023-10-12 10:32:58 <Ermine> you definitely don't need this with gnome
2023-10-12 10:33:00 <frojnd> after running setup-disk mnt/ I get grub efi no such folder: https://termbin.com/2y8p
2023-10-12 10:33:10 <Ermine> gnome launched pipewire and wireplumber
2023-10-12 10:33:29 <Ermine> if it didn't, gnome wouldn't be able to identify any sound devices
2023-10-12 10:36:04 <linfan> 0. Audio/Sink    auto_null when running wpctl status
2023-10-12 10:38:39 <Ermine> Please send me full output of pw-dump and wpctl status (through pastebin
2023-10-12 10:38:42 <Ermine> )
2023-10-12 10:40:01 <Ermine> where pastebin means 'any pastebin service instead of pasting output here'
2023-10-12 10:40:02 <Sofia> (Mine with the same crackling issues.) http://ix.io/4IOK 
2023-10-12 10:40:20 <linfan> pw-dump not found
2023-10-12 10:40:53 <Sofia> linfan: Already looking for that :p -> https://pkgs.alpinelinux.org/packages?name=pw-dump&branch=edge&repo=&arch=&maintainer=
2023-10-12 10:40:58 <Sofia> (no results)
2023-10-12 10:40:58 <Ermine> it is in pipewire-tools package
2023-10-12 10:41:05 <Sofia> Er wrong search method
2023-10-12 10:41:26 <Ermine> pro tip: apk search cmd:cmdname will show you relevant packages
2023-10-12 10:42:05 <Sofia> http://ix.io/4IOL 
2023-10-12 10:42:12 <Sofia> Ermine: Thanks <3
2023-10-12 10:43:26 <frojnd> Hmm strange, after making it vfat it worked: https://gitlab.alpinelinux.org/alpine/alpine-conf/-/issues/10555
2023-10-12 10:44:06 <Sofia> frojnd: Oh ext4 /boot? EFI? vfat is kinda the thing.
2023-10-12 10:44:46 <Sofia> EFI needs to know the filesystem to read the efi.
2023-10-12 10:45:02 <Sofia> I mean read the efi-stub at least*
2023-10-12 10:48:06 <Ermine> it is needed for ESP
2023-10-12 10:59:42 <frojnd> What am I missing :s https://termbin.com/zgr5 still trying to install system on mnt now zfs is not loaded
2023-10-12 10:59:47 <frojnd> Odd since I modprobe it
2023-10-12 11:51:57 <frojnd> lsmod |grep zfs shows zfs
2023-10-12 11:55:56 <frojnd> Any ideas?
2023-10-12 11:57:07 <Sofia> frojnd: Idea: use btrfs if you really only wanted snapshots; and wait for bcachefs. :)
2023-10-12 11:57:14 <Sofia> Good luck either way.
2023-10-12 11:58:06 <frojnd> Well if I stay with btrfs I have to also implement luks2 :)
2023-10-12 12:06:16 <Sofia> Oh didn't know zfs supported it directly.
2023-10-12 12:06:16 <Sofia> Can't speak for its design, but a friend is happy with at least how bcachefs does this.
2023-10-12 12:06:16 <Sofia> Notably bcachefs allows the in-place encryption and authentication tag to be a single operation, an AEAD (or stronger, an MRAE).
2023-10-12 12:17:45 <frojnd> Hm maybe this is issue with grub-probe impelmentation
2023-10-12 12:17:58 <frojnd> I tried to manually reproduced the error message grub-probe gives
2023-10-12 12:18:37 <frojnd> Like this: grub-probe rpool/ROOT/alpine and I get: grub-probe: error: failed to get canonical path of `rpool/ROOT/alpine'. the message produced with setup-disk /mnt
2023-10-12 12:19:04 <frojnd> So I checkd rpoo/ROOT/alpine to be actually present: df -h shows: https://termbin.com/jvpt
2023-10-12 12:19:45 <quinq> https://openzfs.github.io/openzfs-docs/Getting%20Started/Alpine%20Linux/Root%20on%20ZFS.html says “GRUB installation will fail and will be reinstalled later. The error message about ZFS kernel module can be ignored.”
2023-10-12 12:20:52 <frojnd> quinq: Oh! Thank you
2023-10-12 12:21:17 <quinq> I assume you're following that guide as you're using this strange “rpool/ROOT/alpine” layout
2023-10-12 12:23:43 <frojnd> No actually that one: https://wiki.alpinelinux.org/wiki/Root_on_ZFS_with_native_encryption
2023-10-12 12:25:01 <frojnd> I guess it didn't work just executing: dd if=/usr/share/syslinux/mbr.bin of=/dev/nvme0n1 after that zfs kernel message
2023-10-12 12:25:16 <frojnd> since I'm in GRUB prompt
2023-10-12 12:25:51 <quinq> I suppose one's copied from the other (or from each-other)
2023-10-12 12:26:40 <minimal> frojnd: why would you write the syslinux BIOS MBR to a disk when you're using UEFI booting?
2023-10-12 12:27:39 <frojnd> minimal: are you telling me that part is redundant?
2023-10-12 12:28:27 <minimal> I saying that the /usr/share/syslinux/mbr.bin file is a Syslinux MBR that you've written to the 1st sector of the NVME device
2023-10-12 12:28:40 <minimal> the MBR is used for BIOS booting, not for UEFI booting
2023-10-12 12:28:47 <frojnd> Hm
2023-10-12 12:28:53 <frojnd> Didn't know that
2023-10-12 12:29:19 <minimal> and you're trying to install Grub so why would you use that Syslinux file even for BIOS booting?
2023-10-12 12:31:52 <frojnd> Didn't occur to me while following wiki
2023-10-12 12:31:53 <minimal> that Wiki article you are following is written for BIOS booting, it doesn't cover UEFI booting
2023-10-12 12:32:00 <frojnd> Ah
2023-10-12 12:32:19 <frojnd> I was just saying I was able to make it work on one of my older laptops x230
2023-10-12 12:32:24 <minimal> as it doesn't include creating a ESP partition/filesystem
2023-10-12 12:33:34 <quinq> x230 doesn't need UEFI for booting it supports the simpler MBR boot
2023-10-12 12:33:44 <frojnd> So for my use case I should rather look at: https://openzfs.github.io/openzfs-docs/Getting%20Started/Alpine%20Linux/Root%20on%20ZFS.html
2023-10-12 12:34:16 <quinq> What's your use-case?
2023-10-12 12:34:27 <frojnd> t14 with efi
2023-10-12 12:35:48 <quinq> Yeah, if your hardware limits you to EFI boot, you have to use EFI :/
2023-10-12 12:36:18 <quinq> Note that this openzfs guide makes you install *both* MBR and EFI boot xD
2023-10-12 12:36:27 <quinq> So I'd advice to use what you need
2023-10-12 12:36:39 <quinq> But that implies knowing a bit what you're doing
2023-10-12 12:39:49 <minimal> that 2nd article also contains some "hacks" on things like Grub files, which is messy
2023-10-12 12:47:01 <frojnd> Remind me again, why is swap used for?
2023-10-12 12:47:20 <frojnd> So can suspend to disk?
2023-10-12 12:47:37 <ikke> No, that's not the purpose
2023-10-12 12:47:55 <ikke> The kernel swaps memory to disk that it does not need right now (and can be used for more useful things)
2023-10-12 12:48:51 <frojnd> And 4GB of swap is enough?
2023-10-12 12:48:57 <quinq> Well, not the *main* purpose, but that's still one
2023-10-12 12:49:06 <quinq> Enough for what?
2023-10-12 12:49:11 <frojnd> For swap
2023-10-12 12:49:21 <frojnd> Set swap size in GB, set to 1 if you don’t want swap to take up too much space
2023-10-12 12:49:26 <quinq> Yes, 4GB of swap is enough for having swap.
2023-10-12 12:49:35 <quinq> 1kB of swap is also enough for having swap
2023-10-12 12:49:53 <quinq> Size is irrelevent, having swap is enough for having swap
2023-10-12 12:50:07 <frojnd> Ok.
2023-10-12 12:50:12 <quinq> But I suppose you actually want to do something with the swap, the goal isn't just to have swap, is it?
2023-10-12 12:50:17 <ikke> But you ofcourse want to have sufficient swap for the kernel
2023-10-12 12:50:28 <frojnd> Yeah, 4GB enough?
2023-10-12 12:50:36 <quinq> Enough for *what*
2023-10-12 12:50:40 <frojnd> For the kernel
2023-10-12 12:50:51 <quinq> Only for the kernel, that's way more than enough
2023-10-12 12:51:01 <ikke> An old rule of thum was 1/2 of available memory
2023-10-12 12:51:04 <ikke> thumb*
2023-10-12 12:51:08 <ikke> not sure if that still counts
2023-10-12 12:51:16 <quinq> Again, that depends for *what*
2023-10-12 12:51:36 <frojnd> I thought that rule was for suspending to "disk" swap ?
2023-10-12 12:51:47 <quinq> no
2023-10-12 12:51:56 <ikke> You don't actively use swap for something
2023-10-12 12:52:09 <ikke> the kernel will use swap space if available
2023-10-12 12:52:33 <frojnd> Ok. Don't care about space too much currently so gonna reserve 1/2 of memory
2023-10-12 12:54:39 <ikke> (the kernel will use swap space even if there is still plenty of memory available)
2023-10-12 12:54:50 <ikke> so it's not just spare ram
2023-10-12 12:55:30 <ikke> There is a kernel parameter to tune this, but for most cases, that's not necessary
2023-10-12 12:57:53 <minimal> frojnd: *If* you do want to suspend the machine to disk then swap will need to be bigger than the "memory image" it needs to save, which could as large as the size of RAM in the machine
2023-10-12 12:58:08 <ikke> https://unix.stackexchange.com/questions/2658/why-use-swap-when-there-is-more-than-enough-free-space-in-ram
2023-10-12 13:00:45 <minimal> now the data suspended to disk (swap) may or may not be compressed (I think the default is to compress it)
2023-10-12 13:04:29 <frojnd> Ah man
2023-10-12 13:04:54 <frojnd> I've missed about bios legacy
2023-10-12 13:05:04 <frojnd> Repeat the process
2023-10-12 13:06:43 <frojnd> set 5 bios_grub on \
2023-10-12 13:06:45 <frojnd> ANd 
2023-10-12 13:06:51 <frojnd> set 5 legacy_boot on
2023-10-12 13:06:59 <frojnd> This can be left out for my use case right?
2023-10-12 13:07:22 <frojnd> https://openzfs.github.io/openzfs-docs/Getting%20Started/Alpine%20Linux/Root%20on%20ZFS.html System Installation 1.
2023-10-12 13:08:47 <quinq> Then you don't need either  mkpart BIOS 1MiB 2MiB \
2023-10-12 13:09:12 <quinq> But note that the calculation of the other partitions is based on that (starting after 2MB)
2023-10-12 13:13:12 <Sofia> ikke: Huh, I've heard the rule of 2x your RAM as a suitable rule of thumb.
2023-10-12 13:13:46 <quinq> Do you understand why?
2023-10-12 13:25:02 <elly> I also remember hearing that but it was many years ago
2023-10-12 13:25:37 <elly> I don't know if there was ever a principled reason for it
2023-10-12 13:26:05 <ikke> hearing what? twice or half?
2023-10-12 13:26:17 <ikke> I was too inexperienced then to really understand it
2023-10-12 13:26:22 <ikke> but I do still remember it
2023-10-12 13:26:27 <elly> the twice thing
2023-10-12 13:26:35 <ikke> ah ok
2023-10-12 13:27:02 <ikke> I suppose it means you could do a complete swap for whatever reason
2023-10-12 13:27:11 <ikke> swap one set out, one set in
2023-10-12 13:27:13 <invoked> probably for hibernation but twice is too much
2023-10-12 13:27:17 <ikke> yeah
2023-10-12 13:28:13 <elly> disk vs ram ratios used to be a lot different back then though so the tradeoffs were different
2023-10-12 13:28:20 <ikke> yes
2023-10-12 13:28:31 <ikke> and disk speeds as well
2023-10-12 13:28:42 <Habbie> i remember the 'twice' recommendation from the 4MB-8MB RAM times :)
2023-10-12 13:29:54 <quinq> Recommendation is to adapt swap to your memory usage
2023-10-12 13:29:59 <quinq> There's no rule of the thumb
2023-10-12 13:30:19 <quinq> Memory usage being dependant on the hardware you have at hand
2023-10-12 13:31:32 <elly> I just had a look and on my current laptop, I have 32 GB of RAM, 512 GB of disk, and... 4 GB of swap
2023-10-12 13:31:36 <elly> how's that for a ratio :P
2023-10-12 13:31:47 <Habbie> probably fine
2023-10-12 13:31:54 <Habbie> if not, you'd know :)
2023-10-12 13:31:59 <quinq> That's a 12% ratio
2023-10-12 13:32:09 <Habbie> although it is possible your desktop would be faster with more swap
2023-10-12 13:32:18 <elly> empirically it works just fine
2023-10-12 13:32:46 <quinq> Habbie, I highly doubt it
2023-10-12 13:32:54 <quinq> But again, that depends on your usage
2023-10-12 13:32:57 <Habbie> yeah
2023-10-12 13:33:02 <Habbie> disk vs. nvme matters too
2023-10-12 13:33:10 <Habbie> nvme doesn't benefit that much from caching; spinning metal does
2023-10-12 13:33:36 <elly> I don't have any spinning metal left in my life thankfully
2023-10-12 13:34:03 <quinq> I have ~15TB of it
2023-10-12 13:34:14 <quinq> And it's great
2023-10-12 13:34:21 <Habbie> i have 100TB of metal but it's not currently spinning
2023-10-12 13:34:25 <elly> that is like, 10 times more total storage than I have :P
2023-10-12 13:34:29 <elly> the 15TB I mean
2023-10-12 13:34:50 <quinq> On the other hand, I don't have 32GB of RAM around
2023-10-12 13:34:55 <invoked> 
ACTION does a drive-by degauss on quinq

