2026-03-01 01:28:27 <WhyNotHugo> The wiki says both "COSMIC is intended to be highly portable" and then that it requires eudev and elogind. That seems contradictory.
2026-03-01 07:13:19 <staceee> We have lutris on Alpine but why for? is there any game that launch?
2026-03-01 07:26:16 <staceee> it fails to decompress using zlib, it fails to detect vulkan
2026-03-01 08:17:46 <Saijin_Naib[m]> Because I think game play/preservation is important, and I wanted to try getting it working
2026-03-01 08:17:58 <Saijin_Naib[m]> Most games assume glibc, and gcompat doesnt provide the needed symbols
2026-03-01 08:18:05 <Saijin_Naib[m]> It is in testing for a reason
2026-03-01 08:18:41 <Saijin_Naib[m]> Same goes for minigalaxy
2026-03-01 10:19:54 <WhyNotHugo> staceee: latest wine implements WoW64 without needing multilib. Shouldn't 32bit games run fine now?
2026-03-01 10:20:13 <WhyNotHugo> Are the zlib and vukan issues "lack of support", or packaging issues?
2026-03-01 10:20:47 <staceee> lutris package doesn't depends on zlib, but I already had it. And lutris or py3 subpackages still fails to use it
2026-03-01 10:21:02 <staceee> not sure if there is some linux runtime in between that cause a problem
2026-03-01 10:21:28 <staceee> vulkan detection has been somehow fixed recently, but wine dxvk still can't be enable
2026-03-01 10:21:51 <staceee> I wasn't aware that wow64 some this 32 bit issue, but that seems great!
2026-03-01 11:11:24 <quinq> WINE works fine most of the time
2026-03-01 11:11:29 <quinq> 
ACTION launches Constance

2026-03-01 11:12:48 <staceee> at least the GOG installer fails
2026-03-01 11:13:46 <vv221> staceee, forget about actually running the GOG installers, innoextract is what you need.
2026-03-01 11:14:06 <quinq> Never had an issue with GOG installer
2026-03-01 11:14:20 <quinq> How does it fail?
2026-03-01 11:38:30 <staceee> zig failure
2026-03-01 11:38:45 <staceee> can't find zig.so.1 iirc
2026-03-01 12:15:45 <quinq> I don't have zig installed
2026-03-01 13:13:30 <staceee> ah! my bad, zlib
2026-03-01 13:14:37 <staceee> vv221: I'm not sure to follow, I meant the gog installer based game, in lutris
2026-03-01 13:22:11 <vv221> staceee, innoextract is a tool to extract the content of GOG installers without the need to run them through Wine. But I guess you were actually talking about the game, not the installer itself?
2026-03-01 13:22:44 <staceee> I know what innoextract does, but it is not the tool used when installing a game with lutris right?
2026-03-01 13:22:58 <vv221> Probably not, I don’t use lutris.
2026-03-01 18:51:29 <frag> when making a virtual package can one add tags? trying this but it says its not a valid package dependency : apk add --virtual surfv surf@testing vimb@testing mesa-gles@testing gst-plugins-good@testing
2026-03-01 18:55:53 <frag> guess have to add version instead like it says...
2026-03-01 19:01:29 <achill> frag: https://gitlab.alpinelinux.org/alpine/apk-tools/-/issues/10902
2026-03-01 19:09:03 <frag> oh, neat
2026-03-01 19:23:19 <invoked> apologies. is there a release plan for v3 index and packages? i probably was asleep when that discussion took place
2026-03-01 19:23:46 <invoked> just curious
2026-03-01 19:24:48 <ikke> It would probably involve first a side-by-side generation of both v2 and v3 indexes
2026-03-01 19:25:25 <invoked> ok. sounds like it's not moving right now
2026-03-01 19:26:02 <invoked> which is understandable with the usr-merge and other stuff happening
2026-03-01 19:26:10 <ikke> Problem is that we have some systems that rely on the v2 index format
2026-03-01 19:32:07 <invoked> ah
2026-03-01 19:51:09 <RoadRoller01[m]> hello
2026-03-01 19:53:24 <RoadRoller01[m]> for some reason now my pipeline errors earlier then usual with this error:... (full message at <https://matrix.org/oftc/media/v1/media/download/AZ3teGGPghimjkGoloRKfbKQCVRk-dFQ2lDXmHHa36L3N2L_oInHuiqXYNcVW9-AzmJVbRZkSk9thvRJscoNAPxCec8w97qQAG1hdHJpeC5vcmcva05sWGVQWHR5SW5nUUV2cE1RVG9zZXN6>)
2026-03-01 19:53:42 <RoadRoller01[m]> https://gitlab.alpinelinux.org/RoadRoller01/aports/-/jobs/2241154#L83
2026-03-01 19:55:52 <RoadRoller01[m]> it wasn't erroring with this error the last night https://gitlab.alpinelinux.org/RoadRoller01/aports/-/jobs/2237760#L2330 
2026-03-01 20:13:13 <ikke> RoadRoller01[m]: glslang has been updated, but some packages still need to be rebuilt
2026-03-01 20:13:19 <ikke> RoadRoller01[m]: Nothing wrong with your MR
2026-03-01 23:32:39 <Saijin_Naib[m]> @WhyNotHugo linux native games fail because gcompat is missing needed symbols. Wine does indeed work better
2026-03-01 23:33:01 <Saijin_Naib[m]> @staceee @vv221 help welcome for lutris, minigalaxy, and reviving dxvk
2026-03-01 23:34:05 <vv221> Sorry, I play games through Wine *a lot*, but not on Alpine (Debian for now, switching to Gentoo), and I do not use neither lutrix nor minigalaxy (I hate all Steam clones).
2026-03-01 23:35:49 <Saijin_Naib[m]> Okay
2026-03-01 23:38:06 <vv221> I plan on giving a try on Alpine for gaming, but I need to update all my tools first, it’s going to take a while (several months at least).
2026-03-01 23:39:30 <vv221> When I do, I’m going to need dxvk for a couple games, so at that point I would gladly join its maintenance effort (keeping in mind that unlike many people, I do not use it by default, only for games actually requiring it).
2026-03-01 23:40:52 <invoked> hope you're ready to patch a lot of things.
2026-03-01 23:43:06 <Saijin_Naib[m]> I have not been able to get it to build properly yet
2026-03-01 23:43:20 <Saijin_Naib[m]> I'll restore my broken aport so you can hack on it if you wish
2026-03-01 23:43:34 <Saijin_Naib[m]> We might also benefit from d7vk
2026-03-01 23:45:22 <vv221> I’m not sure to get the benefit of d7vk, but I’m not using dxvk for direct3d 8 and 9 already, only select games on D3D10 and 11.
2026-03-01 23:45:43 <vv221> (most of my D3D9 → D3D11 I play with regular wined3d)
2026-03-01 23:46:14 <vv221> No hurry in restoring the aport you worked on, I probably won’t have opportunities to help before at least a couple months.
2026-03-01 23:46:51 <vv221> invoked, I’ve been tinkering with running games on Linux before Wine 1.0, before even multiarch, so I’m used to many things requiring a lot of effort ;)
2026-03-01 23:48:50 <invoked> sure. i'm just saying, if that's with glibc then you never had to patch the way you'll have to now
2026-03-01 23:49:11 <invoked> i'm assuming you stubbed out a bunch of things with wine 1.0 and this will be ... different
2026-03-02 00:19:11 <orion> hello
2026-03-02 00:23:20 <madtypewriter[m]> vv221: I certainly don't want to discourage more native approaches, or indeed other approaches in general, but I can tell you that if you just want things to work, Alpine runs Flatpak fine, and the Flatpak of Steam works quite well, and for a long time that stack was my gaming PC. (And once I finish setting up my freshly-reinstalled Steam Deck on Alpine, it might be again...:P)
2026-03-02 00:31:04 <vv221> madtypewriter[m], I did not use Steam in 2003, I’m not going to use it in 2026 ;)
2026-03-02 00:31:58 <vv221> (actually I might not gave known about it before 2004, when it was enforced on Half-Life player)
2026-03-02 00:33:08 <vv221> My gaming setup has been full Linux since ~2009 (Linux + Windows dual boot for 2007-2008), and fully DRM-free since that time too.
2026-03-02 00:33:27 <madtypewriter[m]> That is a good goal:)
2026-03-02 00:34:02 <madtypewriter[m]> Although FWIW there are also decent native+FOSS games in flatpak, too
2026-03-02 00:34:29 <vv221> Flatpak here is because it includes an embedded copy of glibc?
2026-03-02 00:34:38 <vv221> Or does it come with some other benefit?
2026-03-02 00:35:04 <madtypewriter[m]> That's certainly the big one
2026-03-02 00:35:21 <madtypewriter[m]> I also - especially for Steam! - like the sandboxing
2026-03-02 00:35:48 <madtypewriter[m]> There's no reason random video games should be able to see arbitary files in ~
2026-03-02 00:36:07 <vv221> I mostly rely on unshare for sandboxing. But I don’t run Steam, so my attack surface is much smaller.
2026-03-02 00:36:29 <madtypewriter[m]> That's fair
2026-03-02 00:36:46 <madtypewriter[m]> I've also been meaning to try https://flathub.org/en/apps/com.heroicgameslauncher.hgl for GOG
2026-03-02 00:36:48 <vv221> I understand that it is a very atypical way to handle games ;)
2026-03-02 00:37:02 <madtypewriter[m]> huh
2026-03-02 00:37:04 <madtypewriter[m]> *heh
2026-03-02 00:37:36 <madtypewriter[m]> This is true. But not bad.
2026-03-02 00:39:13 <madtypewriter[m]> I remain confused at the number of people who don't bother to sandbox their browser. Y'know, the thing that represents 90% of the actual attack surface on their machine.
2026-03-02 00:39:51 <vv221> Here I tend to use an atypical defense, once again: do as few things as possible in a Web browser.
2026-03-02 00:40:10 <vv221> With, of course, very strong content blocking rules on top.
2026-03-02 00:40:47 <madtypewriter[m]> :) Likewise, that also works
2026-03-02 00:41:50 <vv221> Back to games, I think I finally fixed Heroes of Might and Magic V current GOG.com build \o/
2026-03-02 00:42:36 <vv221> For some reason, their build from last July broke the base game with Wine (nobody complains, I guess everyone plays the extensions now, or maybe it only happens with Wine default OpenGL renderer).
2026-03-02 01:50:12 <invoked> vv221: that is generally the most workable approach. freebsd basically takes this approach, having the same basic problem as musl linux. it has a linux emulator and just imports glibc + friends
2026-03-02 01:51:10 <vv221> Sorry, I’m not sure to get what "that" is referring to in "that is generally the most workable approach."
2026-03-02 01:51:15 <vv221> Libraries embedding?
2026-03-02 01:51:38 <invoked> that = flatpak
2026-03-02 01:51:52 <vv221> OK, got it.
2026-03-02 01:52:20 <vv221> But I’m not sure what flatpak actually does outside of libraries embedding and some sandbox.
2026-03-02 01:52:23 <vv221> (I never used it)
2026-03-02 01:53:46 <invoked> it handles file distribution and configuring the sandbox on a per-app basis, and has an interface for portals
2026-03-02 01:53:51 <invoked> that's roughly it
2026-03-02 01:54:45 <invoked> i'm definitely not advocating for flatpak, though, or having compat layers
2026-03-02 01:54:49 <vv221> Right, I skipped the distribution part because I was mostly focusing on what it does when focusing on a specific app.
2026-03-02 01:55:20 <vv221> Well, I’m mostly interested in what flatpak does to see what could be replicated without it ;)
2026-03-02 01:56:20 <invoked> primarily as it concerns what you're trying to do, it imports glibc and the freedesktop platform
2026-03-02 01:56:43 <invoked> so, you know, mesa and all of it.
2026-03-02 01:57:38 <vv221> Are LD_* env variables something specific to GNU tooling, or does Alpine have them too?
2026-03-02 01:57:57 <invoked> ubuntu binaries iirc
2026-03-02 01:58:14 <invoked> (i don't have a flatpak install in front of me currently)
2026-03-02 01:58:55 <invoked> mainly you need to think in terms of musl vs glibc
2026-03-02 01:59:01 <invoked> what musl doesn't have
2026-03-02 02:00:55 <invoked> there could be some userland/busybox things too. but coreutils is available
2026-03-02 02:02:02 <vv221> I’m too early in my use of musl to get a good idea of the differences. That’s why I asked above about ld.so stuff, and am probably going to keep asking questions about really basic stuff in the coming weeks.
2026-03-02 02:02:34 <invoked> i don't have the answers, i'm just pointing you in a direction that is probably correct
2026-03-02 02:02:56 <vv221> Since I’ve only used GNU + Linux for the last 20 years, I’m still hazy about the limits of each part of these.
2026-03-02 02:03:06 <invoked> musl is really the issue here, and we're a musl distro
2026-03-02 02:03:26 <vv221> That part I got ;)
2026-03-02 02:03:34 <invoked> and you're not going to get musl to change, so, you'll need to patch everything to work with musl somehow
2026-03-02 02:04:04 <vv221> Do you mean that musl is closed to contributions?
2026-03-02 02:04:21 <invoked> no, but it's considered finished for the most part
2026-03-02 02:04:55 <invoked> afaik
2026-03-02 02:05:06 <vv221> Nice, I’m not used to a lot of devs who accept the idea of finished software, I wish I would see that more often.
2026-03-02 03:12:53 <Saijin_Naib[m]> Maybe contributing needed functions and symbols to gcompat could be a path
2026-03-02 03:20:13 <vv221> Yes, that’s what sounds like the more appealing option to me for now.
2026-03-02 03:54:49 <WhyNotHugo> Saijin_Naib[m]: I hate to say it, but when it comes to proprietary games, wine is better than native builds. 
2026-03-02 03:55:09 <WhyNotHugo> vv221: can you share that unshare setup? I've tried using docker, but also have weird glitches which I can never narrow down.
2026-03-02 03:59:30 <Saijin_Naib[m]> @WhyNotHugo yeah, as others have (not really joked) win32 is the only stable linux ABI
2026-03-02 04:00:25 <Saijin_Naib[m]> Still, it would be nice to have better native linux glibc support for games and other proprietary tools we may need to make use of
2026-03-02 04:10:22 <WhyNotHugo> Given that you'd want to sandbox them anyway, why not use something like docker/podman?
2026-03-02 04:40:17 <Saijin_Naib[m]> Pain in the ass factor plus on lower-end devices, the storage hit isnt tenable. Exactly what motivated me to try with minigalaxy and lutris
2026-03-02 04:40:53 <Saijin_Naib[m]> Needing tens of gbs of flatpak runtimes to run a handful of mbs of programs got old on 64GB eMMC
2026-03-02 04:42:50 <Saijin_Naib[m]> As for sandboxing, eh. Not really my threat model to worry about running zoom or something outside of a sandbox. The account/service is already insanely invasive 🤷
2026-03-02 04:54:33 <WhyNotHugo> Sent a draft MR with a dxvk aport. Seems to work
2026-03-02 05:17:23 <Saijin_Naib[m]> Sweet!
2026-03-02 06:59:07 <staceee> I'm completely in favor of improving gaming support in aports. I use flatpak for years, and I don't like this. Will help as I can
2026-03-02 07:05:40 <staceee> about the glibc issue, if I'm right lutris comes with a linux runtime. Maybe it is okay enough
2026-03-02 07:29:36 <hazardaj_nombroj> bottles is also nice, though its devs outright advise people to use flatpak
2026-03-02 08:50:35 <hoel> vv221: https://wiki.musl-libc.org/functional-differences-from-glibc
2026-03-02 09:27:21 <quinq> Also, musl is not closed at all for changes
2026-03-02 09:27:41 <quinq> But the changes have to make sense according to the project goals
2026-03-02 09:28:19 <quinq> (of which being bug-compatible with GNU libc isn't one)
2026-03-02 10:42:20 <monotux> I'm looking to setup raspberry pi headless, what are people using for this? I've tried a few tools like alpine-linux-headless-bootstrap and apkovl-builder, but the first tool broke upon restart and the second can't seem to generate a network/interfaces file no matter what I do :D
2026-03-02 10:43:05 <monotux> I have a rpi 4b already setup (manually with kbd/monitor) but would like to just build a apkovl and put it on a sd card, when ssh and deal with the rest
2026-03-02 12:27:10 <vv221> WhyNotHugo, I don’t have a common unshare setup, for now I’m still experimenting with it. I use it mostly to work around bugs, like this clock-related one: http://git.dotslashplay.it/games-point-and-click/tree/games/play-return-to-mysterious-island-2.sh#n179
2026-03-02 12:29:14 <vv221> My baseline is usually `unshare --map-current-user --net`, to prevent network access.
2026-03-02 12:44:08 <invoked> quinq: yeah. i should have mentioned that changes to posix would go in. and "widely agreed-upon extensions" but i doubt glibc stuff to support gaming would be agreeable
2026-03-02 12:48:57 <quinq> yeah
2026-03-02 19:48:25 <karmmah[m]> monotux: I set up a pi zero 2 headlessly with alpine recently over ssh using the "installation on a headless host" guide on the alpine linux wiki. It worked well with the instructions however I got hung up on the restart after the initial setup. The reason was that during "setup-alpine" I did not set up the wlan0 interface since I thought I didn't have to do it since I was already connected over wifi. However if you don't set it up
2026-03-02 19:48:25 <karmmah[m]> again it won't be able to connect to wifi after the reboot.
2026-03-03 00:46:39 <RoadRoller01[m]> hello
2026-03-03 00:47:11 <RoadRoller01[m]> i would love to feature request something with CI
2026-03-03 00:47:50 <RoadRoller01[m]> there should be a way to enable ccache for a weak interval in and be disabled by default 
2026-03-03 00:49:03 <RoadRoller01[m]> This could be utilized  by big pkgs like in example:
2026-03-03 00:49:19 <RoadRoller01[m]> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/98116
2026-03-03 00:50:31 <RoadRoller01[m]> i have ran the pipe line over ten times in a weak
2026-03-03 00:50:41 <RoadRoller01[m]> and each one take 5 hour 
2026-03-03 00:54:32 <RoadRoller01[m]> its resource consuming ccache for one weak (or 1 or 2 days after the last pipe line had stoped) could really optimize that; it would release a lot of vcpu to other pipe lines to use
2026-03-03 00:55:20 <RoadRoller01[m]> and release a lot of time for the developer(: to other tasks he could do 
2026-03-03 01:10:01 <RoadRoller01[m]> s/in//
2026-03-03 01:10:44 <RoadRoller01[m]> s/pipe/pipeline/, s/line//
2026-03-03 01:11:46 <RoadRoller01[m]> s/or/-/, s/pipe/pipeline/, s/line//
2026-03-03 01:11:56 <RoadRoller01[m]> * its resource consuming ccache for one weak (or 1-2 days after the last pipeline had stoped) could really optimize that; it would release a lot of vcpu to other pipelines to use
2026-03-03 13:47:53 <monotux> karmmah[m]: thank you!
2026-03-03 15:18:45 <elagost> I've never written a man page before but I'm aware there are a few different formats or tools that can generate them. Is there a preferred tool?
2026-03-03 15:39:01 <Habbie> whatever tool you pick, the output will be nroff+mandoc
2026-03-03 15:39:06 <Habbie> so some people prefer to just write that
2026-03-03 15:52:47 <vv221> Having tried, I have difficulties understanding these pople ;)
2026-03-03 15:53:00 <vv221> (but I have no good alternative to suggest yet)
2026-03-03 16:14:16 <paulgupta> I've done "setup-interfaces"
2026-03-03 16:15:20 <paulgupta> It set up ipv4 with a static IP and netmask. How do I get dhcp to snatch an IPV6 address from my VPS provider?
2026-03-03 16:15:54 <paulgupta> I'm kind of new to running ipv6 please be kind. I'm trying to join this century.
2026-03-03 16:18:27 <paulgupta> It has set up a fe80 address, which I believe is just a link-local address.... Not sure why it didn't try and snag a DHCP lease for ipv6 as well. 
2026-03-03 16:20:06 <ikke> For DHCPv6 you need something like dhcpcd
2026-03-03 16:20:23 <ikke> Did they indicate they use dhcpv6?
2026-03-03 16:20:57 <paulgupta> does udhcpd included with busybox not work?
2026-03-03 16:21:17 <paulgupta> I'm honestly not 100% that they even offer ipv6, but I'd like to try and get an address without shutting myself out of my server?
2026-03-03 16:21:49 <paulgupta> I'm also not sure why my public internet address ipv4 has a netmask of 255.0.0.0. ...I feel like I set that up wrong but it's somehow working?
2026-03-03 16:21:53 <paulgupta> is that even possible?
2026-03-03 16:22:18 <ikke> Technically it would be possible, but unlikely.
2026-03-03 16:26:02 <ikke> paulgupta: I think at least on alpinelinux, it only invokes udhcpc, which only does ipv4
2026-03-03 16:32:47 <paulgupta> aahhhh, it's in a location that doesn't offer ipv6!!! Silly racknerd...lol
2026-03-03 16:40:04 <invoked> 6over4 is also a possible thing. it might be on our wiki, dunno
2026-03-03 19:12:01 <frag> to paste PRIMARY into librewolf adresse line, you have to move the mouse there and click middle button?
2026-03-03 19:44:39 <ssm> shift insert works maybe?
2026-03-03 20:44:05 <frag> nope
2026-03-03 20:45:25 <invoked> shift-ins works for me
2026-03-03 20:45:44 <invoked> foot -> librewolf
2026-03-03 20:51:25 <invoked> i take that back. tmux was putting selections into PRIMARY. with foot (no tmux) i have to control-shift-c the selection and then shift-insert in librewolf
2026-03-03 20:52:35 <invoked> i still do everything in tmux, fortunately or not.
2026-03-03 20:53:16 <dwfreed> shift-insert in firefox (I assume librewolf would be the same here) is pasting from the clipboard buffer, not the primary selection buffer
2026-03-03 20:53:27 <invoked> yeah i was just about to correct myself (again)
2026-03-03 20:53:29 <vv221> Oh, that might be why I never noticed these problems seemingly related to Wayland. I to do use tmux all the time.
2026-03-03 20:53:37 <vv221> (automatically spawned with my sehll)
2026-03-03 20:56:28 <invoked> tmux in the middle is useful but it's not really how things should work.
2026-03-03 21:00:39 <invoked> frag: back when i did a bunch of stuff in xephyr, i had a few helper binds that used xsel(1)
2026-03-03 21:01:13 <invoked> you could kludge something that way
2026-03-03 21:03:30 <invoked> something like xsel -o -p | xsel -i -b
2026-03-04 00:01:44 <ffoss_> elagost: maybe scdoc https://git.sr.ht/~sircmpwn/scdoc
2026-03-04 00:05:54 <vv221> I assume this is an example of the scdoc syntax? https://git.sr.ht/~sircmpwn/scdoc/blob/master/scdoc.5.scd
2026-03-04 00:39:38 <elagost> ffoss_: I decided to use scdoc. Thanks. It's a very clean format!
2026-03-04 00:56:29 <vv221> To reduce the load on Alpine mirrors, is there some APK cache server I could run for my LAN, similar to apt-cacher-ng for APT/Debian?
2026-03-04 00:56:54 <dogg0> i use nginx 
2026-03-04 00:58:19 <vv221> I’m not sure to get it. nginx set as a proxy, and the real mirror as the backend?
2026-03-04 00:58:30 <vv221> Then relying on nginx cache options?
2026-03-04 01:01:11 <dogg0> i couldnt figure out how to use apt-cacher-ng for alpine-apk, so i just used nginx to do the same caches the packs on my lan
2026-03-04 01:01:33 <dogg0> sorry i dont know the proper terms
2026-03-04 01:06:40 <dogg0> my working config if it helps you https://paste.debian.net/hidden/d53671db
2026-03-04 01:08:42 <vv221> it helps a lot, thanks for sharing it!
2026-03-04 01:08:52 <vv221> It’s a smart way to handle that, I like it.
2026-03-04 01:09:16 <dogg0> if you can improve share please
2026-03-04 01:12:27 <vv221> Sure, when I set that up (taking inspiration from your config), I’m probably going to share the config afterwards.
2026-03-04 01:14:32 <dogg0> and i put this into nginx.conf to see hit/miss to make sure its work https://paste.debian.net/hidden/a8adff1d
2026-03-04 01:14:45 <dogg0> gl
2026-03-04 01:15:35 <dogg0> im pretty sure we could add apk to acng, but i couldnt figure it out
2026-03-04 01:15:52 <dogg0> gave up easy heh
2026-03-04 01:16:35 <vv221> To be fair, you nginx-based setup seems more interesting than apt-cacher-ng.
2026-03-04 01:16:55 <dogg0> 'log format' must come before 'access log' or it doesnt work
2026-03-04 01:17:23 <dogg0> i like the acng stats tho
2026-03-04 01:17:45 <vv221> I used to like them… then realised I did not check them in years ;)
2026-03-04 01:18:03 <dogg0> ya true
2026-03-04 01:18:19 <dogg0> nice thing nginx works good for bsd too or whatever really
2026-03-04 01:19:00 <dogg0> in alpine setup-desktop works in 5seconds now 
2026-03-04 01:19:37 <vv221> Is that the script installing/configuring stuff to run Alpine as a desktop?
2026-03-04 01:19:59 <vv221> I learned about the setup-* stuff a bit late, after having done most stuff I wanted by hand.
2026-03-04 01:20:30 <dogg0> yes and it works soo nicely too.  1 command, then reboot and youre on the desktop
2026-03-04 01:21:02 <dogg0> im using plasma
2026-03-04 01:21:54 <vv221> Here I run Alpine on servers, so with more of a focus on a minimal set of packages installed. And it seems to be really good at that, especially with the subpackages, something I had not seen outside of Debian (and derivatives).
2026-03-04 01:22:36 <dogg0> oh ya i love alpine for my vms, just recently needed some DE and alpine done good
2026-03-04 01:24:09 <vv221> For the desktop/laptop I’m most probably going with Gentoo, I’ve been waiting half a dozen years for an opportunity to give it a try.
2026-03-04 01:24:20 <vv221> (I’ve been using nothing but Debian in the last ~17 years)
2026-03-04 01:27:39 <dogg0> debian o7
2026-03-04 09:49:12 <frag> hmm plugged in a memstick, sdb, and it says i dont have a valid partition table, after a while i also checked an already mounted device, sda, and it says the same :S last entries in dmesg https://0x0.st/Pjz9.txt
2026-03-04 09:50:05 <frag> its and diskless rpi, and dmesg also says i should run fsck on mmcblk0p1 i found while investigating...
2026-03-04 10:38:47 <hoel> frag: I think these entries are for your mouse.
2026-03-04 10:48:23 <longnoserob[m]> what does lsblk or fidks -l report?
2026-03-04 12:21:26 <mrbubze> Does Slackware current isos from alienbob have plasma 6 or plasma 5
2026-03-04 12:22:23 <longnoserob[m]> why do you ask this on an alpine-linux related channel?
2026-03-04 12:22:44 <mrbubze> Shit wrong channel sorry
2026-03-04 13:11:23 <f_> lopid: vv221: IIRC openrc has experimental parallel starting
2026-03-04 13:11:35 <f_> oh wait nevermind I was scrolled up
2026-03-04 13:11:41 <lopid> indeed
2026-03-04 13:16:55 <vv221> f_, right, I plan on giving it a try.
2026-03-04 13:17:53 <vv221> I’d still like to have other option for ordering than editing the daemon scripts themselves, especially for things like "start as soon as possible" or "start after everything else".
2026-03-04 13:18:20 <vv221> I thought about using custom runlevels, but these do not seem as hackable as I hoped.
2026-03-04 13:19:22 <vv221> (I did not really want to replace "default" with another runlevel, but to add extra runlevels in the chain)
2026-03-04 13:57:13 <Shiz> vv221: you can at least override specific fields using an /etc/conf.d/ file without needing to edit the init.d script
2026-03-04 14:00:23 <vv221> Sure, and I really like that system. But it is not going to help here.
2026-03-04 14:01:23 <vv221> Due to my LXC containers taking longer to start/stop than any other daemon, and mostly relying on the rest of the system being ready, I would like to have them as their own runlevel, started after / stopped before the "default" runlevel.
2026-03-04 14:01:51 <vv221> But I don’t really see how I can add a runlevel, instead of replacing the default runlevel with another one.
2026-03-04 14:02:16 <vv221> (keeping in mind that in this specific context, I’m using SysV as init, not openrc-init nor busybox init)
2026-03-04 15:14:46 <invoked> chromium and chrome switching to releases every 2 weeks in september.
2026-03-04 15:21:19 <lopid> ought to release when necessary, otherwise it seems releasing a release's sake
2026-03-04 15:21:28 <lopid> for
2026-03-04 15:22:18 <invoked> current is every 4
2026-03-04 16:49:05 <vv221> lopid, community-driven software tend to release "when it’s done", corporate software on the other hand, well, not so much ;)
2026-03-04 16:49:17 <lopid> mmhmm
2026-03-04 16:49:32 <vv221> Remember when Firefox started to number its releases as "Chrome + 1" just so it looked more modern?
2026-03-04 16:50:00 <lopid> no but yeah
2026-03-04 16:51:12 <vv221> I think I stopped believing version numbers mean anything when I started working with commercial video games… I have seen such horrors!
2026-03-04 16:52:18 <lopid> just look at the number of oss projects that haven't even reached version 1
2026-03-04 16:52:28 <lopid> people are scared of it
2026-03-04 16:52:40 <vv221> That is another kind of silliness, probably unrelated to the other one ;)
2026-03-04 16:52:48 <lopid> 
ACTION nods

2026-03-04 16:52:54 <vv221> But you’re right, 0 A.D. just left "alpha".
2026-03-04 16:53:15 <vv221> But they still are in 0.x, just no longer on 0.0.x
2026-03-04 16:53:36 <vv221> For my own software I did something very simple: the first build I share with the world, is, by definition, 1.0.
2026-03-04 16:53:51 <vv221> And I shared the first build that could actually be used to do something.
2026-03-04 16:54:25 <lopid> yes
2026-03-04 16:55:06 <vv221> I do have some 0.x too… but these are restrospective numbering, of archives I found after the fact and only shared for historic curiosity.
2026-03-04 16:55:24 <vv221> When I did actually work on them, they had no version number at all.
2026-03-04 23:20:45 <deplorable_pretzel> hi, is there a guide for installing alpine with detached luks header? if not, would appreciate general guidance/advice on doing so
2026-03-04 23:25:01 <deplorable_pretzel> also, im currently on edge and im considering switching back to stable, is there a way for me to know which packages are edge/testing only?
2026-03-04 23:43:27 <longnoserob[m]> deplorable_pretzel: there is probably a clever apk syntax for it
2026-03-05 01:26:31 <WhyNotHugo> deplorable_pretzel: stable will also have older package versions which yuo'd need to downgrade. Maybe `apk upgrade -a` would roll back to older versions?
2026-03-05 01:27:43 <shrizza> asdfa
2026-03-05 01:37:18 <vv221> With a stable release twice a year, the smartest move might be to stay on edge until next stable, then stick to that stable release.
2026-03-05 01:37:41 <vv221> That would prevent problems related to downgrades.
2026-03-05 02:22:05 <Saijin_Naib[m]> Use the latest-stable tag unless you need edge packages?
2026-03-05 02:49:57 <deplorable_pretzel> @WhyNotHugo @Saijin_Naib[m] I will be reinstalling alpine on a new laptop and wanted to determine if i'll be dropping any packages when switching to stable
2026-03-05 02:51:48 <deplorable_pretzel> ive manually checked one package and it seems its in testing, if i pin a package, will the deps be pinned too? the wiki seems to be down btw
2026-03-05 07:08:46 <mxbob> Hey good morning. Is anyone else experiencing issues with wiki.alpinelinux.org the last days? NGINX reports 504 errors quite often.
2026-03-05 07:13:22 <longnoserob[m]> was having no issues lastnight
2026-03-05 07:17:05 <jvvv> just loaded it. was slow to load the main page, but after seems responsive enough.
2026-03-05 07:21:40 <mxbob> Yeah, right now it works for me as well but yesterday and this morning I had issues. Nevermind.
2026-03-05 08:15:48 <vv221> Sounds like the never ending LLM bots DDoS…
2026-03-05 10:37:13 <frag> is there an ai in repo that you can use on local txt/pdf etc?
2026-03-05 10:38:59 <ssm_> ollama works if python works. not sure why you'd want to do that to yourself or your information integrity though
2026-03-05 10:39:25 <ssm_> one must imagine a boot with 7 toes stomping on a face with 2 noses forever
2026-03-05 10:40:09 <courge_musque[m]> vllm always has failed me on Alpine, but it could be a skill issue on my side
2026-03-05 10:42:43 <vv221> (ugh, I thought I would have avoided that LLM crap when joining the current channel…)
2026-03-05 10:43:10 <frag> abondon all hope, there is no escape
2026-03-05 10:44:06 <vv221> Oh, if all goes to worst, I still have my own channel where I can ban on a whim ;P
2026-03-05 10:44:35 <vv221> But I’ve already heard that TINA thing, and for some reason never really believed in it even the first time.
2026-03-05 11:02:01 <f_> there's a very nice LLM built entirely on natural resources that can be used on local txt/pdf, it's called the human brain
2026-03-05 11:02:57 <vv221> Please don’t reduce the human brain to some basic token approximation system ;)
2026-03-05 11:03:05 <f_> oh it's more than that
2026-03-05 11:03:29 <f_> it's very, very advanced for its time
2026-03-05 11:03:42 <courge_musque[m]> My best SUPER AI TOOL for txt files is grep
2026-03-05 11:03:55 <courge_musque[m]> It works wonder every time I need something
2026-03-05 11:03:58 <vv221> Yet, for some reason, we have companies (and invidiuals) trying everythin to prevent us from using this Brain technology…
2026-03-05 11:05:10 <courge_musque[m]> Of course, because you don't need to pay a subscription to use it
2026-03-05 11:05:33 <courge_musque[m]> Else, they would be promoting the brain as the newest cool thing
2026-03-05 11:05:38 <vv221> In my country I do have to pay one. They call it "health insurance".
2026-03-05 11:06:41 <courge_musque[m]> And your brain stops when you stop to pay it? What a wonderful country
2026-03-05 11:07:31 <vv221> Not immediately, but if I stop paying I’m cut off from the support contract. Meaning that if anything start to dysfunction, I’m fucked.
2026-03-05 11:10:08 <achill> i think this is getting quite offtopic
2026-03-05 11:12:22 <courge_musque[m]> @vv221 I just remembered that we live in the same country, so why did I ask? Sorry
2026-03-05 11:12:31 <courge_musque[m]> [@_oftc_achill:matrix.org](https://matrix.to/#/@_oftc_achill:matrix.org) you're right
2026-03-05 11:17:40 <vv221> Oh, right. Is your nickname related to a Cucurbitaceae sporting a mask?
2026-03-05 11:17:53 <vv221> (that I think would not be well-known outside of the country)
2026-03-05 11:20:46 <courge_musque[m]> I guess that people in the north don't know the courge musqué, but it's a common courge to eat in autumn in the south, maybe because there's more sun (musqué, not masqué)
2026-03-05 11:22:44 <vv221> Oh, I did not know about it, it is most probably the inspiration for the comics character I had in mind.
2026-03-05 11:23:08 <vv221> (https://fr.wikipedia.org/wiki/Le_Concombre_masqué)
2026-03-05 11:23:15 <Sertonix[m]> #alpine-offtopic
2026-03-05 16:18:11 <Quantum_3[m]> quick question but why is samurai still being used for ninja? its quite old at this point (1.9.0 vs the latest 1.13.2)
2026-03-05 16:18:42 <lopid> this is a rather outdated and off topic question
2026-03-05 16:19:59 <achill> lopid: no it is not?
2026-03-05 16:20:28 <lopid> you expect ninjas to use uzis?
2026-03-05 16:26:19 <ikke> https://lists.alpinelinux.org/~alpine/devel/%3CCAAOiGNwnzTomzOQ6YjH3t7VHmbfMRZUNXap8GM0Gy_mGS7aKhg%40mail.gmail.com%3E
2026-03-05 16:26:27 <ikke> https://lists.alpinelinux.org/~alpine/devel/%3CBYZDB5LH4OVY.9FDVF5I3J954%40homura%3E
2026-03-05 16:28:42 <achill> but it actually not being updates is a very very good point
2026-03-05 16:29:08 <achill> e.g. for building mesa missing dyndep support is really not great
2026-03-05 16:29:24 <achill> and the PR for that is now open for almost 6 years
2026-03-05 16:29:57 <achill> and there are a lot of other users who have seen broken behaviour between samurai/ninja
2026-03-05 16:30:35 <achill> i plan to bring that up for the TSC
2026-03-05 16:31:12 <ikke> One thing that samurai supports and ninja not is the equivalent of SAMUFLAGS
2026-03-05 16:31:24 <Quantum_3[m]> ikke: i see
2026-03-05 16:34:42 <achill> ikke: too bad, looks like they actually dont want to support that https://github.com/ninja-build/ninja/issues/1482
2026-03-05 16:35:09 <ikke> yes, exactly, which was one of the drivers switching to samurai
2026-03-05 16:36:13 <achill> maybe we can help with the maintainance of samurai instead
2026-03-05 18:04:12 <Sertonix[m]> Yes, I think improving samurai is a better way forward
2026-03-05 21:32:10 <yvs> Hi, it's about APKBUILD's syntax for "depends" - is there a way to use OR operation? Like for some dependency on one of two packages - depends='package1 OR package2'
2026-03-05 21:32:32 <ikke> yvs: no
2026-03-05 21:32:39 <ikke> yvs: well, not directly
2026-03-05 21:33:05 <ikke> if both package1 and package2 provide the same virtual package, you can depend on the virtual package
2026-03-05 21:36:07 <yvs> okay, is there some more direct workaround? (unfortunatelly I cannot convert one of them in a virtual because other packages depend on that one)
2026-03-05 21:39:06 <ikke> yvs: you don't need to convert it, it only need to provide it (
2026-03-05 21:39:24 <ikke> (you should add a provider_priority though)
2026-03-05 21:57:29 <yvs> @ikke anyway thanks for hinting
2026-03-06 00:03:11 <fabricionaweb> Openwrt 25.12 released with apk
2026-03-06 00:14:18 <Habbie> hooray
2026-03-06 01:12:45 <Saijin_Naib[m]> So excited for it, but it bricks my router whereas 24.x was perfect 😭
2026-03-06 07:23:12 <fabricionaweb> Mine is x86 but im too lazy to upgrades
2026-03-06 10:43:56 <frag> are you experiencing extreme slowness when playing youtube links with mpv?
2026-03-06 10:45:31 <frag> way more buffering than playing ... got 7s into the video prob been on 5m
2026-03-06 11:01:51 <achill> frag: im watching a yt lecture over mpv right now, and everything looks fine
2026-03-06 11:02:27 <achill> maybe its the internet connection or your geographic yt server that are slowed down?
2026-03-06 11:18:20 <frag> possibly its been a bit on/off lately
2026-03-06 11:18:42 <frag> only noticed it with yt tho
2026-03-06 14:07:37 <mxbob> Hey. I just installed alpine-sdk in the alpine:3.23.3 Docker image which depends on abuild which depends on tar-1.35-r4. CVE-2026-23950
2026-03-06 14:09:16 <mxbob> sorry. When scanning the image with Docker scout it says that there is a CVE issued for this version. 
2026-03-06 14:09:54 <mxbob> Is that something to take care for or is this a false positive?
2026-03-06 14:13:22 <ikke> False positive, it's about nodejs tar
2026-03-06 14:15:38 <mxbob> ty
2026-03-06 22:03:57 <mrsvcd> How do I set a swedish keymap in sddm? and in lxqt?
2026-03-06 22:11:39 <quinq> mrsvcd, the same way you set it in Xorg, it seems
2026-03-06 22:28:19 <mrsvcd> Just for the people that reads the archives: installing setkbdmap fixes LXQt keymap problem. No idea what to do about sddm.
2026-03-06 22:29:07 <quinq> I gave you one :)
2026-03-07 02:19:18 <Saijin_Naib[m]> Was it here that someone was talking about the django dev sloperating on the codebase now? I saw vim 🙄
2026-03-07 03:17:11 <jpp1> I no longer use vim
2026-03-07 03:20:01 <jpp1> I use vis much better
2026-03-07 03:32:04 <jpp1> https://wiki.alpinelinux.org/ is down?
2026-03-07 03:35:23 <Saijin_Naib[m]> I use BusyBox vi. I like it?
2026-03-07 03:38:11 <fipoac> jpp1: seems that way
2026-03-07 03:49:30 <longnoserob[m]> 504
2026-03-07 04:28:04 <astella> Hey everyone. I'm popping in to ask a generalized question regarding manual installation on an ARM-based laptop. I tried the standard install and it failed to be recognized by my HP 11a Chromebook as bootable. I imagine that this is probably a bootloader or driver situation, but I don´t know. What are the general steps required to get a generic U-boot installer to work? I can usually figure out the details,
2026-03-07 04:28:06 <astella>  I just don't know where to start for this. Right now, I have it running PostMarket so that will be the environment used to set things up.
2026-03-07 04:28:51 <astella> It's a kukui model if that helps
2026-03-07 04:39:27 <astella> Maybe I can adjust things in this guide to port to kukui/kappa: https://wiki.alpinelinux.org/wiki/Alpine_on_the_Aopen_Chromebase_or_Chromebox_Mini_with_Mainline_Kernel
2026-03-07 04:39:33 <astella> ?