2023-10-12 13:35:17 <quinq> 
ACTION goes living naked into the woods

2023-10-12 13:35:26 <Habbie> ah, the 20-20-20 rule
2023-10-12 13:35:42 <Habbie> after 20 minutes of irc, stare at something in the distance for 20 seconds, then live naked in the woods for 20 years
2023-10-12 13:35:47 <elly> lol
2023-10-12 13:35:48 <ikke> haha
2023-10-12 13:35:56 <elly> we're drifting offtopic now but that is hilarious
2023-10-12 13:36:10 <Habbie> i stole it somewhere but it is very adaptable
2023-10-12 13:36:18 <quinq> haha Habbie, perfect
2023-10-12 13:36:52 <quinq> “For every 20 minutes a person looks at a screen, they should look at something 20 feet away for 20 seconds.”
2023-10-12 13:36:58 <Habbie> that's the original
2023-10-12 13:37:03 <quinq> Who the heck still uses feet for measuring things…
2023-10-12 13:37:09 <quinq> I have two.
2023-10-12 13:37:09 <Habbie> then somebody modified it to say '20 years in the woods'
2023-10-12 13:38:09 <quinq> So that's 20(min)-20(sec)-6(meter)
2023-10-12 13:39:00 <quinq> That made me sneeze twice, not sure that's really healthy!
2023-10-12 13:46:18 <minimal> I grew up learning a mix of metric and non-metric measurements so I tend to use different ones for different purposes
2023-10-12 13:46:22 <Habbie> it is 20 feet, not "at least 20 feet", so the sun doesn't qualify
2023-10-12 13:46:40 <quinq> Habbie ;)
2023-10-12 13:47:34 <quinq> minimal, I would understand using feet for American Football
2023-10-12 13:48:09 <minimal> quinq: I don't understand American Football ;-)
2023-10-12 13:48:25 <quinq> :D
2023-10-12 13:48:39 <minimal> I use a fix of metres and miles for length/distances, i.e. here speed is in MPH
2023-10-12 13:49:18 <minimal> I've no idea how long a "yard" is though (so a GPS saying turn left in X yards has me confused)
2023-10-12 13:49:21 <quinq> nautic miles?
2023-10-12 13:49:39 <quinq> I think that the length of a yard depends on how big the house is
2023-10-12 13:49:44 <Habbie> a yard is a meter (-10%)
2023-10-12 13:49:46 <Habbie> quinq, lol
2023-10-12 13:50:36 <quinq> Nice mnemonic
2023-10-12 13:51:22 <Habbie> a brit once taught me 'a liter of water is a pint and three quarters' which has also been useful
2023-10-12 13:52:04 <quinq> Hummm
2023-10-12 13:52:15 <minimal> no idea how much a gallon (whether Imperial or American) is, I think in terms of litres and also pints :-)
2023-10-12 13:52:29 <quinq> unit(1) tells me that's 2.11
2023-10-12 13:53:15 <quinq> Ah, right, again with the UK/US
2023-10-12 13:53:29 <Habbie> quarters, not quarts.. honestly i mostly use it to remember a pint is roughly 0.5L
2023-10-12 13:54:00 <quinq> ah haha, right
2023-10-12 13:54:19 <quinq> Yeah
2023-10-12 13:54:27 <quinq> Better ordering 2 pints than 1L of beer
2023-10-12 13:55:43 <minimal> a pint is 568ml from memory
2023-10-12 13:56:16 <Habbie> not on wolfram alpha
2023-10-12 13:56:23 <quinq> Because UK/US
2023-10-12 13:56:29 <Habbie> ah there we go https://www.wolframalpha.com/input?i=1.75+uk+pint+in+liter
2023-10-12 13:56:31 <Habbie> yes
2023-10-12 13:56:53 <minimal> remember the "fun" of ordering beer in Australia.....schooners and tinnies, didn't know what size those were
2023-10-12 13:57:21 <Habbie> i once witnessed a tourist at an airport, 7:30 am, order a pint, and the waitress goes "large or small"
2023-10-12 13:57:31 <quinq> haha
2023-10-12 13:57:38 <Habbie> now, in some parts of .nl, pint just means beer, it's not a size
2023-10-12 13:57:48 <Habbie> that wasn't this
2023-10-12 13:57:56 <minimal> maybe she was asking about the glass type? i.e. Weizen glass is tall and thin lol
2023-10-12 13:58:09 <Habbie> no, i think they had two sizes of beer; i suspect one of them was close to a pint
2023-10-12 13:58:12 <Habbie> but the tourist took it in stride
2023-10-12 13:58:23 <elly> over to the -offtopic channel, y'all :)
2023-10-12 13:58:32 <Habbie> yeah sorry
2023-10-12 13:59:07 <quinq> How many #alpine-linux is #alpine-linux-offtopic?
2023-10-12 14:35:47 <frojnd> re: quinq | But note that the calculation of the other partitions is based on that (starting after 2MB) I don't understand why? bios boot is last in that example.
2023-10-12 14:36:49 <quinq> That's just the command-line order (and its implicit ID)
2023-10-12 14:37:02 <quinq> The layout, as you can read, is actuall 1MB-2MB
2023-10-12 14:37:09 <quinq> s/actuall/&y/
2023-10-12 14:57:15 <Test111> I have gnome installed through setup-desktop and I'd like to replace pulseaudio with pipewire, but I can't uninstall pulseaudio due to a lot of gnome packages, including gdm itself.
2023-10-12 14:57:51 <Test111> In a video guide, they uninstall pulseaudio along with gnome-core-apps, but the latter is not found when I try to run the command.
2023-10-12 15:09:31 <quinq> Isn't pipewire supposed to be a drop-in replacement for pulseaudio?
2023-10-12 15:19:50 <Test111> Yes
2023-10-12 15:20:00 <Test111> I'm not sure what that implies however
2023-10-12 15:25:29 <Test111> I still don't know how to install pipewire
2023-10-12 15:25:33 <leitz> Newb question. How much of Alpine is written in Go?
2023-10-12 15:25:49 <frojnd> quinq: just don't know one thing. Are those commands that use ${MNT} meant to be using only ${MNT} that's being declared once or for each zpool create or partion function? Since MNT=$(mktemp -d) and it creates a temp dir?
2023-10-12 15:26:10 <frojnd> I was using it for each function command I was calling
2023-10-12 15:26:24 <frojnd> I mean I was reusing it and therefor creating multiple /tmp/dirs
2023-10-12 15:27:12 <ptrc> leitz: what do you mean by "Alpine"? package manager, init system, etc.?
2023-10-12 15:27:36 <ptrc> the stuff that runs on users' computers, infra?
2023-10-12 15:27:46 <ptrc> there's a lot of components that make up "Alpine"
2023-10-12 15:30:34 <leitz> ptrc, I don't know what I don't know, in this case. I saw a commit message about "community/*: rebuild with Go 1.20.10", and it looked interesting. I like both C and Go, and could use an excuse to play with them more.
2023-10-12 15:31:18 <ptrc> ah, this just means rebuilding packages that are made with Go
2023-10-12 15:31:50 <leitz> So the plan is to stick with C, for the most part?
2023-10-12 15:32:01 <ptrc> because it builds statically, so every stdlib update ( and especially one that affects security ) requires you to rebuild everything
2023-10-12 15:32:13 <ptrc> i don't think there's a "plan" per se
2023-10-12 15:32:18 <ikke> leitz: these are 3rd party packages
2023-10-12 15:32:22 <ptrc> ^
2023-10-12 15:32:42 <ikke> So whatever upstream decides to use
2023-10-12 15:32:46 <leitz> Cool. Thanks!
2023-10-12 15:32:50 <ptrc> as for Alpine-specific stuffs, apk-tools is written in C, a lot of things are shell scripts, but we do also use Go for some infra things
2023-10-12 15:33:20 <leitz> ptrc, Infra in the build parts, or in what goes out to the user?
2023-10-12 15:33:54 <ptrc> currently it's just small tools for package maintainers / developers, such as https://gitlab.alpinelinux.org/kdaudt/apkgquery
2023-10-12 15:33:56 <ptrc> and the gitlab bot
2023-10-12 15:36:40 <leitz> Cool.
2023-10-12 15:37:55 <frojnd> I guess it matters :s
2023-10-12 15:37:58 <frojnd> Back to square one\
2023-10-12 15:38:49 <somercet1> what is the Alpine meta package to pull in all -doc pkgs? ty
2023-10-12 15:38:54 <ikke> docs
2023-10-12 15:39:19 <somercet1> ty
2023-10-12 15:49:08 <somercet1> hmm, these packages are not covered by the docs pkg: alpine-doc gnupg-doc libsodium transmission-doc
2023-10-12 15:50:33 <somercet1> sorry, not libsodium, obvs
2023-10-12 15:52:14 <somercet1> ah, okay, which should I install? gnupg or gpg?
2023-10-12 15:55:27 <ikke> somercet1: it only installs -doc packages for packages that you have installed
2023-10-12 15:55:46 <somercet1> I'm looking through gnupg now, ty
2023-10-12 16:51:08 <quinq> 17:38:54 @ikke$ docs
2023-10-12 16:51:10 <quinq> oh!
2023-10-12 16:51:12 <quinq> TIL
2023-10-12 16:53:18 <elly> TIL also
2023-10-12 16:57:22 <tyler[m]123> Hello guys. I have just heard about ZRAM, is it worth it? I have very little memory on my PC (4GB).
2023-10-12 16:57:25 <frojnd> Hm
2023-10-12 16:57:48 <frojnd> Why when I try to paste into terminal sshed to alpine pasted command is "previous" But only this terminal
2023-10-12 16:57:48 <lopid> is it not enough?
2023-10-12 16:58:09 <frojnd> Maybe it's my alacritty to blame?
2023-10-12 16:58:24 <tyler[m]123> Not enough, some apps use too much memory
2023-10-12 16:58:52 <tyler[m]123> I use alpine as desktop OS and I need some Java apps for example
2023-10-12 16:59:20 <ikke> RIP
2023-10-12 16:59:23 <quinq> Ask the entity forcing you to use Java to buy you the required memory
2023-10-12 16:59:33 <frojnd> Man, it's the alacritty. Works fine in foot
2023-10-12 17:00:24 <tyler[m]123> Is it worth it though? ZRAM
2023-10-12 17:00:30 <tyler[m]123> Or it's too slow
2023-10-12 17:00:37 <elly> if you are desperately short on RAM and okay with the speed drop, it's worth trying
2023-10-12 17:00:43 <elly> whether it's too slow is up to you
2023-10-12 17:01:08 <tyler[m]123> Hmm OK
2023-10-12 17:01:15 <tyler[m]123> Do I have to disable my swap?
2023-10-12 17:01:20 <quinq> tyler[m]123, I configured zram on a computer I have with 4GB, and got a lot of system lockup when getting close to limits
2023-10-12 17:01:42 <tyler[m]123> uh oh
2023-10-12 17:01:47 <quinq> Last time I tried was a ~year ago, YMMV
2023-10-12 17:02:25 <quinq> But otherwise I don't recall having slowness from it
2023-10-12 17:02:59 <tyler[m]123> I think I will try, but only give it 1GB
2023-10-12 17:03:39 <tyler[m]123> Do I just sudo apk add zram-init and it's done?
2023-10-12 17:19:06 <frojnd> So I'm inside chroot and trying to isntall grub as per docs: https://openzfs.github.io/openzfs-docs/Getting%20Started/Alpine%20Linux/Root%20on%20ZFS.html step 6. And I get: grub-install: error: will not proceed with blocklists. https://termbin.com/dqdi Any ideas?
2023-10-12 17:20:42 <frojnd> ${DISK} is: /dev/disk/by-id/nvme-SKHynix_HFS256GDE9X081N_FYB4N018013704E31
2023-10-12 17:20:47 <frojnd> If that plays any role
2023-10-12 17:21:23 <consus> Alpine does not support /etc/network/interfaces.d or am I missing something?
2023-10-12 17:21:54 <ikke> consus: it does not by default, but you can add it
2023-10-12 17:22:08 <consus> Oh, how do I do that
2023-10-12 17:22:22 <consus> Can't find anything in man interfaces
2023-10-12 17:22:34 <consus> Ah
2023-10-12 17:22:35 <consus> Of course
2023-10-12 17:22:36 <consus> include
2023-10-12 17:22:38 <consus> Sorry :)
2023-10-12 17:22:49 <ikke> :)
2023-10-12 17:23:50 <minimal> frojnd: why are you trying to install Grub for BIOS on a UEFI system?
2023-10-12 17:24:06 <minimal> also why are you using a loop? Are you installing to more than a single disk?
2023-10-12 17:24:36 <frojnd> minimal: I just noticed I was trying to use the wrong i386-pc
2023-10-12 17:24:55 <frojnd> minimal: ur right don't need loop I only have one disk
2023-10-12 17:28:44 <frojnd> Wow step 8. For both legacy and EFI bootimg
2023-10-12 17:28:46 <frojnd> Zey
2023-10-12 17:29:16 <frojnd> Which find command is for efi
2023-10-12 17:30:46 <frojnd> Is it 1) find /boot/efi/ -maxdepth 1 -mindepth 1 -type d -print0 | xargs -t -0I '{}' cp -r '{}' "${espdir}" or 2) find "${espdir}" -maxdepth 1 -mindepth 1 -type d -print0 | xargs -t -0I '{}' sh -vxc "find /boot/efis/ -maxdepth 1 -mindepth 1 -type d -print0 | xargs -t -0I '[]' cp -r '{}' '[]'"
2023-10-12 17:31:25 <frojnd> I guess first command
2023-10-12 17:33:17 <frojnd> I have both /boot/efis and /boot/efi
2023-10-12 17:34:33 <minimal> frojnd: it seems like maybe you should read up on things like Grub and UEFI booting to understand how they work rather than just cutting and pasting commands from "random" articles
2023-10-12 17:37:14 <frojnd> I have a feeling I am gonna have to sadly :s What seemed like an easy task to do at first looks like I'm gonna have to understand it a bit more. Luckily it will get handy the second time I'm gonna have to setup on a similar hw
2023-10-12 17:40:45 <frojnd> Strangely after reboot I was presented with grub
2023-10-12 17:42:17 <asimovsh> hi
2023-10-12 17:42:18 <frojnd> But after loading I get mounting rpool/alpinelinux/root on /sysroot failed: No such file or directory Monting root failed. So I must have made a mistake somewhere during the chroot process... back to chroot
2023-10-12 17:42:22 <asimovsh> just question
2023-10-12 17:42:47 <asimovsh> someone have ever completed a build of hyprland in alpine?
2023-10-12 17:46:29 <ikke> not succesfully
2023-10-12 17:46:40 <ikke> it's a lesson in futility
2023-10-12 17:50:53 <quinq> The videos on their website make you feel like you're on a drunken boat
2023-10-12 18:06:59 <idkrn[m]> You mean the boat is drunk?
2023-10-12 18:07:57 <quinq> idkrn[m], they all are!
2023-10-13 08:33:55 <dcreator> I'm trying to use alpine linux on an embedded system with an raspberry pi. I have issues with the serial port /dev/ttyAMA2 not working properly. I discovered for some reason that a getty is configured use it in /etc/inittab. If I comment this out, it is just added again with next boot.
2023-10-13 08:34:55 <dcreator> How do I get rid of this.. and generally, how do I get rid of it randomly editing configuration files?
2023-10-13 08:39:28 <aron> dcreator: define "not working properly"
2023-10-13 08:39:58 <Habbie> my guess (but nothing more than a guess) is "getty listening on it makes it really hard to use the port to talk to some device"
2023-10-13 08:40:40 <dcreator> yes, that is correct.. but how do I get rid of this setting. If I comment it out, it just comes back at next reboot.
2023-10-13 08:41:11 <dcreator> I want to use this serial port to communicate with a microcontroller, I don't want that getty is using it
2023-10-13 08:42:15 <dcreator> I use lbu commit before rebooting.. my old commented out lines are there still. I tried to make the file read only, but that doesn't seem to prevent it for adding the lines again
2023-10-13 08:43:28 <aron> what kernel parameters do you have?
2023-10-13 08:44:18 <dcreator> The ones in cmdline.txt? modules=loop,squashfs,sd-mod,usb-storage quiet console=tty1 ?
2023-10-13 08:45:02 <quinq> More like /proc/cmdline to be sure, but should be equal yeah
2023-10-13 08:45:04 <aron> `cat /proc/cmdline` is the final one
2023-10-13 08:45:30 <dcreator> coherent_pool=1M 8250.nr_uarts=0 snd_bcm2835.enable_headphones=0  smsc95xx.macaddr=E4:5F:01:83:5C:DC vc_mem.mem_base=0x3ec00000 vc_mem.mem_size=0x40000000  modules=loop,squashfs,sd-mod,usb-storage quiet console=tty1
2023-10-13 08:46:28 <dcreator> In usercfg.txt I added dtoverlay=uart2 to enable the uart. maybe that triggers this behaviour?
2023-10-13 08:48:44 <aron> I don't think so
2023-10-13 08:48:55 <aron> must be a getty-related issue
2023-10-13 08:49:06 <aron> just to be sure, what happens, if you kill getty on AMA1?
2023-10-13 08:49:48 <quinq> Have you checked that /etc/inittab is part of the LBU commit?
2023-10-13 08:50:00 <quinq> aron, it should be respawned
2023-10-13 08:50:19 <aron> should be, yes
2023-10-13 08:50:49 <dcreator> quinq, I've commented out the lines multiple times, and it just appends them at the bottom repeatedly
2023-10-13 08:51:36 <dcreator> https://pastebin.com/qY1qqBYm
2023-10-13 08:51:49 <quinq> dcreator, that's not really what I asked though ;)
2023-10-13 08:52:32 <quinq> modify the file, then issue an lbu status
2023-10-13 08:53:28 <dcreator> quinq, since it remembers the changes I do every time I do a lbu commit before rebooting, I assumes it is part of the lbu commit
2023-10-13 08:53:40 <aron> dcreator: what   `ls /etc/init.d/agetty*`  yields?
2023-10-13 08:54:05 <quinq> Don't assume, check
2023-10-13 08:55:33 <dcreator> quinq, I get it mentioned as follows at lbu status: D etc/inittab
2023-10-13 08:56:30 <dcreator> ls: /etc/init.d/agetty*: No such file or directory
2023-10-13 08:56:51 <quinq> wut
2023-10-13 08:56:57 <quinq> You should have at least /etc/init.d/agetty
2023-10-13 08:58:43 <aron> also, "console=tty1" is also interesting to me
2023-10-13 08:58:48 <aron> why not tty0?
2023-10-13 08:58:56 <aron> for instance
2023-10-13 08:59:52 <dcreator> I followed the instructions at https://wiki.alpinelinux.org/wiki/Raspberry_Pi until lbu_commit -d; reboot right before WiFi drivers sections. Everything I need else works on the system except this serial port.
2023-10-13 09:01:48 <aron> >> To enable the UART console: enable_uart=1
2023-10-13 09:01:53 <aron> let's investigate this
2023-10-13 09:02:18 <dwfreed> aron: there can be some value in using console=tty1, like that you can switch to tty2 and not have to worry about kernel messages flooding your screen
2023-10-13 09:03:28 <aron> dwfreed: sure, my idea was that tty1 somehow turned into ttyAMA1 -- bc why not :P
2023-10-13 09:03:59 <dcreator> I'm trying to enable_uart=1 in usercfg.txt
2023-10-13 09:04:20 <dcreator> or should i add it as kernel cmdline ?
2023-10-13 09:04:46 <aron> usercfg.txt has this enable_uart option. the question is that: was it enabled or not?
2023-10-13 09:05:01 <dcreator> initially it was not enabled
2023-10-13 09:05:34 <aron> you enabled it, and ttyAMA1 got occupied?
2023-10-13 09:06:03 <aron> try with enable_uart=0
2023-10-13 09:06:23 <dcreator> ok.. I just enabled it and rebooted... will change it to =0 and reboot again
2023-10-13 09:07:52 <dcreator> Seems that setting enable_uart=1 didn't start getty with ttyAMA2 allthough it is appended to the inittab file
2023-10-13 09:09:21 <aron> if getty isn't started then just ignore inittab
2023-10-13 09:10:29 <dcreator> Seems that the user doesn't have access to the serialport as well. Is that something that can be configured with udev rules.d config files?
2023-10-13 09:10:30 <dcreator> crw--w----    1 root     root      204,  66 Oct 13 11:09 /dev/ttyAMA2
2023-10-13 09:12:12 <aron> yeah, it seems AMA is not handled by default
2023-10-13 09:13:01 <aron> for ttyUSB, I got this:
2023-10-13 09:13:01 <aron> crw-rw---- 1 root dialout 188, 2 Oct 13 10:55 /dev/ttyUSB2
2023-10-13 09:13:19 <aron> and add yourself to dialout group
2023-10-13 09:14:27 <aron> I can't help you with udev, tho
2023-10-13 09:14:47 <aron> at least your AMA is now free :)
2023-10-13 09:19:09 <dcreator> I need to get the correct permissions on ttyAMA2 still
2023-10-13 09:22:43 <aron> I think something like this could work:
2023-10-13 09:22:57 <aron> KERNEL=="ttyS[0-9]*", NAME="tts/%n", SYMLINK+="%k", GROUP="uucp", MODE="0666"
2023-10-13 09:22:58 <aron> just
2023-10-13 09:23:04 <aron> replace ttyS with ttyAMA
2023-10-13 09:23:42 <aron> oslt
2023-10-13 09:24:02 <aron> just search for it on the web
2023-10-13 09:24:13 <aron> (this is what I did)
2023-10-13 09:29:25 <dcreator> I've already added SUBSYSTEM=="tty", GROUP="dialout", MODE="0666" 
2023-10-13 09:29:29 <dcreator> but thanks anyway
2023-10-13 09:30:37 <aron> yw
2023-10-13 09:41:16 <minimal> dcreator: the getty is being added automatically by the initramfs' init during boot: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in?ref_type=heads#L107
2023-10-13 10:09:59 <frag> links -g is pretty neat tbh .. havent tried js on it tho, coooould maybe replace a normal browser(?)
2023-10-13 10:14:56 <dcreator> hmm.. it seems that 2535 root      0:00 /sbin/getty -L 0 ttyAMA2 vt100 getty is still running.. but the serial port is sending a lot of data so it crashes occasionally.. and the permissions isn't correct for ttyAMA2 as well
2023-10-13 10:15:47 <frag> how/when do you use these minimalist browser (lynx/links/dillo etc)?
2023-10-13 10:21:05 <minimal> dcreator: basically that "autodetect serial console" functionality seemed like a good idea, but it obviously detects serial devices not just consoles and there's (currently) no way to disable it
2023-10-13 10:22:15 <dcreator> seems like alpine linux isn't a good choice for this system then.. I wanted to use it in favour of raspbian since it is more lightweight and have this lbu commit and can easily run from ram
2023-10-13 10:22:56 <dcreator> unless I can edit this initramfs code and remove it
2023-10-13 10:23:35 <minimal> dcreator: It would make sense for a cmdline option to be implemented to disable this autodetection
2023-10-13 10:25:52 <dcreator> Still I think it generally is a bad idea to have the system randomly modify random configuration files. If I would need to have a getty on that serial port I could just add it myself
2023-10-13 10:26:01 <dcreator> Now I've spent several hours debugging this issue
2023-10-13 10:26:53 <minimal> I assume the reason it was added was for "headless" systems with only a serial console
2023-10-13 10:27:22 <dcreator> If I manually change the permissions on ttyAMA2 something is setting it back to wrong settings some few seconds later.. suspect it is getty
2023-10-13 10:27:37 <minimal> yes, you need to stop the getty
2023-10-13 10:28:32 <dcreator> I could replace it by a program that does nothing except sleep or something.. but than it might be difficult to log on later if needed
2023-10-13 10:29:40 <minimal> dcreator: raise a mkinitfs issue for the serial detection to have cmdline option to disable?
2023-10-13 10:30:44 <Inga> Is anybody else using LXD here? Mine worked just fine for a year, but then everything broke after one of the latest updates (IIRC I updated my Alpine Edge installation around end of August or beginning of September and it worked after that; it no longer works after I updated again a couple of days ago). Even the simplest commands, like `doas lxc launch images:alpine/edge simple-lxd-test`, fail now with weird errors; am I th
2023-10-13 10:31:32 <Inga> * am I the only one experiencing this?
2023-10-13 10:31:51 <Inga> (Not sure what is the best way to send LXD logs to IRC)
2023-10-13 10:34:20 <minimal> Inga: tpaste.us, or pastebin.com, or similar
2023-10-13 10:35:30 <minimal> dcreator: in general things are unlikely to get fixed if no Issues are opened
2023-10-13 10:36:25 <Inga> uploaded logs to https://pastebin.com/LZz9fZTS
2023-10-13 10:37:03 <Inga> I'm just not sure if this is a packaging issue or LXD issue or something else. So not sure if this belongs to aports issue tracker?
2023-10-13 10:37:49 <Inga> (or maybe there is something with my system only, and it works for everybody else)
2023-10-13 10:38:36 <minimal> Inga: that looks similar to #15290
2023-10-13 10:41:03 <Inga> minimal: thank you, it surely does. I should have checked the issue tracker first, instead of irc logs...
2023-10-13 10:41:54 <minimal> I noticed that openrc was updated in the past few days and its changelog mentioned a change in cgroup mode to unified
2023-10-13 10:45:44 <Inga> after `doas umount /sys/fs/cgroup/systemd`, the containers do start. I'm a total linux newbie though and have no idea what this umount actually does, or whether it has any unintended consequences.
2023-10-13 10:46:04 <Inga> will open an issue on aports now, thank you for the support @minimal
2023-10-13 10:46:28 <minimal> Inga: which Alpine version are you using?
2023-10-13 10:47:40 <minimal> that previous fix was only applied to LXD in Alpine Edge, not to Alpine 3.18.x
2023-10-13 10:55:08 <Inga> minimal: I'm using edge
2023-10-13 10:56:13 <Inga> minimal: the bug was fixed in 5.0.2-r16, while I currently have 5.0.2-r17 installed
2023-10-13 10:56:47 <minimal> right, which contains that "fix"
2023-10-13 10:57:52 <minimal> so do you have the /sys/fs/cgroup/ directory?
2023-10-13 10:58:41 <Inga> ...yet it still doesn't work for me, not unless I do `doas umount /sys/fs/cgroup/systemd` prior to trying to launch / start. So apparently the issue was not fixed after all, or only fixed for some configurations, or maybe it was reintroduced by some other package update. So I'm not sure whether I should create a new issue, or post a comment to 15290 (which is now closed)?
2023-10-13 10:59:23 <Inga> yes I do have /sys/fs/cgroup, with a lot of stuff inside of it
2023-10-13 10:59:39 <minimal> Inga: I'm not sure why you have /sys/fs/cgroup/systemd mounted in the first place
2023-10-13 11:00:25 <minimal> the fix was to stop it trying to mount that (when in "unified" mode)
2023-10-13 11:01:26 <minimal> what's the value of rc_cgroup_mode in your /etc/rc.conf file?
2023-10-13 11:01:38 <Inga> (according to my year-old install log, I enabled cgroups service (`rc-update add cgroups`), apparently lxd wouldn't work without that. not sure if that's relevant here)
2023-10-13 11:02:55 <Inga> the only line with `rc_cgroup_mode` in my `/etc/rc.conf` is commented out `#rc_cgroup_mode="hybrid"`
2023-10-13 11:03:10 <dcreator> minimal, I wrote an issue now. I know of several embedded systems where this UART is used and that I would prefer to use Alpine Linux on.
2023-10-13 11:03:49 <minimal> Inga: with the current openrc on Edge that value (although commented out) ia "unified", not "hybrid"
2023-10-13 11:05:14 <Inga> hmm, my openrc version is 0.51-r0, same as on edge
2023-10-13 11:06:50 <Inga> ah, turns out I have edited my rc.conf long time ago, in order to enable rc_logger.
2023-10-13 11:07:36 <Inga> but the only differences between /etc/rc.conf and /etc/rc.conf.apk-new are enabled rc_logger, and "unified" instead of "hybrid" in the commented out line. not sure that this should affect anything?
2023-10-13 11:07:42 <linfan> Sooo, got sound working in gnome by installing 3.17 out of the box without any pipewire or wireplumber, but cannot get musl-locales to work, hehe
2023-10-13 11:10:07 <linfan> which leads me to think 3.18 gnome setup-desktop or new kernel is not correctly modified
2023-10-13 11:11:45 <linfan> or maybe firefox, for now uninstalled firefox and flatpaked it instead
2023-10-13 11:13:19 <minimal> Inga: with init.d/cgroups run here I don't see a /sys/fs/cgroup/systemd mount so don't know why it was there in the first place, the current /etc/init.d/cgroups isn't doing it and the current /etc/init.d/lxd isn't doing it
2023-10-13 11:15:52 <Inga> I don't have `unified` in my `/sys/fs/cgroup` though. Not sure if that's relevant
2023-10-13 11:17:36 <Inga> hmm, tried ripgrepping systemd in my /etc, and found this one in `conf.d/lxc`: `systemd_container=yes`.
2023-10-13 11:18:48 <Inga> apparently I did edit conf.d/lxc at some point in the distant past. this line is the only difference between lxc and lxc.apk-new (lxc.apk-new has `#systemd_container=no` instead)
2023-10-13 11:19:26 <Inga> maybe I did that in order to get docker to work inside containers, no idea now
2023-10-13 11:19:46 <minimal> try changing that to "no"?
2023-10-13 11:19:51 <Inga> will try that now
2023-10-13 11:20:19 <Inga> (I wish `apk upgrade` produced some warnings about not updating some files because their local copy was changed)
2023-10-13 11:20:44 <Inga> (not the first time that I get bitten by inconsistent state produced by apk upgrade keeping old files with my tiny changes in place)
2023-10-13 11:21:23 <minimal> it just creates new files with a suffix (forget what it is)
2023-10-13 11:21:41 <Inga> yeah, `.apk-new`
2023-10-13 11:21:47 <Inga> the problem is that it does that silently
2023-10-13 11:22:25 <Inga> I wish `apk update` would tell me "I didn't update these 7 files because they had local changes"
2023-10-13 11:22:44 <Inga> instead of me figuring it out much later when stuff breaks and it takes so much time to diagnose it
2023-10-13 11:22:59 <Inga> *apk upgrade
2023-10-13 11:30:59 <frag> is webkit (from repo) working well now on apline?
2023-10-13 11:34:17 <Test111> I have gnome installed through setup-desktop and I'd like to replace pulseaudio with pipewire, but I can't uninstall pulseaudio due to a lot of gnome packages, including gdm itself. In a video guide, they uninstall pulseaudio along with gnome-core-apps, but the latter is not found when I try to run the command.
2023-10-13 11:34:23 <Test111> What do I do
2023-10-13 11:34:57 <Inga> Replaced /etc/conf.d/lxc with /etc/conf.d/lxc.apk-new (removing `systemd_container=yes` from it), rebooted, and now everything works as before
2023-10-13 11:35:07 <Inga> Thank you @minimal for your help
2023-10-13 11:35:37 <Inga> I wish there was an easy way for me to see that my system is in inconsistent state after upgrade, and that this might cause issues...
2023-10-13 11:36:00 <frojnd> Hm if for example after installation wants to change hostname and enters: setup-hostname and enters a new hostname. What else does it needs to be done so new hostname is taken in effect?
2023-10-13 11:36:02 <Ermine> Hello, have bb crond command line syntax changed recently?
2023-10-13 11:36:29 <ikke> Inga: run update-conf after an upgrade 
2023-10-13 11:38:26 <Inga> yeah, just found about it :)
2023-10-13 11:38:29 <Inga> thank you
2023-10-13 12:04:14 <singham> How do I set permanent nameservers?
2023-10-13 12:04:33 <singham> Setting all my queries to be resolved by a certain nameserver
2023-10-13 12:04:52 <ikke> Are you using dhcp?
2023-10-13 12:05:32 <singham> Yes, think so. The network is autoconfigured
2023-10-13 12:10:01 <minimal> singham: so switch to static network configuration rather than DHCP? With DHCP you're deciding to use what the DHCP tells you (you  may be able to configure your DHCP client to ignore any announced nameservers).
2023-10-13 12:11:51 <singham> How do I do that?
2023-10-13 12:11:59 <singham> I use pppoe 
2023-10-13 12:12:08 <singham> And I have a static ip
2023-10-13 12:13:28 <minimal> so you're not using DHCP then, so you can't reconfigure the DHCP client (that you're not using) to ignore announced nameservers
2023-10-13 12:15:36 <singham> Sothe conf is iface eth0 inet dhcp
2023-10-13 12:16:35 <singham> Pretty sure dhcp means dhcp client
2023-10-13 12:17:15 <minimal> singham: but you just said you use pppoe
2023-10-13 12:17:58 <minimal> and you also said you have a static IP address
2023-10-13 12:18:02 <singham> Yes but pppoe works over eth0 right
2023-10-13 12:18:15 <minimal> so that's 3 different conflicting answers
2023-10-13 12:18:20 <singham> Yes, public static ip
2023-10-13 12:18:42 <singham> Naah. the port by default is eth0
2023-10-13 12:18:59 <singham> Not port but whatever
2023-10-13 12:23:34 <singham> What are other options for auto eth0
2023-10-13 12:29:44 <singham> Done
2023-10-13 12:29:51 <singham> minimal: Thanks
2023-10-13 12:29:55 <singham> :)
2023-10-13 12:54:29 <frag> xterm and w3m from repo cant show images?
2023-10-13 12:55:31 <lopid> neither are image viewers?
2023-10-13 12:56:29 <frag> today they are :]
2023-10-13 12:56:48 <singham> Is there a mail user agent which does this
2023-10-13 12:57:18 <singham> Keeps a gpg keyring. Checks if a message is encrypted with public key of it
2023-10-13 12:57:36 <singham> If encrypted, it labels secure, if not, unsecure
2023-10-13 12:57:49 <lopid> labels?
2023-10-13 12:58:16 <lopid> iirc mutt has a flag that tells you if a message is pgp encrypted
2023-10-13 12:58:18 <singham> Yes, so, like a letter at end, say like in mutt, u/s
2023-10-13 12:58:40 <singham> Does it automatically decrypt with the key?
2023-10-13 12:58:57 <lopid> if you open it
2023-10-13 12:59:08 <singham> Yes
2023-10-13 12:59:22 <singham> So, heuvxuwubxioe
2023-10-13 12:59:25 <singham> open it
2023-10-13 12:59:29 <singham> hello lopid
2023-10-13 12:59:37 <lopid> hello
2023-10-13 13:00:07 <lopid> it'll remember your key if you've opened already and/or are using an agent
2023-10-13 13:00:41 <lopid> as would any, that being the point of an agent
2023-10-13 13:01:55 <singham> hmm... gpg and pgp are different right?
2023-10-13 13:02:10 <singham> I saw this superb app, deltachat
2023-10-13 13:02:18 <singham> It allows chat over email
2023-10-13 13:02:41 <lopid> gpg is an implementation of pgp
2023-10-13 13:14:08 <Test111> have gnome installed through setup-desktop and I'd like to replace pulseaudio with pipewire, but I can't uninstall pulseaudio due to a lot of gnome packages, including gdm itself. In a video guide, they uninstall pulseaudio along with gnome-core-apps, but the latter is not found when I try to run the command.
2023-10-13 13:14:19 <Test111> Sorry I kind of had to repeat myself
2023-10-13 15:11:15 <frag> how do you execute py3-telegram ? i supposed to be a client for telegram iiuc...
2023-10-13 15:18:59 <ikke> that package is just a library
2023-10-13 15:31:47 <frag> so there is no terminal client for telegram in repo?
2023-10-13 15:31:58 <frag> think i saw one in there in the past, might be wrong ..
2023-10-13 15:41:59 <quinq> Is there any actually?
2023-10-13 15:42:09 <quinq> I mean outside Alpine packages
2023-10-13 15:42:57 <frag> i remember i used some pythong thing called 'tg' once ..
2023-10-13 15:43:16 <frag> it had a github url too, but dont think i remember it
2023-10-13 15:43:27 <frag> maybe peter something
2023-10-13 15:43:50 <quinq> Yeah, just saw https://github.com/vysheng/tg
2023-10-13 15:44:00 <Ermine> tg is from maths, not python :]
2023-10-13 15:44:31 <quinq> In C/LUA!
2023-10-13 15:47:46 <frag> oh cool
2023-10-13 15:55:19 <frag> huh, there is py3-tg in repo
2023-10-13 15:55:23 <frag> paul
2023-10-13 16:07:16 <frag> seems to work
2023-10-13 16:16:52 <quinq> py3-tg?
2023-10-13 16:30:27 <frag> quinq: yeah
2023-10-13 16:32:53 <quinq> noice
2023-10-13 16:49:26 <frag> it seems to store credentials in .cache :S
2023-10-13 17:06:34 <quinq> in plaintext?
2023-10-13 17:11:57 <frag> no
2023-10-13 17:36:29 <Cyanat> Hey! Is there a repository from wich I can install php 7.4 on alpine 3.16?
2023-10-13 17:39:14 <minimal> Cyanat: unlikely
2023-10-13 17:39:32 <fluix> is getting it from the 3.15 repos even worth a try?
2023-10-13 17:39:51 <ikke> large chance that you'll get dependency issues
2023-10-13 17:39:59 <fluix> of course
2023-10-13 17:40:00 <minimal> fluix: why? you shouldn't mix different repo versions
2023-10-13 17:40:13 <ikke> I'd use docker
2023-10-13 17:40:36 <fluix> that's why I asked if it was "worth a try", depending on what your requirements are
2023-10-13 17:42:43 <Cyanat> ikke: but with docker, I will have php 7.4 inside a container, right?
2023-10-13 17:42:53 <ikke> ofcourse
2023-10-13 17:46:20 <Cyanat> mmh, ok, thanks for your feedbacks. Seems I have to find another solution to composer install my old project ;)
2023-10-13 18:14:18 <frojnd> I tried to setup wireless network after alpine installation with setup-interfaces and I've set wlan0 but I get: ERROR: Unable to lock database: Permission denied Do I have to put user into some groups?
2023-10-13 18:16:03 <frojnd> I've also installed networkmanager-cli and after firing nmcli dev wifi list I get: rror: Could not create NMClient object: Could not connect: No such file or directory.
2023-10-13 18:23:47 <frojnd> iwctl finds network though
2023-10-13 18:43:00 <linfan> I have now installed alpine on vmware player and here pipewire works. On my previous vox install it seems it does not find the correct snd module
2023-10-13 18:44:04 <linfan> vox=vbox*
2023-10-13 18:58:34 <frag> uninstalled my normals browsers qutebrowser/firefox/vieb, mpv couldnt play webm from the web.. i guess its because of mesa-dri-gallium or something ..
2023-10-14 01:24:37 <lew-adv-sol> Hey, new here and with alpine. I have an alpine VM setup and am attempting to install something with apk but i get unable to lock database permission denied and the usual second one of failed to open apk database permission denied. I have attempted to add the to the admin group but could not get any confirmation or a successful install. 
2023-10-14 01:25:14 <lew-adv-sol> odoo is the user, and odoo+dependencies is the installation
2023-10-14 01:34:56 <lew-adv-sol> Apologies if this is not the right spot for this.  
2023-10-14 01:50:42 <fluix> you need something to elevate your privileges, be that logging in as root or using doas or sudo
2023-10-14 01:53:20 <idkrn[m]> <minimal> "fluix: why? you shouldn't mix..." <- Is there a way to get just the up to date kernel?
2023-10-14 01:58:17 <lew-adv-sol> The install says install under the odoo user. doas and sudo failed. "no such package"
2023-10-14 01:59:02 <lew-adv-sol> I attempted to use doas, i added to wheel group and stil had no success
2023-10-14 02:00:20 <lew-adv-sol> https://wiki.alpinelinux.org/wiki/Odoo is the guide used and it gives that error after attempting install the packages. No mention of adding additional privileges. Did I miss something in the install?
2023-10-14 02:02:15 <lew-adv-sol> I did apk update, added user, attempted to install sudo and doas, but can't get passed this permission issue.
2023-10-14 02:05:55 <lew-adv-sol> ....I sorted it out. back to the basics. Thanks
2023-10-14 05:35:20 <singham> What is inetd?
2023-10-14 05:35:28 <singham> Does alpine have it?
2023-10-14 10:01:46 <sewn> why is xinit part of xorg-server?
2023-10-14 10:03:17 <ikke> as opposed to..?
2023-10-14 10:04:50 <ikke> xinit is it's own package btwe
2023-10-14 10:13:53 <sewn> ikke: nothing
2023-10-14 10:14:03 <sewn> xinit or startx isnt required to start X, sx can be used
2023-10-14 10:14:13 <sewn> i cant really force remove a package from apk so im stuck with xinit
2023-10-14 10:20:19 <ikke> you can try apk add -t xinit sx
2023-10-14 10:26:46 <quinq> TIL signal 0
2023-10-14 10:27:07 <dwfreed> one way to find out if a pid still exists
2023-10-14 10:27:09 <quinq> Though how it's used there doesn't really avoid race conditions
2023-10-14 10:27:15 <quinq> Yep, dwfreed
2023-10-14 10:27:39 <dwfreed> but if the pid is reused, it's not very useful
2023-10-14 10:27:54 <quinq> Or even kill after the command succeeds :)
2023-10-14 10:28:07 <quinq> *killed (or exited)
2023-10-14 10:29:32 <dwfreed> which is why pidfd is a thing
2023-10-14 11:10:59 <|cos|> What's the preferred way to detect from a shell script whether running on alpine (or preferably musl in general)?
2023-10-14 11:11:53 <|cos|> I found https://stackoverflow.com/q/58177815/how-to-actually-detect-musl-libc which pipes ls through ldd and stuff. Guess that works, but is there some cleaner way?
2023-10-14 11:12:51 <ikke> Depends on what you want to achieve, generally it's better to detect specific features / issues rather than assuming alpine / musl means something is supported or nto
2023-10-14 11:12:56 <ikke> not*
2023-10-14 11:15:24 <|cos|> Real issue: I'm attempting to put my alpine laptop to actual use, sharing my config files to be the same i use with libc linux, illumos and freebsd.
2023-10-14 11:16:07 <|cos|> The first thing which hits me is that locale support in musl seems a bit limited compared to those others, right?
2023-10-14 11:16:38 <|cos|> I didn't go down the rabbit hole, but settled with that LC_ALL=en_US.UTF-8 works on my alpine system.
2023-10-14 11:17:34 <|cos|> Fine-grained locale control of paper-sizes, currencies and what not would be nice to have, but I can live without it.
2023-10-14 11:19:01 <ikke> /etc/os-release is a way to detect what distribution is running
2023-10-14 11:19:32 <|cos|> So, I'm after something pragmatic. `uname` is never concerned about libc, `lsb_release` appears unavailable, parsing ldd output seems pone to future breakage. The /etc/os-release seems like a great way. Thanks!
2023-10-14 13:00:19 <p00f> how do i skip bootloader installation during setup-alpine?
2023-10-14 13:02:29 <p00f> i followed https://wiki.alpinelinux.org/wiki/Dualbooting but t still tries to install grub
2023-10-14 13:02:31 <minimal> p00f: BOOTLOADER=none
2023-10-14 13:02:38 <p00f> minimal: thanks
2023-10-14 13:02:53 <minimal> p00f: that wiki article is not very good
2023-10-14 13:03:07 <minimal> i.e. it doesn't cover UEFI booting scenario
2023-10-14 13:03:27 <minimal> it also isn't very clear in general
2023-10-14 13:04:29 <minimal> p00f: what are you trying to dualboot?
2023-10-14 13:50:45 <p00f> minimal: existing archlinux
2023-10-14 13:51:11 <p00f> btw it still tries to install grub
2023-10-14 13:51:25 <p00f> and it doesn't recognize my existing ESP
2023-10-14 13:51:42 <p00f> so it puts /boot in the same partition
2023-10-14 13:52:28 <p00f> and because of this, my existing systemd-boot can't launch it
2023-10-14 13:57:49 <p00f> for the latter i can try mounting the esp at /mnt/boot before running setup-alpine
2023-10-14 13:58:02 <p00f> just please don't install grub 🤮
2023-10-14 13:58:12 <minimal> p00f: did you run "BOOTLOADER=none setup-disk -m /mnt" ?
2023-10-14 13:58:17 <p00f> oh
2023-10-14 13:58:28 <p00f> i ran BOOTLOADER=none setup-alpine
2023-10-14 13:58:56 <minimal> I thought you were following that Wiki article
2023-10-14 13:59:01 <p00f> yeah
2023-10-14 13:59:30 <minimal> so then why would Grub be installed?
2023-10-14 13:59:53 <minimal> they tell you to answer "none" when you run setup-alpine and it asked which disk to use...
2023-10-14 14:01:23 <p00f> it didn't ask which disk to use, it said "couldn't find xx, use dm-0" or something
2023-10-14 14:01:31 <p00f> and i said no
2023-10-14 14:01:40 <p00f> iirc dm-0 is the tpm?
2023-10-14 14:01:52 <minimal> so then why would it try and install grub?
2023-10-14 14:01:58 <p00f> idk
2023-10-14 14:02:07 <minimal> no, dm-0 is a device mapper device
2023-10-14 14:02:23 <p00f> oh
2023-10-14 14:02:31 <minimal> well setup-alpine cannot BOTH not find a disk to use AND try and install Grub on a disk it cannot find...
2023-10-14 14:03:10 <minimal> are you trying to install with encryption and/or LVM?
2023-10-14 14:03:14 <p00f> no
2023-10-14 14:03:29 <p00f> i'll try installing again, noting down the messages this time
2023-10-14 16:21:14 <neutrocyte> Hello there, is it possible to make permanent changes when booting from usb stick/iso?
2023-10-14 16:21:50 <ikke> neutrocyte: with lbu you can commit changes that get restored after reboot
2023-10-14 16:22:57 <neutrocyte> So i could switch out the bootloader and then commit those changes with lbu? 
2023-10-14 16:24:08 <ikke> no, lbu only covers the changes in the tmpfs
2023-10-14 16:24:25 <ikke> the bootloader is part of the boot medium
2023-10-14 16:25:20 <neutrocyte> So i would have to make my own iso with an other bootloader
2023-10-14 16:27:09 <minimal> neutrocyte: if it actually is a ISO (ISO-9660) then typically nothing can be changed on it as it is read-only
2023-10-14 16:29:00 <minimal> that's different however from the situation if you actually installed Alpine on a USB stick using setup-bootable ("diskless" mode)
2023-10-14 16:29:18 <neutrocyte> minimal: thanks I guess that would have answered that question on the spot, still climbing the learning curve on asking the right question
2023-10-14 16:30:47 <minimal> neutrocyte: why do you want to change to bootloader? and to which other bootloader?
2023-10-14 16:33:41 <neutrocyte> I have a librebooted x200 and wanted to do fulk disk encryption and have some problemes getting alpine to boot so wanted to see if it was just bootloader related
2023-10-14 16:35:34 <minimal> neutrocyte: but the Alpine ISO is not encrypted, and also setup-bootable doesn't create an encrypted disk/USB stick either
2023-10-14 16:35:57 <minimal> typically for encryption you do a Sys mode install, i..e run-from-disk, not run-from-ram
2023-10-14 16:43:56 <neutrocyte> I know the alpine iso is not encrypted, i know it will require lots of manual intervention to make a alpine install run from the disk. Libreboot on the x200 either runs grub or seabios as the first payload and i had some problemes with grub parsing the syslinux config
2023-10-14 16:46:45 <minimal> you said you wanted to do full disk encryption - so that would require you to install Alpine in some fashion but your earlier questions seems to be about the install "media", not about an installed Alpine and so that's why I'm confused about what you are actually trying to achieve
2023-10-14 16:57:32 <neutrocyte> minimal: I am sorry for the confusing wording, my problem is that is i basically have a grub payload as the"only" means of loading the iso from the usb. Since grub seams to have a probleme loading the syslinux config from the alpine iso, I wanted to change the bootloader on the iso / create an iso that has grub as default bootloader o just load the grub config 
2023-10-14 17:04:47 <minimal> nuetrocyte: from memory the ISO has *isolinux* (part of Syslinux), not syslinux, configured as this is booting from an ISO
2023-10-14 17:07:04 <minimal> setting up Grub for booting on an ISO is different than setting up Grub for booting of a disk/usb stick etc
2023-10-14 17:11:12 <android2> Can you help our opensource project. We want the unreal map file types.
2023-10-14 17:11:15 <android2> http://sf.net/p/chaosesqueanthology/tickets/2/
2023-10-14 17:11:20 <android2> .t3d and .unr file formats (C project, GPL licensed)
2023-10-14 17:11:30 <ikke> android2: please don't spam channels with your question
2023-10-14 17:11:47 <android2> ikke: code what we want
2023-10-14 17:11:53 <android2> ikke: why would we obey you?
2023-10-14 17:11:56 <android2> tell me
2023-10-14 17:12:31 <qaqland> 👀
2023-10-14 17:13:50 <ikke> asking the same question in many channels is what is considered spam, and spam is not welcome
2023-10-14 17:17:49 <minimal> ikke: now in devel channel...
2023-10-14 17:20:32 <aron> what an attitude...
2023-10-14 17:21:56 <aron> hm, maybe s/he is a droid afterall
2023-10-14 18:12:19 <linfan> I would like to change the Default Sample Spcification for PulseAudio (on PipeWire 0.3.79). Where is that .conf or is it done with pactl commands?
2023-10-14 20:20:35 <WhyNotHugo> Firefox keeps freezing after a system update today.
2023-10-14 20:20:39 <WhyNotHugo> I see [24360] Sandbox: seccomp sandbox violation: pid 24360, tid 24360, syscall 25, args 140721801326592 4096 8192 0 0 0.
2023-10-14 20:20:47 <WhyNotHugo> ... in its output.
2023-10-14 20:45:40 <WhyNotHugo> Package depending on libdav1d.so.6 seem to be failing to build: https://gitlab.alpinelinux.org/WhyNotHugo/aports/-/jobs/1144380#L71
2023-10-14 20:45:48 <WhyNotHugo> I have the same issue locally.
2023-10-14 20:46:01 <WhyNotHugo> I think something in the dependency chain needs a rebuild.
2023-10-14 20:46:10 <ikke> ark is keeping the builders hostage
2023-10-14 20:46:34 <WhyNotHugo> ark? What's that?
2023-10-14 20:46:39 <ikke> kde archiving tool
2023-10-14 20:47:32 <WhyNotHugo> Is the status of builders publicly queriable somewhere?
2023-10-14 20:48:18 <elly> I'll do the chromium 118 upgrade today, I figured out earlier what patches are needed
2023-10-14 20:49:24 <ikke> WhyNotHugo: https://build.alpinelinux.org/
2023-10-14 20:49:32 <ikke> WhyNotHugo: also, there is the #alpine-commits channel
2023-10-14 21:07:48 <WhyNotHugo> How do builders avoid concurrently editing the package index?
2023-10-14 21:08:34 <ikke> there is only one builder per arch
2023-10-14 21:08:42 <ikke> so no concurrency
2023-10-14 21:22:21 <WhyNotHugo> Oh, the three logs shown for the x86_64 builder are the three most recent, not the three current.
2023-10-14 21:22:41 <WhyNotHugo> Does this mean that you need one very large build server and can't use lots of little ones?
2023-10-14 21:22:54 <ikke> WhyNotHugo: yes
2023-10-14 21:23:34 <WhyNotHugo> So support for multiple build servers might enable using many small devices for some arches (e.g.: riscv)?
2023-10-14 21:23:34 <ikke> It's a simpler architecture
2023-10-14 21:24:03 <ikke> Doing things concurrently will complicate things a lot
2023-10-14 21:26:07 <WhyNotHugo> "a lot"?
2023-10-14 21:26:13 <ikke> yes
2023-10-15 05:56:32 <Test111> I installed apache2 but a2enmod doesn't work. Why?
2023-10-15 05:57:23 <Hello71> it's for debian
2023-10-15 06:55:00 <Test111> A program required libresolv.so.2, which I got by installing gcompat. Now it asks for libcrypt.so.1 but it's deprecated. Arch has libxcrypt-compat for this, but it's unavailable on Alpine. What can I do here?
2023-10-15 06:56:21 <Test111> Nevermind, libc6-compat has it
2023-10-15 06:58:16 <Test111> Now I get a different error: `isnan: symbol not found`.
2023-10-15 06:58:34 <Test111> I thought gcompat would solve all glibc compatibility issues
2023-10-15 06:59:19 <quinq> Oh no
2023-10-15 06:59:34 <quinq> To be compatible with glibc you need to be (the correct version of) glibc
2023-10-15 07:00:30 <Test111> I see
2023-10-15 07:00:54 <Test111> Is there anything I can do to make this work? `isnan: symbol not found`
2023-10-15 07:04:33 <quinq> Not really :/
2023-10-15 07:05:55 <quinq> isnan is supposed to be a macro, not a function
2023-10-15 07:07:49 <quinq> Though gcompat do provide a wrapper for it
2023-10-15 07:08:15 <Test111> If gcompat provides a wrapper for it, I'm not sure why I'm getting this error
2023-10-15 07:11:15 <quinq> What are you trying to run in the first place?
2023-10-15 07:12:05 <Test111> I'm trying to use Radicale-DecSync
2023-10-15 07:12:51 <Test111> It's an addon for radicale that you can use to sync your calendar without servers (can use syncthing etc)
2023-10-15 07:13:51 <quinq> Isn't that just a python script?
2023-10-15 07:13:58 <quinq> https://github.com/39aldo39/Radicale-DecSync
2023-10-15 07:14:58 <Test111> It is
2023-10-15 07:15:37 <Test111> It uses libdecsync, which has some libraries that have to do with glibc I think
2023-10-15 07:15:46 <Test111> It looks like the issue stems from Kotlin itself
2023-10-15 07:15:53 <quinq> Ah, so the problem is decsync
2023-10-15 07:16:20 <Test111> Yeah radicale runs fine on its own
2023-10-15 07:17:10 <Test111> I don't have a local network, so I kind of have to use it
2023-10-15 07:17:25 <dwfreed> well that's a rabbit hole I didn't want to fall into
2023-10-15 07:17:29 <dwfreed> it's *kotlin*
2023-10-15 07:17:32 <quinq> I don't see a call to isnan in libdecsync
2023-10-15 07:17:33 <dwfreed> https://github.com/39aldo39/libdecsync
2023-10-15 07:17:53 <dwfreed> aka java, but slightly better
2023-10-15 07:18:27 <quinq> aka not
2023-10-15 07:18:41 <quinq> Test111, how did you build libdecsync?
2023-10-15 07:18:51 <Test111> I didn't build it myself
2023-10-15 07:19:11 <quinq> erf
2023-10-15 07:19:23 <Test111> I just installed ran `pip3 install radicale_storage_decsync`
2023-10-15 07:19:25 <quinq> That's the issue then :)
2023-10-15 07:19:47 <quinq> pip distributes third-party based built libraries? oO
2023-10-15 07:19:59 <quinq> the horror
2023-10-15 07:20:13 <dwfreed> yes, wheels can contain built native libs
2023-10-15 07:20:34 <dwfreed> but alpine is not manylinux compatible out of the box, so it shouldn't use them on alpine
2023-10-15 07:20:53 <quinq> Well, there's no way to know what's on the target system, unless you scan it
2023-10-15 07:21:00 <Test111> On github.com/39aldo39/Radicale-DecSync, the installation never mentions anything about building libdecsync
2023-10-15 07:21:11 <Test111> so I'm not sure why I need to build it
2023-10-15 07:21:11 <quinq> It's 2023 and this windows vendor mentality is still here…
2023-10-15 07:21:18 <quinq> Well, it's actually increasing
2023-10-15 07:22:24 <dwfreed> oh, wow, it's worse
2023-10-15 07:22:28 <dwfreed> https://github.com/39aldo39/libdecsync-bindings-python3/tree/master/libdecsync/libs
2023-10-15 07:22:28 <Test111> In fact, installing Radicale-DecSync should install libdecsync aswell
2023-10-15 07:22:32 <quinq> It seems to also be using some “python-libdecsync” https://github.com/39aldo39/libdecsync-bindings-python3/tree/master/libdecsync/libs
2023-10-15 07:22:36 <quinq> fear
2023-10-15 07:22:43 <dwfreed> he committed the built libs to git
2023-10-15 07:22:47 <quinq> :D
2023-10-15 07:22:53 <dwfreed> and just ships them as "package data"
2023-10-15 07:22:55 <Test111> .dll?
2023-10-15 07:23:15 <dwfreed> I mean, it probably "works" on windows too
2023-10-15 07:23:18 <dwfreed> python does, after all
2023-10-15 07:23:39 <Test111> I take it decsync is messy and I shouldn't use it, then
2023-10-15 07:23:57 <dwfreed> that's probably the easiest answer
2023-10-15 07:24:00 <Test111> probably can't
2023-10-15 07:24:02 <Test111> anyway
2023-10-15 07:24:15 <dwfreed> I mean, if you wanted to get your hands dirty, you could probably get it working
2023-10-15 07:24:38 <quinq> I suppose you can still build/package this libdesync, and Radicale-DecSync
2023-10-15 07:24:48 <Test111> I would need to figure out a way to build libdecsync for a musl system
2023-10-15 07:25:33 <quinq> oh, you also actually have the “source code” for the bindings https://github.com/39aldo39/libdecsync-bindings-python3/blob/master/libdecsync/__init__.py
2023-10-15 07:25:40 <quinq> So you can build that too
2023-10-15 07:25:43 <dwfreed> just need the kotlin toolchain, and it should "just work"
2023-10-15 07:26:40 <quinq> Sounds like a fun Sunday (I hope it's rainy over-there ;))
2023-10-15 07:26:41 <dwfreed> but like I said, easiest answer is to run away screaming bloody murder
2023-10-15 07:26:44 <Test111> hmm very simple indeed
2023-10-15 07:27:30 <Test111> Well I better look into tailscale then
2023-10-15 08:25:43 <Test111> Tailscale works but it complicates things on my phone, which I don't prefer.
2023-10-15 08:25:53 <Test111> I really need to get this thing to work
2023-10-15 10:03:29 <singham> Folks, is it possible to install alpine on AllWinner H3?
2023-10-15 10:03:35 <singham> Orange Pi?
2023-10-15 10:14:28 <WhyNotHugo> So package chromium bundles its own copy of a slightly older xdg-mime binary?
2023-10-15 10:15:49 <WhyNotHugo> singham: The aarch64 builds seem to support an rk3588 (which is the chip that the orange pi uses).
2023-10-15 10:16:07 <singham> Some others have H3 chip
2023-10-15 10:16:28 <singham> There's RK too and then there's H3
2023-10-15 10:18:35 <WhyNotHugo> The u-boot-sunxi package seems to wokr on 'orangepi_3'.
2023-10-15 10:18:46 <WhyNotHugo> At least that's what the APKBUILD indicates.
2023-10-15 10:26:55 <singham> https://www.crazypi.com/single-board-computers/orange-pi-india/orange-pi-latest-boards/orange-pi-zero256-india?sort=p.price&order=ASC&limit=100
2023-10-15 10:31:41 <singham> H2 chip
2023-10-15 11:21:15 <singham> I was afk
2023-10-15 11:21:28 <singham> Any comment on support for H2 chip
2023-10-15 11:22:19 <lopid> no
2023-10-15 11:22:42 <singham> 4?
2023-10-15 11:22:47 <lopid> or was there…
2023-10-15 11:38:44 <singham> H3 is supported
2023-10-15 11:38:53 <singham> no info on H2
2023-10-15 11:55:23 <frag> youtube-viewer-3.10.9-r0 - Lightweight YouTube client # installs 28 packages ........
2023-10-15 11:55:55 <dwfreed> package count does not define weight
2023-10-15 11:56:20 <dwfreed> xfce installs a bunch of things, and yet as DEs go, it's quite lightweight
2023-10-15 11:57:02 <dwfreed> I would imagine youtube-viewer is a lot lighter than a full web browser
2023-10-15 12:00:04 <frag> everyone can interpret it as they want ofc, i dont find it light weight hehe
2023-10-15 12:01:21 <dwfreed> youtube-viewer has 6 direct dependencies, and they're all perl modules
2023-10-15 12:06:26 <frag> i think the lightweigh label is a bit overused tbh
2023-10-15 12:23:38 <frag> shouldnt : socks_proxy "socks://localhost:9050"  set up links to use tor?
2023-10-15 12:25:54 <frag> ah, found out in menues ...
2023-10-15 12:25:57 <frag> so weird using menues..
2023-10-15 12:34:20 <witcher> does `abuild package` also install subpackages into the `pkg` dir?
2023-10-15 13:56:55 <ikke> no
2023-10-15 14:11:00 <witcher> how does one verify that the files are installed correctly for a subpackage?
2023-10-15 14:12:19 <ikke> 1) use abuild rootpkg, which runs the split function
2023-10-15 14:12:28 <ikke> 2) after building the package, run checkapk
2023-10-15 14:57:35 <Test111> I'm not sure what has gone wrong but there are serious issues regarding sound on my Alpine system. I had none of it on Void (also musl). Speakers go silent for a short moment every few miliseconds, and bluetooth earbuds play music for a sec, then go silent for another second.
2023-10-15 14:57:46 <Test111> Microphone doesn't work at all
2023-10-15 15:15:21 <Saijin_Naib[m]> Test111, what sound server are you using? pulse? pipewire?
2023-10-15 15:15:32 <Test111> Pipewire
2023-10-15 15:15:53 <Test111> (Unable to uninstall pulseaudio due to dependencies)
2023-10-15 15:16:20 <linfan> yeah, sound is a pain
2023-10-15 15:17:26 <invoked> i don't think libc has anything to do with it.
2023-10-15 15:17:56 <invoked> anyway, it would probably help if you explained your hardware/etc
2023-10-15 15:18:24 <Test111> Lenovo thinkpad e14 gen 2
2023-10-15 15:19:04 <Saijin_Naib[m]> What pipewire session manager? wireplumber?
2023-10-15 15:19:10 <Test111> Wireplumber
2023-10-15 15:19:29 <Saijin_Naib[m]> Are you using TLP or anything else to force power management on the audio chipset?
2023-10-15 15:19:43 <Test111> Yeah that might be one of the culprita
2023-10-15 15:19:44 <Test111> s
2023-10-15 15:19:54 <Saijin_Naib[m]> Yep
2023-10-15 15:20:01 <Test111> okay just stopped tlp
2023-10-15 15:20:09 <Test111> let me try playing a video
2023-10-15 15:20:16 <Saijin_Naib[m]> Not sure stopping it is the same as undoing the changes...
2023-10-15 15:20:17 <Saijin_Naib[m]> Hmm
2023-10-15 15:20:33 <Test111> yeah it did nothing
2023-10-15 15:20:50 <Test111> should i disable the service and reboot to see if it works
2023-10-15 15:22:56 <invoked> i have alpine on a couple thinkpads (unfortunately neither are with me rn) and i don't recall doing anything special fore pipewire.
2023-10-15 15:23:04 <Test111> According to https://github.com/ramaureirac/thinkpad-e14-linux/blob/main/tweaks/ , installing firmware-sof-signed helps. I did that (installed sof-firmware) on Void and sound worked just fine, both input and output.
2023-10-15 15:23:31 <invoked> ah this may be specific to your thinkpad then.
2023-10-15 15:23:52 <Test111> Yeah but both on Mint and Void it works
2023-10-15 15:24:50 <Test111> I wonder if it has to do with the kernel. Not sure if alpine has a stripped down kernel
2023-10-15 15:25:09 <invoked> it is lts. i don't know what void is running.
2023-10-15 15:26:20 <Hello71> alpine has sof firmware too
2023-10-15 15:26:28 <Test111> Yeah I have it installed
2023-10-15 15:26:32 <quinq> The pain in the sound system *is* pulseaudio/pipewire
2023-10-15 15:26:43 <quinq> alsa works ootb without problem
2023-10-15 15:26:48 <Hello71> aiui those sound cards aren't really expected to work without the firmware
2023-10-15 15:27:17 <Test111> On https://docs.voidlinux.org/config/kernel.html, it says "if you'd like to use the linux-lts or linux-mainline kernel series instead of the default linux"
2023-10-15 15:27:17 <quinq> Try layer by layer, disable those
2023-10-15 15:27:44 <quinq> Then you can more easily see if that's a driver issue, or power management one, or something else
2023-10-15 15:27:53 <linfan> Is the sof-firmware added to the kernel modules?
2023-10-15 15:28:24 <Hello71> quinq: as long as you don't want per-application volume controls, runtime device selection, sandboxing, or any other features added in the past decade since nobody writes alsa plugins anymore
2023-10-15 15:28:59 <Test111> It's not entirely on tlp because also no input device is found
2023-10-15 15:29:14 <Test111> neither the internal nor my earbuds' mic works
2023-10-15 15:29:50 <Test111> linfan: It does everything automatically upon install afaik, so there's no need to do anything yourself
2023-10-15 15:31:32 <linfan> cool have to reboot and see, coz i need to change float32le 2ch 48000Hz to s16le 2ch 44100 Hz
2023-10-15 15:32:30 <quinq> I have no trouble controlling several per mpv instance volume control
2023-10-15 15:32:30 <linfan> and there is no /etc/pulse/daemon.conf to edit
2023-10-15 15:32:34 <quinq> ^-- Hello71 
2023-10-15 15:32:38 <Test111> Okay something has happened
2023-10-15 15:32:50 <Test111> nevermind it's bad again
2023-10-15 15:33:13 <quinq> I can also switch from output device with a single key
2023-10-15 15:33:25 <quinq> s/from//
2023-10-15 15:37:26 <Test111> it's not tlp
2023-10-15 15:37:51 <quinq> You driver would only cut sound when no device is using it
2023-10-15 15:38:09 <quinq> So somehow the added layer doesn't tell correctly the driver that it's still using it
2023-10-15 15:43:46 <linfan> nae, sof-firmware dit not add anything, sound is as crackling as it is
2023-10-15 15:46:02 <linfan> no problem with sound on arch or slackware or void or fedora etc., but here the lack of configurability is annoying
2023-10-15 15:56:55 <p00f> how do i change where the kernel, initramfs etc are placed?
2023-10-15 15:57:02 <p00f> i'm sharing the ESP with another distro
2023-10-15 15:57:05 <witcher> ikke: thanks! checkapk is very helpful, too
2023-10-15 15:57:52 <ikke> witcher: indeed
2023-10-15 15:57:55 <p00f> the filenames of the kernel and initramfs are slightly different so there is no problem as such but i'd still like to put it in a different directory under the ESP (which is mounted at /boot)
2023-10-15 16:22:21 <minimal> p00f: the kernel package puts the kernel in /boot so the only way to change that would be to change and rebuild your own "local" kernel package
2023-10-15 16:24:19 <minimal> p00f: the mkinitfs package's trigger creates the initramfs in /boot so the only way to change that would be to modify the mkinitfs package and rebuild your own "local" mkinitfs package
2023-10-15 16:24:58 <p00f[m]> oh no
2023-10-15 16:25:13 <minimal> p00f: of course you could just install with the ESP's fileysstem mounted at /boot/efi rather than /boot...
2023-10-15 16:25:13 <p00f[m]> people must be dual booting right? is there no elegant solution
2023-10-15 16:25:36 <minimal> then the kernel and initramfs would not be in the ESP filesystem
2023-10-15 16:25:43 <quinq> Use a separate /boot
2023-10-15 16:25:55 <quinq> Why using the same partition for all systems
2023-10-15 16:27:40 <p00f[m]> systemd-boot can't binaries from other partitions
2023-10-15 16:27:55 <p00f[m]> and i really don
2023-10-15 16:28:03 <p00f[m]> don't want to install another bootloader
2023-10-15 16:28:31 <p00f[m]> i might just use efistub for alpine i guess
2023-10-15 16:28:44 <p00f[m]> thanks
2023-10-15 16:28:54 <quinq> Yeah, better keeping using the same broken boot-loader xD
2023-10-15 16:29:09 <minimal> p00f[m]: what's wrong with the separate /boot and /boot/efi suggestion?
2023-10-15 16:29:41 <p00f[m]> minimal: systemd-boot can't load kernels from other partitions
2023-10-15 16:29:45 <p00f[m]> quinq: whatever
2023-10-15 16:29:49 <minimal> I thought you were using Grub
2023-10-15 16:29:58 <p00f[m]> i was not
2023-10-15 16:30:02 <quinq> Good luck with your whatever problem then
2023-10-15 16:30:10 <p00f[m]> i explicitly was not using grub :)
2023-10-15 16:32:25 <Ermine> What's wrong with mounting esp as /boot ?
2023-10-15 16:32:26 <minimal> p00f[m]: ok, so you're hitting a systemd-boot limitation
2023-10-15 16:34:25 <minimal> Ermine: there's nothing "wrong" with mounting ESP as /boot, but it makes more sense to keep /boot filesystem (which has kernel, initramfs etc) separate from ESP
2023-10-15 16:35:16 <Ermine> I concur, but they're dead set on using systemd-boot...
2023-10-15 16:36:58 <minimal> Ermine: they're not having a problem, they asked how to move kernel and initramfs to a different subdirectory of /boot (where /boot is the ESP filesystem)
2023-10-15 17:02:33 <witcher> does alpine have a rough "rule" on when to add a -dbg subpackage to a package?
2023-10-15 17:04:59 <Hello71> when it's "not too big" and is actually useful for debugging. ideally all (compiled) packages would have it but with the current builder architecture all packages need to be stored on builders with limited disk space
2023-10-15 17:06:53 <witcher> hm alright. I guess I won't add it right now. thanks :)
2023-10-15 19:31:25 <WhyNotHugo> If I open this PDF file in chromium, copy some text from it and paste into firefox or foot, the pasted text is garbled up. Looks like LE/BE encoding mix-up.
2023-10-15 19:31:38 <WhyNotHugo> The problem is this specific file is a visa application, so not quite keen on sharing it.
2023-10-15 19:31:48 <WhyNotHugo> I've seen this before when dragging files too.
2023-10-15 19:31:59 <WhyNotHugo> Any ideas of known issues there/
2023-10-15 19:34:24 <quinq> Have you tried mupdf instead?
2023-10-15 19:35:48 <Ermine> If encoding is the problem, mupdf is unlikely to help
2023-10-15 19:35:59 <quinq> Why not?
2023-10-15 19:36:23 <quinq> It should be able to manage how it displays it
2023-10-15 19:36:40 <quinq> Also this “encoding” problem is just a supposition
2023-10-15 19:37:45 <Ermine> If this document was generated on windows, it could fuck up encodings. But yeah, it's a hypothesis
2023-10-15 19:38:03 <Hello71> it's almost certainly a custom "security" font. I don't think there's such thing as "LE/BE encoding mix-up"
2023-10-15 19:39:27 <Hello71> theoretically there's utf-16be but I don't think anyone actually uses it
2023-10-15 19:51:55 <bleb> I have a low dpi monitor and high dpi monitor.  when I start alpine, the console is shown on both displays, and the bottom of the console text is not visible on the low dpi monitor because the high dpi is much taller
2023-10-15 19:52:11 <bleb> any way to have the console only be displayed on the low dpi monitor?
2023-10-15 19:54:10 <bleb> or, define a maximum character height and width for the console?
2023-10-15 19:58:37 <quinq> iirc you can pass some parameter to the vga console at startup that asks you for a mode
2023-10-15 19:58:46 <quinq> like vgaask or something
2023-10-15 19:59:18 <quinq> You should also be able to give it a size directly
2023-10-15 19:59:31 <quinq> But as to chose output devices, don't know
2023-10-15 20:00:27 <quinq> vga=ask (https://www.kernel.org/doc/Documentation/arch/x86/boot.rst)
2023-10-15 20:01:47 <quinq> https://www.kernel.org/doc/Documentation/admin-guide/svga.rst
2023-10-15 20:14:41 <Hello71> i think it's fb driver specific. post boot most people probably aren't using vesafb
2023-10-15 20:17:43 <quinq> oh right, I still live in the 00's it seems
2023-10-15 20:18:55 <quinq> https://www.kernel.org/doc/Documentation/fb/modedb.rst
2023-10-15 20:18:58 <quinq> DRM drivers also add options to enable or disable outputs:
2023-10-15 20:21:40 <bleb> thanks yall I will take a look
2023-10-15 20:22:17 <quinq> Good luck bleb :)
2023-10-16 03:27:40 <Allak> To be a ghost with a pulse, get in a situation where the mind gives the perceived death signal, by watching a scary movie alone in the dark, and when it does just ignore it or pay no attention to it.- Daniel Alan Gabriel- ghostpeople@gmx.com
2023-10-16 05:35:58 <bananicorn> Hello, has anyone here played with incus, the recently released lxd fork?
2023-10-16 05:36:46 <bananicorn> Would love to have it packaged in Alpine or at least exchange notes on getting it working :)
2023-10-16 06:43:58 <danci1973> Good morning...
2023-10-16 06:46:16 <danci1973> Lately my Alpine Linux router is reporting some kernel messages in the log file that I can't find an explanation for: kernel: Revocation check: IN=br0 OUT=eth7 MAC=00:1f:29:5e:4d:81:60:14:b3:74:aa:19:08:00 SRC=172.16.0.88 DST=193.77.14.145 LEN=52 TOS=0x00 PREC=0x00 TTL=127 ID=9454 DF PROTO=TCP SPT=58938 DPT=80 WINDOW=64240 RES=0x00 SYN URGP=0
2023-10-16 06:47:44 <danci1973> I've downloaded kernel sources, iptables tools sources, but I couldn't find any 'Revocaton check' referrence...
2023-10-16 06:48:11 <danci1973> Any ideas what 'kernel: Revocation check' means?
2023-10-16 06:51:04 <aron> "Revocation check" seems to be SSL/TLS related
2023-10-16 06:51:27 <dwfreed> looks more like an iptables logging line
2023-10-16 06:51:31 <dwfreed> pastebin 'iptables-save'
2023-10-16 07:00:51 <danci1973> dwfreed: Thank you, I found it - there actually is a --log-prefix 'Revocaton check: ' line in the iptables setup... :D
2023-10-16 07:00:59 <dwfreed> :D
2023-10-16 07:01:08 <dwfreed> seen many an iptables log line
2023-10-16 07:02:26 <danci1973> So did I, but 'Revocation check: ' isn't something I'd typically use for that, so I obviously forgot about it. I have no idea what I meant with it. :/
2023-10-16 07:02:55 <danci1973> Thanks anyway, I just needed a nudge in the right direction.
2023-10-16 10:21:40 <frojnd> I Would like to bind systemctl levels to xf86BrightnessMonUp/Down keys. I've put: `bindsym --locked XF86MonBrightnessUp exec --no-startup-id brightnessctl +10%` inside swayconfig but after pressing fn+XF86MonBrightnessUp All I see is `~` inside terminal but no brightness change. I made sure that brightnessctl +10% works from terminal. Any ideas?
2023-10-16 10:22:01 <frojnd> As if fn key is not recognized at all?
2023-10-16 10:27:04 <frojnd> Figured it out. It was my command :s Should be just: `bindsym XF86MonBrightnessUp exec brightnessctl set +10%`
2023-10-16 11:51:27 <frojnd> Another issue? I've setup XDG_RUNTIME_DIR inside ~/.profile like this: https://termbin.com/xm56 I start my sway session like this: dbus-run-session -- sway All good right? yet when I do pactl list I get: Connection failure: Connection refused. And on the pipewire wiki https://wiki.alpinelinux.org/wiki/PipeWire says probably my XDG_RUNTIME_DIR is not set properly? 
2023-10-16 11:53:43 <frojnd> I use seatd 
2023-10-16 11:57:11 <donoban> hmm, could you check if pipewire is running? 'ps aux | grep pipewire'
2023-10-16 11:57:43 <frojnd> No haven't installed pipewire yet
2023-10-16 11:58:05 <donoban> so do you want to use pulseaudio?
2023-10-16 11:58:13 <frojnd> Thought pactl was independant?
2023-10-16 11:58:29 <frojnd> Yes I need pipewire
2023-10-16 11:58:58 <donoban> you need pipewire and a it's pulseaudio plugin/extension
2023-10-16 11:59:15 <donoban> maybe it runs with the default version
2023-10-16 11:59:35 <frojnd> I've installed now pipewire but pactl list gives me the samme Connection refused
2023-10-16 12:00:22 <frojnd> I think dbus is actually not configured properly? Even though I start sway like: dbus-run-session -- sway
2023-10-16 12:01:12 <donoban> you installed and executed it?
2023-10-16 12:01:47 <donoban> you should have 'pulse' dir in your XD_RUNTIME..
2023-10-16 12:02:48 <frojnd> I have following inside ls "$XDG_RUNTIME_DIR" https://termbin.com/slz9 including pulse
2023-10-16 12:03:44 <frojnd> My $XDG_RUNTIME_DIR path is: /tmp/1000-runtime-dir
2023-10-16 12:05:49 <donoban> uhm, you have pulse dir but not pipewire sock
2023-10-16 12:05:58 <donoban> what is there inside 'pulse'?
2023-10-16 12:06:22 <donoban> I think that this stuff has no relation with dbus
2023-10-16 12:06:41 <frojnd> Could be I have no idea
2023-10-16 12:07:15 <donoban> but have you checked that pipewire is running?
2023-10-16 12:07:15 <frojnd> donoban: pulse dir is empty
2023-10-16 12:07:20 <donoban> how did you started it?
2023-10-16 12:07:41 <frojnd> Didn't start it actually 
2023-10-16 12:07:51 <frojnd> ps aux |grep pipewire is empty
2023-10-16 12:07:55 <donoban> ok
2023-10-16 12:08:01 <donoban> try to execute just 'pipewire'
2023-10-16 12:08:21 <donoban> I think that its the right way but I don't now why it doesn't work for me on edge since some time
2023-10-16 12:08:41 <donoban> and I have to run /usr/libexec/pipewire-launcher
2023-10-16 12:09:02 <donoban> if some of them works ok, you should add it to your sway config
2023-10-16 12:09:25 <donoban> and it will start automatically
2023-10-16 12:09:50 <frojnd> pipwire is now running ps aux shows it. However pactl list still shows connection refused
2023-10-16 12:10:08 <frojnd> I've started pipewire manually inside terminal 
2023-10-16 12:11:17 <frojnd> This is my first installation of Alpine. I didn't installed any alsa or such tools yet?
2023-10-16 12:13:23 <donoban> https://wiki.alpinelinux.org/wiki/PipeWire maybe it's a litle outdated
2023-10-16 12:13:36 <donoban> "Install pipewire-pulse package, which provides a daemon so PulseAudio applications could use PipeWire as backend."
2023-10-16 12:13:44 <donoban> did you installed this?
2023-10-16 12:13:54 <donoban> I got to go frojnd , see you later
2023-10-16 12:14:30 <frojnd> donoban: allright take care!
2023-10-16 12:22:36 <frojnd> Having pipewire-pulse package installed didn't help connection refused by pactl list
2023-10-16 12:31:51 <Ermine> donoban: pipewire-pulse is PulseAudio compatibility layer
2023-10-16 12:33:22 <frojnd> strace shows I don't have pulseaudio: open("/usr/lib/pulseaudio/libpulse.so.0", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
2023-10-16 12:33:37 <Ermine> frojnd: ls $XDG_RUNTIME_DIR
2023-10-16 12:34:13 <frojnd> Ermine: https://termbin.com/m3st
2023-10-16 12:36:15 <Ermine> do you run pipewire or pulseaudio?
2023-10-16 12:36:39 <frojnd> Trying to run pipewire. pactl is mentioned in wiki
2023-10-16 12:36:50 <frojnd> I know wiki is not for n00bs
2023-10-16 12:37:23 <Ermine> We'll see if it can be improved
2023-10-16 12:37:40 <Ermine> So, you are running sway, right?
2023-10-16 12:38:16 <Ermine> if so, you need to use pipewire-launcher
2023-10-16 12:38:42 <Ermine> / And I need to get better script merged
2023-10-16 12:39:34 <frojnd> Let me try and add `exec /usr/libexec/pipewire-launcher` to sway config
2023-10-16 12:40:55 <Ermine> try this in the terminal for beginning
2023-10-16 12:41:17 <Ermine> So you'll be sure that this works
2023-10-16 12:41:35 <frojnd> ps aux |grep pipewire shows: 30349 frojnd    0:00 /usr/bin/pipewire (I've added to sway config)
2023-10-16 12:41:42 <frojnd> So it ran automatically
2023-10-16 12:43:12 <Ermine> and check wireplumber as well
2023-10-16 12:44:49 <frojnd> Ermine: I did
2023-10-16 12:45:04 <Ermine> does it run, I mean?
2023-10-16 12:45:12 <frojnd> wireplumber?
2023-10-16 12:45:21 <Ermine> yes
2023-10-16 12:45:25 <frojnd> Yes.
2023-10-16 12:45:37 <frojnd> How do I control volume levels? I was used to do it with pavucontrol
2023-10-16 12:45:58 <Ermine> pavucontrol should work with pipewire as well
2023-10-16 12:46:32 <frojnd> Shows no devices :)
2023-10-16 12:46:34 <frojnd> Only dummy device
2023-10-16 12:47:58 <Ermine> try installing pipewire-alsa and restarting your sway session
2023-10-16 12:48:09 <asimovsh> Yo
2023-10-16 12:48:31 <asimovsh> Someone have succesfully builded hyprland for alpine linux?
2023-10-16 12:49:33 <Ermine> frojnd: also, do you run on bare metal or in VM?
2023-10-16 12:49:44 <Ermine> asimovsh: abandon hope, all ye who enter here
2023-10-16 12:53:08 <frojnd> Ermine: bare metal
2023-10-16 12:53:15 <frojnd> I see movement inside pavucontrol but no voice
2023-10-16 12:53:43 <Ermine> So, devices are detected this time?
2023-10-16 12:53:48 <frojnd> alsamixer shows - (default) enter device name
2023-10-16 12:53:54 <frojnd> Ermine: it recognizes firefox
2023-10-16 12:54:01 <frojnd> Tried to play some yt video
2023-10-16 12:54:22 <frojnd> It's like no other sound card as dummy is currently detected
2023-10-16 12:54:37 <Ermine> firefox is not a device, it's a web browser. Does pavucontrol show any devices?
2023-10-16 12:55:08 <frojnd> Ermine: no just Dummy output says
2023-10-16 12:55:39 <Ermine> Did you add yourself to audio group?
2023-10-16 12:56:02 <frojnd> Yes: https://termbin.com/hbfr
2023-10-16 12:56:22 <PureTryOut> (and relogged after doing that?)
2023-10-16 12:56:30 <frojnd> No
2023-10-16 12:56:34 <frojnd> Man just killed sway 
2023-10-16 12:56:38 <frojnd> :s
2023-10-16 12:56:40 <frojnd> Good poiunt
2023-10-16 12:56:44 <PureTryOut> Then do so. group changes are not applied when not relogging
2023-10-16 12:57:40 <frojnd> Hm no luck still only dummy
2023-10-16 12:58:12 <frojnd> /proc/asound/cards shows --- no soundcards ---
2023-10-16 12:59:24 <frojnd> ls /sys/class/sound/ shows only time
2023-10-16 12:59:27 <frojnd> s/time/timer
2023-10-16 12:59:44 <Ermine> Which kernel do you run and which sound card do you have?
2023-10-16 13:00:20 <frojnd> Kernel: Linux t14 6.1.57-0-lts #1-Alpine SMP PREEMPT_DYNAMIC Wed, 11 Oct 2023 07:01:12 +0000 x86_64 Linux Sound card? Not sure what's lspci equivalent on alpine?
2023-10-16 13:00:51 <Ermine> lspci is in pcitools package iirc
2023-10-16 13:01:29 <Ermine> err, pciutils
2023-10-16 13:04:22 <frojnd> 00:1f.3 Audio device: Intel Corporation Alder Lake PCH-P High Definition Audio Controller (rev 01)
2023-10-16 13:05:34 <frojnd> Hm dmesg https://termbin.com/vuxm shows some missing firmware files
2023-10-16 13:05:43 <frojnd> Brb
2023-10-16 13:06:02 <aron> looks like your kernel does not supporting the sound card
2023-10-16 13:08:39 <Ermine> frojnd: you have 2 options: 1) try linux-edge kernel, 2) you're provided with a link to firmware file
2023-10-16 13:09:11 <aron> maybe you need CONFIG_SND_SOC_SOF_ALDERLAKE, which is not set in linux-lts
2023-10-16 13:09:44 <aron> but it is set to "m" in linux-edge
2023-10-16 13:10:22 <aron> (for you case, it means "enabled")
2023-10-16 13:13:23 <openlinux21[m]> 
ACTION uploaded an image: (125KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/ZlVoWLrZKUfFAfrQmikyEarF/VirtualBox_linux_16_10_2023_20_59_23.png >