2026-03-07 04:41:55 <bash> hi
2026-03-07 04:42:16 <bash> wifi stopped working on upgrade
2026-03-07 04:43:01 <bash> i am getting error, nl80211 could not set interface p2p dev wlan0 up
2026-03-07 04:43:42 <bash> wlano is set DOWN and does not come up with ip link set
2026-03-07 04:46:25 <astella> I don't know enough to actually provide a solution, but I can certainly come up with some questions that might lead you there if that helps. Do you have a pre-update backup that you could restore and compare with the new system? You might be able to run all of the usual info-gathering commands on both versions and hopefuly spot a difference
2026-03-07 05:26:11 <jpp1> https://wiki.alpinelinux.org/ works :)
2026-03-07 06:28:13 <longnoserob[m]> astella: out of curiosity, why do you want native alpine?  is pmos not good enough?
2026-03-07 06:44:44 <idkrnx[m]> jpp1: Neovim is the goat
2026-03-07 06:48:17 <jpp1> no way 😁 
2026-03-07 07:07:59 <jpp1> Neovim uses AI integration, so it's not for me.
2026-03-07 07:14:41 <jpp1> idkrnx[m]
2026-03-07 07:59:05 <frag> i assume you can have the same tag (eg @t) for edge main, community and testing, while having stable main/community untagged in repo?
2026-03-07 08:01:24 <idkrnx[m]> jpp1: ?
2026-03-07 08:04:42 <achill> frag: this is unsupported and will cause issues in the end
2026-03-07 08:05:25 <frag> achill using the same tag for those three edge repos or using tags at all?
2026-03-07 08:05:54 <achill> mixing stable and edge on the same system
2026-03-07 08:06:44 <frag> mhmhm, kinda nice to see if stuff is in edge and test them tho
2026-03-07 08:20:51 <jpp1> idkrnx[m]?
2026-03-07 13:13:14 <mlody> Hello, how is Alpine Linux for daily driving? I want to try it out, but I wonder about busybox.
2026-03-07 13:14:54 <longnoserob[m]> why that?
2026-03-07 13:15:36 <mlody> I like really minimalism, but I heard that it's slow and miss a lot of features.
2026-03-07 13:16:01 <ssm_> if you don't like busybox, chimerautils (bsdutils fork for chimeraOS, dunno why bsdutils isn't packaged) and coreutils are packaged. for games or anything that needs glibc, you can use a chroot (harder) or runtime manager like flatpak. there's also gcompat though I haven't used it
2026-03-07 13:16:41 <ssm_> note not all utils are replaced by chimerautils or coreutils, but at least you get a much more usable ed and vi ;)
2026-03-07 13:17:08 <longnoserob[m]> just install proper vim or so
2026-03-07 13:17:25 <mlody> I prefer vi instead of vim.
2026-03-07 13:17:34 <longnoserob[m]> and even on older HW, its still snappy
2026-03-07 13:18:04 <ssm_> vim author drank slopcode koolaid, codebase now compromised, no thanks. also feature bloat etc etc
2026-03-07 13:18:25 <ssm_> if vi feels limiting learn motion commands better
2026-03-07 13:18:41 <mlody> Vim sucks.
2026-03-07 13:19:16 <rnkn> ssm_: slopcode? AI?
2026-03-07 13:19:20 <ssm_> yes
2026-03-07 13:19:26 <ssm_> claud
2026-03-07 13:19:28 <ssm_> e
2026-03-07 13:20:49 <mlody> I'm looking for minimal, like some people say "suckless" setup. At the moment I'm using Gentoo and it's okay for me as daily driver, but compilation times take long time and it's using bash and python a lot which I don't like.
2026-03-07 13:21:28 <ssm_> gentoo/portage is kid wheels openbsd and alpine teach you to write your own patches ;)
2026-03-07 13:22:06 <mlody> I really wanted to try out OpenBSD, but it's missing a lot of features and I'm not sure about choosing it as main operating system.
2026-03-07 13:22:41 <rnkn> mlody: what features is it missing?
2026-03-07 13:23:39 <mlody> Virtualization features for example.
2026-03-07 13:24:19 <mlody> Also as I heard it have problems with performance and file system.
2026-03-07 13:24:30 <f_> ssm_: I use neovim :P
2026-03-07 13:24:50 <rnkn> there is https://man.openbsd.org/vmd.8
2026-03-07 13:24:53 <ssm_> I've been daily driving it for 8-9 years, only "complaints" are the lack of gaming. you can mitigate that with the games/moonlight-qt port which will let you use a liberated gamestream protocol to stream audio video and input from a device that can run the sunshine server. And yes, virtualization is poor due to lack of kvm, though if you don't need a framebuffer vmm(4) is very speedy
2026-03-07 13:25:14 <rnkn> ssm_: alpine or obsd?
2026-03-07 13:25:16 <ssm_> obsd
2026-03-07 13:25:29 <ssm_> alpine can game you just use flatpak or a chroot lol
2026-03-07 13:25:59 <longnoserob[m]> doesnt obsd also have xen support?
2026-03-07 13:26:07 <mlody> I wonder also about hardware compatibility. I'm forced to use wifi to access internet.
2026-03-07 13:26:55 <ssm_> there is a xen driver for openbsd guests, vmware too, but for hosting vms openbsd doesn't have much for advanced usecases
2026-03-07 13:27:34 <ssm_> openbsd has qemu, but only with very slow tcg acceleration since no kvm
2026-03-07 13:28:37 <mlody> Also I need to learn Android Studio, Java and C# Windows apps for school.
2026-03-07 13:30:02 <ssm_> find /usr/ports/!(pobj) -name DESCR | xargs grep -li 'android studio' returns nothing :/  there's adb though :)
2026-03-07 13:40:46 <mlody> Chimera Linux looks like interesting option, but I wonder about security. Operating systems like Alpine Linux and OpenBSD are trusted.
2026-03-07 13:44:04 <mlody> By the way what is wrong with editors shipped with busybox?
2026-03-07 13:47:24 <jvvv> wrong?
2026-03-07 13:48:37 <mlody> People are saying that they're bad, missing some features.
2026-03-07 13:48:46 <mlody> I heard that about vi.
2026-03-07 13:49:30 <ssm_> mlody: missing features compared to normal vi/nvi (no, nvi is not neovim). can't remember what specifically since I haven't used it in ages. A and E were missing I think, maybe no buffer execution. ed is missing fork-to-shell if I remember
2026-03-07 13:50:13 <ssm_> someone test if !} to execute on paragraph works? I can't live witout that one
2026-03-07 13:55:29 <ssm_> -s/A and E/A and I/
2026-03-07 16:22:46 <dogg0> set nu is one that i miss
2026-03-07 16:23:43 <ssm_> apk add oksh
2026-03-07 16:24:12 <ssm_> oh you mean some vi capability? set -nu is also a thing in posix sh
2026-03-07 16:24:33 <ssm_> ah it enables line numbers
2026-03-07 16:25:18 <dogg0> ya oops i was scrolled up
2026-03-07 16:52:27 <io1> @astella i dont have pre-update backup
2026-03-07 16:54:08 <io1> nl180211 could not set interface p2p dev wlan0 up
2026-03-07 16:54:40 <io1> is it a problem with the kernel wifi module ?
2026-03-07 16:54:54 <io1> or udhcpc
2026-03-07 17:02:37 <io1> ?/
2026-03-07 17:37:24 <jpp1> mlody: I prefer vis to vim
2026-03-07 17:38:28 <penguinz_rule[m]> vis is good
2026-03-07 17:39:17 <jpp1> https://pkgs.alpinelinux.org/package/v3.23/community/aarch64/vis
2026-03-07 17:43:16 <jpp1> :)
2026-03-07 18:12:07 <frag> minimal weather site, almost never loads x.X https://wttr.in/
2026-03-07 18:24:46 <failxontour[m]> frag: loads for me
2026-03-07 18:25:08 <courge_musque[m]> For me too with curl
2026-03-07 18:25:35 <frag> im cursed then ;(
2026-03-07 18:25:54 <penguinz_rule[m]> 0x0.st is blocking older curl so mayhaps wttr is also
2026-03-07 18:30:22 <frag> just kinda ironic, the minimal solution just takes longer...
2026-03-07 18:37:54 <monotux[m]> it's probably being hammered from others
2026-03-07 19:24:31 <jpp1> vv221 said about IRC #debian-easy: To my knowledge, there's no IA generated code yet in vim. In all this not in the latest version I used. On the other hand it is financed, and therefore controlled, by
2026-03-07 19:24:33 <jpp1> general AI sellers.
2026-03-07 19:24:35 <jpp1> Change that happened, O surprise, just after the death of the original vim author.
2026-03-07 19:27:56 <f_> as far as I know vim and neovim are not vibecoded yet
2026-03-07 19:28:13 <f_> and one can hope that with the backlash they will consider not doing so
2026-03-07 19:31:37 <abby> there are claude-coauthored commits in the vim repo that have been ported to neovim
2026-03-07 19:33:58 <f_> ah
2026-03-07 19:34:01 <f_> ugh
2026-03-07 19:46:57 <Noisytoot> solution: use emacs
2026-03-07 19:47:13 <Noisytoot> I doubt AI generated code will be accepted there
2026-03-07 19:47:24 <invoked> people probably doubted that about vim too
2026-03-07 19:48:02 <invoked> (what would bram have done? i guess that's irrelevant now.)
2026-03-07 20:06:05 <p_6f3Ik7Suw> i was informed that "claude-coauthored commits" - aka vibe coded commits are widely known as vommits
2026-03-07 20:32:16 <f_> Noisytoot: I cannot
2026-03-07 20:32:32 <f_> vim destroyed my brain already, it would take me quite a bit of time to join emacs
2026-03-07 20:34:44 <frag> vi is kinda sweet, but i dunno if any of them support all the stuff i want, eg utf8
2026-03-07 20:35:18 <frag> and easy run a shell command and insert the output
2026-03-07 20:35:37 <frag> some have one of those and not the other etc
2026-03-07 20:35:55 <frag> i had to make my own in posix shell script
2026-03-07 20:36:51 <frag> and soft-wrapping
2026-03-07 20:43:17 <invoked> we could pull in nvi but i don't think it's there currently
2026-03-07 20:45:07 <invoked> https://repo.or.cz/nvi.git
2026-03-07 20:46:30 <quinq> vis is nice too
2026-03-07 20:47:06 <p_6f3Ik7Suw> f_ you know that emacs has better vi keybinding emulation than vim?
2026-03-07 20:47:31 <p_6f3Ik7Suw> and a much saner scripting language
2026-03-07 20:47:35 <Noisytoot> f_: evil-mode
2026-03-07 20:47:47 <p_6f3Ik7Suw> indeed, even the name is cool!
2026-03-07 20:48:11 <Noisytoot> nvi should move to less-shitty git hosting that doesn't geoblock the UK
2026-03-07 20:48:35 <invoked> maybe, but it's good enough for debian
2026-03-07 20:48:52 <frag> emacs doesnt even work in the terminal(?)
2026-03-07 20:49:24 <Noisytoot> frag: yes it does
2026-03-07 20:49:29 <Noisytoot> what are you talking about?
2026-03-07 20:49:34 <frag> okay then :]
2026-03-07 20:50:14 <Noisytoot> emacs -nw, or terminal is the default if you compiled it without graphics support
2026-03-07 20:50:51 <Noisytoot> alpine packages the terminal-only version as emacs-nox
2026-03-07 20:51:04 <invoked> i'm not interested in using any editor that has a pretend-vi mode, and is not built around ex. but that's me
2026-03-07 20:51:11 <quinq> WITNESSSSS
2026-03-07 20:52:07 <Noisytoot> there's always busybox vi
2026-03-07 20:52:33 <invoked> indeed
2026-03-07 20:52:48 <invoked> don't know if that's fully utf8 though
2026-03-07 20:52:54 <frag> its hella big too..
2026-03-07 20:53:22 <p_6f3Ik7Suw> 
ACTION was the poor soul porting emacs to musl/alpine back in the days of the undump horrors, and yes, the gtk2 and nox variants were hi prio

2026-03-07 20:54:03 <frag> oh no, what have i done, how do you quit emacs?
2026-03-07 20:54:23 <invoked> stepped on a rake
2026-03-07 20:54:40 <ikke> frag: you cannot exit emacs, it has become your operating system now
2026-03-07 20:55:22 <frag> brb, have to unplug!
2026-03-07 20:56:38 <frag> oh had some notes on emacs
2026-03-07 20:57:21 <p_6f3Ik7Suw> emacs itself is only like 7-9MB big, the packages it comes with and that give it super powers like M-x butterfly are 90MB, but you could in theory strip that down to a minimum.
2026-03-07 20:57:54 <frag> this is why i like vi, its like almost 0bytes
2026-03-07 20:58:09 <frag> just (a bit less) than what you need
2026-03-07 20:58:48 <frag> you can actually master it in a lifetime
2026-03-07 20:58:56 <frag> or probaby a month
2026-03-07 21:01:01 <invoked> most people who use vim would be lucky to use 10-20% of what it does
2026-03-07 21:01:15 <invoked> [citation needed]
2026-03-07 21:01:41 <quinq> A bit like teh human brain
2026-03-07 21:02:04 <frag> nobody has explored 100% of vim, so no one knows how big it is
2026-03-07 21:02:17 <yeti> 
ACTION . o O ( I demand that the unused parets be called The dark Brain! )

2026-03-07 21:02:25 <quinq> Unless you're professor Xavier
2026-03-07 21:06:24 <invoked> something like nvi won't make some people happy, those who want to turn vi into a full-fat ide
2026-03-07 21:06:48 <invoked> but those people won't be put off by vibecoding and the rest
2026-03-07 21:06:51 <yeti> I've no problem switching between mcedit, (n)vi and emacs.  sometimes I'm using even ed.
2026-03-07 21:07:09 <quinq> I love you inside ed
2026-03-07 21:07:29 <yeti> ed > acme
2026-03-07 21:08:44 <yeti> sam is said to have a pure text mode (on 9front) but ed has too and ed is the standard editor.  :-P
2026-03-07 21:10:08 <DaKnig[m]> I config'd emacs enough to be vscode basically (but without the LSP, this shit's eating all my compute)
2026-03-07 21:10:12 <DaKnig[m]> Took a month
2026-03-07 21:10:19 <invoked> i'll put together a wip nvi aport monday or so, if nobody else does.
2026-03-07 21:10:24 <DaKnig[m]> But now I have Ctrl c Ctrl v
2026-03-07 21:11:53 <frag> invoked i dunno if ive tried it, but sounds cool!
2026-03-07 21:12:21 <Noisytoot> is there a mirror on nvi hosted on an actually usable (doesn't geoblock the UK) git hosting site?
2026-03-07 21:14:01 <invoked> not aware of any
2026-03-07 21:18:19 <Habbie> freebsd and gentoo keep (or they did at least) distfiles mirrors
2026-03-07 21:19:15 <quinq> Noisytoot, you're talking about repo.or.cz?
2026-03-07 21:19:20 <Habbie> i never realised alpine didn't have an nvi aport. nvi used to be my preferred vi for decades
2026-03-07 21:19:33 <Habbie> (but now it's vis. nvi still second)
2026-03-07 21:19:54 <Noisytoot> quinq: yes
2026-03-07 21:20:09 <quinq> :(
2026-03-07 21:21:02 <invoked> i'm not sure if it's good form for us to source= on someone's distfiles
2026-03-07 21:21:20 <invoked> :)
2026-03-07 21:23:33 <Noisytoot> it's surprisingly hard to find a free software public git hosting site that allows pull mirrors. codeberg does not, notabug used to but that's dead, framagit it appears does not
2026-03-07 21:23:47 <Habbie> what are pull mirrors?
2026-03-07 21:24:24 <Noisytoot> mirrors that automatically update by pushing
2026-03-07 21:24:26 <Habbie> invoked, i've seen ports (maybe not alpine) do it, but it's not great, no
2026-03-07 21:24:33 <Noisytoot> s/push/pull/ I mean
2026-03-07 21:24:41 <Noisytoot> framagit allows push mirrors, which is in the opposite direction
2026-03-07 21:24:43 <Habbie> Noisytoot, oh, like, codeberg does the pulling?
2026-03-07 21:24:47 <Noisytoot> yes
2026-03-07 21:25:03 <Habbie> you could CI that, would even fit their small free runners (although, do them a favour, bring your own runner)
2026-03-07 21:25:39 <Noisytoot> can't really bring my own runner since it would be in the UK
2026-03-07 21:25:54 <dwfreed> geography has no impact here
2026-03-07 21:26:07 <dwfreed> or just put a small hook in the repo you push to that then pushes to codeberg
2026-03-07 21:26:15 <Noisytoot> it does, because I can't clone repos from repo.or.cz with a UK IP address
2026-03-07 21:26:19 <Noisytoot> that's the entire problem
2026-03-07 21:26:53 <quinq> That sounds weird, send them an emaid
2026-03-07 21:26:55 <quinq> email
2026-03-07 21:27:17 <Noisytoot> they know, they intentionally geoblocked the UK: https://repo.or.cz/uk-blocked.html
2026-03-07 21:28:10 <invoked> prima facie, seems reasonable
2026-03-07 21:28:32 <invoked> at least they're explaining why, it's not for shit arbitrary reasons
2026-03-07 21:29:12 <quinq> I don't get it
2026-03-07 21:29:30 <quinq> They're hosting from UK, but don't want to serve UK citizens because of their weird government laws?
2026-03-07 21:29:54 <Habbie> they are not hosting from UK
2026-03-07 21:30:09 <quinq> So why would they care about UK laws
2026-03-07 21:30:14 <Habbie> sorry. let me avoid confusion
2026-03-07 21:30:18 <quinq> ah
2026-03-07 21:30:19 <Habbie> repo.or.cz is not hosting from UK
2026-03-07 21:30:25 <quinq> ok so no worries
2026-03-07 21:30:31 <Habbie> you are not their lawyer
2026-03-07 21:30:35 <quinq> wat
2026-03-07 21:30:44 <Habbie> they chose a different risk assessment than you are making here
2026-03-07 21:30:47 <Habbie> i sadly understand their choice
2026-03-07 21:30:54 <quinq> oO
2026-03-07 21:31:00 <quinq> What risk are you talking about
2026-03-07 21:31:15 <Habbie> the risk described on https://repo.or.cz/uk-blocked.html
2026-03-07 21:31:21 <invoked> it runs on hetzner
2026-03-07 21:31:27 <quinq> No it's not
2026-03-07 21:31:51 <Habbie> what's not?
2026-03-07 21:32:36 <quinq> There's no risk that they'd be exposed to described on that page
2026-03-07 21:32:49 <Habbie> in your opinion
2026-03-07 21:32:56 <dwfreed> individual fines of up to 18 million GBP is not a risk?
2026-03-07 21:33:12 <quinq> Well, like in what law is xD
2026-03-07 21:33:17 <Noisytoot> I disagree on principle with preemptive compliance with unjust laws
2026-03-07 21:33:26 <quinq> You told yourself they're not FROM UK
2026-03-07 21:33:28 <invoked> czech has extradition treaty with the uk
2026-03-07 21:33:29 <Habbie> not a risk they'd -be exposed to-, is quinq's interpretation, which i can follow
2026-03-07 21:33:30 <quinq> They're not hosting FROM UK
2026-03-07 21:33:37 <Habbie> there is no need to shout
2026-03-07 21:33:39 <Noisytoot> they're not in the UK and ofcomnadzor would contact them first anyway
2026-03-07 21:33:41 <quinq> Habbie, stop telling nonsense
2026-03-07 21:33:58 <Habbie> i hate to do this, but, please point out where i am telling nonsense
2026-03-07 21:34:00 <quinq> They're not exposed to UK law if they're not uk -_-
2026-03-07 21:34:07 <invoked> not true.
2026-03-07 21:34:08 <dwfreed> [citation needed]
2026-03-07 21:34:09 <Habbie> i understand you, but i don't see where i am telling nonsense in your view
2026-03-07 21:34:22 <Habbie> do i need to clarify that i did not write that page?
2026-03-07 21:34:24 <quinq> I've been explaining for several lines already
2026-03-07 21:34:37 <Habbie> and i'm not following. so please feel free to point out my nonsense explicitly
2026-03-07 21:34:41 <invoked> germany will extradite non-german people (that covers hetzner) and czech has a treaty to extradite its own people to the uk (but IANAL)
2026-03-07 21:34:44 <quinq> I did…
2026-03-07 21:35:06 <Noisytoot> geoblocking the UK because of this law legitimizes the ability of countries to pass laws that apply extraterritorially
2026-03-07 21:35:32 <invoked> so, my understanding is, in the presence of extradition agreements, that allows for prosecution cross border on just about anything.
2026-03-07 21:36:01 <invoked> but i don't really care about this fight too much.  we can't do anything about it.
2026-03-07 21:37:00 <invoked> everyone has to look after themselves, ultimately
2026-03-07 21:37:06 <quinq> Sure
2026-03-07 21:37:25 <quinq> Lock yourself home, call the police if somebody's trying to talk to you on the rare occasion you have to go out on the street
2026-03-07 21:38:09 <invoked> that's taking it further than what we're talking about here
2026-03-07 21:38:11 <quinq> This way you're pretty sure a New Zealand judge will not fine you with 14 billions $money
2026-03-07 21:38:14 <invoked> it's an inconvenience to uk people, that's it.