2023-10-16 13:13:33 <openlinux21[m]> running spigotmc1.13.2 in alpine linux
2023-10-16 13:26:08 <PureTryOut> So? 🤔
2023-10-16 13:26:18 <openlinux21[m]> sO
2023-10-16 13:28:01 <openlinux21[m]> SO₂
2023-10-16 13:30:11 <openlinux21[m]> * <del> sO <\del>
2023-10-16 13:30:34 <openlinux21[m]> * <del> sO <del>
2023-10-16 13:30:50 <openlinux21[m]> * <del> sO
2023-10-16 13:31:23 <openlinux21[m]> * <del> sO <\del>
2023-10-16 13:43:54 <john3d> Hi. I'm currently testing the latest Alpine Edge with VirtualBox Guest Additions, and now the clipboard finally works without the previous need to call VBoxClient with su. Unfortunately, VBoxClient isn't starting automatically at the moment. However, I've already figured out how it was usually resolved via the install script from the VirtualBox ISO. The last time, there was a developer in the chat... maybe I can speak with 
2023-10-16 13:43:54 <john3d> you again?
2023-10-16 13:45:29 <minimal> "a developer" ?
2023-10-16 13:49:32 <john3d> minimal: Did we last talk about the clipboard issue in Virtualbox?
2023-10-16 13:49:55 <minimal> briefly, I don't use graphics in VirtualBox
2023-10-16 13:49:59 <frojnd> aron: let me first try to set CONFIG_SND_SOC_SOF_ALDERLAKE=m how do I set this?
2023-10-16 13:50:19 <minimal> have you enabled the virtualbox-drm-client init.d in the VM?
2023-10-16 13:51:20 <minimal> that runs VBoxDRMClient on boot, I don't see an init.d script to run VBoxClient specifically
2023-10-16 13:52:23 <aron> frojnd: I think it's easier to replace linux-lts with linux-edge
2023-10-16 13:52:51 <aron> building new kernel is not necessarily easy nor fast
2023-10-16 13:53:35 <john3d> VBoxClient need to be startet by /etc/xdg/autostart/vboxclient.desktop and /etc/X11/Xsession.d/98vboxadd-xclient. the is made by the original virtualbox-iso-script
2023-10-16 13:55:01 <john3d> vboxclient.desktop will start VBoxClient-all (shellscript starts VBoxClient --clipboard...)
2023-10-16 13:56:19 <john3d> maybe we can add 98vboxadd-xclient and vboxclient.desktop to the virtualbox-x11 package
2023-10-16 13:58:48 <frojnd> aron: ah ok. What's the "recommended" way to upgrade to linux-edge? To switch all repos to /*edge ?
2023-10-16 14:01:40 <aron> I think (1) switch repos from v3.xx to edge; (2) apk add linux-edge; covers it
2023-10-16 14:02:10 <frojnd> And then I can switch back to repos v3.xx ? Or is this not recommended?
2023-10-16 14:02:10 <aron> and boot edge instead of lts (depends on your bootloader)
2023-10-16 14:02:35 <aron> you can switch back anytime you want
2023-10-16 14:03:19 <aron> if you want to invest more time in this, you can also build your own kernel, but to try it out, it's better to use linux-edge
2023-10-16 14:05:07 <frojnd> I've installed linux-edge but uname -a still shows lts
2023-10-16 14:05:17 <aron> did you reboot your system?
2023-10-16 14:05:20 <frojnd> Or if I reboot it will automatically replace the lts?
2023-10-16 14:05:31 <frojnd> Or I must *select* the edge one?
2023-10-16 14:05:50 <aron> your bootloader will have 2 options now
2023-10-16 14:05:52 <linfan> add the new kernel and del the old one and reboot
2023-10-16 14:05:56 <aron> you should see them
2023-10-16 14:06:09 <aron> yeah, del old kernel also works, if you don't neet it anymore
2023-10-16 14:06:19 <aron> s/neet/need/
2023-10-16 14:06:30 <frojnd> So just apk del linux-lts ?
2023-10-16 14:06:35 <linfan> yep
2023-10-16 14:06:38 <aron> that will work
2023-10-16 14:06:51 <frojnd> Let me first see if zfsbootmenu shows new kernel option
2023-10-16 14:07:26 <aron> sure, better to check it first
2023-10-16 14:10:18 <linfan> I am still struggling with the desktop audio. After gnome install one snd module shows up, but after installing pipewire and wireplumber, suddenly there is a new sound module. For five days I have nonstop scanned the net, w/o finding any way to change back to original setup.
2023-10-16 14:13:23 <linfan> Some pages refers to a /etc/pulse/daemon.conf, but alpine only lists client.conf and it does not do anything, coz pulseaudio is not running as it says.
2023-10-16 14:14:00 <linfan> So the question is, where do I edit sound properties?
2023-10-16 14:16:18 <frojnd> Well it shows which is good. After selecting it zfs modules are not loaded so that is something I have to solve
2023-10-16 14:17:53 <aron> frojnd: try to add zfs to /etc/mkinitfs/mkinitfs.conf
2023-10-16 14:18:48 <frojnd> aron: it is there (after I boot from linux-lts)
2023-10-16 14:19:45 <aron> linfan: sry, can't help you with that, I use plain alsa, nothing else
2023-10-16 14:20:18 <frojnd> Do I have to install specific version of zfs for linux-edge? Or Reinstall it?
2023-10-16 14:20:40 <aron> frojnd: hm, maybe run mkinitfs for linux-edge as well? (but I presume apk did that)
2023-10-16 14:20:49 <frojnd> aron: it did
2023-10-16 14:21:03 <frojnd> /etc/mkinitfs/mkinitfs.conf shows: https://termbin.com/fpd5
2023-10-16 14:21:15 <linfan> i'd like that but as for gnome, it depends on pulseaudio and pipewire is a replacement for pulseaudio, so if i remove any of them it rips out most of gnome
2023-10-16 14:21:28 <aron> then initramfs should contain it -- weird
2023-10-16 14:21:57 <aron> mkinitfs.conf looks okay
2023-10-16 14:22:48 <frojnd> aron: it seems zfs package for stable main and edge is different
2023-10-16 14:23:38 <aron> you mean it's version?
2023-10-16 14:24:39 <aron> ** its
2023-10-16 14:25:20 <frojnd> I'm looking at https://pkgs.alpinelinux.org/packages?name=zfs&branch=v3.18&repo=&arch=&maintainer= and
2023-10-16 14:25:22 <frojnd> Yes
2023-10-16 14:25:31 <frojnd> So I think I also have to "resinstall" zfs package now
2023-10-16 14:25:54 <aron> or, you need the kernel modules only
2023-10-16 14:26:02 <frojnd> which package is that?
2023-10-16 14:26:09 <aron> https://pkgs.alpinelinux.org/package/edge/main/x86_64/zfs-lts
2023-10-16 14:26:15 <aron> it says "ZFS Linux kernel modules"
2023-10-16 14:26:27 <aron> version number seems to match with linux-edge
2023-10-16 14:26:39 <aron> or not :D
2023-10-16 14:27:03 <aron> hah
2023-10-16 14:27:09 <aron> no kernel modules for edge?
2023-10-16 14:28:40 <frojnd> hook 22
2023-10-16 14:28:44 <linfan> oh, nice Upgrading pipewire-pulse (0.3.79-r0 -> 0.3.82-r0)
2023-10-16 14:28:55 <aron> yep, zfs kernel modules are off-tree
2023-10-16 14:29:26 <frojnd> Maybe I can get a way with just reinstalling edge zfs package?
2023-10-16 14:29:37 <aron> I don't think so
2023-10-16 14:29:56 <aron> version difference in zfs tools should not be a huge problem to boot
2023-10-16 14:30:03 <aron> what is really need is the zfs kernel modules
2023-10-16 14:30:26 <aron> I mean, "what you really need are the zfs kernel modules"
2023-10-16 14:31:02 <aron> this seems to be a problem
2023-10-16 14:31:17 <minimal> john3d: open an Issue describing the problem and then it will be assigned to the package maintainer
2023-10-16 14:32:24 <john3d> minimal. thank you. i will do it. :-)
2023-10-16 14:33:15 <frojnd> aron: basically that means: one cannot use zfs with edge
2023-10-16 14:33:52 <aron> frojnd: I see 2 possible solutions now: (1) open an issue, and request zfs-edge (2) rebuild linux-lts with CONFIG_SND_SOC_SOF_ALDERLAKE=m
2023-10-16 14:34:09 <aron> option #1 might be useful for others in the future
2023-10-16 14:35:22 <frojnd> aron: I'm just wondering. Noone else did use zfs with kernel-edge before here? That seems quite strange, unless it's hard to keep the modules aligned with kernel
2023-10-16 14:36:06 <aron> well, for every linux-edge build, you'd need to build a fresh zfs-edge, otherwise the modules won't be loaded
2023-10-16 14:36:53 <aron> so, you can choose now: no sound, or no filesystem ^^
2023-10-16 14:38:32 <aron> but yeah, it's a bit strange
2023-10-16 14:38:41 <frojnd> Or build it myself. Which is an entire new beast for me
2023-10-16 14:38:51 <aron> (although, I use the good ol' ext4)
2023-10-16 14:39:08 <aron> yes, this is also an option
2023-10-16 14:39:33 <aron> but I think, you'll need a running linux-edge, if you want to spare some hacking ;)
2023-10-16 14:40:29 <aron> if you go into build yourself direction, maybe option #2 above will be easier
2023-10-16 14:42:51 <aron> just a few steps: (1) prepare pkg build environment, see wiki; (2) download linux-lts' APKBUILD + patches (i.e. with git, or whatever); (3) append CONFIG_SND_SOC_SOF_ALDERLAKE=m to the x86_64 config; (4) abuild checksum; (5) abuild -r; (6) apk add /path/to/your/home/packages/.../linux-lts-<version>.apk
2023-10-16 14:43:15 <aron> this is basically easy, but will take some time, kernel is large ;)
2023-10-16 14:46:38 <aron> but make sure you build the same linux-lts version you have
2023-10-16 18:32:04 <Ermine> if you slim config down to your hardware only, it might take like 30 mins to build
2023-10-16 19:57:09 <linfan> phew, finally got sound working
2023-10-16 19:57:50 <linfan> it only took me some 7 days and countless reinstalls
2023-10-16 20:06:20 <frojnd> aron: have you had that wiki in mind? https://wiki.alpinelinux.org/wiki/Custom_Kernel
2023-10-16 20:11:44 <aron> frojnd: nah, I just figured out by myself
2023-10-16 20:11:59 <aron> I see an error in the large table, tho
2023-10-16 20:12:16 <aron> it says "ARCH x86_64" is "bitness 32", which is odd
2023-10-16 20:12:17 <frojnd> aron: what do you mean?
2023-10-16 20:12:42 <aron> should be 64
2023-10-16 20:13:45 <frojnd> You mean the part 3?
2023-10-16 20:13:58 <frojnd> I am at step (1) prepare build env see wiki
2023-10-16 20:14:01 <frojnd> :)
2023-10-16 20:14:02 <aron> in the table, named "Vanilla targets and tuning"
2023-10-16 20:14:31 <aron> by step 1, I meant https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package
2023-10-16 20:14:32 <aron> ;)
2023-10-16 20:16:06 <frojnd> Oh thank you, gonna try to do it now before I go to sleep
2023-10-16 20:16:25 <aron> sleep time for me as well
2023-10-16 20:20:34 <Ermine> linfan: how did you manage to get it working?
2023-10-16 20:21:59 <frojnd> linfan | it only took me some 7 days and countless reinstalls > whole system reinstalls? Or just some apks?
2023-10-16 20:23:44 <Ermine> It's worth documenting
2023-10-17 05:08:29 <frojnd> aron: you were right that part of building was busy. But after I did apk add ~/coding/alpine/packages/coding/x86_64/linux-lts-dev-6.1.58-r0.apk it installed kernel, but I did not see any mkinit hooks
2023-10-17 05:08:52 <frojnd> Also after rebooting I didn't see linux-lts-dev as na option
2023-10-17 05:10:47 <aron> frojnd: did you installed ~/coding/alpine/packages/coding/x86_64/linux-lts-6.1.58-r0.apk  (not -dev pkg) as well?
2023-10-17 05:13:47 <frojnd> aron: err no, just dev
2023-10-17 05:14:49 <frojnd> So I need to install both?
2023-10-17 05:16:23 <aron> linux-lts is the kernel package itself; linux-lts-dev is the package you need if you want to build off-tree modules for example
2023-10-17 05:16:47 <aron> did you appended that extra config before build, right?
2023-10-17 05:17:00 <frojnd> Yes
2023-10-17 05:17:03 <aron> cool
2023-10-17 05:17:21 <frojnd> So I don't need dev since I just added that extra config
2023-10-17 05:17:34 <frojnd> Ok will remove it and installed linux-lts with extra config
2023-10-17 05:27:50 <frojnd> aron: ok now initrams line was present
2023-10-17 05:28:07 <frojnd> Booted inside of it, just don't know yet how to check if this kernel has the config set to m
2023-10-17 05:28:27 <frojnd> CONFIG_SND_SOC_SOF_ALDERLAKE=m
2023-10-17 05:30:07 <aron> probably that module is loaded by now
2023-10-17 05:31:57 <frojnd> grep ALDERLAKE /proc/modules  gives me empty line hm
2023-10-17 05:32:52 <frojnd> After I did apk install ~/coding/alpine/packages/coding/x86_64/linux-lts-6.1.58-r0.apk  <- that kernel should replace current linux-lts ?
2023-10-17 05:33:43 <aron> yes, it replaced the old one
2023-10-17 05:34:24 <aron> grep ALDERLAKE /boot/config-lts   <-
2023-10-17 05:34:40 <aron> you get the actual config there
2023-10-17 05:39:14 <frojnd> CONFIG_SND_SOC_SOF_ALDERLAKE=m
2023-10-17 05:39:32 <aron> also:  "modprobe configs", and you'll have /proc/config.gz
2023-10-17 05:40:13 <aron> should be the same as /boot/config-lts, tho
2023-10-17 05:41:32 <aron> okay then, you have sound now? :)
2023-10-17 05:42:57 <frojnd> Sadly no
2023-10-17 05:43:10 <aron> bummer
2023-10-17 05:43:14 <frojnd> Trying to extract /proc/config.gz
2023-10-17 05:43:25 <frojnd> From here: https://git.alpinelinux.org/aports/tree/main/linux-lts?h=3.18-stable
2023-10-17 05:43:35 <frojnd> I skipped virt.* files
2023-10-17 05:43:52 <aron> virt is not important
2023-10-17 05:44:17 <frojnd> Also inside APKGBUILD I deleted virt.* and lts.* except lts.x86_64.config lines
2023-10-17 05:44:20 <frojnd> If that matters
2023-10-17 05:45:39 <aron> probably not
2023-10-17 05:46:07 <frojnd> aron: I don't see ALDERLAKE inside config
2023-10-17 05:46:28 <aron> well, that's a problem
2023-10-17 05:46:32 <aron> okay, so
2023-10-17 05:46:34 <frojnd> After extracting /proc/config.gz
2023-10-17 05:46:59 <aron> when you installed the linux-lts-6.1.blah.apk, a new initramfs is created
2023-10-17 05:47:13 <aron> then you reboot your system
2023-10-17 05:47:20 <aron> right?
2023-10-17 05:47:24 <frojnd> right
2023-10-17 05:48:49 <aron> and the config is not there
2023-10-17 05:49:09 <frojnd> Apperantly
2023-10-17 05:49:51 <aron> and I don't see the module name yet
2023-10-17 05:50:01 <frojnd> Module name?
2023-10-17 05:50:20 <aron> yes, it was set to "m", which means "module"
2023-10-17 05:50:36 <frojnd> Where should the name bee seen?
2023-10-17 05:50:55 <frojnd> lsmod ?
2023-10-17 05:51:11 <aron> the loaded modules are listed by lsmod, yes
2023-10-17 05:51:21 <aron> ok, I see something: snd-sof-pci-intel-tgl
2023-10-17 05:51:26 <aron> (don't ask)
2023-10-17 05:54:12 <frojnd> For some reason module isn't there
2023-10-17 05:54:36 <frojnd> But it's inside lts.x86_64.config that APKBUILD took for building
2023-10-17 05:55:28 <aron> and you have CONFIG_SND_SOC_SOF_INTEL_TGL in /proc/config.gz ?
2023-10-17 05:55:57 <frojnd> CONFIG_SND_SOC_SOF_INTEL_TGL=m
2023-10-17 05:56:18 <aron> should be ok, I think
2023-10-17 05:56:26 <aron> do you have lspci, right?
2023-10-17 05:56:40 <frojnd> Yes
2023-10-17 05:56:52 <aron> lspci -v   will list the kernel driver as well
2023-10-17 05:57:23 <aron> search your audio device in the output, and "Kernel driver in use:" will tell
2023-10-17 05:57:27 <frojnd>  Kernel driver in use: sof-audio-pci-intel-tgl
2023-10-17 05:58:03 <aron> cool
2023-10-17 05:58:17 <aron> have alsa installed?
2023-10-17 05:58:23 <frojnd> Yes
2023-10-17 05:58:44 <frojnd> But alsamixer shows: This sound device does not have any controls. Let me check cards
2023-10-17 05:58:54 <frojnd> No cards to select
2023-10-17 05:58:58 <frojnd> HHHm
2023-10-17 05:59:17 <aron> wait a sec
2023-10-17 05:59:21 <aron> aplay -l   <--
2023-10-17 05:59:34 <frojnd> aplay: device_list:277: no soundcards found...
2023-10-17 06:00:05 <aron> hmm
2023-10-17 06:00:21 <aron> did you followed this, right? https://wiki.alpinelinux.org/wiki/ALSA
2023-10-17 06:03:45 <aron> do you have anything in /dev/snd/ ?
2023-10-17 06:10:06 <aron> apk add sof-firmware  <- also this
2023-10-17 06:15:33 <aron> it seems alderlake is "the same" as tigerlake (in the audio aspect)
2023-10-17 06:15:45 <aron> something to read: https://thesofproject.github.io/latest/getting_started/intel_debug/introduction.html
2023-10-17 06:17:29 <aron> it seems this alderlake audio stuff is handled by sof
2023-10-17 06:17:45 <aron> sadly, I never seen one before
2023-10-17 06:18:56 <aron> for all of my devices, I managed to get them work with the kernel only (even on allwinnder SoCs)
2023-10-17 06:19:13 <winny> i have a tigerlake and it won't do anything until you get sof
2023-10-17 06:20:24 <aron> winny: did you managed to get it work?
2023-10-17 06:21:06 <winny> i've only tested gentoo and nixos on the device.  NixOS auto-detects it, gentoo it was a matter of installing the sof-firmware package (I also updated it for some fixes)
2023-10-17 06:21:17 <winny> I think alpine may be the same, unless the kernel driver is missing.
2023-10-17 06:22:11 <aron> alpine has sof-firmware pkg
2023-10-17 06:22:29 <winny> I have this also which is alderlake, same story https://linux-hardware.org/?probe=9bb1b8fbca
2023-10-17 06:22:44 <winny> maybe i should test it 🤔
2023-10-17 06:23:49 <aron> ok, it uses the same "sof_audio_pci_intel_tgl" module for that
2023-10-17 06:24:01 <winny> hm.  At the risk of derailing discussion - if i wanted to test desktop alpine on a random box, is there a livecd/USB image appropriate for this usage
2023-10-17 06:24:37 <winny> this is the XPS with tigerlake for comparison https://linux-hardware.org/?probe=99232ffba3
2023-10-17 06:25:29 <aron> download "extended" from https://alpinelinux.org/downloads/ ; then configure apk, and install what you want :)
2023-10-17 06:25:51 <aron> you can get working X11 this way
2023-10-17 06:25:51 <winny> ;) can do
2023-10-17 06:26:20 <aron> I did this a year ago, or so
2023-10-17 06:28:23 <winny> cool, frojnd if you're still stuck by wednesday, i'll test on my alderlake laptop
2023-10-17 06:30:55 <aron> I hope installing sof-firmware will do something at last
2023-10-17 07:17:51 <frojnd> aron: my electricith went off... no I haven't followed wiki ALSA just pipewire
2023-10-17 07:17:56 <frojnd> aron: will check that first now
2023-10-17 07:20:59 <frojnd> aron: nothing inside /dev/snd even after installing (not rebooting) sof-firmware
2023-10-17 07:22:06 <aron> try rebooting
2023-10-17 07:22:10 <frojnd> Ok
2023-10-17 07:22:25 <frojnd> I mean content currently inside /dev/snd/ is just timer
2023-10-17 07:22:51 <aron> that's insufficient
2023-10-17 07:24:23 <frojnd> there is a card! Default one: sof-hda-dsp
2023-10-17 07:24:55 <aron> nice!
2023-10-17 07:25:26 <frojnd> Wondering. Was that because of the sof-firmware or also recompiled kernel with extra module?
2023-10-17 07:25:57 <aron> after investigating that ALDERLAKE config, my guess is sof-firmware is the key
2023-10-17 07:26:16 <aron> the ALDERLAKE config forwars to sof_audio_pci_intel_tgl
2023-10-17 07:26:35 <aron> which should there in the upstream linux-lts
2023-10-17 07:26:36 <frojnd> So let me retest this by reinstalling linux-lts 
2023-10-17 07:26:43 <aron> yes, definitely
2023-10-17 07:26:46 <frojnd> upstream linux-lts
2023-10-17 07:27:03 <frojnd> Can I force it apk add linux-lts just says ok
2023-10-17 07:27:43 <aron> apk fix linux-lts   I think
2023-10-17 07:28:01 <frojnd> That would be the one
2023-10-17 07:28:23 <frojnd> Nope
2023-10-17 07:28:42 <frojnd> https://termbin.com/41iu
2023-10-17 07:29:55 <aron> ok, then just hack it:  in /etc/apk/world, linux-lts will have a suffix; remove it
2023-10-17 07:30:00 <aron> then   apk fix linux-lts
2023-10-17 07:30:05 <frojnd> There is --force
2023-10-17 07:30:09 <frojnd> as global option
2023-10-17 07:30:27 <aron> you could try, I have less hope on that
2023-10-17 07:30:56 <frojnd> Didn't work with --force
2023-10-17 07:31:01 <aron> yeah
2023-10-17 07:31:26 <frojnd> in /etc/apk/world there is just linux-lts
2023-10-17 07:31:52 <frojnd> Can't see any suffix
2023-10-17 07:32:03 <aron> hm
2023-10-17 07:32:19 <aron> the  apk fix linux-lts  should work
2023-10-17 07:32:28 <aron> maybe the repo changed meanwhile
2023-10-17 07:32:34 <aron> apk update && apk fix linux-lts
2023-10-17 07:33:22 <frojnd> I mean it reinstalls it witwh ok but can't see nay initramfs messages
2023-10-17 07:33:40 <aron> ah, ok
2023-10-17 07:33:45 <aron> that's a different thing
2023-10-17 07:34:12 <frojnd> https://termbin.com/oigt
2023-10-17 07:34:21 <aron> you can do it manually:   mkinitfs
2023-10-17 07:34:56 <aron> that's not good, still didn't reinstalled linux-lts
2023-10-17 07:35:59 <frojnd> Well I can reboot and see if there is ALDERLAKE inside /boot/config-lts
2023-10-17 07:36:16 <aron> apk fix --reinstall linux-lts   <- maybe this?
2023-10-17 07:36:17 <frojnd> Right? And if there is no ALDERLAKE then I'm on upstream?
2023-10-17 07:36:59 <frojnd> https://termbin.com/8h1y
2023-10-17 07:37:08 <aron> upstream does not have the ALDERLAKE config, but the module itself comes from TIGERLAKE
2023-10-17 07:37:41 <aron> I don't get it...
2023-10-17 07:38:42 <aron> oh, I know
2023-10-17 07:38:52 <aron> did you changed back to latest-stable
2023-10-17 07:39:05 <aron> your specific kernel version is available only in the edge repo
2023-10-17 07:39:59 <aron> the "edge" repo is a different thing to "linux-edge"
2023-10-17 07:40:42 <aron> if you want your old-old linux-lts back, then keep at latest-stable, and "apk upgrade"
2023-10-17 07:41:08 <aron> which will downgrade the kernel back to latest-stable's
2023-10-17 07:41:42 <frojnd> Not sure what linux-edge did when I installed it before
2023-10-17 07:41:51 <frojnd> And remove it since it didn't have modules for zfs
2023-10-17 07:42:09 <aron> yes, forget linux-edge, it's not needed now
2023-10-17 07:42:15 <frojnd> aron: I removed it
2023-10-17 07:42:23 <frojnd> I've switched my repos to latest-stable
2023-10-17 07:42:34 <aron> do "apk upgrade"
2023-10-17 07:42:50 <aron> it should do what we want
2023-10-17 07:44:03 <frojnd> After reboot no sound card in alsamixer
2023-10-17 07:44:05 <frojnd> Hm
2023-10-17 07:45:19 <aron> yeah, then that kernel config is really a must
2023-10-17 07:46:12 <aron> put your handmade kernel back, and use that
2023-10-17 07:46:27 <aron> so we learned, that CONFIG_SND_SOC_SOF_INTEL_TGL=m and sof-firmware is what you need
2023-10-17 07:47:07 <aron> if you want, open an issue, and request CONFIG_SND_SOC_SOF_INTEL_TGL=m to be in linux-lts, it seems to be important
2023-10-17 07:47:15 <aron> or, i mean
2023-10-17 07:47:22 <aron> CONFIG_SND_SOC_SOF_ALDERLAKE=m
2023-10-17 07:48:41 <aron> I saw the TIGERLAKE config in linux-edge, and I though it was set in linux-lts as well, my bad
2023-10-17 07:56:18 <frojnd> Ha wait
2023-10-17 07:56:24 <frojnd> Maybe kernel doesn't play a role man
2023-10-17 07:56:31 <frojnd> Because I have to select default card
2023-10-17 07:56:37 <frojnd> In alsamixer
2023-10-17 07:56:45 <frojnd> Let me repeat the test
2023-10-17 07:59:22 <aron> sure
2023-10-17 08:03:29 <frojnd> Yes, upstream linux-lts is enough as is (provided that I'm actually on upstream linux-lts)
2023-10-17 08:04:11 <aron> and you got sound with sof-firmware only?
2023-10-17 08:04:43 <frojnd> Yes
2023-10-17 08:04:47 <aron> perfect
2023-10-17 08:05:18 <frojnd> Such an amazing help here really guys. And I even learnt a bit about building apks
2023-10-17 08:05:32 <frojnd> Thank you aron so much
2023-10-17 08:05:51 <aron> yw
2023-10-17 08:06:06 <aron> I'm glad it works now
2023-10-17 09:03:06 <frojnd> Hm... out of the blue question. How can I find system calls on linux for ethernet card if it's in duplex, full, half duplex, or how can I find out cpu_max_load or even memory total or memory average usage
2023-10-17 09:07:39 <Habbie> for ethernet, look at the ethtool source perhaps
2023-10-17 09:07:50 <Habbie> for cpu, memory, etc., dig around in /proc or look at the source of ps or top
2023-10-17 10:17:28 <frojnd> Hm
2023-10-17 10:17:50 <frojnd> Not sure why but when I open alsamixer default card is -
2023-10-17 10:18:44 <aron> follow the steps here to set the default card: https://wiki.alpinelinux.org/wiki/ALSA
2023-10-17 10:20:18 <frojnd> When I select card within alsamixer I can see: https://termbin.com/g44y
2023-10-17 10:21:17 <frojnd> So the snd has 0 and inside /usr/share/alsa/alsa.conf I have defaults.ctl.card 0; defaults.pcm.card 0
2023-10-17 10:25:31 <aron> is alsa service running?
2023-10-17 10:26:09 <frojnd> I've just added to openrc
2023-10-17 10:26:14 <frojnd> rebooted and it works
2023-10-17 10:26:32 <frojnd> But first time I didn't start it with openrc so weird why it worked in the first place
2023-10-17 11:31:53 <frojnd> When I have an option I plug in ethernet cable for the internet. So I've setup my network interface eth0 with setup-interfaces. Now I've also setup wlan0 following this wiki: https://wiki.alpinelinux.org/wiki/Wi-Fi using iwd. I've also added iwd to start at boot. All workrs great except 2 things: 1) If my ethernet cable is not plugged during boot, I'm waiting for udhcpc to
2023-10-17 11:31:55 <frojnd> figure out there is no network cable and seconds are losts. 2) even though iwd connects to my network after I login, I have to manually execute: udhcpc -i wlan0. Best scenario would be if there would be a mechanism to recognize is ethernet cable is plugged and if it's not no need to start udhcpc for eth0 during boot. I think issue is because currenctly issuing dhcp is within:
2023-10-17 11:31:57 <frojnd> /etc/network/interfaces for eth0 and can also be done for wlan0 looking at the wiki. Would be great if there were any mechanism to detect during boot if lan cable is not plugged don't issue udhcpc
2023-10-17 11:35:49 <frojnd> Well not sure why but wlan0 does't get lease during boot process. Bummer
2023-10-17 12:42:46 <sewn> This might not be related to alpine, but how can I grab the dmesg log before openrc finishes? I have a weird graphics issue where it freezes up and openrc is stuck on "waiting for uevents"
2023-10-17 12:52:27 <mercenary> serial console?
2023-10-17 12:55:26 <hodapp> that's the only way I can think of
2023-10-17 12:58:18 <townsfolkPravda> I wonder if setting init=/bin/sh in Linux kernel cmdline could help?
2023-10-17 13:01:08 <sewn> mercenary: meaning to do that, how on earth do I do that
2023-10-17 13:01:26 <sewn> do I use an arduino or
2023-10-17 13:01:48 <sewn> townsfolkPravda: I dont think by then the initramfs would function
2023-10-17 13:02:13 <sewn> and also no input because of that
2023-10-17 13:02:16 <aron> netconsole?
2023-10-17 13:02:54 <sewn> let me try that with my phone
2023-10-17 13:03:55 <townsfolkPravda> sewn, I thought initramfs is used before running init. But I could be wrong.
2023-10-17 13:04:32 <mercenary> sewn: anything that can run cu/miniterm/... The trick is to have something serial on the broken machine that the kernel sees, and the right magic console=/dev/ttyS0,115200 type incantation
2023-10-17 13:06:40 <sewn> apparently netconsole is some form of service and required syslog which openrc won't be able to reach
2023-10-17 13:07:53 <sewn> how will I even do serial, all I have is an android phone with root access
2023-10-17 13:10:19 <mercenary> ah. that could be a problem. I am so used to having piles of old computers lying around that I forget not everybody is a hoarder
2023-10-17 13:14:00 <sewn> well, how can I blacklist a module from kernel command line? since the issue is in amdgpu
2023-10-17 13:24:12 <minimal> sewn: "blacklist=<module name>"
2023-10-17 13:25:01 <sewn> ok then i guess amdgpu isnt the problem
2023-10-17 13:28:24 <aron> try with "nomodeset=1" kernel parameter
2023-10-17 13:35:21 <minimal> it's just "nomodeset"
2023-10-17 14:34:23 <frojnd> sewn: not sure how but usbotg, drivers for usb to serial. Your Android kernel should be compiled agains serial console modules though
2023-10-17 14:35:00 <frojnd> sewn: didn't see messages related to kernel params... ignore me
2023-10-17 14:49:32 <vkrishn> seeing this msg in alpine-3.15.4 logs (nginx+http2, letsencrypt + php7), any guide, guessing i should upgrade to newer al ? <- https://tpaste.us/WxzZ 
2023-10-17 15:02:45 <minimal> vkrishn: yes, upgrade
2023-10-17 15:13:32 <bleb> al... pine?
2023-10-17 15:15:00 <minimal> vkrishn: especially as support for Alpine 3.15.x ends in 14 days time...
2023-10-17 15:15:51 <minimal> indeed why are you even on 3.15.4 when the latest 3.15.x release is 3.15.10?
2023-10-17 15:16:13 <sewn> frojnd: i need serial output anyways, i get kernel panics often and i wont know why without any sort of logging
2023-10-17 15:19:58 <vkrishn> minimal: sadly, i have some php7 apps, would move to v3.18.x asap
2023-10-17 15:21:11 <vkrishn> was wanting to know the msg reason, and so if i could delay the upgrade
2023-10-17 15:21:28 <vkrishn> thanks
2023-10-17 15:21:59 <minimal> any specific reason why you are not on 3.15.10?
2023-10-17 15:23:01 <vkrishn> hmmm, will that solve ? just becoz i replicated setup from another device
2023-10-17 15:23:28 <minimal> no I'm not saying it will solve anything, just that you are not even on the latest 3.15.x release
2023-10-17 15:23:37 <vkrishn> ok
2023-10-17 15:23:51 <minimal> look at nginx 1.23.1 Changelog
2023-10-17 15:24:02 <vkrishn> ok
2023-10-17 15:55:37 <alyssa_rose> hi there, trying to set up pppoe connection on an alpine box and i'm getting zero logs from pppd anywhere
2023-10-17 15:55:58 <alyssa_rose> this is unhelpful because i'm trying to figure out why it doesn't automatically reconnect after the modem is rebooted
2023-10-17 15:56:54 <alyssa_rose> i can restart it myself and get back online, but i don't want to have to do that (or write a script to do that for me)
2023-10-17 15:57:03 <alyssa_rose> what am i doing wrong?
2023-10-17 16:12:58 <alyssa_rose> ah, i can set post-down on the ethernet interface to poff
2023-10-17 16:13:05 <alyssa_rose> let's see if that works
2023-10-17 16:16:43 <alyssa_rose> hmm, post-down doesn't seem to be called when i unplug my modem and the interface goes down
2023-10-17 16:42:34 <linfan> whooa!!! unplugged my ancient speaker system and now sound works superbly throug hdmi display port
2023-10-17 17:02:06 <alyssa_rose> agh, even with `persist` and `maxfail 0` in /etc/ppp/peers/dsl-provider pppd just refuses to even try to reconnect
2023-10-17 17:04:22 <dalias> alyssa_rose, well first you probably need to figure out how to setup logging and see if it's even using the options you give it
2023-10-17 17:04:32 <dalias> how is pppd being launched?
2023-10-17 17:05:15 <alyssa_rose> i'm launching pppd with `/usr/sbin/pppd debug call dsl-provider`
2023-10-17 17:05:21 <alyssa_rose> it is indeed logging to /var/log/messages
2023-10-17 17:08:36 <alyssa_rose> i think it is using the options, because sometimes when i launch it it fails to authenticate, closes the connection, and tries again
2023-10-17 23:51:25 <armin> so why is the standard image 17mb smaller than the netboot image? what sense exactly does thit make?
2023-10-17 23:51:35 <armin> s/thit/this/
2023-10-18 00:11:35 <minimal> armin: well the standard image is an ISO filesystem, it has both Isolinux (for BIOS) and Grub (for UEFI) present etc
2023-10-18 00:11:46 <minimal> whereas the netboot is just a tarball
2023-10-18 00:12:30 <minimal> are you asking why the netboot is not smaller than the standard ISO?
2023-10-18 00:15:09 <dwfreed> that's what the question seems to be
2023-10-18 00:15:46 <minimal> the netboot includes BOTH linux-lts and linux-virt kernels (and also initramfs-lts and initramfs-virt)
2023-10-18 00:16:03 <minimal> so that would account for the size difference
2023-10-18 00:17:14 <minimal> armin: I assume the reason being that you unpack the netboot tarball and then if you are netbooting a VM you use the linux-virt and initramfs-virt whereas for a physical machine you use the linux-lts and initramfs-lts
2023-10-18 00:18:52 <minimal> oh and there's also modloop-lts & modloop-virt and config-lts & config-virt in the tarball
2023-10-18 02:25:58 <somercet1> Github now requires 2FA. Can anyone rec a site that doesn't demand I upload all my passwords and is FOSS-compatible?
2023-10-18 02:54:51 <elly> what do you mean, "doesn't demand I upload all my passwords"?
2023-10-18 03:28:40 <fluix> ^, but to answer the second part: gitlab, codeberg, sourcehut, to name a few
2023-10-18 03:29:04 <fluix> TOTP isn't uploading passwords or anything sooo yeah
2023-10-18 03:29:20 <elly> plug for sourcehut, personally
2023-10-18 03:29:30 <elly> (but also, you should use 2fa for everything you can anyway...)
2023-10-18 08:34:02 <Habbie> lots of alpine inspiration here it seems :) https://thenewstack.io/meet-wolfi-the-linux-distro-designed-to-shrink-your-supply-chain/
2023-10-18 08:42:43 <Habbie> actually that makes more sense on offtopic, so pretend i didn't post it here
2023-10-18 08:42:51 <witcher> where is it documented what patterns apk-search takes? e.g. "cmd:grep"
2023-10-18 08:44:34 <Ermine> Does man apk-search show anything?
2023-10-18 08:44:54 <witcher> nope, nothing :(
2023-10-18 08:44:56 <Habbie> i think cmd:grep is not really an apk feature though? more an alpine packaging rule?
2023-10-18 08:47:31 <witcher> hm i see
2023-10-18 08:56:53 <aron> according to: https://docs.alpinelinux.org/user-handbook/0.1a/Working/apk.html
2023-10-18 08:56:58 <aron> "Further, globbing is supported. As such, here are a few examples of searching:"
2023-10-18 08:57:05 <aron> so, it's globbing
2023-10-18 08:57:21 <aron> you can read about "cmd:" as well
2023-10-18 09:01:58 <ikke> cmd: is implemented as a provides vakue
2023-10-18 09:02:01 <ikke> Value
2023-10-18 09:02:57 <ikke> So you ask apk to install something that provides the 'cmd:grep' virtual package 
2023-10-18 09:03:27 <ikke> And as Habbie implied, it's abuild that automatically adds those provides to packages 
2023-10-18 09:06:11 <witcher> thanks ikke 
2023-10-18 09:09:24 <ikke> apk info -P <pkg> shows you what each package provides 
2023-10-18 09:10:42 <ikke> And if you check /etc/apk/world, you'll see that apk just literally adds cmd:grep, just like you installed another package 
2023-10-18 09:23:08 <frojnd> I installed chromium and after running it, I get: FATAL:credentials.cc(111)] Check failed: . : Invalid argument (22) :FATAL:zygote_host_impl_linux.cc(215)] Check failed: . : No such file or directory (2)
2023-10-18 09:23:23 <frojnd> Maybe I'm missing something obvious. It's recent alpine with sway installation
2023-10-18 09:56:11 <frojnd> So as a workaround I
2023-10-18 09:56:21 <frojnd> I'm trying to install flatpak https://wiki.alpinelinux.org/wiki/Flatpak under sway
2023-10-18 09:57:02 <frojnd> It says to start dbus session explicitly if running wm from .xinitrc (well I'm running it without .xinitrc) like this: dbus-run-session -- sway
2023-10-18 09:57:08 <frojnd> So this should be covered already
2023-10-18 09:57:51 <frojnd> So after adding remote repo: flatpak remote-add --user --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo and trying to search for app: I  get error: While opening repository /var/lib/flatpak/repo: opening repo: opendir(tmp): No such file or directory
2023-10-18 09:58:12 <Ermine> Create that dir lol
2023-10-18 09:58:43 <frojnd> Ermine: I've created that dir
2023-10-18 09:59:22 <frojnd> drwxr-sr-x    3 root     flatpak          3 Oct 18 11:48 repo Maybe it does'nt like root
2023-10-18 10:00:22 <frojnd> https://termbin.com/fp4g doesn't help
2023-10-18 10:00:58 <frojnd> Maybe I have to add /var/lib/flatpak to XDG_DATA_DIR
2023-10-18 10:01:52 <frojnd> Still no good
2023-10-18 10:01:54 <Ermine> You can try strace'ing it and look for ENOENT stuff
2023-10-18 10:01:58 <frojnd> After sourcing ~/.profile
2023-10-18 10:04:58 <frojnd> open("/etc/ld-musl-x86_64.path", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
2023-10-18 10:05:01 <frojnd> And list goes on
2023-10-18 10:05:24 <Ermine> try looking for the last one
2023-10-18 10:06:06 <frojnd> openat(5, "tmp", O_RDONLY|O_NOCTTY|O_NONBLOCK|O_LARGEFILE|O_CLOEXEC|O_DIRECTORY) = -1 ENOENT (No such file or directory)
2023-10-18 10:06:48 <frojnd> Also around Permission denied
2023-10-18 10:07:04 <frojnd> https://termbin.com/qx96
2023-10-18 10:08:39 <frojnd> Oh
2023-10-18 10:08:46 <frojnd> Maybe I have to add myself to flatpak group
2023-10-18 10:11:53 <frojnd> Nope that didn't help either
2023-10-18 10:33:26 <Ermine> create tmp dir in repo dir
2023-10-18 10:40:04 <frojnd> Still no good hm hm: https://termbin.com/8u2wu
2023-10-18 10:43:25 <Ermine> Seems like it doesn't want to create its own structure
2023-10-18 11:07:18 <frojnd> Maybe even first command isn't working as it should? flatpak remote-add --user --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo
2023-10-18 11:08:57 <frojnd> First command didn't work untill I added export XDG_DATA_DIR=/var/lib/flatpak/exports/share:/home/frojnd/.local/share/flatpak/exports/share:/var/lib/flatpak/ to ~/.profile
2023-10-18 11:55:19 <skyenosaur> hello i would like to try out alpine but i can’t because i get stuck at keymaps. i managed to unearth a command to convert a .map into the binary format that loadkmap wants but it doesn’t work
2023-10-18 11:56:19 <skyenosaur> it is really no fun trying to wrangle a new system when i have to search every key on the keyboard and can’t type any of my passwords because the characters aren’t on it
2023-10-18 12:42:31 <frojnd> skyenosaur: setup-keymap
2023-10-18 12:53:55 <fluix> frojnd: do you mean XDG_DATA_DIR*S*?
2023-10-18 13:00:20 <skyenosaur> frojnd: my keymap is not in there
2023-10-18 13:01:57 <frojnd> fluix: no. "XDG_DATA_DIR" Should it be "XDG_DATA_DIRS" ?
2023-10-18 13:02:13 <frojnd> currently in ~/.profile I have: export XDG_DATA_DIR=/var/lib/flatpak/exports/share:/home/frojnd/.local/share/flatpak/exports/share:/var/lib/flatpak/
2023-10-18 13:02:34 <frojnd> Arr le me try to add S
2023-10-18 13:04:50 <frojnd> My ~/.profile looks like this now: https://termbin.com/73y5h I then sourced ~/.profile like this: source ~/.profile and still flatpak search chromium gives me: error: While opening repository /var/lib/flatpak/repo: opening repo: opendir(objects): No such file or directory
2023-10-18 13:05:06 <frojnd> skyenosaur: sorry don't know how to help you with that one
2023-10-18 13:15:56 <fluix> XDG_DATA_DIR isn't a thing (https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html) and you have multiple paths in there, hence the plural
2023-10-18 13:17:00 <fluix> not saying this is the actual issue for you or anything, I just noticed the multiple paths and thought "that's strange" so looked it up quickly
2023-10-18 13:18:57 <frojnd> fluix: yeeah I fixed it nevertheless
2023-10-18 13:19:22 <frojnd> But issue still persists... I'm gonna remove /var/lib/flatpak and remove package and try to isntall it again
2023-10-18 13:21:22 <frojnd> Oh that's interesting. apk add flatpak after purging I see this in installation logs: * To enable access to XDG directories (e.g. ~/Desktop, ~/Downloads, ...) in Flatpak apps, run xdg-user-dirs-update (from xdg-user-dirs package) as a regular user that will use Flatpak.
2023-10-18 13:21:45 <frojnd> So I just run xdg-user-dirs-update first
2023-10-18 13:21:47 <frojnd> ?
2023-10-18 13:21:58 <frojnd> Nah
2023-10-18 13:23:25 <frojnd> Not sure what that means actually
2023-10-18 13:54:02 <frojnd> I think apk del -r flatpak installed a few more packages that I intended to
2023-10-18 13:54:55 <frojnd> My /var/log is empty. That's trange
2023-10-18 15:19:44 <skyenosaur> okay i figured out that you have to transform the keymap into binary format with loadkeys -bu, and then it took me forever to find out where on earth you put that setting to make me permanent, because there is NO documentation
2023-10-18 15:20:18 <skyenosaur> *make it permanent, not me. i mean i hope i am permanent too but my brain just melted i think
2023-10-18 15:21:07 <ikke> 
ACTION makes skyenosaur permanent

2023-10-18 15:21:21 <skyenosaur> appreciate it
2023-10-18 15:22:22 <ikke> sadly I don't know the answer either
2023-10-18 15:26:10 <skyenosaur> i did find out where to change it, it’s in /etc/conf.d/loadkmap, but the binary file needs to also be gzipped for some reason, i was just complaining that i am apparently the first person who tried to do that because there’s not a single word in the wiki
2023-10-18 15:26:33 <skyenosaur> which i am probably going to change later today unless someone stops me
2023-10-18 16:08:19 <rtyler> have any of you folks done much work with nbd-server and nbd-client for mounting block devices over the network. I've got this netbooted alpine host and after mounting it all I/O operations seem to block infinitely which has me perplexed
2023-10-18 16:14:06 <qaqland> how can I commit to aport anonymously
2023-10-18 16:15:24 <ikke> Pseudonomysly does not suffice?
2023-10-18 16:48:03 <qaqland> thanku, maybe gitlab can also hide my email
2023-10-18 17:17:14 <rtyler> what is quite interesting about this is that a `blockdev --getsize64 /dev/nbd0` command does return the size of the sparse file exported by the nbd-server. Writes hang for an incredibly long time until an I/O error occurs, which makes me wonder if there's some firewall issues here, or something the server side log isn't telling me
2023-10-18 17:21:21 <ikke> qaqland: the commit itself also has an e-mail adres, but there is nothing enforcing it to be an existing address (it should just look like one)
2023-10-18 17:45:19 <krzno[m]> Hi,... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/HQPlLJYoEEOGJESBKmvhweER>)
2023-10-18 17:45:53 <krzno[m]> * Hi,... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/PvLpkzFvXILIKwpnptqzHMIe>)
2023-10-18 17:46:36 <lopid> yeah sure
2023-10-18 17:50:10 <rpi-user> Hi, I noticed that the Alpine Linux download for Raspberry Pi does not include the iwd package. Is it possible to somehow add this to sdcard after extracting the Alpine Linux installation?
2023-10-18 18:05:45 <krzno[m]> * Hi,... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/ebZQLmyjIPUxNpBJftaeCExI>)
2023-10-18 18:06:02 <lopid> maybe if you actually posted your message…
2023-10-18 18:10:40 <quinq> It'll remain a mystery
2023-10-18 18:14:07 <krzno[m]> <lopid> "maybe if you actually posted..." <- Aee you perhaps talking to me?
2023-10-18 18:14:56 <lopid> yes, you appear to be posting a link to your message rather than the message itself
2023-10-18 18:16:11 <krzno[m]> What do you mean? Can you screenshot please?
2023-10-18 18:17:46 <lopid> https://imgur.com/Tzqdbc2.png
2023-10-18 18:19:25 <krzno[m]> Thanks. That's strane. I sent normal message.
2023-10-18 18:19:51 <krzno[m]> s/strane/strange/
2023-10-18 18:23:15 <minimal> that's how the IRC-Matrix gateway works when a Matrix user posts a long message
2023-10-18 18:24:00 <ikke> More like a multi-line message 
2023-10-18 18:24:09 <ikke> Just send messages as a single line 
2023-10-18 18:27:49 <consus> irc at it's best
2023-10-18 18:31:09 <quinq> Yeah, using simple tools teaches one how to communicate efficiently
2023-10-18 18:31:11 <Inga> Hi folks, I think I remember that I've seen the button to flag packages as outdated somewhere on pkgs.alpinelinux.org before, but I cannot find it now. How do I flag / report package as outdated?
2023-10-18 18:33:38 <minimal> Inga: you view the inof of that package (for any arch) and click the red "Flag out of date" button?
2023-10-18 18:33:43 <minimal> s/inof/info/
2023-10-18 18:35:53 <Inga> minimal: I don't get it, I go to e.g. https://pkgs.alpinelinux.org/package/v3.18/community/x86_64/chromium (just as an example) and there are no buttons anywhere on this page. Or do you mean some other info page?
2023-10-18 18:37:01 <Inga> ah OK I figured it out, there is "Flag out of date" button on package info pages in edge, but not in stable releases
2023-10-18 18:37:11 <minimal> hnn, maybe it's only on Edge packages that the button appears
2023-10-18 18:37:23 <Inga> how do I report that the package in release received a security patch then?
2023-10-18 18:37:46 <minimal> open an Issue?
2023-10-18 18:37:50 <quinq> You email the maintainer
2023-10-18 18:44:38 <Inga> minimal: thank you, just did that
2023-10-18 18:54:45 <elagost> Is it acceptable/possible to have a package in community optionally depend on something in testing? I'd like to get the 'kanshictl' portion of kanshi working but that requires 'libvarlink' which isn't in alpine (yet)
2023-10-18 19:02:08 <ikke> No, that's not possible
2023-10-18 19:05:49 <elagost> Ok. So it'd have to go into testing, then community after some amount of time if it's acceptable?
2023-10-18 19:31:15 <ikke> elagost: dependencies can in these cases be directly added in community (add it in the same MR as adding the dep to the package)
2023-10-18 19:35:29 <elagost> thanks, I'll do that.
2023-10-18 21:52:08 <jaz> hello o/
2023-10-18 21:52:30 <jaz> i got the following question, maybe somebody can enlighten me:
2023-10-18 21:54:17 <jaz> so far when i did release upgrades i incremented the minor version in /etc/apk/repositories, apk update & apk upgrade and reboot... can i skip minor versions? like, lets say from 3.14 straight to 3.18? do i need to reboot? 
2023-10-18 21:55:09 <jaz> also, nice irc channel :) reminds of of xkcd.com/1782/
2023-10-18 22:37:22 <Saijin_Naib[m]> I have only jumped two releases, but I just used latest-stable as my branch 
2023-10-18 22:37:44 <Saijin_Naib[m]> Reboot needed for kernel and a few other things, yeah 
2023-10-18 22:41:02 <jaz> ty 
2023-10-18 22:43:59 <dari> jaz, if you upgrading to another release then better use: upgrade -a 
2023-10-18 22:45:43 <dari> "The --available switch is used to force all packages to be upgraded, even if they have the same version numbers. Sometimes changes in uClibc require doing this."
2023-10-18 22:46:15 <Saijin_Naib[m]> Yep 
2023-10-18 22:47:10 <Saijin_Naib[m]> I have found stepwise doing apk upgrade, then apk upgrade -a, then apk upgrade -al to get everything without too many stoppages I need to resolve 
2023-10-18 22:52:26 <jaz> ty, i guess i'll have to read up on -a and -l  and see what it does and how it works, thanks a bunch
2023-10-18 22:53:24 <Saijin_Naib[m]> Alpine/apk is pretty much bulletproof, so you should get there eventually 
2023-10-18 22:55:01 <jaz> i am so happy this distro exists, i dont enjoy dealing with bloat, systemd and whatnot... alpine is fantastic 
2023-10-18 22:56:32 <jaz> and apk seems pretty robust, its great
2023-10-18 23:02:12 <jaz> anyway, i gotta go, thank you, bye bye o/
2023-10-19 09:31:43 <grawity> now that I've built my first .apk, how do I actually install it? (on another system)
2023-10-19 09:32:16 <grawity> doesn't look like `apk add` takes URLs
2023-10-19 09:32:23 <grawity> oh, d'oh, it does take local paths at least
2023-10-19 09:32:27 <aron> grawity:  apk add path/to/blah.apk
2023-10-19 09:32:48 <aron> optionally  --allow-untrusted , if your other system does not have your keys
2023-10-19 09:48:38 <grawity> aha, looks like I can `apk add -X https://foo ms-sys`
2023-10-19 12:35:51 <hograthm> I have a question related to Docker instalation on Alpine 3.18. I need to setup insecure registry and for some reason default config in /etc/docker/daemon.json doesn't work, so seems like config is in another place. There is no info about it on wiki.alpinelinux.org, archwiki or official docker wiki (at least I wasn't able to find). If anyone here familiar with this please help :)
2023-10-19 12:37:12 <ikke> hograthm: /etc/docker/daemon.json should work
2023-10-19 13:18:04 <hograthm> ikke: Yes, the problem was related to docker daemon not restart properly due to some files still in use. After proper restart docker was able to read config from default location. Thanks a lot!
2023-10-19 13:18:15 <ikke> no problem
2023-10-19 16:29:35 <lonjil> is apk's commit hook support documented anywhere? I couldn't find anything.
2023-10-19 16:31:14 <ptrc> lonjil: doesn't seem to be, but it's pretty simple: place an executable into /etc/apk/commit_hooks.d/, first parameter you get is either `pre-commit` or `post-commit`
2023-10-19 16:32:29 <lonjil> ptrc: thank you. I saw the same thing in a random blog post, but I wasn't sure if there was more to it or not. (e.g. additional parameters perhaps). 
2023-10-19 16:32:52 <ptrc> tbf it was probably my blog post
2023-10-19 16:33:06 <ptrc> or well, mine was the only one last time i checked
2023-10-19 16:33:22 <lonjil> so it seems 
2023-10-19 16:33:29 <ptrc> anyway, yeah, there wasn't much to it last time i checked the source
2023-10-19 16:33:35 <ikke> Just curious, what's the use case? Custom trigger logic?
2023-10-19 16:34:30 <lonjil> I'm going to make it do a zfs snapshot so I can rollback in case of a bad upgrade.
2023-10-19 16:34:46 <ikke> aha
2023-10-19 19:06:10 <Saijin_Naib[m]> Within I want to say the last 6mo, I recall reading discussions of some fuckery with Google around Go and the Go modules CDN
2023-10-19 19:06:31 <Saijin_Naib[m]> Does anyone recall? I apparently did not internalize that very well 
2023-10-19 19:07:20 <ikke> I guess you refer to googls go proxies ddossing sr.ht
2023-10-19 19:08:51 <fluix> https://sourcehut.org/blog/2023-01-09-gomodulemirror/
2023-10-19 19:15:26 <Saijin_Naib[m]> Oooh, I think that looks right. Thanks! 
2023-10-19 19:17:29 <Saijin_Naib[m]> Trying to evaluate software, and top candidate is Go based, and that was kicking around in my memory as a possible negative for Go ecosystem 
2023-10-19 19:19:47 <invoked> alternative is hermetic builds
2023-10-19 19:20:12 <invoked> but i mean, i guess it depends on what software you're choosing and what they're doing
2023-10-19 19:20:47 <invoked> if it's your own stuff, then do what you want. you don't have to use modules
2023-10-19 19:21:08 <Saijin_Naib[m]> I believe Ariadne just tooted about not using Bazel, but I don't get why 
2023-10-19 19:21:49 <Saijin_Naib[m]> invoked: Looking to dump WordPress site and move to static site toolkit, and Hugo comes very highly recommended 
2023-10-19 19:22:08 <Saijin_Naib[m]> Just trying to research and advise 
2023-10-19 19:22:19 <invoked> bazel is just a tool, hermetic builds are a generic term
2023-10-19 19:22:25 <invoked> aka vendoring
2023-10-19 19:22:36 <Saijin_Naib[m]> Oooh, okay. I need to dig more, then 
2023-10-19 19:23:53 <invoked> take a step back. what are you trying to achieve?
2023-10-19 19:24:08 <Lord> some months ago there was blog posts about building a new init/service system. Is there any news somewhere ?
2023-10-19 19:25:33 <invoked> Lord: it's not alpine specific.  see: s6
2023-10-19 19:25:55 <invoked> there is no timeline afaik about moving alpine to s6
2023-10-19 19:28:28 <ikke> skarnet is still working on improving the s6 suite to be a good replacement for openrc
2023-10-19 19:29:35 <Lord> ok ok
2023-10-19 19:30:04 <Lord> i see there already are many s6-* packages
2023-10-19 19:31:46 <ikke> It's not about the amount of packages
2023-10-19 19:32:29 <Lord> i know i know, but i didn't know the project was already packaged and probably usable already
2023-10-19 19:32:47 <Lord> i thought it was a very long term project with nothing yet
2023-10-19 19:33:06 <ikke> This is old, but still should give you the idea: http://skarnet.com/projects/service-manager.html#plan
2023-10-19 19:40:31 <Ariadne> Saijin_Naib[m]: i am dealing with packaging something that uses bazel and it’s rather unpleasant :)))
2023-10-19 19:41:26 <frojnd> I'm doing another sway installation on alpine on t14 and for some reason I get: wlr Atomic commit failed: Permission denied.... I've installed intel drivers and started dbus at boot
2023-10-19 19:42:10 <frojnd> I'm literally following the wiki for sway
2023-10-19 19:44:11 <invoked> you get that error after doing what, exactly?
2023-10-19 19:46:13 <frojnd> staring sway. I just see background and can't start foot (installed) can't see bemenu (installed configured)
2023-10-19 19:46:47 <invoked> how are you starting sway?  by typing "sway" as your user account?
2023-10-19 19:47:07 <Lord> this blogpost lacks a date somewhere
2023-10-19 19:47:17 <frojnd> I type dbus-run-session -- sway (After I login myself)
2023-10-19 19:47:58 <invoked> you are probably missing your tmp directory for dbus
2023-10-19 19:48:33 <sewn> frojnd: are you in the correct groups and setup mdev properly?
2023-10-19 19:50:09 <frojnd> invoked: you mean XDG_RUNTIME_DIR ?
2023-10-19 19:50:10 <invoked> oh wait you said you run the dbus service.  you shouldn't use dbus-run-session then.
2023-10-19 19:50:20 <Saijin_Naib[m]> <invoked> "take a step back. what are you..." <- Increase privacy of visitors to our site, increase security of our site, lower maintenance and cognitive load of managing our site. Lighter and faster pages to better serve folks on limited devices and connections also good 
2023-10-19 19:50:38 <frojnd> invoked: I've started dbus session AFTER I saw that error
2023-10-19 19:50:55 <frojnd> invoked: even if I just type sway it's the same error
2023-10-19 19:51:02 <Saijin_Naib[m]> <Ariadne> "Saijin_Naib: i am dealing with..." <- Ahhh, okay, thank you. Didn't know if it was more fundamental or deep of an issue 
2023-10-19 19:51:12 <frojnd> swapgs: like video and input?
2023-10-19 19:51:20 <frojnd> sewn: like video and input?
2023-10-19 19:51:36 <frojnd> swapgs: mdev?
2023-10-19 19:51:42 <frojnd> Agh
2023-10-19 19:51:45 <frojnd> sewn: mdev?
2023-10-19 19:52:32 <invoked> Saijin_Naib[m]: i mean, you should only care about the gomodule issue to the extent that it matters to you as a programmer contributing back to hugo, if you plan to do that.
2023-10-19 19:52:59 <invoked> modules themselves aren't controversial, it's how google went about it that was.
2023-10-19 19:53:34 <Saijin_Naib[m]> invoked: We likely would not be contributing code there, so that's good to know. 
2023-10-19 19:53:58 <ikke> And in the end, it was resolved in a satisfactory way for ddevault
2023-10-19 19:56:46 <Saijin_Naib[m]> Awesome. Sorry, I'm not super knowledgeable in this area, but I'm fortunate to be a fly on the wall where folks are, so just trying to upstream that knowledge to decision makers
2023-10-19 19:57:07 <Saijin_Naib[m]> So, thank you too everyone, again, for educating me 
2023-10-19 19:57:08 <frojnd> sewn: mdev is configured 
2023-10-19 19:58:35 <Saijin_Naib[m]> My plan is to pay it back to Alpine 😎 just gotta get some more ducks in rows and up my packaging game 
2023-10-19 19:59:32 <frojnd> I mean I've ran setup # setup-devd mdev
2023-10-19 20:00:25 <frojnd> Now sway won't start at all :) I see more errors. libinput initialization failed, no input devices, Failed to initialize backend, wl)display_roundtrip failed
2023-10-19 20:02:11 <frojnd> It's because wiki says I should configure udev
2023-10-19 20:02:19 <frojnd> So now I have two and they are in clash probably
2023-10-19 20:07:30 <sewn> for sway yeah i did have to use udev
2023-10-19 20:07:43 <sewn> i am using dwl and i dont need udev there for some odd reason
2023-10-19 20:23:58 <skyenosaur> hey alpiners, i am having a problem with keymaps and dmcrypt.  i cannot for the life of me figure out how to make openrc load my keymap before asking me for the password and this is somewhat of a problem
2023-10-19 20:26:19 <skyenosaur> the script to mount file systems does not to its job when i put the dmcrypt in the default instead of boot, but the keymap somehow does not get loaded at boot, but also i dont exactly know how the keymap stuff works because i had to do horrible things to add my keymap to the system at all
2023-10-19 20:26:23 <fluix> do you have keymap in /etc/mkinitfs.conf
2023-10-19 20:27:44 <skyenosaur> no, but i can look up how to do that
2023-10-19 20:29:21 <fluix> you add "keymap" to the "features" list there
2023-10-19 20:29:44 <fluix> I think order matters so put it at least before cryptsetup
2023-10-19 20:31:02 <ikke> skyenosaur: do you mean the password for your linux account, or for disk encryption?
2023-10-19 20:32:13 <fluix> oh yeah ^ is important
2023-10-19 20:32:17 <skyenosaur> ikke: the luks password for my home partition
2023-10-19 20:34:01 <skyenosaur> i don't have full disk encryption, i am just struggling with openrc and the order of these scripts a little. and the root of the problem is that i can't just use setup-keymap because my keymap is not in there
2023-10-19 20:40:52 <skyenosaur> i did try adding loadkmap to runlevel sysinit but that just results on it not mounting the partition because the whole order gets messed up
2023-10-19 22:21:43 <minimal> skyenosaur: that's because the dmcrypt init.d defines "before fsck" and the "loadkmap" init.d defines "need localmount" and the "loadmount" int.d defines "need fsck"
2023-10-20 07:21:33 <hronro> Hi
2023-10-20 07:24:11 <hronro> Does anyone know when Alpine 3.19 will be released?
2023-10-20 07:25:16 <ikke> Somewhere next month or so
2023-10-20 07:27:57 <hronro> Which kernel will 3.19 use? 6.1 or 6.5?
2023-10-20 07:32:44 <tzeentch[m]> <hronro> "Does anyone know when Alpine 3.1..." <- which zfs zfs 2.1 or 2.2
2023-10-20 07:32:49 <tzeentch[m]> s//,/
2023-10-20 07:38:41 <ikke> Most likely 6.1
2023-10-20 10:23:33 <CC0422> I have an Arm SBC running Alpine in diskless mode and for some reason it keeps adding swclock to the boot runlevel, completely on its own, every time it boots. I already have hwclock added to boot, so the clock dependency should be already satisfied.
2023-10-20 11:39:57 <singham> Has anyone managed to get perdition running
2023-10-20 11:40:33 <singham> How does it function? It interacts directly with mailbox directory?
2023-10-20 13:41:37 <invoked> https://projects.horms.net/projects/perdition/docs.shtml
2023-10-20 15:39:33 <jrcresawn> Hello. I need help with OpenRC dependencies. I have an init script named openconnect.CORP (VPN client) that works perfectly when called with "sudo rc-service openconnect.CORP start". I added openconnect.CORP to the default runlevel, but it fails to start at boot time and I'm not sure why. Obviously openconnect.CORP needs the network to be up so I suspect that I may need to configure it as a 
2023-10-20 15:39:40 <jrcresawn> dependency. I added "rc_need=net" to /etc/conf.d/openconnect.CORP but that didn't help. I also added "need net" to the "depend()" function in /etc/init.d/openconnect.CORP, but that didn't help. I'm stuck and I'd appreciate help. Thanks in advance.
2023-10-20 16:18:48 <Kladky> Hello, I recently upgraded my system by running apk upgrade, but while kde plasma packages were being replaced with plasma 5 packages, I noticed the following error: (178/357) Installing kactivities5-libs (5.111.0-r0)
2023-10-20 16:19:00 <Kladky> ERROR: kactivities5-libs-5.111.0-r0: trying to overwrite usr/lib/libKF5Activities.so.5 owned by kactivities-libs-5.110.0-r0
2023-10-20 16:19:37 <Kladky> Now whenever running apk upgrade, I am told there was 1 error.
2023-10-20 16:19:55 <Kladky> Is this something to be concerned by?
2023-10-20 16:20:21 <Kladky> And if it is, how do I fix this?
2023-10-20 16:22:51 <Saijin_Naib[m]> Have you tried running apk fix? 
2023-10-20 16:23:23 <Saijin_Naib[m]> Or, apk update && apk upgrade && apk upgrade -a && apk upgrade -al
2023-10-20 16:23:36 <Kladky> Just ran apk fix, and it worked, thanks!
2023-10-20 16:23:41 <Saijin_Naib[m]> Very very infrequently, you may have to remove the offending package and re-add
2023-10-20 16:25:17 <Kladky> apk fix re-installed the package, I tried removing and re-adding manually before but apk warned me that it would remove all of plasma.
2023-10-20 16:45:54 <ptrc> PureTryOut: ^
2023-10-20 16:46:01 <ptrc> missing a `replaces=` probably?
2023-10-20 16:49:11 <PureTryOut> Oh yeah subpackages sadly need their own replaces
2023-10-20 17:55:03 <linfan> if i want to install a pkg from testing, do i have to enable the repo or is it possible to install it manually after download?
2023-10-20 17:55:40 <linfan> that is to say, is local install of .apk possible
2023-10-20 17:55:50 <ikke> yes, but you might be missing dependencies
2023-10-20 17:56:23 <linfan> so:libc.musl-aarch64.so.1 (no such package)
2023-10-20 17:56:50 <ikke> are you on an aarch64 system?
2023-10-20 17:57:02 <linfan> no, shit, lol
2023-10-20 17:57:28 <linfan> must have gotten the wrong file, i am on 86_64 edge
2023-10-20 17:58:09 <linfan> and i was thinking if maybe gst-rtsp-server might fix radio listening
2023-10-20 17:59:10 <linfan> but maybe alpine is not for desktop, haven't yet figured out who or what it is meant for
2023-10-20 17:59:31 <ikke> It's mainly a server OS, but plenty of people are running it on the desktop for all kinds of purposes
2023-10-20 17:59:59 <ikke> and even on phones (through postmarket os)
2023-10-20 18:00:13 <donoban> it's the year of alpine... :D
2023-10-20 18:00:20 <linfan> and i run it on vbox, and a lot of stuff is not working flawlessly, a lot of freezes and stuff
2023-10-20 18:00:33 <ikke> Try qemu
2023-10-20 18:00:37 <donoban> what kind of stuf?
2023-10-20 18:01:22 <linfan> freezes, sound, malfunctioning gnome, which is a devil by itself
2023-10-20 18:02:11 <donoban> well gnome will not probably run smoothly if you don't have harwdware 3d acceleration
2023-10-20 18:02:23 <donoban> no idea how good it's supported on vbox
2023-10-20 18:02:26 <linfan> every other log in fails and i need to cold reboot
2023-10-20 18:02:53 <donoban> that could be some bug
2023-10-20 18:02:54 <linfan> screen tearing and freeze if i use flatpak etc. etc.
2023-10-20 18:02:56 <donoban> do you have ssh access?
2023-10-20 18:03:06 <linfan> openssh
2023-10-20 18:03:41 <donoban> maybe you can check what happens if it's not fully freeze/panic
2023-10-20 18:04:16 <minimal> linfan: did you do stuff like check the "Enable 3D Acceleration" Display setting for your VM?
2023-10-20 18:04:18 <linfan> if i have the patience to wait it out, yes
2023-10-20 18:04:52 <linfan> i have enabled 3d yes, maybe i should not since i don't have 3d glasses :-)
2023-10-20 18:05:31 <minimal> Also have you tried different VBox graphics controllers?
2023-10-20 18:06:56 <minimal> 3D acceleration is nothing to do with 3D glasses
2023-10-20 18:07:12 <linfan> but the last 2 weeks i have almost constantly non-stop scanned web pages which has been aweful. there si so much inaccurate information and even alpine wike does not apply to latest version
2023-10-20 18:08:22 <minimal> I don't think many Alpine users make use of Virtualbox for Alpine graphical VMs
2023-10-20 18:08:27 <linfan> i tried on vmplayer and it ran smoothly but do not like the interface of it, it would like me to buy it
2023-10-20 18:08:47 <linfan> vmware player*
2023-10-20 18:09:14 <linfan> quemu is awefully complicated to set up on win 11
2023-10-20 18:09:29 <minimal> Hyper-V then?
2023-10-20 18:09:57 <linfan> hyper-v is enabled in bios
2023-10-20 18:10:09 <linfan> otherwise i could not run vbox
2023-10-20 18:10:15 <minimal> I mean use Hyper-V rather than VirtualBox?
2023-10-20 18:10:23 <ikke> But then you need to make sure that vbox uses it
2023-10-20 18:11:29 <ikke> System > Acceleration > Paravirtualization interface
2023-10-20 18:11:35 <ikke> make sure it's set to Hyper-V
2023-10-20 18:11:38 <linfan> i do not have a clear picture of those things, i am not techy, only used linux since 98, probably tried douzens of distros and know which i like and don't like
2023-10-20 18:11:55 <minimal> Can't you just create an Alpine hyper-v VM from the Hyper-V manager?
2023-10-20 18:12:13 <linfan> dunno, first have to study it so i know what i am doing
2023-10-20 18:12:46 <linfan> and besides, both arch linux and slackwar, which i currently have on vbox don't have any problems
2023-10-20 18:13:47 <linfan> slack runs plasma and arch xorg gnome, coz i hate gdm, want to land in command line when booting up
2023-10-20 18:14:37 <linfan> alpine does not seem to give me the freedom I's like, but still I like the idea of alpine
2023-10-20 18:16:09 <linfan> customizing alpine is a pain, especially sound, no default setting in /etc and whatever changes done in pulse.conf files make no difference
2023-10-20 18:16:56 <linfan> they are not linked to anything that can be modified
2023-10-20 18:17:46 <linfan> but ok, will look what hyper-v is, if it is a program, why not give it a spin
2023-10-20 18:19:54 <ikke> It's a hypervizor, one that comes by default with windows (except for the consumer version)
2023-10-20 18:19:57 <linfan> oh, it is that microsoft thing, once tride
2023-10-20 18:20:19 <linfan> i have home edition
2023-10-20 18:21:55 <linfan> hyper - greek over, above vizor prob. latin vider to view, see, i.e. something to view on top of something else
2023-10-20 18:22:22 <ikke> A hypervizor is something that runs virtual machines
2023-10-20 18:22:46 <linfan> i know, just kidding
2023-10-20 18:23:03 <linfan> or ridiculing the name
2023-10-20 18:23:44 <linfan> ok, tx for the input, but strange, though, that arch and slack work ;-)
2023-10-20 18:24:30 <minimal> linfan: not necessarily strange, e.g. arch and slack are glibc-based, Alpine is musl-based,
2023-10-20 18:24:43 <ikke> minimal: they left
2023-10-20 18:25:01 <minimal> ikke: as I was typing lol
2023-10-20 18:31:35 <dwfreed> ikke: consumer windows 10+ comes with hyper-v, it's just not exposed :D
2023-10-20 18:31:45 <dwfreed> (wsl2 is just a hyper-v VM)
2023-10-20 18:32:43 <Ermine> Why running alpine if it's so annoying
2023-10-20 18:33:24 <minimal> I was just thinking that the Alpine hyperv_drm kernel module might be more performant than the virtualbox (well actually VMware) module
2023-10-20 18:36:39 <Hello71> qemu supports hyperv
2023-10-20 18:54:19 <Saijin_Naib[m]> I was going to say AlpineWSL and X410 is how I started, and it worked great 
2023-10-20 18:54:35 <Saijin_Naib[m]> Pretty much seamless
2023-10-20 19:02:07 <jrcresawn> I posted a message about four hours ago and no one replied. Is it rude to repost it?
2023-10-20 19:02:42 <ikke> At least I can look back in my history
2023-10-20 19:03:04 <dwfreed> need net doesn't mean the network's going to actually be functional by that point
2023-10-20 19:03:51 <jrcresawn> Thanks dwfreed. Is rc_need a better solution?
2023-10-20 19:03:56 <dwfreed> they do the same thing
2023-10-20 19:04:05 <dwfreed> one just doesn't require editing the initscript
2023-10-20 19:04:21 <dwfreed> (which should generally be avoided anyway)
2023-10-20 19:05:27 <jrcresawn> Can you recommend a way to solve this problem?
2023-10-20 19:06:08 <dwfreed> you could make a "vpn-online" initscript that waits for the vpn server to be reachable, and then make openconnect depend on that
2023-10-20 19:07:05 <invoked> late to the party, but what was linfan even trying to achieve?  "arch and slack work fine"  ok, then use those.
2023-10-20 19:07:21 <ikke> invoked: they wanted to get alpine working, which is a valid question
2023-10-20 19:07:43 <invoked> it's topical but doesn't answer what the goal is
2023-10-20 19:09:14 <jrcresawn> dwfreed: Do you propose the "vpn-online" init script have a ping/wait loop or something similar?
2023-10-20 19:09:24 <fluix> there's some more documentation on net stuff here: https://github.com/OpenRC/openrc/blob/master/service-script-guide.md#be-wary-of-need-net-dependencies
2023-10-20 19:09:27 <dwfreed> jrcresawn: yeah
2023-10-20 19:10:54 <jrcresawn> dwfreed: Ok got it. So if the VPN server responds to a ping then run "rc-service openconnect.CORP start". Right?
2023-10-20 19:11:32 <dwfreed> no, just depend on vpn-online, and then openconnect.CORP won't get started until vpn-online "starts"
2023-10-20 19:12:35 <dwfreed> dunno if alpine ships the net-online initscript with openrc, but you could use that for inspiration (or just instance it)
2023-10-20 19:12:51 <jrcresawn> Where would the dependency be defined?
2023-10-20 19:13:05 <dwfreed> with rc_need ?
2023-10-20 19:13:29 <jrcresawn> dwfreed: net-online is included.
2023-10-20 19:14:29 <dwfreed> then I'd probably just instance it (ln -s net-online /etc/init.d/net-online.vpn) and make /etc/conf.d/net-online.vpn filled with appropriate values, then make openconnect.CORP depend on net-online.vpn
2023-10-20 19:19:31 <jrcresawn> dwfreed: What does "provide network-online" do? It's in the depend() function of net-online.
2023-10-20 19:20:58 <dwfreed> virtual services; for generic things, like "dns resolver", multiple initscripts could do the job, and you generally don't want dependent initscripts to depend on a specific one
2023-10-20 19:21:27 <dwfreed> not useful for your VPN situation
2023-10-20 19:22:47 <jrcresawn> dwfreed: Ok thanks. I will follow your proposal.
2023-10-20 20:41:01 <jrcresawn> When I add net-online to the default runlevel and reboot the init script is listed as "stopped" by rc-status. Are there logs that might reveal why?
2023-10-20 20:57:58 <frag> is it possible to pipe html (mail) to 'links'?
2023-10-20 20:59:04 <caskd[m]> if you can extract the html part only, then yes
2023-10-20 20:59:58 <frag> caskd[m]: how?
2023-10-20 21:00:32 <caskd[m]> that depends on your email client
2023-10-20 21:01:58 <frag> i mean like 'cat file.html|links'
2023-10-20 21:02:22 <frag> guess its not possible that way tho
2023-10-20 21:08:36 <caskd[m]> i think you should be able to use the file as a parameter
2023-10-20 21:09:12 <caskd[m]> or use /dev/stdin
2023-10-20 22:44:09 <linfan> soo, upgraded to window 11 pro and installed hyper-v, but it seems it lacks the ability to use advanced settings, so no audio etc. just networking, keyboard and mouse
2023-10-20 22:51:08 <jrcresawn> I installed monit, modified /etc/monitrc, checked the syntax, and verified that it does what I need it to do. I ran "sudo rc-update monit add" and rebooted. The monit init script reports "stopped" after the system boots. Why?
2023-10-20 23:06:20 <dwfreed> does it show up in 'rc-update show' ?
2023-10-20 23:07:09 <dwfreed> or 'rc-status -a'
2023-10-20 23:08:50 <jrcresawn> dwfreed: Yes, monit is listed in the default runlevel by both commands.
2023-10-20 23:09:14 <dwfreed> what does rc-status -r say
2023-10-20 23:09:37 <jrcresawn> default
2023-10-20 23:10:34 <dwfreed> enable rc_logger in /etc/rc.conf and reboot again
2023-10-20 23:11:57 <jrcresawn> dwfreed: done. Where should I look for the output?
2023-10-20 23:13:41 <jrcresawn> dwfreed: /var/log/rc.log, but it's not there.
2023-10-20 23:15:47 <dwfreed> are you actually rebooting the right system
2023-10-20 23:17:37 <jrcresawn> dwfreed: :-) Fair question. Yes, I am. I know this because I have one VM that runs Alpine Linux that I am running with Lima (https://lima-vm.io/).
2023-10-20 23:18:25 <jrcresawn> I'm either rebooting my laptop or the VM. I'd know if I reboot my laptop.
2023-10-21 00:11:28 <Saijin_Naib[m]> Anyone using any form of Intel Arc? I just got my A380 and 64GB RAM. Anything that requires 3D accel seems to crash under linux-lts edge, but works under linux-edge edge. But, and I guess I have to diff the configs, I don't have my normal Realtek HDA under linux-edge edge, just the HDMI/DP from the Arc A380
2023-10-21 00:23:10 <Saijin_Naib[m]> Nope, that's not it. Something about the way the A380 is loaded is killing my onboard... maybe module load order? Ugh, damnit
2023-10-21 02:37:55 <Saijin_Naib[m]> -_-
2023-10-21 02:38:15 <ryuji> ^_^
2023-10-21 02:38:39 <Saijin_Naib[m]> A helpful motherboard feature where on-board gets disabled when it detects a soundcard during POST, and something about the Arc must be special since it presented as a soundcard to my UEFI, apparently, though no GPU I had prior ever had
2023-10-21 02:38:49 <Saijin_Naib[m]> And that's why I had no audio after I installed my Arc. Thank you, thank you
2023-10-21 02:54:18 <dwfreed> I mean, basically every graphics card these days includes a "soundcard"
2023-10-21 02:58:44 <dwfreed> only thought would be maybe the A380 has the soundcard first in the PCI tree ?
2023-10-21 03:00:34 <Saijin_Naib[m]> That's a possibility. My prior cards are GT 240, GTX 650 TI, and GTX 1050TI, so perhaps NVIDIA does it differently, or they're just so old it doesn't matter?
2023-10-21 03:03:58 <dwfreed> if they've got an HDMI port or DisplayPort, they have a soundcard
2023-10-21 03:04:24 <dwfreed> the GT 240 *maybe* won't, the others definitely will
2023-10-21 03:04:57 <Saijin_Naib[m]> Well, you learn something new every day, right? That was a head-scratcher haha
2023-10-21 03:05:27 <Saijin_Naib[m]> Now to see how to get this thing to actually accelerate AV1/AVIF encode
2023-10-21 03:05:27 <dwfreed> check 'lspci'; what does it list first, the actual video card or the sound card on the video card
2023-10-21 03:06:28 <Saijin_Naib[m]> If the order matters, then the VGA compatible controller first, then the Audio device for the Arc
2023-10-21 03:06:54 <dwfreed> huh, so not that then
2023-10-21 03:07:01 <dwfreed> weird
2023-10-21 03:11:01 <Saijin_Naib[m]> It's entirely possible my ASRock board doesn't handle things properly. It was cheap, and it is now like, 10 years old
2023-10-21 03:24:24 <invoked> some years back i switched to usb dacs and disable the onboard audio forever now.
2023-10-21 03:24:43 <invoked> (ymmv)
2023-10-21 03:31:32 <Saijin_Naib[m]> That's a whole rabbit hole my wallet and my space don't need 🤣
2023-10-21 03:32:10 <Saijin_Naib[m]> My friends all extoll the virtues, and one even built a tube-based USB DAC, because why not
2023-10-21 13:46:12 <knolle> what is the 'installed' file in the apk cache directory for?
2023-10-21 13:47:19 <quinq> I don't have such thing
2023-10-21 13:47:56 <knolle> probably related to setup-apkcache
2023-10-21 13:48:49 <quinq> Possibly, I don't use it indeed
2023-10-21 13:49:40 <ikke> I have it as well, not sure what it's fore
2023-10-21 13:50:36 <knolle> trying to share a cache dir between lxc containers so i don't have to download every update 10 times
2023-10-21 15:01:46 <invoked> if you work it out, that'd be an excellent wiki page. at a certain scale, i just have my own internal mirror for distfiles (as i have done with other distros)
2023-10-21 15:02:33 <invoked> that might be the way to do it even for 10.
2023-10-21 16:29:38 <ikke> can't you bind mount the cache dir from the host into the containers?
2023-10-21 16:30:24 <satmd> technically, you can
2023-10-21 16:31:50 <ikke> That sounds like a tautology
2023-10-21 16:32:36 <satmd> I checked how I solved this: I have a common directory mounted at the appropiate place in each container
2023-10-21 16:32:59 <ikke> We do a similar things for the distfiles directories on the builders
2023-10-21 16:33:04 <ikke> s/things/thing
2023-10-21 16:33:20 <satmd> I use both DISTDIR and PORTAGHGE_RO_DISTDIRS
2023-10-21 16:33:30 <satmd> -HG
2023-10-21 16:33:40 <satmd> I fatfingered
2023-10-21 17:25:55 <knolle> i just have the same cache dir bind mounted to a bunch of containers. so far it works, just don't know what that file does
2023-10-21 17:28:08 <satmd> things get complicated when UIDs don't match :D
2023-10-21 17:28:53 <ikke> files there are owned by root
2023-10-21 17:28:57 <ikke> which is generally UID 0
2023-10-21 17:29:02 <knolle> they all have the same uid for root
2023-10-21 17:31:51 <satmd> my containers use subuids and root in the container isn't the same everywhere :P
2023-10-21 17:32:10 <quinq> Everybody's allowed to shot themselves in the foot
2023-10-21 18:08:07 <singham> There is sun8i-h2-plus-orangepi-r1.dtb 
2023-10-21 18:08:27 <singham> Has someone managed to run things smooth with it on Alpine?
2023-10-21 22:06:02 <owouwu> hello, I tried to run lvresize with --resizefs today and it threw me an error, saying /sbin/fsadm: execvp failed: No such file or directory, upon googling i only found reports from 10 15 years ago, anyone knows what
2023-10-21 22:06:07 <owouwu> what's up* fs is ext4
2023-10-21 22:08:20 <Habbie> owouwu, apk add cmd:fsadm, perhaps
2023-10-21 22:10:00 <owouwu> uh Habbie i'm a silly goose, i'm too used to mksh output :< 
2023-10-21 22:10:24 <owouwu> but still, I had ran that command many times in the past and it worked, I didn't uninstall anything too
2023-10-21 22:12:54 <invoked> quinq: you can do that only so long as you still have feet.
2023-10-21 22:15:56 <quinq> ^^
2023-10-21 22:17:37 <Habbie> a corrolary to "do not look into beam with remaining eye": "do not aim footgun at remaining foot"
2023-10-21 22:18:10 <Habbie> re cache sharing between containers, i wonder if apk checks hashes when using that cache
2023-10-22 08:51:26 <quinq> meh
2023-10-22 08:51:38 <quinq> postegresql transparent upgrade strikes again
2023-10-22 08:51:50 <quinq> Any hint on how to migrate?
2023-10-22 08:52:12 <quinq> If I run: pg_upgrade -b /usr/libexec/postgresql15/ -d /var/lib/postgresql/15/data/ -D /var/lib/postgresql/16/data/
2023-10-22 08:52:21 <quinq> It tells me could not read permissions of directory "/var/lib/postgresql/16/data": Permission denied
2023-10-22 08:57:51 <vkrishn> quinq: /usr/libexec/postgresql15/ or /usr/libexec/postgresql16/ ?
2023-10-22 09:01:04 <quinq> What I pasted
2023-10-22 09:01:51 <quinq> Problem is that the data is owned by postgres user, switching to that user fixes *this* issue
2023-10-22 09:02:04 <ikke> as what user are you running this?
2023-10-22 09:02:42 <quinq> Now pg_upgrade complains that the new format requires checksums and can't upgrade.
2023-10-22 09:03:26 <vkrishn> ok
2023-10-22 09:05:15 <quinq> I'll ask an AI to do the migration
2023-10-22 09:07:55 <vkrishn> it might try to add user:AI to it first, borking the whole again
2023-10-22 09:08:38 <quinq> :D
2023-10-22 09:14:06 <quinq>  * rc-service: service `postgresql15' does not exist
2023-10-22 09:14:07 <quinq> wat
2023-10-22 09:14:20 <quinq> apk info -L postgresql15-openrc
2023-10-22 09:14:20 <quinq> postgresql15-openrc-15.4-r1 contains:
2023-10-22 09:14:26 <quinq>  
2023-10-22 09:14:27 <quinq> wat
2023-10-22 10:30:24 <username25> Hi! I'm having some trouble getting yubikeys to work on alpine as a non-root user, and am looking for troubleshooting tips. gpg --card-status and fido2/webauth works as root, but not otherwise. I am not seeing any output in the default alpine logs or with udevadm monitor and am at a bit of a loss.
2023-10-22 10:31:01 <username25> I am not running pcscd (it seems to interfere with scdaemon and browsers)
2023-10-22 10:38:29 <satmd> you will need to adjust permissions for device nodes so that your user has access to it or add yourself to the correct groups
2023-10-22 10:39:11 <satmd> see for example https://support.yubico.com/hc/en-us/articles/360013708900-Using-Your-YubiKey-with-Linux (1st hit with my search)
2023-10-22 10:39:52 <satmd> https://github.com/Yubico/libfido2/blob/main/udev/70-u2f.rules can be adapted if your device isn't yet supported
2023-10-22 10:40:28 <satmd> I'm not an alpine user, but if there's a group plugdev, you should add yourself to it, then run `udevadm control --reload-rules` and re-plug your key
2023-10-22 10:41:43 <ikke> that group does exist
2023-10-22 10:42:32 <satmd> I'm using so many different distributions that I sometimes feel unsure which thing applies where.
2023-10-22 10:43:31 <satmd> but I think, I can add 'plugdev' to my "universally applicable" list
2023-10-22 10:45:51 <alethkit> Is there unified documentation on how to do a fully-automatic install of alpine? I can only find documentation for a semi-automatic install
2023-10-22 10:47:56 <ikke> https://lists.alpinelinux.org/~alpine/devel/%3CZR7Z4SZFDqbZVXEw%40z.c.quaqua.net%3E
2023-10-22 10:50:28 <username25> satmd, thanks though there doesn't seem to be a plugdev group
2023-10-22 10:58:53 <username25> as for the device nodes, I have tried adding permissions on /dev/hidraw[0-1]. This lets the non-root user use ykman when pcscd is running, but does not change gpg/browsers. (Also the yubikey pgp applet won't work when pcscd is running)
2023-10-22 11:01:41 <ikke> username25: ftr, I don't recall having to do anything special to let my yubikey work with browsers
2023-10-22 11:01:47 <alethkit> ikke: Thank you!
2023-10-22 11:10:42 <Sofia> Huh! That was neat. So I did some `apk add` things which failed the other day (network errors). Just installed something else, and noted the other items marked the other day for install, were installed now. Was deferred !
2023-10-22 11:25:13 <alethkit> is the tiny-cloud configuration missing some parts?
2023-10-22 11:25:35 <alethkit> I don't see a place to configure networking in https://gitlab.alpinelinux.org/alpine/cloud/tiny-cloud/-/blob/main/lib/tiny-cloud/tiny-cloud.conf
2023-10-22 11:30:49 <ikke> alethkit: That's supposed to come from userdata
2023-10-22 11:32:24 <ikke> https://cloudinit.readthedocs.io/en/latest/reference/datasources/nocloud.html
2023-10-22 11:33:22 <ikke> The idea is that you have an image with tiny-cloud installed, and then use user-data to provide the per-instance details
2023-10-22 11:35:30 <alethkit> Oh, so is the only thing one is supposed to do for the image is provide the hostname?
2023-10-22 11:36:55 <ikke> no, the cloud provider, which is `nocloud` if you are self-hosting
2023-10-22 11:37:01 <WhyNotHugo> 'abuild checksum' fails for any package which has sources on github: curl: (35) Recv failure: Connection reset by peer
2023-10-22 11:37:06 <WhyNotHugo> Is this just me?
2023-10-22 11:37:15 <ikke> WhyNotHugo: Sounds like github issues
2023-10-22 11:37:30 <ikke> WhyNotHugo: I've noticed it as well
2023-10-22 11:37:32 <WhyNotHugo> https://gitlab.alpinelinux.org/WhyNotHugo/aports/-/jobs/1152421#L288
2023-10-22 11:37:36 <WhyNotHugo> Seems like a global issue.
2023-10-22 11:37:39 <ikke> yeah
2023-10-22 11:37:50 <ikke> nothing on githubstatus.com though (who would've thought)
2023-10-22 11:41:12 <quinq> “Hello, technical support, what can I do for you? Humm, a problem on the servers, no I can't see any. *sneakly reboot* Can you retry now? Ah, see it works”
2023-10-22 11:44:43 <Habbie> githubstatus.com is pretty much useless
2023-10-22 11:45:03 <ikke> yup
2023-10-22 11:45:15 <ikke> Only when they really can't deny it it will be updated
2023-10-22 11:52:22 <alethkit> So user-data.in is supposed to be https://gitlab.alpinelinux.org/alpine/cloud/tiny-cloud/-/blob/main/lib/tiny-cloud/tiny-cloud.conf ?
2023-10-22 11:52:42 <alethkit> They look like they have different formats
2023-10-22 11:53:24 <Sofia> How does one remove any packages not depended by world?
2023-10-22 11:53:44 <Sofia> Or is this already the case?
2023-10-22 11:58:43 <ikke> yes
2023-10-22 11:59:19 <ikke> apk only keeps packages that are pulled in via world
2023-10-22 11:59:27 <username25> satmd, thanks for the help with troubleshooting! The solution was indeed to create the system group 'plugdev' and add proper udev rules for hidraw `KERNEL=="hidraw*", SUBSYSTEM=="hidraw", ATTRS{idVendor}=="1050", ATTRS{idProduct}=="0113|0114|0115|0116|0120|0121|0200|0402|0403|0406|0407|0410", TAG+="uaccess", GROUP="plugdev", MODE="0660"`
2023-10-22 11:59:34 <Sofia> 
ACTION nods, thanks for confirming ikke.