2026-03-07 21:38:37 <quinq> It's more like a political statement against some country law
2026-03-07 21:38:47 <Noisytoot> in practice the risk is very low (ofcomnadzor would contact them first, at that point geoblocking is more reasonable)
2026-03-07 21:39:12 <quinq> Than a “I protect myself in case the UK takes over the world and apply they're own laws to the rest of the world”
2026-03-07 21:39:36 <invoked> that's already happening in a lot of cases. does it suck? yes
2026-03-07 21:39:47 <invoked> CRA/NIS2 for instance
2026-03-07 21:39:54 <invoked> age verification
2026-03-07 21:40:14 <invoked> i could go on, but, it's not like everyone can sit back and invite the fight.
2026-03-07 21:40:18 <quinq> No
2026-03-07 21:40:26 <quinq> You're mixing individuals with companies
2026-03-07 21:40:38 <Noisytoot> normalizing geoblocking as a solution to countries trying to apply their laws extraterritorially is bad and the result of that is eventually going to be companies geoblocking everywhere by default
2026-03-07 21:40:51 <invoked> i agree
2026-03-07 21:41:16 <Noisytoot> anyway, I finally found a public free software git hosting site that allows for pull mirrors: https://git.gay/noisytoot/nvi-mirror
2026-03-07 21:41:23 <quinq> A country can't judge people that are not citizens of their own countries -_-
2026-03-07 21:41:39 <invoked> lol, that happens all the time sir.
2026-03-07 21:41:49 <quinq> Of course, it happens sometimes by breaking the law, but then that's a political fight
2026-03-07 21:41:56 <f_> btw are there still plans to drop armhf from alpine
2026-03-07 21:42:26 <f_> some kind of roadmap maybe?
2026-03-07 21:42:33 <invoked> you don't need to break the law. there are multinational agencies that will kick your door in by just running the wrong software.
2026-03-07 21:42:46 <dwfreed> How about we stick to #alpine-linux topics
2026-03-07 21:43:00 <quinq> Well, at least around here there are actual laws
2026-03-07 21:43:10 <quinq> Companies don't have their own police yet
2026-03-07 21:43:22 <quinq> dwfreed, right
2026-03-07 21:44:14 <invoked> Noisytoot: noted, thanks
2026-03-07 21:46:26 <f_> Wonder how many still fly armhf TBH
2026-03-07 21:46:36 <invoked> rpi3
2026-03-07 21:46:41 <invoked> probably a lot
2026-03-07 21:46:58 <Noisytoot> alpine should add armv5 support so I can run it on my BMC
2026-03-07 21:47:36 <Habbie> what's alpine armhf? armv5? v6?
2026-03-07 21:48:06 <Noisytoot> iirc it's v6 with hard-floats (hence the hf), whereas debian's armhf is v7 with hard-floats
2026-03-07 21:48:22 <aelin> yup, v6 with hardfloats
2026-03-07 21:48:23 <Habbie> yes, debian's is v7
2026-03-07 21:48:27 <Noisytoot> my BMC is v5 with no FPU
2026-03-07 21:48:27 <Habbie> and raspi's hf is v6
2026-03-07 21:48:30 <Habbie> making them painfully incompatible
2026-03-07 21:49:49 <Habbie> invoked, rpi3 can run arm64 though
2026-03-07 21:49:53 <Habbie> *aarch64
2026-03-07 21:50:04 <f_> pi2 onwards are armv7 or armv8
2026-03-07 21:50:15 <f_> only pi0/pi1 are armv6+hf
2026-03-07 21:50:39 <Noisytoot> and pi zero 2w is armv8
2026-03-07 21:51:18 <f_> I just looked, on pmOS at least there really is only a handful of devices on armhf, and I suspect a few of them to actually be armv7
2026-03-07 21:51:32 <Habbie> on a sidenote, the difference between tar and img in the pi section on https://alpinelinux.org/downloads/ is invisible outside of hovering the link button
2026-03-07 21:51:48 <f_> (some device ports are armhf on armv7 hw because of legacy reasons or downstream hackery)
2026-03-07 21:52:44 <invoked> well there are two rpi3 versions, right? there was the 3b, then later they added the 3a, which makes total sense
2026-03-07 21:52:53 <invoked> the raspberry pi stuff is mental
2026-03-07 21:53:11 <Habbie> 3b and 3a are both v8
2026-03-07 21:53:32 <f_> invoked: all pi3 are aarch64
2026-03-07 21:53:41 <f_> and all pi2 are armv7
2026-03-07 21:53:59 <invoked> 32-bit isa
2026-03-07 21:54:36 <Habbie> f_, pi2 v1.2 and up are armv8
2026-03-07 21:56:46 <Habbie> but, in the early 64 bit pi days, some people preferred 32 bit OSes to save some memory
2026-03-07 23:47:32 <Renan[m]> when we will have Cosmic in the stable alpine?
2026-03-08 02:07:13 <WhyNotHugo> Renan[m]: it's already in community, so should be part of the next stable release.
2026-03-08 02:07:34 <WhyNotHugo> Uptream stability is a separate matter, of course.
2026-03-08 02:18:12 <Renan[m]> WhyNotHugo: understood. i will install alpine again after that. how often is the compositor update in stable version? sorry my poor english
2026-03-08 11:37:19 <realroot[m]> <Renan[m]> "understood. i will install..." <- i think that it will be updated to new versions in next stable
2026-03-08 12:52:23 <Woelmuizen> Hello! Is the package maintainer for the gonic package around?
2026-03-08 13:10:26 <f_> Woelmuizen: Ask away
2026-03-08 13:10:34 <f_> if they're not there right now they may be later.
2026-03-08 13:11:24 <ikke> I don't think they are on IRC
2026-03-08 13:16:29 <Woelmuizen> Alright, no worries, thank you!
2026-03-08 18:49:29 <Guest4630> test
2026-03-08 18:50:11 <dogg0> check check
2026-03-08 21:56:44 <dogg0> alpine makes a great guest for sure, but anyone using it as host?
2026-03-08 21:57:58 <Lord> yes many people
2026-03-08 21:59:59 <ellyqw> my router runs alpine, my phone runs postmarketOS (nicely-dressed Alpine)
2026-03-08 22:00:21 <dogg0> i meant as host for vms
2026-03-08 22:00:47 <dogg0> hey ellyqw ive wanted to try alpine on actual router hw is that what youre doing?
2026-03-08 22:01:05 <ellyqw> well, i am indeed using alpine as a hypervisor, my router runs vms
2026-03-08 22:01:33 <ellyqw> dogg0: yes
2026-03-08 22:01:37 <dogg0> i have a couple of linksys routers uboot and 512ram wondered if i could actually put alpine on it
2026-03-08 22:02:06 <ellyqw> sure, my router runs x86 though
2026-03-08 22:03:38 <ellyqw> i bought myself a Topton X2E about 2 years ago with Intel N100 and 4x2.5GbE NICs, stuck 16GB of RAM in there (back when it was cheap... good times), SSD, WiFi 6 card
2026-03-08 22:03:48 <ellyqw> then I ported coreboot to it and installed alpine
2026-03-08 22:03:55 <ellyqw> it boots in 4 seconds
2026-03-08 22:04:09 <dogg0> ah coreboot another project i want to do so bad but seems like over my head
2026-03-08 22:05:28 <ellyqw> i'm currently working on adding support for Ryzen 7000 and Framework 16
2026-03-08 22:07:04 <ellyqw> but that's OT :P
2026-03-08 22:11:28 <courge_musque[m]> @ellyqw when you say "boots in 4 seconds", do you include the bootloader and the initrd?
2026-03-08 22:11:48 <ellyqw> courge_musque[m]: from pressing a power button to device appearing on the network
2026-03-08 22:12:43 <courge_musque[m]> Wow, it seems worth the hassle
2026-03-08 22:16:41 <ellyqw> well, i wouldn't trust a device with firmware from a Chinese vendor to work as a core of my network either way
2026-03-08 22:32:20 <Ermine> it reminds me of meme about firewalling your networks so it is immune to all sorts of backdoors
2026-03-09 00:01:04 <mlody> ellyqw: What model of phone do you have?
2026-03-09 00:01:13 <mlody> I was thinking about to buy phone with postmarket os support/
2026-03-09 00:01:30 <ellyqw> mlody: https://wiki.postmarketos.org/wiki/Xiaomi_Mi_9_Lite_(xiaomi-pyxis)
2026-03-09 00:01:53 <ellyqw> you'd need patches for it though, https://f.sakamoto.pl/elly/linux/
2026-03-09 00:05:00 <ellyqw> i've been thinking of buying something with MediaTek Dimensity 6300 and mainlining that SoC from scratch, sounds like a fun challenge
2026-03-09 00:05:35 <ellyqw> and phones with those SoCs are affordable, have features i'm looking for (mSD, headphone jack) and you can get them in regular stores right now
2026-03-09 01:08:16 <Saijin_Naib[m]> Certainly not a weak SoC either
2026-03-09 01:08:37 <Saijin_Naib[m]> says the person with a MediaTek A25 🤣 pen and paper feels fast in comparison
2026-03-09 01:47:19 <el[m]1> what's alpine linux's wondershaper equivalent? is there one? since wondershaper doesn't seem to be available
2026-03-09 02:29:38 <dwfreed> el[m]1: you might find https://www.bufferbloat.net/projects/bloat/wiki/Wondershaper_Must_Die/ interesting
2026-03-09 02:30:08 <dwfreed> if you're running a router with alpine, then https://pkgs.alpinelinux.org/package/edge/testing/x86_64/sqm-scripts is the way to go
2026-03-09 03:42:18 <el[m]1> dwfreed: i just want something that lets me limit a single process
2026-03-09 03:42:58 <el[m]1> i don't really care if somebody thinks it would need to die or not 🫣
2026-03-09 03:47:17 <dwfreed> put it in its own cgroup and use nftables
2026-03-09 03:47:50 <dwfreed> (or if it runs as its own user, you can natch by uid)
2026-03-09 03:47:54 <dwfreed> s/natch/match/
2026-03-09 05:30:21 <el[m]1> dwfreed: i understood some of those words!
2026-03-09 18:10:02 <dogg0> as someone that uses systemd-boot elsewhere, whats the equivalent in allpine
2026-03-09 18:10:36 <Biswa96[m]> grub, maybe?
2026-03-09 18:11:19 <dogg0> efi no grub
2026-03-09 18:11:19 <Biswa96[m]> or, anyone from there https://wiki.alpinelinux.org/wiki/Bootloaders
2026-03-09 18:11:38 <dogg0> i was looking at that page, wondering waht ppl are mostly using
2026-03-09 18:12:01 <dogg0> the systemd-boot equivalent
2026-03-09 18:12:04 <dogg0> ish
2026-03-09 18:35:57 <Sertonix[m]> There are many bootloaders available in alpine including systemd-boot
2026-03-09 18:37:40 <invoked> for efi i'd suspect most people are using grub, or perhaps uki
2026-03-09 19:43:08 <dogg0> thanks maybe uki ill have to read if it needs to be manually edited each kernel update
2026-03-09 19:44:10 <dogg0> thisll give me something to experiment with
2026-03-09 23:33:44 <WhyNotHugo> dogg0:  when using systemd-boot, nothing needs to be edited on each system update, the kernel-hooks script can update the UEFI bundle / UKI.
2026-03-09 23:34:23 <dogg0> i use systemd-boot in debian. i was mostly wondering if there was something similar to it, but for alpine
2026-03-09 23:34:35 <WhyNotHugo> GRUB is so much the opposite of "simple" that I've never understood why it's the default.
2026-03-09 23:34:47 <dogg0> ya im past grub
2026-03-09 23:34:49 <ellyqw> mostly legacy crap
2026-03-09 23:35:14 <dogg0> i have uefi boot in alpine but  i  want to go past grub
2026-03-09 23:35:42 <dogg0> uki and stub i guess is the way? but theyre not auto like you said about systemd-boot above
2026-03-09 23:37:39 <WhyNotHugo> dogg0: the steps for that are in https://wiki.alpinelinux.org/wiki/UEFI_Secure_Boot, although you can skip some sections if you don't need secureboot
2026-03-09 23:38:12 <WhyNotHugo> You mainly want to kernel-hooks bit
2026-03-09 23:38:39 <WhyNotHugo> This is quite underdocumented now that I see
2026-03-09 23:38:57 <WhyNotHugo> ukify-kernel-hook provides the hook to auto-ugprade the bundle
2026-03-09 23:40:52 <dogg0> ooh thans WhyNotHugo 
2026-03-10 02:56:53 <autumn> Hi! I'm setting up Alpine with Plasma as a desktop. I wanted proper American-preferred date/time/weather formats, so I did as the wiki said and installed the lang package, but afterward, SDDM wouldn't load. After removing lang, all seems well, but I would prefer proper en_US locales.
2026-03-10 02:58:41 <autumn> I'm using a MacBookPro10,1.
2026-03-10 03:14:21 <daigennki> Strange issue: DNS seems to be somewhat borked on this fresh install on an x86 machine. "nslookup www.google.com" fails with "connection timed out; no servers could be reached", but it works with "dl-cdn.alpinelinux.org", and APK updates/upgrades and docker image pulls work too. Why?
2026-03-10 03:15:01 <dwfreed> what are you using for resolvers?
2026-03-10 03:15:32 <dwfreed> there's not much actual DNS going on in a Linux install that's not running a resolver
2026-03-10 03:16:07 <daigennki> I dunno, but what's weird about this is it works just fine on a Raspberry Pi 3 on the same network
2026-03-10 03:16:18 <daigennki> Which also runs Alpine
2026-03-10 03:16:21 <dwfreed> cat /etc/resolv.conf
2026-03-10 03:16:41 <dwfreed> also tcpdump will probably be helpful
2026-03-10 03:17:01 <daigennki> resolv.conf has "search lan" followed by "nameserver 192.168.10.1", which is identical to the one on the RPi3
2026-03-10 03:18:49 <dwfreed> router's running either dnsmasq or unbound; the former is just a slightly less dumb forwarder, the latter is a full recursor
2026-03-10 03:18:58 <dwfreed> I would tcpdump the DNS traffic
2026-03-10 03:19:40 <dwfreed> (I'm assuming you're using your ISP's gateway device, and not your own router)
2026-03-10 03:19:43 <daigennki> How do I use tcpdump here? Haven't used it at all before I think
2026-03-10 03:19:52 <daigennki> Also, the router is a RPi4 running OpenWrt
2026-03-10 03:20:14 <dwfreed> ah, so definitely dnsmasq then, unless you've customized it
2026-03-10 03:20:47 <daigennki> The router does have the DoH and Adblock on it, but I haven't done much else to it
2026-03-10 03:22:40 <dwfreed> 'tcpdump -nvvvei eth0 port 53' on the alpine device may provide some clues; otherwise 'tcpdump -w dns.pcap -i eth0 port 53' and upload the dns.pcap to somewhere like 0x0.st; replace 'eth0' in both instances with whatever your network interface is named if it differs
2026-03-10 03:23:10 <dwfreed> while tcpdump is running, repeat the nslookup
2026-03-10 03:26:26 <daigennki> Looks like with www.google.com it never receives a reply, but with www.alpinelinux.org it does get a reply
2026-03-10 03:27:14 <dwfreed> sounds like something you've configured on the router is messing things up then
2026-03-10 03:27:32 <daigennki> I wonder if this network switch is screwing things up again. Lemme try connecting the machine to the other switch closer to the router
2026-03-10 03:30:32 <daigennki> Okay it's not the switch, so I wonder what's wrong with the router config
2026-03-10 03:39:32 <daigennki> Hmm, "dig" instead of "nslookup" works fine, but that doesn't solve the problem of some programs not being able to do DNS lookups
2026-03-10 03:40:03 <dwfreed> dig doesn't use musl's stub resolver, pretty sure nslookup does
2026-03-10 03:40:14 <daigennki> Yeah, figured
2026-03-10 03:40:28 <dwfreed> you might capture the exchanges and compare the packets in wireshark
2026-03-10 03:45:20 <daigennki> Huh, now "nslookup www.google.com" suddenly works, very strange
2026-03-10 04:09:00 <daigennki> "pip install" in a python virtual environment still seems to be failing though. Running tcpdump on both ends, the router receives the query, but never replies to it
2026-03-10 04:09:25 <daigennki> The query for "pypi.org" that is, which works fine with nslookup
2026-03-10 04:11:42 <daigennki> Again, all this works just fine on the Pi 3 running Alpine in the same network, so what's the difference there?
2026-03-10 05:28:34 <dwfreed> daigennki: like I said, consider capturing the different exchanges with tcpdump as a pcap, and then comparing the packets in the pcap with wireshark; differences might be more apparent
2026-03-10 05:28:58 <dwfreed> or capture the pcap and upload it somewhere (like 0x0.st) and I can look at it
2026-03-10 11:33:12 <vlt> Hello everyone :)  How do I configure and then build an Alpine Linux live boot iso?
2026-03-10 12:53:33 <hazardaj_nombroj> good question... i think if your goal is to make a custom image from which alpine can be installed, the "setup-bootable" script does that, but it's aimed at making bootable USBs and SD cards, not disc images
2026-03-10 13:02:56 <ikke> There are scripts in the scripts/ directory in the aports repository for that
2026-03-10 13:30:11 <M0x5ECF4ULT[m]> Hi, I was wondering where I can find the code that actually builds the release files published to the official mirrors. Specifically I'm looking for the part that builds the netboot-able initramfs. Could someone please point me to the code/repo or the channel this question should be asked in?
2026-03-10 13:44:34 <Shiz> M0x5ECF4ULT[m]: https://git.alpinelinux.org/aports/tree/scripts
2026-03-10 13:44:44 <Shiz> specifically mkimage.sh and mkimg.netboot.sh
2026-03-10 14:00:47 <M0x5ECF4ULT[m]> Thank you very much! @Shiz 
2026-03-10 15:11:37 <dsc_> did someone create an alpine package format v3 parser in another language yet
2026-03-10 15:12:15 <dsc_> (other than C)
2026-03-10 15:29:37 <frag> anyone made a nvi package? hm?
2026-03-10 15:43:41 <dogg0> is there something like apts-listbugs (debian), but for alpine?
2026-03-10 15:59:54 <vlt> hazardaj_nombroj: Bootable USB is exactly what I’m looking for. I’ll have a look at setup-bootable, thank you!
2026-03-10 16:07:02 <Biswa96[m]> dogg0: issues can be searched here https://gitlab.alpinelinux.org/alpine/aports/-/issues
2026-03-10 16:14:02 <dogg0> apt-listbugs is a package, wondering if there was something similar in alpine.  thanks tho
2026-03-10 16:16:55 <dwfreed> the problem is that it would have to rely on convention (and that convention quickly breaks down for bugs that affect multiple packages), whereas debian bugs can actually be assigned to a specific package
2026-03-10 16:20:40 <dogg0> isee, thanks dw
2026-03-10 16:56:11 <vlt> hazardaj_nombroj: Hmmm, from the Alpine wiki: “setup-bootable … copies files from an ISO image … onto a USB-Stick”. At first glance that sounds lke a chicken-and-egg problem. I’m rather looking for a tool to configure said ISO image.
2026-03-10 16:56:15 <vlt> Any idea?
2026-03-10 16:57:49 <hazardaj_nombroj> no idea, sorry, i think making an iso image might require special tools
2026-03-10 17:02:25 <ikke> vlt: check the scripts here: https://wiki.gentoo.org/wiki/User:Mjo/GHC_binary_packages
2026-03-10 17:02:29 <ikke> arg, wrong link
2026-03-10 17:02:35 <ikke> https://git.alpinelinux.org/aports/tree/scripts
2026-03-10 17:02:57 <ikke> The mkimage.sh script is what's used to create the ISOs for AL
2026-03-10 17:03:04 <invoked> frag: (and anyone else playing along at home) i did work on nvi (and nvi2) a little bit. but i'm stopping, because i probably won't maintain them and i'll be using vis instead
2026-03-10 17:05:01 <invoked> vis is already in tree and as a compelling alternative to vim (in light of recent events), vis is probably ideal
2026-03-10 17:05:21 <invoked> though not based on ex, it is what it is
2026-03-10 17:09:26 <Saijin_Naib[m]> I'll check out vis. I want something comfy like busybox vi, but with a few more features
2026-03-10 17:10:21 <invoked> there's a lot more upside with vis. just not a universe of readymade stuff yet like with vim.
2026-03-10 17:10:35 <invoked> (lua is much more sane)
2026-03-10 17:11:01 <Sertonix[m]> dsc_: I think the only non-C parser is the one I wrote in shell: https://codeberg.org/sertonix/alpine-scripts/src/branch/main/adbdump
2026-03-10 17:11:24 <dsc_> Sertonix[m]: interesting, I will try it
2026-03-10 17:12:02 <Sertonix[m]> It is mainly a proof of concept, not very usefull :)
2026-03-10 17:13:06 <Saijin_Naib[m]> invoked, sounds good. Basically, I want it to feel like busybox VI (same snytax for basic operations) but have bold/colors/syntax highlight/line numbers in the gutter, etc
2026-03-10 17:13:19 <Saijin_Naib[m]> Ideally, it could have read-only mode so I can use it for less as well
2026-03-10 17:14:11 <frag> pity
2026-03-10 17:14:11 <dsc_> Sertonix[m]: thanks anyway, its useful
2026-03-10 17:14:54 <invoked> Saijin_Naib[m]: i mean, there's still going to be gaps. vim is a whole universe. vimdiff, xxd, so on
2026-03-10 17:15:15 <invoked> but, none of that stuff is incomprehensible to do elsewhere
2026-03-10 17:15:21 <Saijin_Naib[m]> Oh, I didn't use vim, haha. it was overwhelming
2026-03-10 17:15:24 <Saijin_Naib[m]> same for emacs
2026-03-10 17:19:25 <invoked> nvi will take some work, will need someone who will use it full time, and that's not going to be me
2026-03-10 17:20:13 <invoked> and at the end of the day, it's not really a vim replacement and never will be.
2026-03-10 17:22:00 <Saijin_Naib[m]> Is vis-plug something you might add to vis?
2026-03-10 17:23:35 <invoked> personally, i don't know. i don't rely on a lot of vim plugins, so i doubt it
2026-03-10 17:24:30 <invoked> i worry about supply chain attacks, so it's basically a no.
2026-03-10 17:24:47 <Saijin_Naib[m]> Ah, fair point
2026-03-10 17:29:14 <dsc_> Sertonix[m]: do you know where I can find an apkv3 package (too lazy to generate one :P)
2026-03-10 17:31:04 <Sertonix[m]> dsc_: https://gitlab.alpinelinux.org/alpine/apk-tools/-/issues/11192#note_586567
2026-03-10 17:33:03 <dsc_> cheers
2026-03-10 17:52:44 <henrix> Hi! I'm going to re-installing Alpine, and I was wondering how safe would it be to replace my usual ext4-on-luks by zfs-on-luks.
2026-03-10 17:52:54 <henrix> is this a usual setup, or should I stick with a safer option (maybe xfs or btrfs)?
2026-03-10 17:57:22 <invoked> opinions vary. if you want mine, it'd be if safety is #1 then ext4 is near the top already
2026-03-10 17:57:57 <invoked> zfs works best with raw devices, for pools, without luks in the middle
2026-03-10 17:58:24 <invoked> if you need snapshots then yes you'll have to look beyond ext4
2026-03-10 17:59:37 <invoked> (hopefully that helps you narrow it down.)
2026-03-10 18:00:11 <invoked> there's no universal correct answer for everybody.
2026-03-10 18:01:29 <henrix> thanks invoked, yeah I didn't expect an universal answer :-)
2026-03-10 18:01:45 <invoked> ext4 imo remains the sanest default
2026-03-10 18:02:14 <henrix> dropping luks is out of question, and I don't really need snapshots
2026-03-10 18:02:41 <invoked> zfs will also eat your memory, if you're short on it
2026-03-10 18:03:48 <henrix> ext4 is the boring option, that's why I was looking at other options. I've been using zfs on FreeBSD, so it would be nice to use the same FS on both systems
2026-03-10 18:04:25 <invoked> in your case, it makes more sense.
2026-03-10 18:06:39 <henrix> yeah, not sure yet. I'll probably stick with ext4. or try xfs, which is also rock-solid AFAIK
2026-03-10 18:06:47 <henrix> thanks for your feedback invoked 
2026-03-10 18:07:01 <invoked> np good luck
2026-03-10 18:34:02 <dsc_> damn apv3 format is complicated
2026-03-10 18:34:07 <dsc_> apkv3*
2026-03-10 18:34:38 <dsc_> instead I will wait for someone else to write a parser :P
2026-03-10 19:49:37 <Noisytoot> I'd use btrfs
2026-03-10 19:49:55 <dwfreed> does it still explode when it gets full?
2026-03-10 19:50:40 <Noisytoot> my previous laptop's disk filled up a few times and btrfs never exploded
2026-03-10 19:54:09 <invoked> that's a pretty low bar
2026-03-10 19:55:09 <invoked> but fwiw i'd assume everything is somewhat better than anecdotes from 10 years ago
2026-03-10 19:56:15 <invoked> when xfs got imported all i heard for at least 10 years was about the mess it was making of linux and how unsafe it was.
2026-03-10 19:56:20 <invoked> but now people call it rock solid
2026-03-10 19:57:18 <daigennki> dwfreed: Hello again. I produced a pcap file for the DNS issues I was having earlier, which contains the failed `pip install` attempt followed by a successful nslookup, both querying pypi.org: https://daigennki.net/files/tcpdump-m720q.pcap
2026-03-10 20:02:19 <daigennki> The only difference I can see when looking at it myself is that the DNS queries from `pip install` have "Don't fragment" set, while the nslookup queries don't have it set. I would think that doesn't make a difference considering how small these packets are, but who knows...
2026-03-10 20:09:47 <dwfreed> it shouldn't, anyway
2026-03-10 20:09:55 <dwfreed> I don't see anything obvious either
2026-03-10 20:10:04 <Saijin_Naib[m]> @henrix i have found xfs over lvm best for my machines in terms of data safety. ext4 happily kept corrupting files when my laptop lost power, xfs does not
2026-03-10 20:10:47 <dwfreed> maybe take the same packet capture from the router (you can use 'host 192.168.10.27 and port 53' as the filter to see only this device's DNS queries) and make sure the queries are arriving, and whether responses are going out
2026-03-10 20:15:04 <courge_musque[m]> [@saijin_naib:matrix.org](https://matrix.to/#/@saijin_naib:matrix.org) What's the fun if you don't lose something after a fsck?
2026-03-10 20:15:20 <courge_musque[m]> It's like a lottery, fs gambling
2026-03-10 20:19:41 <Saijin_Naib[m]> NTFS and JFS did me well going back to like, 2000. God, I'm old
2026-03-10 20:34:15 <henrix> Saijin_Naib[m]: thanks for the feedback. my experience is also quite good with xfs
2026-03-10 20:35:31 <Saijin_Naib[m]> Just make sure you don't need to shrink 😭
2026-03-10 20:36:13 <Saijin_Naib[m]> First time needing it, I forgot to make my swap vol as big as RAM, so hibernate is spotty under load
2026-03-10 20:45:31 <daigennki> dwfreed: Here's how it looks from the router side http://daigennki.net/files/tcpdump-openwrt-rpi4.pcap
2026-03-10 20:51:02 <shrizza> Any non-ZFS users using tools recover from periodic bitrot?
2026-03-10 20:51:17 <Sertonix[m]> dsc_: Do you have a specific language in mind?
2026-03-10 20:55:47 <dwfreed> daigennki: on the router side, the first 6 packets have bad udp checksums, causing them to be dropped
2026-03-10 20:56:03 <dwfreed> try disabling udp checksum offloading on the device's nic
2026-03-10 21:14:08 <autumn> Afternoon! I'm setting up Alpine on a MacBookPro10,1 with Xfce. Everything that I've wanted to get working so far has been without incident, but I can't seem to adjust the backlight, which wasn't an issue last night when I had Alpine + Plasma installed.
2026-03-10 21:14:50 <autumn> This machine has Intel iGPU and an NVIDIA GT650M dGPU. The iGPU may or may not feed its output through the dGPU, I can't find a conclusive answer.
2026-03-10 21:15:02 <autumn> Here's the 'ls' of /sys/class/backlight/:
2026-03-10 21:15:20 <autumn> gmux_backlight/
2026-03-10 21:15:24 <autumn> which contains
2026-03-10 21:15:46 <autumn> actual_brightness  device             scale              uevent bl_power           max_brightness     subsystem brightness         power              type
2026-03-10 21:15:59 <autumn> and echoing to brightness and actual brightness does nothing.
2026-03-10 21:16:19 <autumn> Using the xfce4-power-manager applet's slider did nothing either, and the keyboard keys are similarly unresponsive
2026-03-10 21:17:40 <dwfreed> weird, it works fine on an 11,1 (though I don't use alpine, but i do use xfce)
2026-03-10 21:17:53 <dwfreed> which has the same setup, backlight control is on the mux
2026-03-10 21:18:17 <autumn> It worked perfectly with no extra setup on Plasma, so I know it's not Alpine weirdness
2026-03-10 21:19:14 <autumn> setup-desktop seemed to install a much more complete Plasma than it did with Xfce (didn't have goodies or plugins of any sort, nor working audio)
2026-03-10 21:19:18 <autumn> maybe I'm missing a package?
2026-03-10 21:20:12 <dwfreed> you might try asking the xfce people, they might have a better idea
2026-03-10 21:20:26 <autumn> where might i find xfce people ^_^'
2026-03-10 21:20:38 <autumn> their IRC?
2026-03-10 21:20:48 <autumn> or is there an alpine-related channel I should be going to
2026-03-10 21:21:30 <ikke> nothing alpine dedicated for xfce
2026-03-10 21:21:56 <dwfreed> the xfce project is on matrix now (unfortunately); #xfce:matrix.org is their room
2026-03-10 21:22:07 <autumn> sigh
2026-03-10 21:22:13 <autumn> I don't *have* Matrix
2026-03-10 21:22:32 <dwfreed> me either
2026-03-10 21:23:05 <dwfreed> you could try #xfce on libera, there might still be some helpful people in there (that's where they were before they moved to matrix)
2026-03-10 21:23:13 <autumn> i'll try there
2026-03-10 21:27:57 <vlt> ikke: Thanks :)
2026-03-10 21:28:20 <daigennki> dwfreed: What do you know, running `ethtool -K eth0 rx off` on the Openwrt RPi4 "fixed" the problem. So is this an Openwrt issue or an Alpine issue?
2026-03-10 21:28:44 <dwfreed> see I meant the alpine device, interesting that fixed it on the rpi
2026-03-10 21:29:05 <daigennki> Hmm I could try it with tx on the alpine device instead and see what happens
2026-03-10 21:32:47 <daigennki> Just doing `ethtool -K eth0 tx off` doesn't fix it, but it does say this when I run it: "Acutal changes: tx-checksum-ip-generic: off; tx-tcp-segmentation: off [not requested]; tx-tcp6-segmentation: off [not requested]"
2026-03-10 21:33:31 <daigennki> Whoops, typo, that's meant to be "Actual changes". Also, there are actually newlines in place of the semicolons.
2026-03-10 21:45:08 <dwfreed> I'm not sure 'rx off' fixes the problem, but just disables the checksum verification; you could of course check with pcaps; if you're loading the pcap in wireshark, right click on the UDP part, and then in protocol preferences there's an option to enable checksum verification
2026-03-10 21:56:07 <daigennki> Huh, `ethtool --show-offload eth0` on the Alpine device says " tx-checksum-ipv4: off [fixed]". I can't seem to find anything UDP-specific though
2026-03-11 01:24:52 <daigennki> Yeah so I checked the incoming packets on the router side with the rx offloading disabled, and the checksum is still incorrect (but gets replied to by dnsmasq anyways), so there's almost definitely something weird going on with the Alpine machine
2026-03-11 01:25:47 <daigennki> I can't seem to find an option to specifically disable UDP offloading on the Alpine machine, so what should I do next?
2026-03-11 01:48:00 <daigennki> I compared the result of doing the same thing on the x86 Alpine machine (ThinkCentre M720q) with the RPi3 running Alpine, and by the looks of it, the x86 machine is adding 26 bytes of "extraneous data" after tcpdump captures the packet, causing the checksum to become invalid. That "extraneous data" isn't present when doing the same thing on the RPi3.
2026-03-11 01:52:05 <daigennki> I've uploaded the pcap files for this newer test, with rx offloading disabled on the RPi4 router, and with captures on both sides being from the same transaction: http://daigennki.net/files/tcpdump-m720q-2.pcap http://daigennki.net/files/tcpdump-rpi4-2.pcap
2026-03-11 01:58:27 <daigennki> It looks as if the queries for the A and the AAAA records are getting merged into one after it's captured on the x86 machine; the "extraneous data" looks identical to the last 26 bytes of the AAAA query. Why...?
2026-03-11 02:31:17 <daigennki> Ah, I should clarify that the last 26 bytes of the AAAA query are getting stuck onto the end of the A query. The RPi4 router only gets the A query, but with the last 26 bytes of the AAAA query stuck onto it; it never receives a separate AAAA query, even though the A and AAAA queries were separate when tcpdump captured them on the Alpine machine.
2026-03-11 02:42:15 <shrizza> Hm, would be curious to see the packet(s) captured on the ingress of whatever is directly connected to Alpine (if it's not the RPi4).
2026-03-11 02:44:29 <shrizza> (thinking there's like a network switch of some kind)
2026-03-11 03:16:46 <daigennki> Yeah, there are two network switches in between. I have a spare USB NIC here, so I'll try and connect the Alpine machine directly to the RPi4 router.
2026-03-11 04:07:34 <daigennki> Interesting, everything works fine if the Alpine machine is connected to the 100M USB NIC (Wii LAN adapter, lol) connected directly to the RPi4, set up as a bridge to LAN in OpenWrt. That still leaves a lot of possibilities though...
2026-03-11 04:08:08 <dwfreed> daigennki: what does 'lspci -k' say the driver is for the nic on the alpine device
2026-03-11 04:08:32 <daigennki> Ah yes, I probably should've said that. It's an Intel I219-V.
2026-03-11 04:08:41 <dwfreed> is the driver 'e1000e' ?
2026-03-11 04:08:54 <daigennki> Yes it is
2026-03-11 04:09:04 <dwfreed> try disabling all the offloading: gso off gro off tso off tx off rx off rxvlan off txvlan off sg off
2026-03-11 04:09:40 <dwfreed> 'ethtool -K <interface> <that^>'
2026-03-11 04:14:23 <dwfreed> the nics that use the e1000e driver tend to be hot garbage, and intel has never fixed them
2026-03-11 04:19:12 <daigennki> Still borked after disabling all the offloading... I tried connecting the USB NIC to the Alpine machine but it doesn't seem to get recognized. Is there some driver I need to install? It's one of those ASIX chips.
2026-03-11 04:19:55 <dwfreed> you probably need the firmware; see what 'dmesg' says
2026-03-11 04:25:22 <daigennki> It shows up in dmesg and lsusb as AX88772. dmesg and `ip link` says it has an interface name of eth1, but the link doesn't seem to go up when a cable is connected.
2026-03-11 04:27:35 <dwfreed> try manually setting it up? 'ip link set eth1 up'
2026-03-11 04:35:39 <dsc_> Sertonix[m]: python, I created this 'universal script for listing and extracting files from exotic archives' 
2026-03-11 04:36:29 <dsc_> any file format I support will require 1) getting a file listing 2) reading files (fetching bytes)
2026-03-11 04:37:22 <dsc_> so it would ignore SIG blocks and whatever else metadata
2026-03-11 04:38:13 <dsc_> anyway its not important because this project is random and not important 
2026-03-11 04:39:37 <daigennki> Oh great, even with the USB NIC the problem still occurs. Guess that rules out a bad NIC or its driver
2026-03-11 04:41:59 <daigennki> I'll test a direct connection to the RPi4's built-in NIC later when I have the time
2026-03-11 04:42:34 <daigennki> I do appreciate the help so far btw, thank you :)
2026-03-11 08:53:40 <frag> can busybox vi search for words in the beginning/end of lines and similar?
2026-03-11 08:53:59 <frag> the usual ^$ doesnt work
2026-03-11 08:55:12 <frag> phaps it doesnt support regex at all
2026-03-11 09:00:57 <jn> frag: i can reproduce this. i think the issue is that CONFIG_FEATURE_VI_REGEX_SEARCH is not set in the busybox config that alpine ships
2026-03-11 09:02:08 <frag> oh, do you think its intentional jn?
2026-03-11 09:02:29 <jn> no, i haven't checked "git blame"
2026-03-11 09:03:30 <jn> what i'm saying is that it has a direct cause in alpine source code, rather than being a busybox bug
2026-03-11 09:04:10 <jn> intention is another question, but it's clear how to change it, or at least try
2026-03-11 09:04:38 <frag> aha, yes, not a big deal i guess
2026-03-11 09:05:06 <jn> frag: what do you mean by that?
2026-03-11 09:05:37 <frag> convenience makes us weak and dumb
2026-03-11 09:06:06 <frag> also easier to search for regex ;p
2026-03-11 09:06:11 <frag> code
2026-03-11 09:06:31 <jn> ¯\_(ツ)_/¯
2026-03-11 09:07:52 <jn> anyway, vi regex search was disabled in commit 25ed9167936cbb8582ca6630b8077535755c75ed (main/busybox: upgrade to 1.19.0 + patches) from 2011. there was no explicit intention behind it.
2026-03-11 09:07:56 <frag> probably they wanted to keep it minimal
2026-03-11 09:08:06 <jn> plausible
2026-03-11 10:52:56 <omni> frag: if you want a bit more capable vi (but not vim/neovim), there's neatvi (with the recent neatvi-vi subpackage providing a vi -> neatvi symlink)
2026-03-11 10:54:03 <omni> (and there's also nextvi in testing, with the similar nextvi-vi subpackage)
2026-03-11 11:56:00 <hoel> There’s vis too.
2026-03-11 11:57:04 <penguinz_rule[m]> i really like vis' multi-cursor
2026-03-11 13:32:32 <jn> honestly, i'd just want a frozen vim. it doesn't need a lot of improvement (being already improved ;P), it just needs to not be at risk of getting worse
2026-03-11 13:35:53 <penguinz_rule[m]> i mean there's evi and icevim, icevim specifically being a more stable nvim so less config munging every update. i wouldnt be surprised if either get updated as often as mainline
2026-03-11 15:40:10 <shrizza> God if vis had just a bit more split management features I'd be set.
2026-03-11 15:41:42 <shrizza> (move splits, maximize split, equalize splits)
2026-03-11 15:45:15 <penguinz_rule[m]> i wondrr if that could be done in lua
2026-03-11 17:00:12 <invoked> probably
2026-03-11 17:00:48 <invoked> the thing about vis is the core won't get bloated with this stuff people want.
2026-03-11 17:02:10 <invoked> vim also had the struggle of trying to remain perfectly backward compatible while getting bloated
2026-03-11 17:02:19 <invoked> so vis is trying to avoid that mistake
2026-03-11 17:03:07 <invoked> so now we have vim+vibecoding and who knows where that will end up.  vis feels a lot better in contrast
2026-03-11 17:19:29 <Sertonix[m]> dsc_: It might not be much easier to understand than the shell code but maybe you can get something working based on this: https://codeberg.org/sertonix/alpine-scripts/src/branch/main/adb.py
2026-03-11 17:21:47 <dsc_> Sertonix[m]: ah you are a life saver, lets try this :)
2026-03-11 17:26:04 <dsc_> Sertonix[m]: awesome :)
2026-03-11 17:27:53 <dsc_> Sertonix[m]: for context, this is the tool I (quickly) made https://git.maemo.org/sanderfoobar/admp so I will integrate this
2026-03-11 17:28:21 <dsc_> your code correctly lists size, mode, so its perfect, thanks :)
2026-03-11 17:52:20 <jpp1> vv221 said about IRC #debian-easy: To my knowledge, there's no IA generated code yet in vim. In all this not in the latest version I used. On the other hand it is financed, and therefore controlled, by
2026-03-11 17:52:22 <jpp1> general AI sellers.
2026-03-11 17:52:24 <jpp1> Change that happened, O surprise, just after the death of the original vim author.I installed vis-spellcheck as a plugin.I installed vis-spellcheck as a plugin.I installed vis-spellcheck as a plugin.I installed vis-spellcheck as a plugin.RœrereRœuY:      
2026-03-11 17:53:48 <jpp1> https://github.com/martanne/vis/wiki/Plugins
2026-03-11 17:54:39 <jpp1> https://gitlab.com/muhq/vis-spellcheck
2026-03-11 17:56:17 <jpp1> It works well :)
2026-03-11 18:39:29 <daigennki> dwfreed: Welp, the issue persists with a direct connection (no network switches) to the RPi4. So this is specifically between any NIC (including USB) on the Alpine x86 machine and the OpenWrt RPi4's built-in NIC. I wonder if I should bring this up with the OpenWrt people.
2026-03-11 18:50:11 <dwfreed> try testing with the m720q directly connected to something not the rpi4 ? (just build a fake network for testing purposes) just to rule out where the problem lies
2026-03-11 19:30:45 <daigennki> I connected it to a Bazzite gaming laptop, and used the "Shared to other computers" option in KDE's Network Settings. It works just fine there. Confirmed through tcpdump and wireshark.
2026-03-11 19:31:49 <daigennki> The gaming laptop's NIC is an RTL8125, if it matters at all.
2026-03-11 19:57:45 <dwfreed> daigennki: so yeah, sounds like a bug with the rpi4 then
2026-03-11 20:29:02 <daigennki> Alright, I'll ask the OpenWrt people about it then. Again, thanks for the help :)
2026-03-11 20:35:42 <rrfd> happens on the latest iso http://gcdnb.pbrd.co/images/COl7cVMq32M2.jpg
2026-03-11 20:36:34 <rrfd> (usb)
2026-03-11 20:38:57 <penguinz_rule[m]> can you reproduce with same image on another usb 
2026-03-11 20:44:58 <rrfd> ok let me try with another usb, sec
2026-03-11 20:58:40 <rrfd> penguinz_rule[m]: not. but what happened to that usb so it boot but then this error?