2023-10-22 11:59:49 <username25> That got browser to work. gpg also required the user to be in the gnupg group.
2023-10-22 12:07:47 <alethkit> Oh, do we need to mount both the alpine install ISO alongisde the auto-install?
2023-10-22 12:07:49 <WhyNotHugo> username25: there's a package that provides the rules.
2023-10-22 12:08:43 <WhyNotHugo> username25: /lib/udev/rules.d/70-u2f.rules is owned by libfido2-1.13.0-r1
2023-10-22 12:10:57 <username25> Thanks, I seem to be missing that package.
2023-10-22 12:38:28 <vkrishn> ikke: tiny-cloud is now stable feature ?
2023-10-22 12:39:09 <vkrishn> coz, it was introduced as exprimental in v3.18.0
2023-10-22 12:39:54 <vkrishn> think i am ready with "Create bootable Alpine Linux images (Part 1)" for my wiki site :)
2023-10-22 12:53:20 <alethkit> its still experimental, I think
2023-10-22 12:53:38 <alethkit> I'm not sure if I still need a separate alpine install ISO
2023-10-22 12:56:02 <vkrishn> oh, you are not in dev channel, https://wiki.insteps.net/AlpineLinux/Create-bootable-images
2023-10-22 12:56:55 <vkrishn> i mostly do RAM install, which also works with tiny-cloud
2023-10-22 12:59:03 <alethkit> oh, thank you!
2023-10-22 16:19:18 <tommy_plugg[m]> Hello everyone in the room... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/YgbWoxQYVtGDiCovOBtKpzCq>)
2023-10-22 16:19:37 <ikke> (multi-line messages do not work on irc)
2023-10-22 16:20:07 <numono> it was spam anyway :D
2023-10-22 16:41:55 <minimal> spam we can't see? That's progress at least lol
2023-10-22 17:27:14 <_shawn> hello, is there an easy way to list the packages on the repo and compare to what I already have installed?
2023-10-22 17:28:04 <_shawn> I figured out search will show me all the packages on the repo, and info will show me all the installed packages, so I just did this (but it is monstrosity) apk info | grep '^fonts' | sort > installed; apk search | grep '^fonts' | sed -E 's/-[[:digit:]].*$//' | sort > avail; diff -Naur installed avail
2023-10-22 17:51:41 <cyclisme24[m]> _shawn you can try 'apk list --upgradeable' after an 'apk update'
2023-10-22 17:57:59 <Saijin_Naib[m]> https://pine64.com/product/rockpro64-4gb-single-board-computer/
2023-10-22 17:58:35 <Saijin_Naib[m]> I am once again asking if anyone is using this board, or if anyone with ARM knowledge sees anything that would make it a non-starter with Alpine 
2023-10-22 17:58:57 <Saijin_Naib[m]> I need to replace my commercial NAS, and I'm thinking of building an Alpine one around this for home
2023-10-22 18:02:38 <minimal> did you check their Wiki? it lists various "issues" regarding booting etc
2023-10-22 18:03:49 <invoked> arm boards are very one-off, so what's true for one arm board may not be elsewhere.
2023-10-22 18:04:05 <minimal> they do mention postmarketOS support
2023-10-22 18:05:35 <invoked> Saijin_Naib[m]: imo if you are looking to build a nas, going with some x86 (intel nuc or similar) would be a much more sure thing.
2023-10-22 18:06:01 <invoked> but if it's adventure you're after, arm is ... adventurous
2023-10-22 18:17:31 <_shawn> there we go , thanks cyclisme24[m] but this is what I was trying to do apk list -a 'fonts*' | grep -v installed
2023-10-22 18:17:42 <_shawn> thanks for taking time to answer me though
2023-10-22 18:18:37 <_shawn> list -a is available, and it says which ones are installed, unlike search which does not
2023-10-22 18:19:52 <_shawn> still seems like there should be a way to do this without filtering it, but this is close enough until I find out there is a better way
2023-10-22 19:51:53 <malmi70[m]> I tried to create a vm with KDE. I followed the wiki twice but when I reach the sddm login manager keyboard and mouse stops working. In console it works normally and the clock in sddm runs normally.
2023-10-22 20:49:19 <satrapes> hello alpinists, I saw that a selinux package exists but no entry in the wiki. Shall I assume that it is usable but no one has written the documentation yet?
2023-10-22 20:56:41 <ikke> And there is no real integration with Alpine Linux (we don't provide any policies ourselves)
2023-10-22 20:57:49 <satrapes> hi @ikke, is there a plan to do so at any point?
2023-10-22 20:58:14 <ikke> Not in the near future, it takes a lot of effort to write comprehensive policies
2023-10-22 20:59:33 <ikke> If the policies are not correct, they just become anoying and sources of issues, resulting in people disabling selinux
2023-10-22 21:00:19 <ikke> The most common advise you'd get dealing with selinux: disable it
2023-10-22 21:00:27 <ikke> selinux issues*
2023-10-22 21:01:19 <satrapes> thank you very much @ikke
2023-10-22 21:04:13 <dwfreed> and the second most common advice you'll get is audit2allow, which is about the dumbest thing in existence
2023-10-22 21:29:52 <Saijin_Naib[m]> @minimal, @invoked , no not adventures haha. Cheap and low-power and able to be built out reasonably. This was that, but last I dug into it it wasn't really listed for Alpine nor PMOS, just a few Pine recommended images 
2023-10-22 21:30:25 <Saijin_Naib[m]> I guess I'll rethink what I'm doing and stick to x86_64... Maybe a Celeron SBC from Seeed or something 
2023-10-22 21:45:38 <invoked> Saijin_Naib[m]: yeah. x86 can run reasonably low power without any of the potential arm weirdness. for something like a nas box, x86 seems like the better choice. unless maybe it was a pi4, which have all the quirks figured out.
2023-10-22 21:45:52 <invoked> or, most...
2023-10-22 21:50:58 <Saijin_Naib[m]> Yeah... I used to use Pi stuff but I'm not sure I'm pleased with their direction, so they are out, hence Pine. I assumed, ignorantly, that it would be a simple swap to just use another ARM SBC
2023-10-22 21:52:27 <invoked> oh it's never quite that simple. the pis just have a lot more eyeballs figuring out their quirks
2023-10-22 21:52:34 <invoked> every arm board has its quirks
2023-10-22 21:53:03 <invoked> some concept of "generic arm" is very elusive.
2023-10-22 22:35:41 <Saijin_Naib[m]> Yeah, I don't think I'm knowledgeable enough to go for it on my own with ARM
2023-10-22 23:02:06 <user12> i am a newbie. how can i run a command, kill it and restart it every 5 minutes?
2023-10-22 23:04:01 <dwfreed> why
2023-10-22 23:04:16 <dwfreed> https://xyproblem.info/
2023-10-22 23:06:34 <dwfreed> I guess we'll never know
2023-10-22 23:12:13 <ahmadraniri[m]> <user12> "i am a newbie. how can i run a..." <- Use cron.
2023-10-22 23:12:43 <dwfreed> they are gone
2023-10-22 23:12:53 <dwfreed> hence my comment about never knowing
2023-10-23 06:07:34 <qaqland> Can I add a README file near the APKBUILD for document?
2023-10-23 07:23:50 <frojnd> Hi there.
2023-10-23 07:24:46 <frojnd> What's the recommended apk for ssl? I think I'm lacking one since I'm not able to connect to vpn using openconnect: error:0A000152:SSL routines:final_renegotiate:unsafe legacy renegotiation disabled:ssl/statem/extensions.c:892: unable to connect to host
2023-10-23 07:25:03 <frojnd> grep ssl /etc/apk/world gives no matches
2023-10-23 07:39:19 <ikke> apk-tools itself depends on libssl3
2023-10-23 07:57:17 <dwfreed> frojnd: you would not have that message if you did not have libssl
2023-10-23 07:57:36 <dwfreed> your vpn is just ancient and busted
2023-10-23 08:00:40 <frojnd> They recommend cisco official client, but I think that client is closed source?
2023-10-23 08:01:21 <dwfreed> it is, and also wouldn't work on alpine out of the box (glibc compat *may* work, but that's a whole can of worms)
2023-10-23 08:01:22 <frojnd> ikke: grep libssl /etc/apk/world returns nothing
2023-10-23 08:01:40 <dwfreed> world only contains user-installed packages
2023-10-23 08:01:41 <dwfreed> not deps
2023-10-23 08:02:53 <frojnd> Hm how do I check if dep is installed then?
2023-10-23 08:03:36 <dwfreed> I can tell you it's installed just based on your error message, but I'd imagine there's an apk subcommand to show it
2023-10-23 08:03:42 <dwfreed> (I don't actually use alpine)
2023-10-23 08:03:54 <frojnd> apk info -l libssl3 shows some info
2023-10-23 08:14:11 <frojnd> It's interesting.. I've upgraded to 9.12 version which is the same I have on Arch (and there it works)
2023-10-23 08:14:21 <frojnd> After upgrading on alpine I still get the same error hmm
2023-10-23 08:30:26 <qaqland> apk del --purge doesn't delete ".apk-new" file
2023-10-23 08:30:41 <mercenary> frojnd: It is not alpine that needs upgrading, it is the VPN server you connect to that needs upgrading
2023-10-23 08:31:13 <qaqland> maybe it should delete (?)
2023-10-23 08:34:23 <frojnd> mercenary: gotcha
2023-10-23 08:38:27 <frojnd> Hm
2023-10-23 08:39:24 <frojnd> My alsamixer shows soundcard. Yet pavucontrol won't show it
2023-10-23 08:41:22 <frojnd> Trying to make pipewire work with dbus enabled at boot
2023-10-23 09:26:46 <frojnd> It is an issue with /usr/libexec/pipewire-launcher not being properly configured with dbus and my sway conf
2023-10-23 09:41:32 <quinq> mount: /mnt/iso: unkown filesystem type 'iso9660'.
2023-10-23 09:41:34 <quinq> Oh come on :p
2023-10-23 09:50:05 <Habbie> quinq, isofs?
2023-10-23 09:50:33 <Habbie> hmm, no
2023-10-23 09:51:11 <dwfreed> iso9660 is correct
2023-10-23 09:53:48 <ikke> Kernel upgrade (ie, no more access to modules)?
2023-10-23 10:41:21 <witcher> Saijin_Naib[m]: Ive used the RockPro64 for the same thing, decided to sell it again in the end. Support is poor as you need a non-mainline kernel. The only viable option was armbian and even there its easy to break your system when installing a package.
2023-10-23 14:51:35 <reypm> hi there, how do I enable Apache modules in Alpine Linux Docker Container?
2023-10-23 15:24:41 <chereskata> rc-update add apache2 boot
2023-10-23 15:26:23 <ikke> 1. they are gone. 2 It was about modules, not apache itself. 3. openrc does not run in docker
2023-10-23 18:32:36 <pjs> Hi all. My wifi connection drops and reconnects randomly through the day (like every 30 minutes or so).. I finally am annoyed enough to check it and I see power save mode is on (iwconfig also confirms this). How do I disable it permanently? 
2023-10-23 19:03:03 <pjs> disabled it using network manager config, ref for anyone else (just in case): https://unix.stackexchange.com/questions/269661/how-to-turn-off-wireless-power-management-permanently
2023-10-23 21:07:01 <_shawn> matrix moment
2023-10-23 21:36:47 <human_with_ques> Hi, all. I am a human with questions. I am trying to run as minimal a system as possible to keep things light and to help further my understanding of each moving part. As such, I currently do not have elogind, consolekit, or polkit installed. It hasn't seemed to be any sort of issue for me, as I run a single user system and run a WM rather than a DE. My two questions are: where can I learn more about these and their alter
2023-10-23 21:39:11 <human_with_ques> A good example: I manually mount USB drives, which sucks as it requires me to interact as root but then I cannot use Calibre to transfer to. I am trying to learn solutions. Udisks seems to require polkit, for example.
2023-10-23 21:42:18 <Habbie> your first line got truncated at "about these and their alter"
2023-10-23 21:43:59 <human_with_ques> Habbie:  Thanks for catching that. "My two questions are: where can I learn more about these and their alternatives/shortcoming? How may this impact me?"
2023-10-23 21:44:16 <Habbie> np
2023-10-23 21:44:21 <Habbie> i have no answers about elogind, polkit, etc.
2023-10-23 21:44:35 <Habbie> but i do know that even if you need to be root to mount (in your current setup), you can tell mount to set a userid
2023-10-23 21:45:21 <_shawn> elogind automatically gives you permissions using ACLs
2023-10-23 21:45:50 <_shawn> That’s one of it’s main features, I like it because it simplifies the configuration of users now
2023-10-23 21:46:28 <human_with_ques> How does this vary from Polkit?
2023-10-23 21:46:32 <_shawn> Now you add a user to wheel or you don’t, instead of audio video tty
2023-10-23 21:47:42 <_shawn> polkit can be like an automated privlidge escalation based on rules. Elogind just adds your username to the device file in the ACLs
2023-10-23 21:48:01 <_shawn> So it’s not like you are becoming another user
2023-10-23 21:48:07 <human_with_ques> Trying to read through the Gentoo wiki, etc. makes finding the overlap and boundaries at times a little tricky. I'd been using Arch for about a decade previously, but so much of this is baked in. I am trying to learn deeper as I go
2023-10-23 21:48:56 <_shawn> Okay so, think about it this way, elogind is just handling all of the things that you would need if you were physically sitting in front of the computer
2023-10-23 21:49:06 <_shawn> Like you need to see, hear,
2023-10-23 21:49:29 <human_with_ques> OK. And what was the solution before systemd?
2023-10-23 21:49:50 <_shawn> Adding users to supplimental groups
2023-10-23 21:50:03 <human_with_ques> And what is the real drawback there?
2023-10-23 21:50:05 <_shawn> video group, audio group, etc
2023-10-23 21:50:11 <human_with_ques> Because I have done so already
2023-10-23 21:52:16 <human_with_ques> I guess what I am confused about is: what is it that elogind is doing that warrants it's creation and bloat if I choose to install it. (I realise it's actually quite small). What will I be forced to do without it?
2023-10-23 21:52:45 <_shawn>  The drawback is, you’ll install a unixy OS, add a user, you’ll forget some group and later on you’ll find out something doesn’t work and then you’ll have to stop what you’re doing and look at permissions, log out and back in again. elogind has already figured out all these scenarios. It’s just effort that we don’t have to do anymore, but it’s your choice if you want to keep doing extra steps
2023-10-23 21:53:28 <human_with_ques> I appreciate that. Thank you for answering my question. When one looks into automounting USB sticks, for instance, most wikis make it seem like you haven't a choice but to install these things.
2023-10-23 21:53:50 <human_with_ques> I just want to _understand_ reason before blindly installing _stuff_, you know?
2023-10-23 21:54:08 <_shawn> For auto mounting usb sticks, I wonder if you can’t just use autofs for that now
2023-10-23 21:54:10 <human_with_ques> It's why I no longer want to use systemd at all
2023-10-23 21:57:02 <human_with_ques> I am sure if I had the time to try LFS, I would better understand all this seatd, elogind, polkit, consolekit, etc. better. 
2023-10-23 21:58:08 <_shawn> Just go get a polished distro with a DE and then delete the polkit rules, then try to add a user from the settings app
2023-10-23 21:58:33 <_shawn> Go through all the settings and see what stopped working
2023-10-23 21:58:45 <_shawn> You will see what kind of things polkit will be useful for
2023-10-23 21:59:10 <human_with_ques> Well, I do have eudev so I think autofs should work. Thanks, I will consider that, too.
2023-10-24 02:02:06 <vindaloo> Does anyone here use a slicer for Alpine? I see Cura in the repos but couldn't get it running.
2023-10-24 02:15:32 <vimproved> i'd use prusaslicer from flathub personally
2023-10-24 05:02:05 <iggy> cura is such a pain to build now, I've just been sticking with the appimage
2023-10-24 05:03:16 <iggy> I've been trying to move over to orcaslicer, but I just don't have the time to setup my printers from scratch
2023-10-24 09:10:41 <WhyNotHugo> How can I have a sh script bail if a command in a pipeline fails.
2023-10-24 09:10:58 <WhyNotHugo> E.g.: `false | echo hello there` <- I want this to fail and exit.
2023-10-24 09:11:23 <aron> pipefail
2023-10-24 09:11:43 <aron> set -o pipefail
2023-10-24 09:12:28 <WhyNotHugo> Isn't that a bash-specific thing?
2023-10-24 09:12:46 <WhyNotHugo> https://github.com/koalaman/shellcheck/issues/2555 Nope :p
2023-10-24 09:13:08 <aron> afaik, it's posix
2023-10-24 09:14:29 <aron> but may I have mistaken
2023-10-24 09:14:46 <WhyNotHugo> It is POSIX, though it's a relatively recent addition.
2023-10-24 09:14:50 <aron> okay, depends in the std
2023-10-24 09:14:50 <WhyNotHugo> It's safe to use in my case tho +1
2023-10-24 09:15:21 <aron> s/in/on/
2023-10-24 09:15:49 <aron> although, it is widely supported, prior posix 2022
2023-10-24 09:17:06 <WhyNotHugo> Doesn't work on sh on Alpine tho
2023-10-24 09:17:18 <WhyNotHugo> `set -o pipefail \n false | echo hola`
2023-10-24 09:17:22 <WhyNotHugo> Oh, I need set -e as well.
2023-10-24 09:36:23 <ncopa> can anyone with a rpi4 test boot it without arm64=1 and enable_gic=1? and add the result: https://gitlab.alpinelinux.org/alpine/aports/-/issues/15324#note_344991
2023-10-24 09:56:04 <ncopa> i got my own rpi4 booting now. thanks
2023-10-24 12:24:23 <singham> Are there any fuzzy search libraries in alpine?
2023-10-24 12:24:44 <singham> Which search for a word in a list of names and words
2023-10-24 12:29:26 <Habbie> fzf?
2023-10-24 12:38:16 <singham> nice
2023-10-24 12:38:23 <singham> thanks a lot Habbie
2023-10-24 12:39:06 <singham> It's written in Go
2023-10-24 12:39:16 <singham> Is there something written in C?
2023-10-24 12:41:43 <singham> There's fzy
2023-10-24 16:27:20 <xgpt> Is there a way to download a single apk from edge and install it in stable? I'd like to install ddclient but not enable the entire edge repo for a server I'd like to keep running the current stable
2023-10-24 16:32:35 <minimal> xgpt: generally no, as software in Edge (or indeed any specific release) will have been compiled against/linked against versions of libraries in Edge
2023-10-24 16:32:51 <kwadroke[m]> you could download the aport files and build it on your system. Looks to be perl scripts
2023-10-24 16:34:13 <minimal> xgpt: in the case of this specific package, who knows
2023-10-24 16:34:27 <numono> agree with minimal, but in some cases you can make it happen with repository pinning (be careful, it can bork your system)
2023-10-24 16:34:31 <numono> see: https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Repository_pinning
2023-10-24 16:34:40 <xgpt> not interested, it's simply to enable 'ddclient'
2023-10-24 16:35:09 <xgpt> sorry, that sounded rude, just saying I'm not interested in pursuing such drastic changes to underlying system stuff for a perl script
2023-10-24 16:35:21 <xgpt> I would like said perl script to automatically get updates from a maintainer 
2023-10-24 16:35:31 <xgpt> as APIs might change in the future and so on
2023-10-24 16:36:54 <xgpt> did I mess up my repo file here: 
2023-10-24 16:36:54 <xgpt> https://dl-cdn.alpinelinux.org/alpine/v3.18/main
2023-10-24 16:36:54 <xgpt> ~ # cat /etc/apk/repositories
2023-10-24 16:36:54 <xgpt> https://dl-cdn.alpinelinux.org/alpine/v3.18/community
2023-10-24 16:36:54 <xgpt> http://dl-cdn.alpinelinux.org/alpine/edge/testing
2023-10-24 16:37:48 <minimal> well you'll have a unsupported system as testing is only supported for Alpine Edge...
2023-10-24 16:39:05 <minimal> you could also raise a Issue asking for ddclient to be moved from "testing" to "community", and if that is done soon then it would appear in the Alpine 3.19 release due out approx within the next month
2023-10-24 16:54:38 <kuna1> Hi there. I am using alpine on a OrangePi zero. I am using the generic-arm tree with u-boot. I want to use an overlay or kernel module for an TFT display. Seems the fbtft_driver should be fine. I copied the overlay from the rpi tree, but trying manually with 'dtoverlay -d . fbtft.drbo' gives me an error of 'mount: mounting none on /config/device-tree failed: No such file or directory
2023-10-24 16:54:38 <kuna1> * Failed to mount configfs - 2'. I have not found this module to use with modprobe. Any ideas how to proceed?
2023-10-24 17:05:09 <xgpt> minimal how do I raise issues? 
2023-10-24 17:08:36 <xgpt> https://gitlab.alpinelinux.org/alpine/aports/-/issues/15395 like this minimal ?
2023-10-24 17:09:40 <ikke> yes
2023-10-24 17:52:59 <ikke> xgpt: it has been moved :)
2023-10-24 19:15:45 <kuna1> Can anyone link me a howto/guide to rebuild the kernel with some small config change? I would like to enable "CONFIG_FB_TFT=m" in the 6.1.55-lts kernel. I need to compile it on different. I need armv7, need to compile on x86_64. Thank You.
2023-10-24 19:23:20 <xgpt> ikke wow! that was easy! Thank you so much!
2023-10-24 19:23:56 <xgpt> going to integrate my caddy and ddclient LXC containers now :)
2023-10-25 00:37:38 <anon8697> hi! im having difficulty setting up znc. the service crashes on start up as "error loading shared library ld-linux-aarch64.so1 (needed by log.so)". which is a factually correct statement, i do miss the lib. however as it seems it does not come with gcc so im out of tips how to get it. its also awefully strange that znc would directly deppend on it. any tips?
2023-10-25 00:40:57 <heat> anon8697, that znc was built against glibc
2023-10-25 00:41:06 <heat> which alpine doesn't have
2023-10-25 00:49:18 <anon8697> oh, right. i did try to fix a missing module error by copying over znc/log.so from my previous debian install. in retrospect, it was clearly foolish.
2023-10-25 05:53:36 <Gb> Hello
2023-10-25 05:55:00 <Gb> Anyone know how to increase system wide limits in Alpine - Linux ? I have already tried installing shadow package, edit /etc/security/limits.conf & add edit /etc/pam.d/base-auth
2023-10-25 05:58:48 <ikke> Gb: Is this for something you run as a user or for a service?
2023-10-25 06:06:13 <Gb> For service
2023-10-25 06:06:21 <Gb> For docker service to be exact
2023-10-25 07:08:27 <Justin[m]123> Why is Squash commits when merge request is accepted. not an option on Alpine GitLab?
2023-10-25 07:09:35 <Gb> @ikke any suggestions ?
2023-10-25 07:27:13 <caskd[m]> squash commits used to be a option
2023-10-25 07:27:29 <caskd[m]> i guess it got removed in a recent update
2023-10-25 07:33:45 <lucidiot> that option can be turned off by a maintainer, and it gets hidden https://docs.gitlab.com/ee/user/project/merge_requests/squash_and_merge.html#configure-squash-options-for-a-project
2023-10-25 07:42:31 <Justin[m]123> <lucidiot> "that option can be turned off by..." <- Anyone know why it was hidden?
2023-10-25 08:49:02 <frojnd> Alright
2023-10-25 08:49:44 <frojnd> So I've found the reason why I can't connect to my corpo's vpn anyconnect. It's because alpine's openconnect is built with openssl. On Arch openconnect works but it uses gnutls
2023-10-25 08:49:53 <frojnd> So how can I build it with gnutls on alpine?
2023-10-25 08:53:21 <frojnd> I really need this peace of software in order to keep the distro haha
2023-10-25 08:55:32 <Habbie> that looks like you'll have to edit the APKBUILD
2023-10-25 08:56:44 <aron> frojnd: how do you found out this? I think openssl and gnutls are just crypto implementations (of the same algorithms) -- so, what's the clue?
2023-10-25 08:56:59 <Habbie> this is a good question
2023-10-25 08:57:03 <Habbie> and it might then be a configuration issue
2023-10-25 08:57:11 <aron> yes
2023-10-25 08:58:19 <frojnd> Well it works on arch built agains gnutls and doesn't on alpine built agains openssl and this thread: https://gitlab.com/openconnect/openconnect/-/issues/43
2023-10-25 09:01:08 <aron> this thread concludes that the openssl they used is very old, and does not state the openssl does not work ;)
2023-10-25 09:01:11 <frojnd> And also I've manually tired it 
2023-10-25 09:01:38 <aron> but anyway, just edit the APKBUILD
2023-10-25 09:01:57 <frojnd> when I do: openssl s_client -connect remote.company.vpn I get: 48ABEA772F7F0000:error:10000067:BIO routines:BIO_connect:connect error:crypto/bio/bio_sock2.c:116:
2023-10-25 09:02:25 <frojnd> With gnutls-cli remote.company.vpn I get cert info
2023-10-25 09:03:16 <aron> errno=?
2023-10-25 09:05:12 <frojnd> https://termbin.com/6jcle
2023-10-25 09:06:14 <aron> openssl errstr 10000067
2023-10-25 09:06:15 <aron> gives me
2023-10-25 09:06:16 <aron> error:10000067:elliptic curve routines:elliptic curve routines:invalid field
2023-10-25 09:06:18 <aron> nasty :P
2023-10-25 09:06:38 <aron> incomplete implementation perhaps
2023-10-25 09:07:52 <aron> as far as I can tell, the root of the problem is an EC that is not known by openssl
2023-10-25 09:08:20 <frojnd> aron: so you are saying that if I were to add EC to openssl this might work with openssl?
2023-10-25 09:08:20 <aron> take the APKBUILD of the openconnect, and s/openssl/gnutls/
2023-10-25 09:08:29 <frojnd> aron: ok 
2023-10-25 09:08:40 <aron> it's not that easy :P
2023-10-25 09:10:54 <ikke> Gb: then you need to set rc_ulimit in /etc/conf.d/docker
2023-10-25 09:25:14 <frojnd> What did I miss? ERROR: /home/frojnd/coding/alpine/packages/coding/x86_64/openconnect-9.10-r0.apk: UNTRUSTED signature
2023-10-25 09:25:19 <frojnd> After building it
2023-10-25 09:26:29 <frojnd> After editing APKBUILD I did ran abuild checksum; abuild -r; doas apk add ...
2023-10-25 09:27:02 <aron> apk add --allow-untrusted /.../...apk
2023-10-25 09:35:01 <frojnd> That's what I am talking about! Thank you aron
2023-10-25 09:35:44 <frojnd> Err so how can I freeze update for openconnect? So I won't wrongly used the upstream version that uses openssl?
2023-10-25 09:35:56 <aron> it won't
2023-10-25 09:36:32 <aron> it's a local package
2023-10-25 09:37:00 <aron> grep openconnect /etc/apk/world
2023-10-25 09:37:06 <aron> you'll see an "=" sign
2023-10-25 09:37:46 <frojnd> openconnect><Q1qd8Tn9RFztMlfKAWAx/6GDwhcnc= 
2023-10-25 09:38:29 <aron> oh, sorry, it's not *that* "=" sign :P
2023-10-25 09:38:45 <frojnd> That's the only openconnect inside world file
2023-10-25 09:38:52 <frojnd> I assume that's my signature
2023-10-25 09:38:53 <aron> but still, it won't be upgraded automatically
2023-10-25 09:40:48 <frojnd> My laptop now needs only suspesnd to ram, some fan control 
2023-10-25 09:41:28 <aron> I think that stuff after >< is the package signature
2023-10-25 09:41:37 <aron> but don't worry about that
2023-10-25 09:42:57 <aron> sus2ram should be easy
2023-10-25 09:44:23 <frojnd> I've also setup sanoid to take auto snapshots since I'm learning by doing on a prod laptop so I have high expectations of auto snaps
2023-10-25 09:50:33 <ovf> ptrc: i'd like to officially pass this firefox build bug to you (careful, this will abort): https://planetk.org/pl.html  it appears to be happening again on the rust/c++ boundary, which has been a fun source of problems with firefox in alpine in the past
2023-10-25 11:03:29 <fooo> hey, a quick question, did sonething in the docker package change in the last couple of weeks? I have a docker container that is failing now for two weeks and the people at drone-io blame Alpine. but i cannot see why alpine would be aat fault
2023-10-25 11:03:51 <Habbie> how is it failing?
2023-10-25 11:04:32 <fooo> i wrote a ticket here after talking to them: https://github.com/drone-plugins/drone-docker/issues/414