2026-03-11 20:59:22 <penguinz_rule[m]> probably could be some silent corruption
2026-03-11 20:59:31 <penguinz_rule[m]> i have had the same thing happen to me so
2026-03-11 21:00:16 <rrfd> anyway thanks for help so it was usb
2026-03-12 04:07:17 <ang1e> How do you set environmental variables to openrc scripts? I added one in line 18 here, and my program keeps crashing: https://p.ac2.no/kerylanter.sh
2026-03-12 04:25:31 <ang1e> I found out it was the -e variable in start-stop-daemon.
2026-03-12 05:53:18 <ikke> ang1e: fyi: https://github.com/deepin-community/openrc/blob/master/service-script-guide.md#dont-write-your-own-startstop-functions
2026-03-12 05:56:23 <ikke> You can export the variable in the init.d script or conf.d file and it should be available to the daemon
2026-03-12 05:57:26 <ikke> If CONFIG in the start function is intended to be the environment variable, you miss the export keyword
2026-03-12 07:00:08 <Imbus> Does anyone here know why alpine doesnt default openrc parallel?
2026-03-12 07:00:49 <Imbus> Im aware there are wanings in the conf about boot lockup, but gentoo doesnt warn about it last i checked, at least in the wiki
2026-03-12 07:01:50 <Imbus> That may be one reason. On a base system the performance advantage is negligeble
2026-03-12 07:02:56 <Imbus> Oh and the boot log also gets absolutely trashed
2026-03-12 07:10:46 <divVerent1> How will Alpine do age verification? Any chance Alpine can just pull "this is no OS, we do not even have a graphical installer"? ;)
2026-03-12 07:15:55 <frag> omni thanks! will check it out, kinda thinking of sticking with busybox vi too, kinda like the simplicity...
2026-03-12 08:29:45 <courge_musque[m]> [@divverent:matrix.org](https://matrix.to/#/@divverent:matrix.org) These people never even knew about Alpine, so I don't they would even care to apply age verification to it
2026-03-12 08:30:03 <courge_musque[m]> Small fishes can go back to the sea
2026-03-12 09:46:28 <divVerent1> hope so
2026-03-12 10:13:01 <kjell[m]> Hello
2026-03-12 11:05:11 <dsc_> Sertonix[m]: zstd, https://paste.debian.net/plainh/04cbcfb9
2026-03-12 13:24:16 <rrfd> penguinz_rule[m]: about that usb i run badblocks and also compared iso when writed again and verified all perfect, maybe it's just because usb is cheap and alpine linux trouble booting
2026-03-12 13:33:25 <penguinz_rule[m]> ive had usb operating systems (nomadBSD) fail due to ports they were in so who knows
2026-03-12 15:35:52 <Sertonix[m]> dsc_: Thanks, applied with a few changes and other fixes
2026-03-12 17:28:58 <mlody> I installed Alpine on my laptop :)
2026-03-12 17:29:54 <mlody> A lot of things must be configured manually, it's different in other Linux distributions.
2026-03-12 17:31:00 <penguinz_rule[m]> setup-desktop was nice to use
2026-03-12 17:34:46 <frag> mlody what do you have to set up that other distros dont?
2026-03-12 17:36:54 <mlody> For me Gentoo was much more automatized or it's illusion because I know how to use it. Right now I'm trying to start xorg.
2026-03-12 17:37:22 <frag> setup-xorg-base
2026-03-12 17:37:43 <frag> startx /usr/bin/dwm
2026-03-12 17:37:52 <frag> ezpz
2026-03-12 17:38:05 <mlody> Yeah, I used that. But it's probalby problem related to drivers.
2026-03-12 17:38:19 <frag> oh no
2026-03-12 17:38:43 <frag> i guess i had that on rpi5 too, had to config it yeah
2026-03-12 17:38:55 <mlody> fbdev vesa modules don't load
2026-03-12 17:41:47 <frag> ms () { printf 'Section "OutputClass"\n  Identifier "vc4"\n  MatchDriver "vc4"\n  Driver "modesetting"\n  Option "PrimaryGPU" "true"\nEndSection\n' > /usr/share/X11/xorg.conf.d/99-vc4.conf; } # sys
2026-03-12 17:42:06 <frag> i made that function to set it up for me.. dunno about you tho
2026-03-12 17:42:58 <frag> didnt have to do this on rpi4 so maybe the sys _should_ set it up(?)
2026-03-12 17:51:08 <mlody> I had to add my user to "video" group lol
2026-03-12 17:51:50 <mlody> But know input don't work.
2026-03-12 17:51:56 <frag> haha, didnt think of that simple stuff ;p
2026-03-12 17:52:38 <frag> add user to input group also
2026-03-12 17:52:48 <mlody> Yeah.
2026-03-12 17:52:54 <frag> and audio
2026-03-12 17:53:34 <frag> you dont need that on gentoo? hm..
2026-03-12 17:53:58 <mlody> Never.
2026-03-12 17:57:41 <frag> hmmm thought vi handlet pasting via insert snd shit+ins well, but i see it can skew it..
2026-03-12 17:59:11 <mlody> By the way how to give my user access to directory and all directories/files in it?
2026-03-12 17:59:21 <mlody> I have one directory outside my /home
2026-03-12 17:59:37 <mlody> I used chown and chmod with recursive but it only worked for directories
2026-03-12 17:59:44 <frag> chown -R user:user <dir>
2026-03-12 18:00:49 <mlody> Well, I know right now - it's not working for files in subdirectories. I had to use that command for every subdirectory.
2026-03-12 18:02:43 <frag> weird
2026-03-12 18:03:18 <frag> should work i guess, maybe use find to feed it input if nothing else
2026-03-12 18:05:15 <frag> i use busybox chown btw.. maybe coreutils is different or something
2026-03-12 18:05:18 <mlody> Everything is okay right now.
2026-03-12 18:05:23 <mlody> I use busybox.
2026-03-12 18:05:47 <mlody> Maybe I did something wrong I don't know.
2026-03-12 18:08:21 <mlody> Also I wonder why when I setup permissions 764 I can't enter directory as user in correct group
2026-03-12 18:08:39 <mlody> I only don't have permissions to execute, why I can't enter?
2026-03-12 18:13:00 <frag> i dunno those specifics, but weird if its different in gentoo, are u pulling my leg??
2026-03-12 18:15:21 <hoel> You need execute permissions on a directory to enter it.
2026-03-12 18:19:41 <mlody> Oh, okay.
2026-03-12 18:20:14 <mlody> I mean Alpine is much more different than every other Linux distro which I tried. It's using busybox and musl.
2026-03-12 18:21:20 <ikke> mlody: fyi, most commands  can be replaced by non-busybox equivalents if you need to
2026-03-12 18:25:13 <Noisytoot1> There's no way alpine requires more manual configuration than gentoo
2026-03-12 18:29:27 <mlody> It is easier to configure, but also it's much more bloated as it hardly depends on coreutils, python and bash.  It's have also much better wiki so in Gentoo you can make everything step by step. In Alpine I had to know what to do to configure system like I wanted.
2026-03-12 18:29:57 <ikke> "much more bloated as it hardly depends on coreutils, python and bash." huh?
2026-03-12 18:30:21 <mlody> I say about Gentoo.
2026-03-12 18:31:05 <mlody> In Alpine if I don't want Python, Bash I just don't install them.
2026-03-12 18:31:33 <hoel> (hardly as in hard dependency I guess)
2026-03-12 18:32:12 <ikke> ah
2026-03-12 18:32:16 <mlody> Well, I have problem with my English. Sorry.
2026-03-12 18:32:21 <ikke> as an adjective it means the opposite
2026-03-12 18:32:33 <mlody> I confused words.
2026-03-12 18:32:36 <ikke> understood
2026-03-12 18:32:56 <penguinz_rule[m]> i thought Python is a alpine-base dep
2026-03-12 18:33:01 <ikke> no, it's not
2026-03-12 18:35:53 <mlody> Apk is my favourite package manager now, I love how simple it is. And Emerge is really slow in its job, when it counting dependencies for package it takes a lot of time.
2026-03-12 18:40:51 <frag> does nextvi/neatvi have ways to choose which line to start editing a file on? like -c on other vi clones?
2026-03-12 18:45:30 <frag> oh nvm, you can do export EXINIT=:10
2026-03-12 18:53:22 <invoked> mlody: besides python, gentoo has a complicated problem to solve (thus, a complicated solver for it) because of USE flags
2026-03-12 18:54:01 <invoked> the flags controlling compile-time features, so it has to backtrack
2026-03-12 18:54:21 <invoked> apk is simpler in just about every way
2026-03-12 22:26:20 <mlody> How can I display time in my terminal in format like this hh:mm:ss ?
2026-03-12 22:27:48 <mid> mlody: `date +%T`?
2026-03-12 22:28:23 <mlody> Thanks!!
2026-03-12 22:28:29 <mid> yw 😉
2026-03-12 22:37:50 <perpendicular> Bit of an issue with emacs-nox (running in xterm, with i3wm & xorg). It seemingly wont detect the Meta key-press (my alt key), but xev will detect it. emacs does detect it outside xorg. I've fiddled a bit with .Xresources, but no luck
2026-03-12 23:24:09 <mid> perpendicular: stupid question, TERM=xterm ?
2026-03-12 23:45:01 <jvvv> It has been a while since I used i3wm, but first thing I would check is the i3 config file(s) to see what the default modifier key is set to. also worth checking to see if ANY i3wm key bindings are configured with for the ALT key.
2026-03-12 23:45:32 <jvvv> If I remember right mod1 is the ALT key
2026-03-12 23:54:12 <jvvv> I use bspwm + sxhkd, and I have all my wm key binding start with the mod4 key to avoid just such problems.
2026-03-12 23:57:28 <invoked> i3 and sway default to the windows key
2026-03-12 23:58:37 <jvvv> True. That was my reasoning for suggesting the config file(s).
2026-03-13 00:49:38 <mlody> How can I replace vi/ex and ed editor with chimerautils?
2026-03-13 00:49:52 <mlody> I made symlinks, but busybox is replacing them by some hook.
2026-03-13 01:18:12 <jvvv> mlody: You could put symlinks in a directory in your user's home directory and set that as the first directory of your PATH environment variable.
2026-03-13 02:01:44 <perpendicular> mid: jvvv: I do have 'set $mod Mod4' in my i3config, and TERM=xterm. I don't see anything in my i3config mention mod1 or alt
2026-03-13 02:02:37 <perpendicular> emacs does recognize that alt is being pressed - it creates combination characters when alt + key is pressed
2026-03-13 02:05:59 <jvvv> perpendicular: OK, not sure then.
2026-03-13 02:08:50 <perpendicular> nano does appear to have the same issue as emacs. ill try another wm later i guess
2026-03-13 02:41:10 <daigennki> Hello once again. Been looking into that DNS weirdness between my Alpine server and OpenWrt RPi4 router. Turns out it works perfectly fine if I specify the router's IPv6 address (instead of IPv4) for the `nameserver` in Alpine's `/etc/resolv.conf`.
2026-03-13 14:49:05 <mlody> Any idea how to run slock after closind lid in Alpine?
2026-03-13 15:00:36 <invoked> mlody: stuff like lid, power button, etc are acpi events
2026-03-13 15:01:06 <invoked> but this will also depend on whether your hardware shows up in /proc/acpi/
2026-03-13 15:01:41 <invoked> but you would configure one of the acpid daemons (you have busybox acpid by default) to do what you want
2026-03-13 15:02:05 <invoked> you can look at /etc/acpi for an example... busybox acpid is only configured for the power button by default
2026-03-13 15:02:37 <invoked> there is acpid in aports which does more stuff
2026-03-13 15:03:20 <invoked> either way, they just receive an acpid event, and run a script that does whatever you want. that's all it is
2026-03-13 15:04:01 <invoked> good luck :)
2026-03-13 15:04:03 <invoked> 
ACTION afk

2026-03-13 18:33:34 <quinq> Hi
2026-03-13 18:33:50 <quinq> I have to run $vendor ugly graphical glibc binary application
2026-03-13 18:34:09 <quinq> How would you recommend using it, besides installing another OS?
2026-03-13 18:34:11 <penguinz_rule[m]> distrobox is probably how you would have to do that
2026-03-13 18:34:14 <penguinz_rule[m]> gcompat is spotty
2026-03-13 18:34:48 <quinq> Yeah, gcompat isn't enough sadly
2026-03-13 18:34:57 <quinq> Looking at distrobox, thanks penguinz_rule[m] 
2026-03-13 18:36:05 <quinq> looks a bit complex :D
2026-03-13 18:36:28 <penguinz_rule[m]> also its not 100% because virtualbox doesnt work because of le thin kernel and not a fat kernel under like qemu 
2026-03-13 18:57:47 <invoked> quinq: iirc people have used a debian chroot before
2026-03-13 18:58:17 <invoked> i haven't done it, but i suppose it should work
2026-03-13 18:58:19 <prabu> quinq, if you prefer a lightweight solution, checkout https://wiki.alpinelinux.org/wiki/Software_management#Bubblewrap_+_Chroot
2026-03-13 19:00:14 <prabu> here are some data on the disk usage required https://tpaste.us/bQRD
2026-03-13 19:01:25 <invoked> cheers prabu
2026-03-13 19:08:45 <quinq> Thank you :)
2026-03-13 19:08:56 <prabu> thanks invoked, these are the installed packages which includes firefox esr  https://tpaste.us/1yeQ 
2026-03-13 19:10:40 <prabu> i tested obsidian, vs code(microsoft), chrome from google and all three worked with my sway(no xwayland) on alpine without issues..
2026-03-13 19:55:58 <mlody> invoked: I added slock to /etc/acpi/LID/00000080 but its don't working.
2026-03-13 19:56:10 <mlody> before echo mem > /sys/power/state
2026-03-13 19:56:15 <mlody> I'm using acpid from busybox/
2026-03-13 20:08:45 <invoked> what does /var/log/messages say when you trigger the lid event?
2026-03-13 20:14:04 <invoked> also, of course, make sure acpid is started, and make sure the script is +x
2026-03-13 20:32:24 <mlody> I'll give you info when I'll sit to my laptop. Script have permissions to execute because laptop is suspending when I closing lid, I was making that file today and gave permissions because it wasn't working at out of the box.
2026-03-13 20:36:37 <invoked> hm, then it sounds like acpid is doing its job. you'll have to figure out how to get slock to work there
2026-03-13 21:55:06 <bigboydriver> hello! ive just finished installing alpine-extended on my thinkpad t60 (i686 cpu) and im not sure how to connect to the internet. internet worked right after the setup script but when i rebooted now its offline and im not sure how to get it back on. thanks for your help in advance!
2026-03-13 21:56:28 <lopid> wifi?
2026-03-13 21:56:49 <bigboydriver> yeah wifi
2026-03-13 21:57:38 <lopid> is it ok after going through setup-interfaces?
2026-03-13 21:58:23 <bigboydriver> ill try it
2026-03-13 21:59:26 <bigboydriver> it says "no avaliable wireless networks"
2026-03-13 22:00:06 <lopid> nothing from `iw list`?
2026-03-13 22:03:24 <bigboydriver> it just prints the help dialouge
2026-03-13 22:05:47 <mlody> To be honest I don't want to help you that much with wifi problems, because I'm really tired with it on different systems. But my advice would be to avoid wpa_supplicant and use iwd instead.
2026-03-13 22:06:31 <bigboydriver> ok well thanks for your help
2026-03-13 22:07:50 <mlody> iwd is just easier to manage, you can easily do it from terminal. With wpa_supplicant is pain in the ass to use it without some gui.
2026-03-13 22:10:26 <bigboydriver> i dont have either
2026-03-13 22:10:28 <invoked> bigboydriver: well it worked once so you know it can work. make sure you haven't done something like toggle wifi off
2026-03-13 22:10:48 <invoked> which is like the f8 key on most thinkpads, i think
2026-03-13 22:11:13 <bigboydriver> yeah mine has that but its a switch on the front
2026-03-13 22:11:18 <bigboydriver> its defenitly in the on position
2026-03-13 22:11:27 <bigboydriver> i do get an error message when i try to initialize wlan0
2026-03-13 22:12:03 <bigboydriver> "ip: ioctl 0x8914 failed: no such file or directory"
2026-03-13 22:12:46 <invoked> you might be missing firmware
2026-03-13 22:13:13 <invoked> look at your dmesg output
2026-03-13 22:13:49 <invoked> you probably have intel wifi, and you should probably see iwlwifi somewhere
2026-03-13 22:14:12 <bigboydriver> yes your absolutley right
2026-03-13 22:14:45 <bigboydriver> "iwlwifi-3945-1.ucode firmware file req failed: -2"
2026-03-13 22:15:05 <invoked> yeah, you need to get that firmware installed
2026-03-13 22:15:45 <invoked> linux-firmware-intel i believe
2026-03-13 22:15:56 <bigboydriver> oh boy
2026-03-13 22:16:10 <bigboydriver> do i have to like get files onto a flash drive and transfer them and install them
2026-03-13 22:16:21 <invoked> you could do it that way
2026-03-13 22:16:41 <bigboydriver> i do also have an ethernet cable
2026-03-13 22:16:42 <lopid> boot it how you booted it when it worked, install or copy the files from then
2026-03-13 22:16:45 <lopid> oho
2026-03-13 22:17:01 <bigboydriver> how do i make that work
2026-03-13 22:17:06 <bigboydriver> it dosent just work when i plug it in
2026-03-13 22:17:21 <lopid> setup-interfaces
2026-03-13 22:18:05 <bigboydriver> didint work
2026-03-13 22:18:16 <bigboydriver> i chose eth0 and dhcp
2026-03-13 22:18:29 <heakbril> lopid: and the joy of networking... did he remember to `sudo apt update && sudo apt install linux-firmware`? dhcp should handle eth0 just fine.
2026-03-13 22:19:12 <bigboydriver> then "cp: cant create '//etc/network/interfaces': file exists"
2026-03-13 22:19:50 <lopid> i don't think that would work on alpine, heakbril…
2026-03-13 22:19:54 <heakbril> rip alpine humor
2026-03-13 22:20:00 <bigboydriver> yeah its apk package manager lol
2026-03-13 22:25:48 <bigboydriver> uh so what do i do
2026-03-13 22:27:06 <lopid> it worked once before
2026-03-13 22:27:20 <bigboydriver> yeah in the live usb enviornment
2026-03-13 22:27:39 <lopid> can you get the firmware apk?
2026-03-13 22:28:35 <bigboydriver> uh
2026-03-13 22:28:36 <bigboydriver> no
2026-03-13 22:28:46 <lopid> why not? 
2026-03-13 22:28:50 <bigboydriver> no internet
2026-03-13 22:29:03 <lopid> go to the live usb env
2026-03-13 22:29:08 <lopid> where it worked
2026-03-13 22:29:28 <bigboydriver> and do the apk command?
2026-03-13 22:29:38 <lopid> apk fetch, yes
2026-03-13 22:30:09 <lopid> you could even chroot into the installed drive
2026-03-13 22:31:30 <bigboydriver> +++++++++++++uggxz
2026-03-13 22:31:41 <bigboydriver> oops cat walked on keyboard lol
2026-03-13 22:32:19 <lopid> all help welcome
2026-03-13 22:33:15 <bigboydriver> lolol
2026-03-13 22:34:24 <bigboydriver> ok so how do i chroot
2026-03-13 22:34:55 <lopid> over to someone else i'm afraid - bed time for me
2026-03-13 22:35:08 <bigboydriver> oh
2026-03-13 22:35:10 <bigboydriver> good night
2026-03-13 23:10:07 <courge_musque[m]> Hello, does anyone how to do architecture detection in a makefile?
2026-03-13 23:10:34 <courge_musque[m]> I tried with uname -m and ifeq, but it didn't work
2026-03-13 23:10:41 <courge_musque[m]> *know how to
2026-03-13 23:10:51 <dwfreed> uname -m will be wrong for cross builds anyway
2026-03-13 23:17:36 <quinq> What do you mean “it didn't work”?
2026-03-13 23:17:48 <quinq> uname -m will give you the host architecture
2026-03-13 23:20:29 <quinq> I've been trying https://wiki.alpinelinux.org/wiki/Software_management#Bubblewrap_+_Chroot, but it just tells me bwrap: execvp /path/to/bin: No such file or directory
2026-03-13 23:20:44 <quinq> Not sure how (if) that's supposed to work
2026-03-13 23:21:31 <quinq> Even for a simple shell script
2026-03-13 23:21:38 <quinq> $ glibc ./test.sh
2026-03-13 23:21:39 <quinq> bwrap: execvp ./test.sh: No such file or directory
2026-03-13 23:21:57 <Ermine> i'd look at strace
2026-03-13 23:23:59 <quinq> Nothing obvious
2026-03-13 23:24:01 <quinq> Maybe prctl(PR_CAPBSET_READ, 0x30 /* CAP_??? */) = -1 EINVAL (Invalid argument)
2026-03-13 23:24:42 <quinq> and then wait4(-1, 0x7ffe64d8b3b0, WNOHANG, NULL) = -1 ECHILD (No child process)
2026-03-13 23:28:00 <invoked> i'm assuming you went through setting up the binds and everything just to make sure the chroot is correct
2026-03-13 23:28:34 <quinq> What binds?