2023-10-25 11:05:01 <fooo> but in pm they say its an alpine problem (but i dont beleave that)
2023-10-25 11:06:58 <WhyNotHugo> nerd-fonts-2.3.3 dropped the primary subpackage: the one with the nerd font symbols itself.
2023-10-25 11:07:13 <WhyNotHugo> It also completely rewrote the package, so can't quite figure out how to restore it.
2023-10-25 11:17:53 <Gb> nyone know how to increase system wide limits in Alpine - Linux ? I have already tried installing shadow package, edit /etc/security/limits.conf & add edit /etc/pam.d/base-auth
2023-10-25 11:18:02 <Gb> This would be for services
2023-10-25 11:28:22 <ikke> Gb: i replied to you 
2023-10-25 11:28:55 <ikke> For services shadow and pam are not relevant 
2023-10-25 11:32:12 <ikke> You need to set rc_ulimit="<ulimit options here>" in /etc/conf.d/docker 
2023-10-25 11:35:11 <Gb> @ ikke... Hey ! Thanks... Actially, I didn't get your message
2023-10-25 11:35:28 <Gb> What about multiple ulimit options ?
2023-10-25 11:35:42 <ikke> You can add more then one
2023-10-25 11:35:47 <Gb> can i put them with a ";" separator ?
2023-10-25 11:35:51 <Gb> or comma?
2023-10-25 11:35:57 <ikke> space 
2023-10-25 11:36:05 <ikke> Just luke the cli args ti unlimit
2023-10-25 11:36:12 <ikke> to ulimit*
2023-10-25 11:36:56 <Gb> eg: rc_ulimit="ulimit -Hn 999; ulimit -n 999; ulimit -Hc 999; ulimit -c 999"
2023-10-25 11:37:02 <Gb> will the above work ?
2023-10-25 11:39:06 <Gb> one more question... so shadow package need not be installed ?
2023-10-25 11:39:20 <Gb> also, the service name should match file-name i guess (just to be sure
2023-10-25 11:41:19 <ikke> Not the command, just the options 
2023-10-25 11:41:34 <ikke> You dont need shadow for this 
2023-10-25 11:41:49 <ikke> And yes, file and service name should match 
2023-10-25 11:46:01 <Gb> So what about kubernetes, I noticed running "cat /etc/<proc id>/limits" the limits are "unlimited", but when I try to run a docker image under that, then the limits become less
2023-10-25 11:47:26 <Gb> Btw, Ikke... Thank you soooo much !
2023-10-25 11:48:25 <ikke> That's something areanged 
2023-10-25 11:48:33 <ikke> In Kubernetes 
2023-10-25 11:50:39 <Gb> Ikke: So the default limits are different in alpine for daemons vs shell ?
2023-10-25 11:54:15 <ikke> The way you change them is different. ulimits are inherited and hard limits can only be increased by root 
2023-10-25 11:56:25 <Gb> So this is what I know... Hard limits are set by root, then daemons/users (like docker) inherit it (otherwise known as soft limits) and soft limits cannot exceed hard limits.
2023-10-25 11:57:00 <Gb> I do have root access and one of the helm charts I installed in kubernetes was complaining of limits
2023-10-25 11:57:05 <ikke> There is nothing in alpine (afaik) that overrides the default limits 
2023-10-25 11:57:43 <Gb> so I checked the limits in bash which were coming out to be lower
2023-10-25 11:58:10 <ikke> installing shadow and adjusting pam only affects processes inherited from login (which does not apply to services)
2023-10-25 11:58:30 <ikke> docker proceses also do not invoke login 
2023-10-25 11:58:40 <Gb> but then since you told me that rc_ulimits are different, I checked "cat /etc/<proc id>/limits" for k3s
2023-10-25 11:58:47 <Gb> that is different from bash
2023-10-25 12:13:55 <singham> I have Orange Pi R1 with H3 chip but the dtb is only available for the one with H2 plus
2023-10-25 12:14:13 <singham> For nano pi, there is one with H3. which dtb to use?
2023-10-25 12:53:31 <frojnd> Hm
2023-10-25 12:54:24 <frojnd> Laptop goes to suspend to ram after  lid close just after installing acpid and starting service. However I can hear the fans still spinning and not stopping spinning
2023-10-25 12:54:40 <frojnd> I mean laptop goes to sleep after lid close
2023-10-25 12:57:44 <vkrishn> i think good config would be rpm goes realy low instead full stop
2023-10-25 13:29:20 <satmd> in standby, fans should stop eventually
2023-10-25 13:30:55 <satmd> frojnd: BIOS makers sadly suck at doing good BIOSes even after many years. Your problem likely is specific to your model and you might find good advice searching for linux suspend fans $model on the internet
2023-10-25 13:41:29 <satmd> Prime candidates: broken bios needing kernel cmdline hacks, a GPU that needs turning off (e.g. switching the VT, ...), using a different poweroff method, ... too many things to list them here
2023-10-25 14:10:41 <kuna1> Can anyone link me a howto/guide to rebuild the kernel with some small config change? I would like to enable "CONFIG_FB_TFT=m" in the 6.1.55-lts kernel. I need to compile it on different. I need armv7, need to compile on x86_64. Thank You.
2023-10-25 14:48:14 <donoban> any idea about what could produce this error when trying to limit cpus on a container?
2023-10-25 14:48:16 <donoban> @ docker run -it --cpus=1.0 alpine
2023-10-25 14:48:19 <donoban> docker: Error response from daemon: failed to create task for container: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: unable to apply cgroup configuration: cannot enter cgroupv2 "/sys/fs/cgroup/docker" with domain controllers -- it is in domain threaded mode: unknown.
2023-10-25 15:16:44 <singham> https://orangepi.com/index.php?route=product/product&path=237&product_id=861
2023-10-25 15:16:58 <singham> What is the right way to access this board?
2023-10-25 15:31:07 <singham> How do I enable uart at the pins before bootup?
2023-10-25 16:03:36 <WhyNotHugo> singham: Check the right pins here: http://www.orangepi.org/orangepiwiki/index.php?title=26_Pin_Interface_Pin_Description&mobileaction=toggle_view_desktop
2023-10-25 16:03:49 <WhyNotHugo> The ones labeled UART
2023-10-25 16:05:11 <singham> That board is different
2023-10-25 16:05:30 <singham> But yeah, so how to proceed?
2023-10-25 16:05:39 <WhyNotHugo> You need to find the manual for that board and figure out which are the right pins.
2023-10-25 16:05:53 <singham> Right pins for?
2023-10-25 16:06:03 <WhyNotHugo> Regarding the software aspect, some notes from this draft article (which covers the rpi) might help: https://git.sr.ht/~whynothugo/whynothugo.nl/tree/main/item/content/blog/2023-04-04-setting-up-an-rpi-via-uart.md?view-source
2023-10-25 16:06:07 <singham> Is uart enabled by default?
2023-10-25 16:06:21 <WhyNotHugo> singham: You need to find the manual for that board, and the manual will indicate which pins expose the UART.
2023-10-25 16:07:04 <WhyNotHugo> singham: I believe it is not enabled by default by Alpine, check that link for how I enabled it. There might be hardware-specific steps involved too.
2023-10-25 16:08:07 <singham> Is it possible to boot via usb?
2023-10-25 16:08:31 <singham> So using powering usb, access board.
2023-10-25 16:09:09 <WhyNotHugo> define "access".
2023-10-25 16:10:31 <singham> Like raspberry pi pico
2023-10-25 16:10:39 <singham> Like a console
2023-10-25 16:10:44 <singham> terminal
2023-10-25 16:11:02 <singham> ash instance
2023-10-25 16:12:17 <minimal> singham: ask the Orange Pi manufacturers? it's their board...
2023-10-25 16:13:00 <singham> I have, as of now, they have nothing!
2023-10-25 16:13:59 <WhyNotHugo> singham: Have you figured out which pins expose the UART?
2023-10-25 16:14:45 <singham> There are special pins for UART man. You can see them on top left in the image
2023-10-25 16:14:56 <singham> TX RX GND
2023-10-25 16:16:45 <WhyNotHugo> And you connected a cable to a serial console on an Alpine host? Did you then connect power to the device and turn it on? How are you trying to connect to the UART?
2023-10-25 16:17:19 <singham> Wait, what problem are you solving, UART or USB
2023-10-25 16:17:59 <singham> Both are different. There are UART pins and there is a usb port
2023-10-25 16:32:34 <ptrc> ovf: hm, it doesn't seem to abort for me
2023-10-25 16:39:56 <bleb> I'm trying to run an old precompiled binary, getting __strdup: symbol not found while executing
2023-10-25 16:40:06 <bleb> any workarounds for that?
2023-10-25 16:40:14 <bleb> does it just mean I need to compile it on alpine?
2023-10-25 16:40:35 <singham> Regarding your link
2023-10-25 16:40:48 <elly> bleb: that kinda sounds like it was linked against glibc
2023-10-25 16:40:53 <elly> you may need an alpine compile
2023-10-25 16:40:55 <singham> WhyNotHugo: Is there one where we can access with usb?
2023-10-25 17:03:05 <ovf> ptrc: fun! this is alpine x86_64, wayland (no xwayland), fresh profile. mozilla's (or archlinux's) build doesn't abort, build from another musl distribution does
2023-10-25 18:09:26 <Gb> With @Ikke 's help I was able to change the hard limits of the daemon by adding rc_ulimit in /etc/conf.d/<servicename>, however, soft limits are not changing even when I mention both -Hc and & -c in the rc-ulimit argument
2023-10-25 18:10:05 <Gb> Can anyone tell me why thats happening? 
2023-10-25 18:33:13 <dwfreed> They're gone, but the answer is it's a limitation of how shell ulimit behaves (as that is what openrc uses for rc_ulimit); in order to specify both hard and soft limits, you cannot provide either -H or -S unless you provide both, and the value given for a specific limit type will apply to both hard and soft limits equally
2023-10-25 18:35:04 <dwfreed> so rc_ulimit="-c unlimited" will set both hard and soft coredump limits to unlimited, as would "-HSc unlimited", but "-Hc unlimited -c unlimited" would only set the hard limit
2023-10-25 19:06:41 <ikke> dwfreed: good to know, didn't know that
2023-10-25 19:18:16 <kuna1> On OrangePi with uboot that uart pins are enabled by default
2023-10-25 19:18:21 <kuna1> 115200
2023-10-25 19:19:44 <kuna1> i just done it few days ago, tool the alpine files for generic arm and copied the u-boot to the expected sectors at the beginning of sd cars. u-boot can be stopped with any key during boot, serial console is active from beginning.
2023-10-25 21:59:11 <rsc> What's the correct way to change the default-jvm if multiple openjdkX-jre-* are installed?
2023-10-25 22:00:12 <rsc> Some other Linux distributions have something like alternatives-config or similar.
2023-10-25 22:24:33 <ovf> rsc: create a symlink /usr/lib/jvm/forced-jvm pointing to the jvm of your choice
2023-10-25 22:26:59 <rsc> ovf: wow! Is this anywhere documented?
2023-10-25 22:27:45 <ovf> rsc: i wondered the same thing. it's 'documented' in the trigger for java-common (https://git.alpinelinux.org/aports/tree/community/java-common/java-common.trigger)
2023-10-25 22:28:44 <ovf> (i wonder if it actually works with the 'cd' in the middle...)
2023-10-25 22:31:02 <Justin[m]123> Anyone know why Squash commits when merge request is accepted. is not an option on Alpine GitLab?
2023-10-25 23:36:30 <rsc> ovf: thanks :)
2023-10-26 03:31:10 <jscript> how do i get ssdm working with user input
2023-10-26 04:05:40 <jscript> ok i got it working
2023-10-26 04:05:51 <jscript> i needed xf86 input
2023-10-26 10:39:59 <WhyNotHugo> Pressing the power button on my device does nothing. `wev` doesn't report any keypress, /var/log/acpid.log is empty, dmesg says nothing. I don't know what to check next.
2023-10-26 10:42:08 <minimal> WhyNotHugo: what is your device?
2023-10-26 10:42:33 <WhyNotHugo> ThinkPad T14s Gen 2i
2023-10-26 10:42:57 <minimal> which acpid are you using? Busybox's? or the "full" acpid?
2023-10-26 10:44:06 <minimal> is the "button" (or "tiny-power-button") kernel module loaded?
2023-10-26 10:44:18 <Ermine> i think the question here is whether kernel got that
2023-10-26 10:44:58 <minimal> Ermine: that's why I asked about the kernel modules
2023-10-26 10:49:27 <WhyNotHugo> minimal: busybox. The small power button between the keyboard and the screen
2023-10-26 10:49:44 <WhyNotHugo> The kernel module "button" is loaded.
2023-10-26 10:51:14 <minimal> WhyNotHugo: have you tried the "full" version of acpid to see if that makes a difference?
2023-10-26 10:51:46 <WhyNotHugo> Doesn't acpi intercept the events?
2023-10-26 10:52:53 <minimal> yeah I'd expect the kernel to directly handle power events, e.g. I've used tiny-power-button on VMs with no acpid at all. I just though the "full" acpid might log something to help
2023-10-26 10:53:24 <WhyNotHugo> I'll try, it might log something.
2023-10-26 10:55:16 <minimal> for tiny-power-button I've had to pass "tiny_power_button.power_signal=12" on cmdline for it to work, but "button" should just work with no config
2023-10-26 10:56:35 <minimal> when you say you're pressing the power button, how exactly? With PCs there's a difference in behaviour between just pressing the button, and pressing and holding for 5+ seconds
2023-10-26 10:57:52 <WhyNotHugo> Using package acpid yields the same results: nothing on /var/log/messages, /var/log/acpid, compositor logs nor wev
2023-10-26 10:58:12 <WhyNotHugo> Just pressing.
2023-10-26 10:58:19 <WhyNotHugo> Like, half a second.
2023-10-26 11:00:38 <WhyNotHugo> Oh, holding the button for 2 seconds sends an event!
2023-10-26 11:01:51 <WhyNotHugo> I tend to not-hold these power buttons too long since they trigger a hardware poweroff.
2023-10-26 11:05:24 <minimal> hmm
2023-10-26 11:05:46 <minimal> so did the 2 seconds press trigger system shutdown via "button"?
2023-10-26 11:25:17 <WhyNotHugo> Apparently it did.
2023-10-26 11:25:30 <WhyNotHugo> I supose that the 2s delay is hardware-related.
2023-10-26 11:25:45 <WhyNotHugo> On my desktop just tapping the button sends the event immediately.
2023-10-26 11:25:54 <minimal> sounds like it
2023-10-26 11:27:22 <WhyNotHugo> Thanks for helping me find the bug (which was between the keyboard and the seatrest).
2023-10-26 11:27:40 <aron> WhyNotHugo: maybe a bios setting affects this?
2023-10-26 14:07:47 <msrd0> hi, my virtual alpine linux started spitting out ERROR: Unable to read database state: database file format error errors whenever I invoke apk. google wasn't helpful. do you know any way to fix this?
2023-10-26 14:23:53 <ncopa> msrd0: sounds like your database state got corrupted. maybe it is a filesystem corruption?
2023-10-26 14:24:06 <ncopa> make sure that you backup your /etc/apk/world
2023-10-26 14:25:54 <ncopa> the db is a textfile found at /lib/apk/db/installed. you could have a look there to try find out what broke
2023-10-26 14:27:09 <ncopa> or you could simply reinstall the system: rm /lib/apk/db/installed && apk fix
2023-10-26 14:27:47 <ncopa> but have a look at dmesg. you may have a broken disk
2023-10-26 14:46:18 <msrd0> hm no idea what I did but after remounting the fs a few times the error disappeared
2023-10-26 16:01:31 <soacb> Please stop spamming other IRC servers immediately. If you don't want your servers to be unavailable.
2023-10-26 16:01:34 <soacb> LAST WARNING
2023-10-26 17:29:04 <sch> Please send a screenshot
2023-10-26 17:29:15 <sch> Some people might not see what you see
2023-10-26 17:29:29 <sch> I'm using an XMPP transport to communicate to IRC
2023-10-26 17:38:17 <Apachez> are there other people here with us?
2023-10-26 17:51:27 <ikke> sch: if you are referring to what soacb said, that's just spam
2023-10-26 17:52:12 <sch> I've received a PM implying I'm spamming
2023-10-26 17:58:39 <invoked> i suggest you set your /mode to +iGj
2023-10-26 17:59:08 <invoked> or maybe even +g until this calms down
2023-10-26 18:01:30 <invoked> eventually they will have to bring a report card home, or something
2023-10-26 19:19:53 <i9j8k7> hi, am I wrong or there is no way to use opencl in void because the open runtimes are missing?
2023-10-26 19:22:05 <i9j8k7> the opencl runtimes*
2023-10-26 20:43:32 <frag> any recommendation for easy hardening of firefox?
2023-10-26 20:45:39 <fluix> librewolf?
2023-10-26 20:46:16 <fluix> the better answer is "what are your concerns/goals?"
2023-10-26 20:46:33 <frag> whats librewolf?
2023-10-26 20:46:47 <frag> i dunno, but ff wants to store my bankid password :|
2023-10-26 20:47:08 <fluix> https://librewolf.net/ and you can disable the password manager
2023-10-26 20:47:48 <frag> yeah im sure there are a million settings to make it more private
2023-10-26 20:47:49 <fluix> to be clear, librewolf probably doesn't disable it by default either, because it's not really an issue
2023-10-26 20:47:58 <frag> i guess privacy is my main concern
2023-10-26 20:48:50 <fluix> disable the external features (Pocket, etc.) and install ublock origin. unfortunately the web kind of sucks w.r.t. privacy
2023-10-26 20:49:02 <fluix> I'll note "privacy" is still very broad, and so these are broad suggestions that may or may not achieve what you want
2023-10-26 20:53:19 <invoked> long story short, librewolf already does a lot of the privacy work for you
2023-10-26 20:53:26 <invoked> if that's your main concern
2023-10-26 20:54:08 <frag> librewolf is not in repo tho?
2023-10-26 20:54:16 <fluix> indeed
2023-10-26 20:54:49 <invoked> it's a very heavy thing to do browser builds. you can run the flatpak version without too much fiddling, though
2023-10-26 20:56:08 <frag> was thinking something easy/native :]
2023-10-26 20:56:54 <frag> i try to use links -g mostly, but sometimes need a normal browser ... firefox seem to be the one that works on everything
2023-10-26 20:57:01 <invoked> the patchset for firefox could be adapted to librewolf but someone has to do the work.
2023-10-26 20:57:13 <vimproved> https://github.com/arkenfox/user.js
2023-10-26 20:57:36 <vimproved> covers pretty much all the same ground as librewolf
2023-10-26 20:58:14 <invoked> it's not just prefs.
2023-10-26 20:58:48 <vimproved> librewolf applies a couple patches but most of them aren't very necessary
2023-10-26 20:58:55 <invoked> says who
2023-10-26 20:59:46 <vimproved> says me, i have looked at the patches
2023-10-26 21:00:27 <invoked> ok well that's an opinion.  as in, it applies to you, not to the people who make librewolf or anyone who share those values
2023-10-26 21:01:05 <vimproved> most of the patches are hardcoding stuff that can also be controlled by prefs or rebranding with the librewolf name
2023-10-26 21:01:12 <vimproved> such as the dbus name
2023-10-26 21:01:13 <fluix> I don't think we need to continue this discussion. All of us are just sharing opinions and they all lack context of what we want in terms of privacy or security
2023-10-26 21:01:46 <fluix> we've all shared helpful options :)
2023-10-26 21:01:54 <invoked> i'm careful to not advocate things one way or another
2023-10-26 21:01:58 <vimproved> https://codeberg.org/librewolf/source/src/branch/main/patches/sed-patches/stop-undesired-requests.patch seems like it would be the only patch that really does anything in terms of privacy
2023-10-26 21:02:14 <vimproved> to be clear i'm not trashing librewolf here
2023-10-26 21:02:44 <vimproved> it's a lot easier than arkenfox and also includes some nice patches for additional functionality
2023-10-26 21:03:15 <vimproved> also it appears they have alpine packages
2023-10-26 21:03:24 <vimproved> https://codeberg.org/librewolf/alpine
2023-10-26 21:04:42 <invoked> someone needs to step forward to do that work... being a browser maintainer is not fun
2023-10-26 21:05:47 <frag> all work and no fun :c
2023-10-26 21:06:10 <invoked> it is one of the most thankless ones
2023-10-26 21:06:18 <invoked> nothing but complaints
2023-10-26 21:07:02 <frag> links -g is beautiful tho ;|
2023-10-26 21:07:12 <frag> and soo fast
2023-10-26 21:07:14 <vimproved> invoked: sounds a lot like being a distro maintainer then
2023-10-26 21:07:20 <vimproved> :)
2023-10-26 21:07:45 <frag> would think any type of maintaioner
2023-10-26 21:08:00 <invoked> no i mean maintaining the port/package/whatever for browsers on a distro.  it's a lot of pain
2023-10-26 21:08:31 <invoked> if you break the build it's just pain.
2023-10-26 21:09:03 <invoked> browsers also the most complex thing in the tree
2023-10-26 21:09:19 <vimproved> ehhh not necessarily
2023-10-26 21:09:37 <vimproved> in terms of packaging ghc is probably worse
2023-10-26 21:10:07 <vimproved> they're up there though
2023-10-26 21:10:23 <vimproved> especially firefox with its godawful build system
2023-10-26 21:11:45 <invoked> i maintained firefox for a little while elsewhere, before "oxidation" started happening
2023-10-26 21:11:52 <invoked> before wasm and everything
2023-10-26 21:12:04 <invoked> that was bad enough
2023-10-26 21:12:20 <invoked> there's a certain person who can tolerate that type of packaging, but it's not me
2023-10-26 21:13:32 <vimproved> https://bugs.gentoo.org/915651 this is a specific kind of stupid bug that could only happen with something like firefox
2023-10-26 21:15:25 <invoked> anyway i hope someone (weirdo masochist) takes up librewolf for alpine.  that would be good.
2023-10-27 00:12:30 <goran> In https://wiki.alpinelinux.org/wiki/Xen_Dom0 the commands 'grub-mkconfig -o /boot/grub/grub.cfg' and 'grub-set-default "$(grep ^menuentry /boot/grub/grub.cfg | grep Xen | cut -d \' -f 2 | head -1)"' are to be invoked manually each time a new kernel or xen update comes in. Can this in some way be automatised? 
2023-10-27 00:29:33 <invoked> grub updates too
2023-10-27 00:42:48 <idkrn[m]> Has anyone here used systemd-boot on Alpine?
2023-10-27 06:45:44 <ptrc> idkrn[m]: yes
2023-10-27 06:45:59 <ptrc> ( but also, don't ask to ask )
2023-10-27 07:14:01 <linfan> morning
2023-10-27 07:14:35 <Justin[m]123> Why is Squash commits when merge request is accepted. not an option on Alpine GitLab?
2023-10-27 07:14:59 <Justin[m]123> ikke do you know?
2023-10-27 07:21:10 <linfan> Happy to inform that alpine runs flawlessly on vmware player out of the box. That is not the case on virtualbox
2023-10-27 07:33:18 <ikke> Justin[m]123: we disabled it. We prefer deliberate / selective squashing to just indiscriminately squashing everything into a single commit
2023-10-27 07:34:27 <Justin[m]123> Ah. That's unfortunate. It's certainly easier for new contributors with that on.
2023-10-27 07:41:56 <qaqland> how should I do with *.apk-new file? just delete them?
2023-10-27 07:42:18 <ikke> Check if they contain relevant changes 
2023-10-27 07:42:33 <ikke> update-conf can help with that
2023-10-27 07:52:33 <qaqland> thx
2023-10-27 09:12:25 <WhyNotHugo> ikke: Thanks, TIL!
2023-10-27 09:14:23 <WhyNotHugo> I expected (e)dit to open vim with conflict markers to resolve conflicts.
2023-10-27 09:14:43 <WhyNotHugo> It just lets me edit the new file but I need to find the conflicts myself?
2023-10-27 09:14:50 <WhyNotHugo> e
2023-10-27 09:19:04 <ikke> If you install vimdiff, then it should be shown as a new option 
2023-10-27 09:50:13 <dennis14e> Hi, I have a question. I installed Alpine as a qemu vm using the alpine-virt iso. Later I switched to the linux-lts kernel and removed linux-virt (because I needed a kernel module, which is apparently not included in virt). Basically the system works, but via the terminal of the hypervisor I have no login prompt anymore. The last message displayed is "OpenRC 0.48 is starting up Linux 6.1.60-0-lts (x86_64)". 
2023-10-27 10:13:59 <WhyNotHugo> > 
2023-10-27 10:13:59 <WhyNotHugo>     Your issue has been recognized as spam. Please, change the content to proceed.
2023-10-27 10:14:26 <WhyNotHugo> Contents of my issue: https://paste.sr.ht/~whynothugo/61f457b7e858b300f1cf295d9f75394a40a790fb
2023-10-27 12:46:32 <john3d> hi. is it possible to manage apks in gnome-software? at the moment i just got the error about: the name dev.Cogitri.apkPolkit2 was not provided by any .service files .i have the same problem on edge
2023-10-27 12:48:05 <Saijin_Naib[m]> john3d yes, you just need the package to give Software that capability 
2023-10-27 12:48:50 <Saijin_Naib[m]> https://pkgs.alpinelinux.org/package/edge/community/x86_64/apk-polkit-rs
2023-10-27 12:49:04 <Hello71> WhyNotHugo: i think you need some permission to post links
2023-10-27 12:50:20 <Saijin_Naib[m]> john3d ```apk-polkit-rs```
2023-10-27 12:50:55 <john3d> apk-polkit-rs & gnome-software-plugin-apk still installed in my system
2023-10-27 12:55:51 <Saijin_Naib[m]> Did you add the OpenRC service for it? 
2023-10-27 12:56:07 <Saijin_Naib[m]> ```rc-service -l```
2023-10-27 12:58:14 <john3d> rc-update add apk-polkit-server # solved the problem!! 
2023-10-27 12:58:54 <john3d> Saijin_Naib: thx for for your help :-)
2023-10-27 13:37:47 <scv> Hi, Does Alpine-linux have full Chinese locale support? I already installed musl-locales, but I didn't see any relevant charset from `iconv -l`.
2023-10-27 13:40:40 <scv> When I use a SSH client to log on a Chinese BBS(Bulletin Board System), the Chinese words (encoded in GB18030/GB2312/GBK) cannot be converted correctly.
2023-10-27 14:01:21 <Saijin_Naib[m]> scv what font are you using? I had to add font-noto-cjk-extra to get good coverage 
2023-10-27 14:13:02 <scv> Saijin_Naib[m]: Hi Saijin, you mean my terminal font?
2023-10-27 14:14:00 <scv> I use iTerm2, my current font is Melso Nerd Font Mono...
2023-10-27 14:15:56 <Saijin_Naib[m]> Yeah, your terminal font. Do you know if that covers your needed characters? 
2023-10-27 14:16:11 <scv> Saijin_Naib[m]: or you mean font encoding in the Alpine system? I'm not sure..
2023-10-27 14:16:52 <scv> Saijin_Naib[m]: Well, it works well for debian, it doesn't work for Alpine
2023-10-27 14:20:27 <Saijin_Naib[m]> Hmm 🤔 Are you writing in Pinyin or otherwise using a non-Latin IME? I don't, so I'm only viewing those characters, and installing musl-locales and font-noto-cjk-extra and font-noto-cjk does it for me to view 
2023-10-27 14:22:19 <scv> actually, I was trying to deploy an app, which use SSH client to log on a Chinese BBS, onto Docker images(Debian, Alpine..), I use luit to convert Chinese encoded text for UTF-8 terminals. And it works fine on Debian systems, but not on Alpine.
2023-10-27 14:23:51 <Saijin_Naib[m]> Ah, I don't know then 
2023-10-27 14:26:09 <scv> Saijin_Naib[m]: thank you all the same. 
2023-10-27 14:56:47 <idkrn[m]> <ptrc> "( but also, don't ask to ask )" <- I wasn't
2023-10-27 14:57:59 <idkrn[m]> <ptrc> "( but also, don't ask to ask )" <- I didn't
2023-10-27 15:21:15 <Saijin_Naib[m]> scv: Sorry, hopefully someone else knows better
2023-10-27 15:25:59 <scv> Saijin_Naib[m]: I just sent a email to "General user support mailing list", hopefully can get some help. Thanks Saijin.
2023-10-27 15:30:49 <singham> Folks, I want to know how to connect to serial console of Orange Pi R1
2023-10-27 16:55:28 <fluix> idkrn[m]: they likely meant it in the sense that if somebody replies "yes" (and if nobody does that doesn't mean people still couldn't help) you're probably going to have follow-up questions, so just ask those questions.
2023-10-27 17:18:10 <telmich> Is it possible that edge is broken?
2023-10-27 17:18:11 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/CSZrQglVhpwojHHylSbMPKpw