2026-03-13 23:28:38 <quinq> The one passed in the command?
2026-03-13 23:28:59 <invoked> you set up a debian chroot, no?
2026-03-13 23:29:15 <heakbril> yea, but the issue is usually not in the chroot setup itself, but in how the chroot is being used or the environment variables. check your shebangs and make sure the binaries are where bwrap expects them to be.
2026-03-13 23:29:32 <Ermine> hmm actually, is that ./test.sh in your home dir?
2026-03-13 23:29:39 <p_6f3Ik7Suw> courge_musque[m]: this works for me well:
2026-03-13 23:29:39 <quinq> No
2026-03-13 23:29:42 <p_6f3Ik7Suw> UNAME := $(if $(UNAME),$(UNAME),$(shell uname -s))
2026-03-13 23:29:44 <p_6f3Ik7Suw> ARCH := $(if $(ARCH),$(ARCH),$(shell uname -m))
2026-03-13 23:29:46 <p_6f3Ik7Suw> ifeq ($(UNAME),Darwin)
2026-03-13 23:30:20 <p_6f3Ik7Suw> or replace UNAME with ARCH and then compare with x86_64 or parisc64 for exmaple
2026-03-13 23:31:03 <Ermine> probably bwrap child trying to execvp has container's as its cwd
2026-03-13 23:31:32 <WhyNotHugo> mlody: I suppose that slock needs to run withing your x session as your user, but /etc/acpi/LID/00000080 runs as root.
2026-03-13 23:32:04 <quinq> The wiki says it can run any command, like glibc google-chrome --enable-features=UseOzonePlatform --ozone-platform=wayland
2026-03-13 23:32:13 <quinq> I doubt that google-chrome is installed in the user's $HOME
2026-03-13 23:32:59 <quinq> And otherwise I assume they'd have documented in the example something like $ glibc $HOME/binary
2026-03-13 23:33:00 <WhyNotHugo> I typically truncate /etc/acpi/LID/00000080 to an empty script, and handle the power button as a regular global keyboard mapping (you get a key event when pressing it, if /etc/acpi/LID/00000080 does not turn off your computer).
2026-03-13 23:34:20 <invoked> quinq: i'm assuming that meant that you can run anything that's installed in the chroot
2026-03-13 23:34:40 <invoked> ie, apt-get chromium
2026-03-13 23:35:04 <courge_musque[m]> [@_oftc_p_6f3Ik7Suw:matrix.org](https://matrix.to/#/@_oftc_p_6f3Ik7Suw:matrix.org) It worked, tysm!
2026-03-13 23:36:49 <quinq> So it can actually only run what's installed in the chroot?
2026-03-13 23:37:07 <quinq> Why would they provide a wrapper then, just chroot inside
2026-03-13 23:37:07 <invoked> i would bet a donut that's the idea
2026-03-13 23:37:33 <invoked> i'm not looking at the page right now, just an assumption
2026-03-13 23:38:01 <quinq> Yeah, it's unclear
2026-03-13 23:38:04 <invoked> the wrapper is what, an alias that runs the chroot?
2026-03-13 23:38:08 <invoked> just convenience
2026-03-13 23:38:48 <invoked> i'm assuming 'glibc' is an alias
2026-03-13 23:41:16 <quinq> $ doas printf '/var/chroots/debian/lib/x86_64-linux-gnu\n/var/chroots/debian/usr/lib/x86_64-linux-gnu\n' > /etc/ld.so.conf
2026-03-13 23:41:26 <quinq> Why would they run printf as root
2026-03-13 23:42:34 <invoked> who is they :)
2026-03-13 23:42:49 <quinq> The author(s) of this wiki page
2026-03-13 23:42:52 <p_6f3Ik7Suw> quinq: whatever you wanna run must be inside the chroot afaik
2026-03-13 23:43:52 <invoked> i'm in the middle of a massive salad, i can look in a little bit after
2026-03-13 23:47:11 <quinq> Can I not just bind the directory instead?
2026-03-13 23:47:40 <quinq> I can see it does --bind /home /home
2026-03-13 23:48:13 <Ermine> if you bind nothing you end up in empty dir
2026-03-13 23:49:19 <quinq> I just follow the wiki
2026-03-14 00:05:07 <quinq> Broken AI
2026-03-14 00:05:18 <Habbie> but make it French
2026-03-14 00:12:44 <invoked> what the
2026-03-14 00:13:45 <longnoserob[m]> IA_22: identify yourself
2026-03-14 00:14:06 <invoked> dwfreed: sorry for the ping
2026-03-14 00:14:25 <quinq> Habbie ;)
2026-03-14 01:27:40 <mlody> WhyNotHugo: So what solution do you have for that?
2026-03-14 01:27:56 <mlody> I have serious problem with pipewire, I can't make user service/
2026-03-14 01:30:52 <mlody> https://pastebin.com/kqv4kiWy
2026-03-14 01:30:59 <mlody> Something like this, I tried everything.
2026-03-14 01:33:42 <WhyNotHugo> mlody: that's a know issue with openrc. One workaround is `mkdir -p $XDG_RUNTIME_DIR/openrc`.
2026-03-14 01:33:53 <WhyNotHugo> Er, mkdir and then `touch $XDG_RUNTIME_DIR/openrc/softlevel`
2026-03-14 01:34:19 <mlody> I should put that somewhere for example in xinit?
2026-03-14 01:34:23 <WhyNotHugo> I think navi had another workarund for that.
2026-03-14 01:34:40 <WhyNotHugo> Yeah, I put it in ~/.zshrc, since it's for a session which doesn't run any grpahical mode
2026-03-14 01:35:11 <mlody> Is there any good alternative for pipewire without this weird stuff like dbus?
2026-03-14 01:37:12 <WhyNotHugo> mlody: really depends what you're using it for. Alsa works without dbus, but I'm not sure which desktop applications support it.
2026-03-14 01:48:47 <invoked> quinq: did you figure it out?
2026-03-14 02:15:19 <mlody> WhyNotHugo: It's working, thank you.
2026-03-14 02:27:51 <mlody> But I have to run manually them everytime, because they're failing before file is created.
2026-03-14 02:31:21 <mlody> I removed them from runlevel and added to .profile
2026-03-14 02:36:38 <mlody> But it's ugly way. If anybody know how to fix it please let me know. I don't know too much about init systems.
2026-03-14 03:39:10 <UnknownArtistBot> Hi all! check this out, An Unknown Artist - https://www.anunknownartist.com/
2026-03-14 10:26:32 <navi> mlody: call `openrc --user default` (or any runlevel) and it'll create that file for you
2026-03-14 11:19:37 <mlody> navi: Not valid runlevel.
2026-03-14 12:02:09 <quinq> invoked, yeah, was a mix of bwrap parameters (just add a --bind), and installing x32 libs in Debian…
2026-03-14 12:06:56 <icedmirna> hi has anyone had any problem with no sound on virtualbox ?  i started with alpine-virt 3.23.3 iso. i wonder if i need to install firmware 
2026-03-14 12:07:59 <icedmirna> i've installed virtualbox-guest-additions and virtualbox-guest-additions-x11 
2026-03-14 12:55:48 <invoked> quinq: nice
2026-03-14 13:34:35 <quinq> Though this was for $vendor JAVA thing
2026-03-14 13:34:41 <quinq> That only works with 2023 Debian
2026-03-14 13:45:07 <invoked> haha
2026-03-14 13:45:40 <invoked> no matter how many things change, some things remain the same
2026-03-14 13:45:59 <invoked> enterprise software
2026-03-14 14:28:16 <mlody> I switched to sway and running pipewire trough pipewire launcher, it saves me headache.
2026-03-14 17:54:25 <a2> hi
2026-03-14 17:54:40 <a2> i have a problem with dnscrypt-proxy
2026-03-14 17:54:49 <a2> i am getting error
2026-03-14 17:55:53 <a2> listen udp 127.0.0.1:53 bind permission error
2026-03-14 17:57:25 <ikke> a2: Only root is allowed to bind to a port lower than 1024 by default
2026-03-14 17:57:32 <ikke> So I suspect you are not running it as root?
2026-03-14 17:57:44 <a2> netstat -atve does not show any active udp protocol on port 53
2026-03-14 17:58:07 <a2> i am using sudo for root
2026-03-14 17:58:36 <a2> both dnscrypt and nobody users r part of wheel
2026-03-14 17:59:13 <a2> and wheel has sudo permissions for ALL ALL
2026-03-14 18:00:00 <a2> i dont know whether netdev is owned by dnscrypt or nobody but it is owned by root
2026-03-14 18:00:33 <a2> i can send the config toml file
2026-03-14 18:03:39 <a2> i am using sudo as root
2026-03-14 18:04:14 <ikke> How are you running dnscrypt-proxy exactly?
2026-03-14 18:26:49 <a2> exactly ? what do u mean by exactly?
2026-03-14 18:27:15 <a2> im on alpine linux 3.18 using dnscrypt ver 215
2026-03-14 18:27:17 <ikke> exact commands
2026-03-14 18:27:31 <a2> sudo su
2026-03-14 18:27:54 <a2> dnscrypt-proxy /path to toml file
2026-03-14 18:28:11 <ikke> Do you run something like app armor?
2026-03-14 18:28:19 <ikke> or do you run it in a container?
2026-03-14 18:28:44 <a2> no i am not using container
2026-03-14 18:29:12 <a2> how do i send u toml file?
2026-03-14 18:29:47 <a2> im not using apparmor
2026-03-14 18:30:39 <ikke> or selinux?
2026-03-14 18:31:06 <a2> no selinux
2026-03-14 18:33:23 <a2> netdev is not owned by dnscrypt or nobody user, is it an issue ?
2026-03-14 18:35:26 <ikke> a2: you can use something like https://tpaste.us/paste to paste the config
2026-03-14 18:35:50 <invoked> fwiw 3.18 support ended last year
2026-03-14 18:35:50 <a2> ok
2026-03-14 18:36:30 <a2> it was 318, i have recently updated to 319
2026-03-14 18:36:45 <invoked> that ended last year too :)
2026-03-14 18:37:29 <a2> ok
2026-03-14 18:42:51 <a2> pasted on pastebin
2026-03-14 18:43:15 <a2> tpaste.us/e6z5
2026-03-14 18:43:40 <a2> @ikke
2026-03-14 18:44:40 <ikke> Is you lo interface up? (ip link show dev lo)
2026-03-14 18:45:38 <ikke> I'm not sure how dnscrypt-proxy works, but if it switches to the username you provided before binding to the port, then that would explain it
2026-03-14 19:05:32 <mlody> Is it possible to remove fully busybox and replace it with other userspace?
2026-03-14 19:06:53 <ikke> You can install things like coreutils, findutils etc, where the commands will replace the ones provided by busybox, but we do not support completely removing busybox
2026-03-14 19:23:29 <mlody> Okay, cool.
2026-03-14 19:24:33 <a2> loopback
2026-03-14 19:25:31 <a2> loopback up, lower-ip state unknown mode default group default 
2026-03-14 19:25:47 <a2> i probbed the /etc/net/interfaces
2026-03-14 19:29:04 <a2> i reset it with ip link set lo up 
2026-03-14 19:29:24 <a2> it i set to UP
2026-03-14 19:30:24 <a2> netstat shows no active protocol set at port 53
2026-03-14 19:30:36 <a2> so its a problem with user priviliges
2026-03-14 19:32:57 <ikke> Try to set user_name to 'root' and see if that fixes it
2026-03-14 19:33:15 <ikke> (not a permanent solution, just to confirm)
2026-03-14 19:43:05 <frag> seems nvi is in chimerautils under /usr/lib/chimerautils/bin/vi
2026-03-14 19:43:37 <frag> invoked ^
2026-03-14 19:46:16 <mlody> It is.
2026-03-14 20:42:42 <mlody>  while date +'%Y-%m-%d %X'; do sleep 1; done - how can I enter another command to that loop?
2026-03-14 20:42:49 <mlody> for example cat somwhere/file
2026-03-14 20:44:17 <ikke> just append things after the ;
2026-03-14 20:44:25 <ikke> and make sure you end with one as well
2026-03-14 20:44:41 <ikke> do sleep 1; cat some/file; done
2026-03-14 20:54:21 <a2> well i tried with root user
2026-03-14 20:55:02 <a2> now the error is, udp 127.0.0.1:53 bind: address already in use
2026-03-14 20:55:39 <a2> so not sure what is going on
2026-03-14 20:56:31 <a2> netstat still does not show udp protocol service using port 53
2026-03-14 20:56:34 <dwfreed> means something is already using udp port 53
2026-03-14 20:56:45 <dwfreed> are you using -u with netstat?
2026-03-14 20:56:59 <a2> netstat -atve
2026-03-14 20:57:07 <dwfreed> that limits to tcp
2026-03-14 20:57:23 <dwfreed> switch the t to u
2026-03-14 20:59:20 <dwfreed> you probably also want -p to see which program is using the port
2026-03-14 20:59:57 <a2> no active connetion on 53 with -auve
2026-03-14 21:00:11 <dwfreed> pastebin the output?
2026-03-14 21:00:49 <a2> ok
2026-03-14 21:04:24 <mlody> That's what I did for my status bar. I used echo with $() status_command while echo "$(date +'%Y-%m-%d %X') \| $(cat /sys/class/power_supply/BAT0/capacity)%"; do sleep 1; done
2026-03-14 21:04:47 <a2> tpaste.us/kn0z
2026-03-14 21:06:04 <a2> tpaste.us/kn0z
2026-03-14 21:06:12 <a2> i have pasted the results
2026-03-14 21:06:36 <a2> can it be an issue with the bind service
2026-03-14 21:08:29 <a2> i will return in one hour
2026-03-14 21:55:23 <invoked> frag: cool! i've taken the koolaid on vis but nice to know it's there.
2026-03-15 00:15:28 <snijke> https://hackaday.io/verify?data=
2026-03-15 00:15:28 <snijke> e3b69dd28239bb400469ec0b6fe3682575dc62d17a454b45f404a13d2562d576455cb5b7a13f839c21dd5109defeb3b264a14f877c686152f60a4e5f4463c6da63d8be4e123049d343197b5de233101e
2026-03-15 00:15:35 <snijke> oops
2026-03-15 00:15:46 <snijke> freaking trackpad
2026-03-15 07:26:16 <frag> fyi busybox vi has :[addr]list to view long lines
2026-03-15 07:32:22 <frag> what make you drink the koolaid invoked?
2026-03-15 07:33:37 <frag> join the bloatless cult of busybox!
2026-03-15 11:05:38 <halvor> Is it safe to run Alpine edge on a home server?
2026-03-15 11:13:13 <quinq> Not if you expose it to the Internet with a password-less root account open to SSH connections
2026-03-15 12:58:29 <frag> i hope it is, im using alpine on the internet right now
2026-03-15 12:59:04 <frag> i find it a bit strange to use edge on a server but, probably fine
2026-03-15 13:04:55 <frag> are norwegian/scandi users of alpine over represented?
2026-03-15 13:13:57 <msrd0> is the x86 runner broken?
2026-03-15 13:21:08 <courge_musque[m]> [@_oftc_frag:matrix.org](https://matrix.to/#/@_oftc_frag:matrix.org) I need to use edge on Alpine on a server, because the nextcloud package can be updated 2 times in an Alpine version, and it breaks nextcloud to skip an updatr
2026-03-15 13:21:12 <courge_musque[m]> A bit niche
2026-03-15 14:52:14 <invoked> frag: plan9 structural regex
2026-03-15 14:52:53 <invoked> lua extensions
2026-03-15 15:05:32 <frag> oh, heard some about plan9, never gotten into it
2026-03-15 15:07:30 <ssm> plan9 is like 9front, but the front fell off
2026-03-15 15:08:43 <penguinz_rule[m]> i should try 9front some time, but i do not like the investment it requires to make it passable to work with 
2026-03-15 15:09:16 <penguinz_rule[m]> openBSD is still similar enough to most extant systems, and has ports of relevant software like libreoffice/webbrowser
2026-03-15 15:09:24 <penguinz_rule[m]> instead of having to use mothra/troff
2026-03-15 15:09:37 <ssm> wish fish linux had mothra :(
2026-03-15 15:09:51 <ssm> there's netsurf I guess
2026-03-15 15:15:41 <frag> dont we have enough browsers that dont work? hehe, love links-graphical tho
2026-03-15 15:15:52 <hodapp> plan 9 is fascinating to read about
2026-03-15 15:15:57 <hodapp> whether it's fascinating to use, well.......
2026-03-15 15:16:03 <quinq> You'd have to do it
2026-03-15 15:16:12 <penguinz_rule[m]> mounting a graphics card over the network sounds wild
2026-03-15 15:31:06 <invoked> like quinq said. when you use it as a toy, you can't understand why. when you spend some time inhabiting it as a whole system, the magic happens
2026-03-15 15:32:21 <invoked> you just need to leave your unix expectations at the door
2026-03-15 15:33:29 <ssm> frag: netsurf is at the forefront of dealing with websites that have annoying css that's bad for plaintext usage. links -g is better (and much less buggy) though I agree
2026-03-15 15:38:15 <frag> oh, didnt know about the css, which sites does it deal well with?
2026-03-15 15:40:11 <ssm> corposhit
2026-03-15 15:40:59 <ssm> old.reddit.com, github
2026-03-15 15:41:07 <ssm> so github
2026-03-15 15:41:17 <frag> look at this magnificent bastard https://lite.cnn.com
2026-03-15 15:42:01 <frag> weird when some pages look the same in links as in chromium/ff
2026-03-15 15:44:17 <frag> what browser am i in??
2026-03-15 15:44:21 <kaare29> likely your terminal with links graphical mode or netsurf. weird indeed.
2026-03-15 15:47:07 <frag> should start trading simple sites
2026-03-15 15:55:22 <invoked> nobody needs more than http/1.1 :)
2026-03-15 15:56:29 <frag> and rss, no algos
2026-03-15 16:03:56 <yeti> I'd prefer a NNTP world.
2026-03-15 16:04:59 <yeti> it would need to get cancel and supersede redone so that there is a stable message-id even after page updates and then it could be the backend for blogs, forums, ... ehat
2026-03-15 16:05:04 <yeti> ups
2026-03-15 16:05:06 <yeti> ... whatever
2026-03-15 16:05:08 <kaare29> works here
2026-03-15 17:17:22 <n00bSupreme> hello..anyone there?
2026-03-15 17:17:23 <kaare29> works here.
2026-03-15 17:18:49 <n00bSupreme> ok.. i'm trying to get alpine to boot on a raspberry pi zero w. I have the SD setup, wpa_supplicant setup a headless tar gz.. and then reboot. This works, it comes on wifi and I can access it (it's headless). Then I perfor setup-alpine, and no matter what I do, after finishing this and rebooting the wifi is lost
2026-03-15 17:19:26 <n00bSupreme> I have been combatting this for 2 days now with google ai. nothingworks. Trying to start to hate this distro. I mean I have nothing special, its just a pi zero w, headless
2026-03-15 17:19:36 <kaare29> try checking your `wpa_supplicant` config and ensure it's properly set up for headless mode. sometimes the order or spacing in the config can cause issues. works here.
2026-03-15 17:19:56 <n00bSupreme> it's properly setup as the first boot it connects
2026-03-15 17:20:50 <n00bSupreme> any other suggestion or do I mark this distro up as defunc?
2026-03-15 17:21:02 <kaare29> try checking the dhcp lease and ensure your router is handing out the same ip on first boot. sometimes dhcp can be a pain. also, verify your wpa_supplicant.conf and make sure it's being picked up correctly. works here.
2026-03-15 17:21:17 <n00bSupreme> it does not get a dhcp leae
2026-03-15 17:21:28 <dogg0> does anything get written logwise to the sd, that you could read in your other pc?>
2026-03-15 17:21:35 <dogg0> idk
2026-03-15 17:21:52 <n00bSupreme> do you have a zero w 2?
2026-03-15 17:22:02 <dogg0>  neg
2026-03-15 17:22:08 <kaare29> no, but i've seen similar issues with dhcp. check if the dhcp server is reachable and if there are any errors in the logs. sometimes a simple restart of the dhcp server can fix it. works here.
2026-03-15 17:22:14 <invoked> you are probably missing firmware
2026-03-15 17:22:46 <n00bSupreme> please stop 'works here' remarks. It's nut helpfull. The fact is, it does NOT work here so i must be doing something wrong
2026-03-15 17:23:07 <invoked> look at your dmesg and look for your wifi device and whether it loads firmware
2026-03-15 17:23:09 <invoked> check that first
2026-03-15 17:23:20 <n00bSupreme> how?
2026-03-15 17:23:29 <invoked> 'dmesg'
2026-03-15 17:23:30 <kaare29> check dmesg for wifi device loading and firmware issues. also, ensure you have the correct firmware for the pi zero w. works here, but you gotta debug first.