2023-10-27 17:19:10 <ikke> telmich:   python3-3.9.4-r0 is not in edge
2023-10-27 17:20:47 <telmich> hmm
2023-10-27 17:21:13 <ikke> do you have some package that keeps python 3.9 installed?
2023-10-27 17:21:24 <telmich> Not that I am aware of
2023-10-27 17:22:04 <ikke> It's strange that it tries to _install_ python3-3.9
2023-10-27 17:22:16 <telmich> it is
2023-10-27 17:22:21 <ikke> telmich: what does apk policy python3 return?
2023-10-27 17:22:45 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/lpkystAnbmoNgcJZWTDNDSBt

2023-10-27 17:23:15 <ikke> so it is still installed
2023-10-27 17:23:24 <ikke> apk version python3?
2023-10-27 17:23:51 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/JJTFdeDpsnTxrWdZLIcBLEqa

2023-10-27 17:24:23 <ikke> try `apk del python3`
2023-10-27 17:24:35 <ikke> It should tell you what keeps python3 installed
2023-10-27 17:24:57 <ikke> python3-3.9*
2023-10-27 17:25:21 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/KxGfsapPybOolhCQdXUFQDAm

2023-10-27 17:26:33 <ikke> so mercurial and httpie
2023-10-27 17:26:59 <telmich> I don't mind deleting them, but I am a bit puzzled as to why/how it is blocking
2023-10-27 17:27:51 <ikke> apk info -R httpie mercurial
2023-10-27 17:28:16 <ikke> apk version httpie mercurial
2023-10-27 17:28:19 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/OsDJbPiekcrhVSYGCDKlXFCr

2023-10-27 17:28:39 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/mXhdyRGvMZFzUXYrGpRgTiDX

2023-10-27 17:29:37 <ikke> apk add 'mercurial>=6.5'
2023-10-27 17:30:04 <telmich> same error:
2023-10-27 17:30:06 <telmich>  `` `... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JdcOUpPrtbPZMjsnMMUCJiMn>)
2023-10-27 17:30:14 <telmich>  * ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/zKpbLBnIzxPsOQPHKKZGGGlA>)
2023-10-27 17:30:42 <ikke> ok, and apk add 'python3>=3.11'?
2023-10-27 17:31:09 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/ZKjRTdxUiCsFXDBpNcHWfDIh

2023-10-27 17:31:27 <telmich> wait... did I not see the libffi issue some months ago already?
2023-10-27 17:31:50 <ikke> llvm11 is no longer in edge
2023-10-27 17:32:07 <ikke> I assume llvm11 is in world?
2023-10-27 17:32:45 <ikke> Oh, something depending on so:libffi.so.7 pulls it in?
2023-10-27 17:33:05 <ikke> telmich: did you provide --available when running upgrade?
2023-10-27 17:33:10 <ikke> `apk upgrade --available`
2023-10-27 17:33:38 <telmich> that is the same as apk upgrade -a, isn't it?
2023-10-27 17:33:48 <telmich> And it does indeed give the same error
2023-10-27 17:34:40 <ikke> ok, and yes
2023-10-27 17:35:39 <ikke> try: apk list -d so:libffi.so.7
2023-10-27 17:36:47 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/hCbfASTJCBlrVnpdhPGikBsG

2023-10-27 17:42:04 <ikke> can you remove llvm11?
2023-10-27 17:43:45 <telmich> yep:
2023-10-27 17:43:49 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/nPoAKyrTmyAUZVCcEmLLeHkU

2023-10-27 17:44:15 <ikke> sorry, llvm11-libs
2023-10-27 17:44:57 <telmich> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/cVVoAvNNkhAWinVWOvGrPIdF

2023-10-27 17:45:09 <telmich> So, no
2023-10-27 17:45:55 <ikke> mesa-dri-intel
2023-10-27 17:46:50 <telmich> uh, nice
2023-10-27 17:47:02 <telmich> after deleting mesa-dri-intel, apk upgrade -a is running through
2023-10-27 17:47:18 <telmich> (or at least running)
2023-10-27 17:47:23 <telmich> I am wondering
2023-10-27 17:47:47 <telmich> Is there something I did wrong on that system or is it a bug that prevents the upgrade to proceed on its own?
2023-10-27 17:54:57 <ikke> I don't think it's either
2023-10-27 17:55:40 <ikke> apk list --orphaned should list you these packages
2023-10-27 17:56:10 <ikke> telmich: the question is basically: is apk upgrade allowed to remove packages from world
2023-10-27 17:57:05 <ikke> because mesa-dri-intel was in world, everything else had to be kept in place to satisfy it's dependencies
2023-10-27 18:20:36 <telmich> Hmm
2023-10-27 18:20:58 <telmich> I'd have expected the error message at least to be that the reason is due to mesa-dri-intel
2023-10-27 18:26:06 <ikke> Yeah, would be nice if apk could tell exactly what package is the cause
2023-10-27 18:57:52 <WhyNotHugo> Can I map caps-lock to escape on a console (e.g.: not a terminal emulator, but the default console you get on installation media).
2023-10-27 19:09:10 <fluix> searching suggests using a keymap should do it
2023-10-27 19:27:51 <WhyNotHugo> I'll look at setup-keymap and see where that leads.
2023-10-27 19:30:15 <WhyNotHugo> Yeah, the files are the text files described in the man doc, simply put into a gz.
2023-10-28 00:03:40 <idkrn[m]> <fluix> "idkrn: they likely meant it in..." <- I don't have any follow up questions
2023-10-28 00:03:54 <idkrn[m]> I just wanted to know if people had done it successfully
2023-10-28 05:07:32 <qaqland> https://pkgs.alpinelinux.org/packages?name=cloudflared&branch=edge&repo=&arch=&maintainer=
2023-10-28 05:08:01 <qaqland> cloudflared can't be searched in both online and apk search
2023-10-28 05:10:20 <qaqland> build is disabled but maybe should showing it🤔
2023-10-28 05:33:38 <ptrc> qaqland: pkgs.a.o works only on built packages
2023-10-28 10:09:05 <frag> why cant links -g show images on nitter.net ? qutebrowser can show them (even without js)
2023-10-28 10:50:16 <dwfreed> works fine in debian
2023-10-28 11:04:22 <frag> someone suggested its because they use webp
2023-10-28 21:33:19 <add8> I set up hardware acceleration on Firefox using amdgpu
2023-10-28 21:34:00 <add8> Vaapi seems to be picked but no video plays, they get stuck on the first frame 
2023-10-28 21:34:20 <add8> I am using h264. What else can be wrong?
2023-10-28 21:34:23 <dwfreed> does anything else work, like mpv ?
2023-10-28 21:35:13 <quinq> Like video group
2023-10-28 21:36:37 <add8> I was about to download a test MP4 to check, but Firefox was able to play it using HA
2023-10-28 21:36:48 <add8> So maybe something to do with YT?
2023-10-28 21:37:31 <add8> quinq: user's in the video group, if that's what you mean to ask
2023-10-28 21:42:10 <add8> dwfreed: it does work with MPV as well
2023-10-28 21:52:37 <add8> Interestingly, the same YouTube video plays in mpv. Seems like an issue with Firefox
2023-10-28 22:54:33 <somercet1> add8, are you in X11 or Wayland?
2023-10-28 23:00:47 <add8> somercet1: X11, but I tried on Wayland too
2023-10-29 00:59:43 <somercet1> ugh, musl iconv can't handle win-1252, and GNU recode is still a mess. What do you guys use?
2023-10-29 01:09:39 <Hello71> tr?
2023-10-29 01:10:05 <Hello71> er, sed
2023-10-29 01:11:16 <urtzo> Hi, I'm having a problem to make X run properly on my RPi b+. Can somebody give me a hand?
2023-10-29 01:12:10 <Hello71> https://www.openwall.com/lists/musl/2013/06/26/5 says that musl supports windows1252
2023-10-29 01:12:57 <urtzo> The problem is: if I try to start X with the video acceleration enabled, I get a black screen; if I disable the video acceleration parameter, X runs fine.
2023-10-29 01:16:54 <urtzo> The video acceleration parameter on the Raspberry Pi is "dtoverlay=vc4-kms-v3d".
2023-10-29 01:20:16 <somercet1> Hello71, 'iconv -l' does not list it. It has ISO-8859-1 through 7, but not win1252
2023-10-29 01:41:25 <Hello71> it has "..."
2023-10-29 02:26:51 <kwadroke[m]> urtzo: I believe you have to use Fake KMS on the 1B+
2023-10-29 03:10:33 <urtzo> kwadroke[m]: Thank you. That worked.
2023-10-29 03:29:45 <snake> wow
2023-10-29 08:03:55 <qaqland> in alpine-script : setup-desktop, here it get all depends of gnome and pass to setup-xorg-base https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-desktop.in?ref_type=heads#L69 
2023-10-29 08:05:29 <qaqland> but `apk add` will auto install all depend packages https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-xorg-base.in?ref_type=heads#L55
2023-10-29 08:06:43 <qaqland> why do this function again :-) 
2023-10-29 08:07:22 <ikke> qaqland: gnome is a meta package, the idea is to explicitly install the packages it pulls in instead of the meta package
2023-10-29 08:07:35 <ikke> ie, the difference is what ends up in world
2023-10-29 08:11:33 <qaqland> maybe harder to see in world, "I install gnome by offical script but not a gnome in world there"
2023-10-29 08:12:53 <ikke> yes, but it's more stable. If someone decides to no longer ship a certain application with gnome by default, apk will automatically remove it from your system even though you still want it
2023-10-29 08:13:08 <ikke> For libraries that's normal, but for applications that's unexpected
2023-10-29 08:16:45 <qaqland> got it
2023-10-29 08:18:46 <qaqland> gnome is going to remove xorg, after that we need something like setup-wayland :p
2023-10-29 08:18:55 <qaqland> https://gitlab.gnome.org/GNOME/gnome-session/-/merge_requests/99#note_1865531
2023-10-29 10:08:25 <A4EVA> Hello, I would like to ask which script on alpine is in charge of cat the /etc/motd file
2023-10-29 10:24:07 <quinq> A4EVA, man motd
2023-10-29 10:56:05 <A4EVA> quinq, man do not tell me which script is actually cat out the /etc/motd 
2023-10-29 10:57:09 <quinq> It tells you login prints it, isn't that what you mean?
2023-10-29 10:57:10 <quinq> “The contents of /etc/motd are displayed by login(1)”
2023-10-29 11:08:03 <A4EVA> quinq:  thanks, you're right
2023-10-29 11:09:32 <A4EVA> I would like to print a coloured motd in the console but login is a binary so I do not think there is an easy way to do it 
2023-10-29 11:09:57 <quinq> You can use console color codes directly in your motd (that's a text file)
2023-10-29 11:11:50 <quinq> Search for which ones the Linux console supports
2023-10-29 12:34:32 <alethkit> So does tiny-cloud follow the user-data and meta-data conventions of cloud-init, or have any changes been made?
2023-10-29 12:49:53 <minimal> alethkit: for user-data I bleieve it uses "#alpine-config" rather than "#cloud-config" and is similar, rather than identical, to cloud-init user-data
2023-10-29 12:51:09 <alethkit> Are the differences documented?
2023-10-29 12:51:16 <alethkit> or is it a matter of trial and error?
2023-10-29 12:52:52 <minimal> I don't see any docs files in the tiny-cloud git repo so I guess the code is the documentation ;-)
2023-10-29 13:05:39 <minimal> alethkit: is there some specific aspect/functionality you're interested in?
2023-10-29 13:05:59 <alethkit> network configuration, primarily
2023-10-29 19:30:55 <linfan> It is a hard nut to crack this Alpine :-)
2023-10-29 19:31:29 <linfan> I do not seem to get changes saved and changing one thing, disables another.
2023-10-29 19:33:26 <linfan> I had a good sound and powered off. A few hour later the good sound was gone. And gtk apps would show up unreadable and all blurred.
2023-10-29 20:05:15 <tyler[m]123> <linfan> "I had a good sound and powered..." <- Are you using pipewire for sound? Works good for me. I also don't have issue with GTK, using Alpine as desktop for almost 2 years I think
2023-10-29 20:07:34 <linfan> i am just running it on virtualbox, have been fighting with it for four weeks daily, but it won't get stable.
2023-10-29 20:08:48 <tyler[m]123> For me it's useful to run `/usr/libexec/pipewire-launcher &` on `.xinitrc`
2023-10-29 20:08:56 <tyler[m]123> It makes audio work fine with pipewire
2023-10-29 20:09:09 <tyler[m]123> Before I added that it wouldn't work for me
2023-10-29 20:09:14 <linfan> well i have wayland so not any use to runs .xinitrc
2023-10-29 20:09:26 <tyler[m]123> As for GTK idk, it always worked for me
2023-10-29 20:09:33 <tyler[m]123> Doesn't wayland have like a start up script
2023-10-29 20:09:37 <tyler[m]123> Like xinitrc
2023-10-29 20:09:46 <linfan> not that i know of
2023-10-29 20:09:48 <tyler[m]123> Something that runs when you start a wayland session?
2023-10-29 20:10:00 <tyler[m]123> try running that manually then when you start wayland
2023-10-29 20:10:02 <tyler[m]123> maybe works
2023-10-29 20:11:15 <linfan> i need to get the start command alpine launches gnome wayland with gdm not set to default
2023-10-29 20:11:53 <linfan> personally i just hate gdm as fuck
2023-10-29 20:12:52 <linfan> but i could try install xorg, but that is not an option during install-desktop
2023-10-29 20:14:46 <tyler[m]123> install-desktop? i didn't even know there is that script
2023-10-29 20:14:47 <tyler[m]123> I just installed the things manually
2023-10-29 20:15:10 <linfan> sorry, setup-desktop
2023-10-29 20:18:41 <linfan> dunno what clashes there would be installing xorg, if setup-desktop:gnome is built around wayland. Maybe it is enough to just rm -rf /home/../*
2023-10-29 20:19:45 <linfan> or maybe there is some apk called xwayland or so
2023-10-29 20:20:56 <tyler[m]123> I never used wayland so I don't know
2023-10-29 20:21:12 <tyler[m]123> Maybe worth testing with X11
2023-10-29 20:21:35 <quinq> Also erasing your entire system doesn't sound very constructive
2023-10-29 20:21:36 <linfan> well, wayland is something gnome pushes
2023-10-29 20:22:03 <linfan> well, setting up a system takes about 5 min?
2023-10-29 20:22:14 <linfan> that is no big deal
2023-10-29 20:22:59 <quinq> But then recreate the fs instead
2023-10-29 20:24:37 <linfan> i dunno, tour weeks and every install results in poor wound quality, gtk apps getting unreadable, hahaaa, but maybe it is vbox. Even repaired windows 11 the host and reinstalled vbox
2023-10-29 20:25:28 <linfan> no problems whatsoever with any other distro on vbox, dunno what goes wrong
2023-10-29 20:25:56 <quinq> Who knows without the precise steps you're taking to install things
2023-10-29 20:26:55 <linfan> usually, setup-alpine, setup-desktop, install musl-locales lang, pipewire etc.
2023-10-29 20:28:20 <linfan> add dbus to default level... there shouldn't need to be anything else done
2023-10-29 20:30:46 <quinq> gnome is a beast though
2023-10-29 20:31:05 <quinq> IIRC they have dependency on systemd nowadays
2023-10-29 20:31:35 <linfan> but it might be the very wayland that gets fucked up, sometimes it does not save settings and session changes
2023-10-29 20:32:04 <linfan> yeah, rc is not linked to gnome as systemd is
2023-10-29 20:33:23 <linfan> but i could do a xfce install and check it. KDE i do not want either
2023-10-29 20:33:46 <linfan> i am on slackware in vbox just now on kde, lol
2023-10-29 20:37:06 <linfan> so i lean towards it is wayland on alpine that is tricky
2023-10-29 20:38:06 <linfan> tx, for the tips, cya
2023-10-29 23:04:59 <yann-kaelig> hello
2023-10-29 23:05:22 <yann-kaelig> Is there any reason that ejaberd is not available for armv7h  ?
2023-10-29 23:16:05 <quinq> There's always a reason
2023-10-29 23:19:06 <kwadroke[m]> from the APKBUILD: arch="all !x86 !armhf !armv7" # limited by elixir
2023-10-29 23:20:15 <kwadroke[m]> from Elixir's APKBUILD: # arm: build fails hundreds of times in a row on builders
2023-10-29 23:21:42 <kwadroke[m]> that is the "limited by elixir" is for EJabberd
2023-10-30 13:18:53 <ayushnix> should one use a single label name ("olympus") or an FQDN ("olympus.home.arpa") as the hostname?
2023-10-30 13:19:19 <ayushnix> does the choice matter in any sense?
2023-10-30 13:20:29 <ayushnix> the only practical consequence I can think of is that musl won't fallback to appending search domain to queries if an FQDN is used
2023-10-30 13:59:10 <lopid> use an fqdn if it's resolvable
2023-10-30 14:55:11 <ayushnix> lopid: it's resolvable only if I setup my stub resolver to resolve it
2023-10-30 14:55:43 <ayushnix> I'm looking for a reason why one choice might be preferable to the other
2023-10-31 12:17:52 <msiism> I've installed Pidgin on 3.18.4 and, apparently, it does not support XMPP. What could I be missing?
2023-10-31 12:24:59 <msiism> What I mean by "apparently doesn't support XMPP" is that you cannot select XMPP from the protocol drop-down menu when setting up a new account.
2023-10-31 12:34:04 <wiktor> msiism: is there a particular reason you're using pidgin? I think it's unmaintained and especially with regards to xmpp. Why not Gajim or Dino.im?
2023-10-31 12:35:11 <msiism> Because I haven't gotten around to sorting upgrading my XMPP usage to current standards and I also like Pidgin's interface.
2023-10-31 12:35:26 <msiism> s/sorting//
2023-10-31 12:35:32 <wiktor> ack, understood
2023-10-31 12:35:57 <msiism> Do Gajim and Dino support OMEMO?
2023-10-31 12:36:17 <wiktor> yes, both of them support it out of the box
2023-10-31 12:36:42 <msiism> And also proxy-ing over Tor?
2023-10-31 12:36:47 <wiktor> for some features you'll need server support too (like http upload)
2023-10-31 12:37:25 <wiktor> I think you need to manually set proxies ( https://dev.gajim.org/gajim/gajim/-/issues/7026 ), but it should work
2023-10-31 12:37:51 <wiktor> as for dino: https://github.com/dino/dino/issues/567
2023-10-31 12:38:00 <msiism> Thanks. I'll have a look.
2023-10-31 12:38:11 <wiktor> no probs
2023-10-31 12:39:07 <msiism> But still, the build log for Pidgin's package shows that XMPP support is built in. So, that's a bit strange.
2023-10-31 14:11:51 <omni> I had a look and, yes, it's not obvious
2023-10-31 14:11:59 <omni> you may need to add purple-xmpp
2023-10-31 14:14:35 <omni> msiism: https://git.alpinelinux.org/aports/tree/community/pidgin/APKBUILD#n29
2023-10-31 14:17:10 <omni> and I woudn't call it unmaintained, the current release is from earlier this year
2023-10-31 14:24:07 <msiism> omni: Okay, thanks.
2023-10-31 14:27:41 <omni> for omemo you may want to add purple-lurch
2023-10-31 14:29:33 <omni> I'm not using this myself btw and the other suggestions, dino or gajim, may work for you too
2023-10-31 14:36:09 <linfan> finally got it right on vbox. Installed xfce4 and uninstalled all of pipewire. Sound and all now perfectly working.
2023-10-31 14:37:17 <msiism> omni: I see. I'm still using OTR, currently.
2023-10-31 22:21:31 <nazar-pc[m]> Hey folks, I have ntpd service enabled and it is running, yet node-exporter reports that estimated error is 16 seconds and clock is not synchronized.
2023-10-31 22:21:31 <nazar-pc[m]> I have several Ubuntu hosts and they all work fine out of the box, with Alpine Linux I had to enable ntpd explicitly and it still doesn't seem to work properly.
2023-10-31 22:21:31 <nazar-pc[m]> What else should I check?
2023-10-31 22:27:38 <nazar-pc[m]> Hm, I just disabled ntpd and enabled chrony through setup-ntp and everything seems to be synced now, no idea why ntpd didn't work though
2023-10-31 22:47:14 <minimal> nazar-pc[m]: which ntpd? Busybox's?
2023-10-31 22:48:21 <nazar-pc[m]> minimal: Yes
2023-10-31 22:54:06 <minimal> it's a pretty basic/limited ntpd, I tend to use chrony all the time myself
2023-10-31 22:55:54 <invoked> same
2023-10-31 23:23:46 <nazar-pc[m]> <minimal> "it's a pretty basic/limited ntpd..." <- Makes sense, but it should be functional if it is an option in the OS, but it didn't seem to.