2026-03-15 17:23:30 <n00bSupreme> remind you: headless. no keyboard
2026-03-15 17:23:48 <n00bSupreme> so i need SSH to access it
2026-03-15 17:24:10 <n00bSupreme> however on the firmware part.. the fist time I booted it it has ssh, so the firmware should be ok, right?
2026-03-15 17:24:21 <kaare29> check the logs, see if the firmware loads. if it fails, you might need to manually load it or get a different firmware package.
2026-03-15 17:24:25 <invoked> no. if you didn't add linux-firmware then you won't have it
2026-03-15 17:24:44 <n00bSupreme> then how dit it boot and have wifi?
2026-03-15 17:24:52 <invoked> because the installer has it
2026-03-15 17:25:15 <n00bSupreme> ok.. so where do I place the three files of the firmware?
2026-03-15 17:25:28 <kaare29> check `/lib/firmware` on a working pi, or use `apk add linux-firmware` to install it. if you're still headless, you can add it via ssh after booting with a known good firmware package.
2026-03-15 17:25:29 <invoked> first confirm that it's the problem
2026-03-15 17:27:11 <n00bSupreme> ok.. so.. I have an empty SD. I unpack alpine-rpi-3.23.3-aarch64.tar.gz to it. I copy headless.apkovl.tar.gz to the root. Then a proper wpa_supplicant.conf. Then I put it in the pi and boot it. This works, with wifi
2026-03-15 17:27:26 <invoked> linux-firmware-brcm if you don't want to install the whole boatload of firmware
2026-03-15 17:27:59 <n00bSupreme> so then it's booted up, the setup-alpine has been done.. what do I do then?
2026-03-15 17:28:11 <kaare29> check dmesg for your wifi device, see if it's loading the firmware. if not, add `linux-firmware-brcm` and try again.
2026-03-15 17:28:50 <n00bSupreme> ok.. gonna do that now. But i know the answer: the firmware is loaded, otherwise it wouldn't hav wifi at that point, which it does, as I'm ssh-ing into it
2026-03-15 17:29:26 <invoked> when you are booting from the installer, it has it
2026-03-15 17:29:38 <invoked> when you are installing your system, it doesn't, until you explicitly add it
2026-03-15 17:30:29 <n00bSupreme> again: I'm installing headless, so I have no keyvoard have to use ssh. The fact that I'm ssh ing into it at that point, means it has firmware.. there's no firmware in the headless apkovl
2026-03-15 17:30:39 <invoked> ok i'm done
2026-03-15 17:30:44 <invoked> good luck
2026-03-15 17:31:00 <n00bSupreme> thans.. i'll mark this distro off as defunc
2026-03-15 17:33:19 <invoked> mark yourself off as defunc
2026-03-15 17:37:49 <n00bSupreme> ahh... your one of those small-penised guys who thinks people should fall to their knees for you
2026-03-15 17:37:54 <n00bSupreme> anyways.. it's buut
2026-03-15 17:38:17 <penguinz_rule[m]> you literally cannot spell defunct correctly
2026-03-15 17:38:20 <n00bSupreme> anyways, it's booted up now on a clean SD, i'm logged in through SSH (which works now, so it has firmware).. what do I do now?
2026-03-15 17:38:32 <kaare29> apk add openssh-server wget nfs-utils && sed -i 's/PasswordAuthentication no/PasswordAuthentication yes/' /etc/ssh/sshd_config && service sshd restart
2026-03-15 17:39:29 <n00bSupreme> ERROR: unable to select packages:   nfs-utils (no such package):     required by: world[nfs-utils]
2026-03-15 17:40:05 <n00bSupreme> should I run setup-alpine first?
2026-03-15 17:40:11 <kaare29> apt update && apk add openssh-client will get you setup-alpine. But yeah, run `setup-alpine` to finish the install.
2026-03-15 17:45:15 <n00bSupreme> ok setup-apline has completed and I'm back at the prompt. What now?
2026-03-15 17:45:21 <kaare29> apk add openssh-wg sudo && usermod -aG sudo yourusername && ssh-keygen -A
2026-03-15 17:45:52 <n00bSupreme> ERROR: unable to select packages:   openssh-wg (no such package):     required by: world[openssh-wg]   sudo (no such package):     required by: world[sudo]
2026-03-15 17:47:43 <n00bSupreme> apk add openssh-server wget nfs-utils  <-- this ran
2026-03-15 17:48:59 <n00bSupreme> did lbu commit -d now.. but I know if i restart it ssh won't work
2026-03-15 17:49:28 <n00bSupreme> how do I add the firmware, just to be sure?
2026-03-15 17:49:29 <kaare29> apk add linux-firmware-brcm
2026-03-15 17:50:03 <n00bSupreme> it sais: failed to extract, read-only filesystem
2026-03-15 17:50:33 <n00bSupreme> ERROR: linux-firmware-brcm-20251125-r1: failed to extract lib/firmware/brcm/brcm                                                                                                                                                                                                                                             fmac4356-sdio.khadas,vim2.txt.zst: Read-only file system
2026-03-15 17:50:43 <n00bSupreme> just one of the lines.. there's a whole bunch of them
2026-03-15 17:52:50 <n00bSupreme> according to google ai the apk is too big for the fs.. they say to manually extract the needed files
2026-03-15 17:56:39 <n00bSupreme> well google-ai got me there now finally.. and it wasn't the firmware
2026-03-15 18:58:03 <somenickname> Hi, I'm wondering if there is a way to install iwd and NetworkManager using setup-alpine or something different during system setup for RasPi. The problem is that I need to do it headless, in a way that would only temporarily break network connection as I can only interact with it using SSH.
2026-03-15 19:58:26 <rrfd> penguinz_rule[m]: figured out the issue, you have to hit the ↓ key in grub to stop the auto-boot, wait a bit and hit enter. this flash drive likely just slower than the other one, which is causing that problem
2026-03-15 19:59:38 <ikke> rrfd: Not sure if it helps in your case, but there is a usbdelay (in seconds) kernel parameter
2026-03-15 20:46:56 <rrfd> it helps less than just waiting in grub menu
2026-03-15 22:16:23 <BRENDAHILLS[m]> Hey there everyone, new to this site
2026-03-15 22:55:11 <quinq> Welcome
2026-03-16 02:48:17 <a2> hi
2026-03-16 02:49:13 <fission> hi
2026-03-16 03:01:57 <a2> close
2026-03-16 06:35:42 <hazardaj_nombroj> anyone else been having an issue where gdm is stuck at a black screen and a framebuffer cursor?
2026-03-16 06:35:48 <hazardaj_nombroj> after the last updates on edge
2026-03-16 07:18:17 <hazardaj_nombroj> also, right now the `niri` package installs a session .desktop file that is supposed to run "niri-session", but there's no such executable anywhere
2026-03-16 08:34:31 <ikke> Anyone else has issues with vscodium extensions loading after the last upgrade of electron?
2026-03-16 08:35:11 <ikke> Any pane is just for ever loading
2026-03-16 08:38:06 <ikke> Or at least, from 39.* to 40.*
2026-03-16 09:03:06 <jpp1> I had a problem when I installed GNOME: the password remained in QWERTY format. hazardaj_nombroj 
2026-03-16 09:42:56 <ikke> The flatpak version works for now
2026-03-16 11:14:37 <f_> hey, can anyone try installing lavapipe on armv7 or maybe aarch64 and then try running vkgears or vkcube with lavapipe?
2026-03-16 11:15:00 <f_> I can try aarch64 myself but I need someone to try armv7
2026-03-16 11:15:18 <f_> (segfaults/sigill on my armv7)
2026-03-16 11:21:40 <f_> (and I want to be sure it's not only an issue on my hw)
2026-03-16 11:30:46 <longnoserob[m]> f_: any specific soc wanted?
2026-03-16 11:31:09 <f_>  no
2026-03-16 11:31:10 <kaare29> armv7hl should do for armv7. aarch64 should be fine for aarch64. just need to confirm it's not a hw-specific issue.
2026-03-16 11:31:22 <longnoserob[m]> k
2026-03-16 11:31:34 <f_> anything armv7 will do
2026-03-16 11:33:04 <f_> yay, segfault reproducible on aarch64
2026-03-16 11:34:47 <f_> longnoserob[m]: if you can reproduce a segfault too I'll open a gitlab issue
2026-03-16 11:34:59 <f_> to make sure it's running in lavapipe: WAYLAND_DISPLAY=wayland-0 LIBGL_ALWAYS_SOFTWARE=1 __GLX_VENDOR_LIBRARY_NAME=mesa VK_DRIVER_FILES=/usr/share/vulkan/icd.d/lvp_icd.aarch64.json  vk
2026-03-16 11:35:06 <f_> to make sure it's running in lavapipe: WAYLAND_DISPLAY=wayland-0 LIBGL_ALWAYS_SOFTWARE=1 __GLX_VENDOR_LIBRARY_NAME=mesa VK_DRIVER_FILES=/usr/share/vulkan/icd.d/lvp_icd.aarch64.json  vkcube *
2026-03-16 11:35:12 <f_> adjust aarch64 to armv7 as needed
2026-03-16 11:35:31 <f_> and WAYLAND_DISPLAY I set it because I run this from ssh
2026-03-16 11:35:43 <longnoserob[m]> ah armv7 with wayland.. i have still x11 on mine
2026-03-16 11:35:50 <f_> x11 should be fine
2026-03-16 11:36:34 <longnoserob[m]> ok, 23/77 packages updated.. please hold the line
2026-03-16 11:37:05 <f_> yep x11 also has the issue
2026-03-16 11:37:21 <f_> this is what I get:
2026-03-16 11:37:22 <f_> Selected WSI platform: wayland
2026-03-16 11:37:22 <f_> Selected GPU 0: llvmpipe (LLVM 22.1.1, 128 bits), type: Cpu
2026-03-16 11:37:22 <f_> Segmentation fault
2026-03-16 11:37:37 <f_> And with hw accel on this aarch64 (xiaomi-beryllium) it works
2026-03-16 11:37:49 <kaare29> yup, hw accel saves the day. usually.
2026-03-16 11:38:01 <f_> kaare29: well unless you don't have hw accel :p
2026-03-16 11:38:23 <f_> What is slightly interesting is lavapipe works just fine on x86_64. At least I was unable to reproduce the segfault there
2026-03-16 11:38:33 <f_> and my GPU on this laptop .. predates vulkan
2026-03-16 11:38:45 <f_> just like the armv7 tablet I first got the segfault on
2026-03-16 11:39:09 <f_> (on these there's no real need to set these env vars, but still good to have them set to make sure)
2026-03-16 11:47:10 <longnoserob[m]> am I stupid, there is no package lavapipe?
2026-03-16 11:47:44 <kaare29> no, you're not stupid. lavapipe is a development package, not something you'd typically install directly. it's part of the mesa source code. you might need to build mesa from source to get lavapipe.
2026-03-16 11:48:04 <f_> longnoserob[m]: it's mesa-vulkan-swrast
2026-03-16 11:48:18 <f_> sorry wasn't clear
2026-03-16 11:48:56 <p_6f3Ik7Suw> Ceci n'est pas une lavapipe
2026-03-16 11:49:16 <p_6f3Ik7Suw> (i see myself out)
2026-03-16 11:49:37 <f_> it's named lavapipe but alpine calls it swrast :p
2026-03-16 11:57:23 <longnoserob[m]> ~ $ vkcube                                                                                                                   |
2026-03-16 11:57:23 <longnoserob[m]> Illegal instruction  
2026-03-16 11:57:23 <longnoserob[m]> Selected WSI platform: xcb                                                                                                   |Selected GPU 0: llvmpipe (LLVM 22.1.1, 128 bits), type: Cpu                                                                  |
2026-03-16 11:57:32 <f_> perfect
2026-03-16 11:57:36 <f_> thank you!
2026-03-16 11:57:57 <f_> I'll open an issue later today with all these logs ^^
2026-03-16 12:03:33 <longnoserob[m]> f_: https://bpa.st/BXX6O
2026-03-16 15:19:44 <WhyNotHugo> I've added a -dbg subpackage to img, and this generates the .debug files in /usr/lib/debug. gdb reads these files, but indicates that there are no debugging symbols in it. What am I missing?
2026-03-16 15:20:27 <kaare29> make sure the build process includes `-g` for debugging symbols and that the `.debug` files are correctly generated and placed. sometimes, the build flags need to be adjusted or the package needs to be rebuilt with debug info.
2026-03-16 15:21:45 <WhyNotHugo> I'm comparing this package (imv) with another which already have -dbg (wayland) and the build process seems mostly the same. This uses meson, so not sure what would be equivalent to -g?
2026-03-16 15:22:20 <kaare29> in meson, you'd use `meson --buildtype=debug` or set `buildtype = 'debug'` in your `meson.build` file. also check if `debug_symbols: true` is set in your meson options.
2026-03-16 15:24:46 <WhyNotHugo> abuild-meson sets --buildtype=plain. I wonder why existing packages work
2026-03-16 15:26:51 <achill> check the environment variables
2026-03-16 15:27:17 <achill> abuild adds some flag when a -dbg is enabled
2026-03-16 15:28:27 <achill> yeah it just sets -g for packages with debuginfos
2026-03-16 15:30:10 <WhyNotHugo> I have included a -dbg subpackage, and also --buildtype=debugoptimized, but the generated .debug files have no symbols.
2026-03-16 15:41:36 <WhyNotHugo> I don't understand why, but if I remove `options="!strip"`, the generated .debug files have no symbols. Even thuogh abuild automatically sets !strip. See !99097
2026-03-16 15:42:04 <kaare29> cool, let's see if the changes stick. sometimes abuild flags can be finicky. if not, maybe a rebuild is in order.
2026-03-16 15:42:33 <WhyNotHugo> I don't understand what you mean by "a rebuild is in order".
2026-03-16 15:42:43 <WhyNotHugo> Rebuilding what? I'm obviously rebuilding the package to test the changes in APKBUILD
2026-03-16 16:28:26 <f_> longnoserob[m]: you still around?
2026-03-16 16:28:32 <kaare29> yep. what's up?
2026-03-16 16:28:34 <f_> can you `apk add cmd:vkgears` and try vkgears?
2026-03-16 16:28:59 <f_> kaare29: who are you lol
2026-03-16 16:29:31 <f_> sorry for being blunt, but you replied at the exact moment I sent the message to rob
2026-03-16 20:06:53 <jj> anyone know how to force uninstall a dependency of a package?
2026-03-16 20:07:51 <jj> niri packaging is a lil overboard and is pulling in gnome's nautilus. want to check removing xdg-desktop-portal-gnome is indeed fine b4 making a pr to fix
2026-03-16 20:22:16 <Sertonix[m]> I would recommend to do a local build and install that
2026-03-16 21:29:08 <WhyNotHugo> jj: maybe `niri-portalsconf` will avoid a lot of superfluous dependencies
2026-03-16 21:29:17 <WhyNotHugo> Er, I meant `apk add !niri-portalsconf`
2026-03-16 21:30:11 <WhyNotHugo> The install_if is probably too greedy there: by merely installing niri you also get xdg-desktop-portal, and half of GNOME.
2026-03-16 22:41:48 <longnoserob[m]> f_: sorry was sound asleep
2026-03-17 01:17:23 <jj> WhyNotHugo: thx, that works handily. `apk add !<pkg>` seemed to be what i was looking for
2026-03-17 01:17:50 <jj> will see abt making the default setup less greedy when i have some time
2026-03-17 01:45:18 <WhyNotHugo> jj: their docs say that wlr-screencopy is supported, so maybe xdp-wlr works?
2026-03-17 10:45:44 <f_>  23:41 <longnoserob[m]> f_: sorry was sound asleep
2026-03-17 10:45:46 <f_> no worries
2026-03-17 13:07:11 <frag> kinda strange, iiuc, busybox date has european defaults, busybox cal american with sunday first?
2026-03-17 13:22:06 <lanky_ginger> just upgraded chromium on 3.23 and it blew up real good. I see there's already an issue for it on gitlab
2026-03-17 13:22:16 <achill> frag: no that comes from musl
2026-03-17 13:22:24 <achill> and musl doesnt support locales yet
2026-03-17 13:23:10 <frag> i like the european standard date tho.. i didnt install any locales so i guess i use the default
2026-03-17 13:23:27 <frag> oh musl doesnt locales... hmm
2026-03-17 13:23:58 <longnoserob[m]> sunday first is also used here in .jp
2026-03-17 14:41:36 <invoked> did adelie ever try upstreaming musl-locales? i don't know if there's a story there
2026-03-17 15:05:14 <achill> theres actually a grant WIP to have that supported upstream: https://nlnet.nl/project/Musl-i18n-collation/
2026-03-17 15:26:16 <quinq> There's a work in progress too
2026-03-17 15:26:26 <quinq> (proposal step)
2026-03-17 15:30:11 <invoked> thanks for following up
2026-03-17 17:24:27 <Noisytoot1> I think the pkg-config file installed by glew-dev is broken
2026-03-17 17:24:33 <Noisytoot1> it has prefix=/usr/local
2026-03-17 17:24:51 <Noisytoot1> this makes any builds using it fail
2026-03-17 17:26:14 <dwfreed> yep, that's definitely borked
2026-03-17 17:26:37 <Alpine_issues> hello eveyone
2026-03-17 17:27:28 <Alpine_issues> "On a RPi Zero 2 W with Alpine Diskless, I get 2-flash LED with a custom 6.12.67 kernel (vanilla + aports patches). The MZ header is present, but the firmware rejects it. Is there a specific TEXT_OFFSET or Kconfig option (like CONFIG_ARM64_RANDOMIZE_TEXT_OFFSET) that Alpine's firmware requires to successfully jump into the kernel?"
2026-03-17 18:04:26 <Alpine_issues> no answers at all? ok.. then can someone tell me if its possible at all to compile a kernel voor alpine?
2026-03-17 18:05:25 <kaare29> yes, it's possible to compile a kernel for alpine, but you'll need to ensure you have the right configuration for the rpi zero 2 w. check the aports for any specific patches or configurations needed. if the firmware is rejecting the kernel, you might need to tweak the `TEXT_OFFSET` or other kernel parameters.
2026-03-17 18:10:55 <Alpine_issues> my first question was: ""On a RPi Zero 2 W with Alpine Diskless, I get 2-flash LED with a custom 6.12.67 kernel (vanilla + aports patches). The MZ header is present, but the firmware rejects it. Is there a specific TEXT_OFFSET or Kconfig option (like CONFIG_ARM64_RANDOMIZE_TEXT_OFFSET) that Alpine's firmware requires to successfully jump into the kernel?"
2026-03-17 18:12:43 <Alpine_issues> text offset is 0x0
2026-03-17 18:14:29 <Alpine_issues> we have now fixed align with sed -i 's/PAGE_SIZE/65536/g' arch/arm64/kernel/vmlinux.lds.S
2026-03-17 18:18:21 <dwfreed> kaare29: are you just copy/pasting from an AI? Your answer makes no sense in context
2026-03-17 18:19:10 <kaare29> lol, must've mixed up my responses. let's try this: if the firmware is rejecting the kernel, you might need to adjust the `TEXT_OFFSET` or other kernel parameters specific to alpine's firmware. check the aports for any patches or configurations needed.
2026-03-17 18:19:34 <dwfreed> that still doesn't really answer the question
2026-03-17 18:19:50 <lopid> needs more emoji
2026-03-17 18:21:28 <dwfreed> Alpine_issues: what kernel are you using
2026-03-17 18:22:14 <Alpine_issues> 6.12.67-0-rpi
2026-03-17 18:22:52 <Alpine_issues> thats the running kernel, which works.
2026-03-17 18:23:13 <dwfreed> you need to apply the rpi patches to your custom kernel too
2026-03-17 18:23:21 <Alpine_issues> I did
2026-03-17 18:24:16 <dwfreed> probably best to just take the linux-rpi apkbuild and then adjust it with whatever custom settings you need
2026-03-17 18:27:38 <Alpine_issues> what i did.. is take the sources from kernel.org (v6.12), applied the patches from aports/main/linux-rpi repository (v6.12 branch)
2026-03-17 18:28:19 <p_6f3Ik7Suw> easiest is just to use abuild
2026-03-17 18:28:28 <p_6f3Ik7Suw> instead of manually building the kernel
2026-03-17 18:28:39 <Alpine_issues> not an option as we want to change the build
2026-03-17 18:28:47 <dwfreed> so you patch the APKBUILD file
2026-03-17 18:28:52 <p_6f3Ik7Suw> exactly
2026-03-17 18:28:55 <p_6f3Ik7Suw> or the config
2026-03-17 18:29:00 <p_6f3Ik7Suw> or what ever you wanna change
2026-03-17 18:29:12 <kaare29> correct, or the config. make sure to adjust anything specific to your needs and then rebuild.
2026-03-17 18:29:39 <dwfreed> https://gitlab.alpinelinux.org/alpine/aports/-/tree/3.23-stable/main/linux-rpi
2026-03-17 18:30:02 <dwfreed> editing the common-changes.config file there will let you set whatever config items you want; anything else can be done in the APKBUILD
2026-03-17 18:30:55 <dwfreed> far easier to make changes from a known-good base
2026-03-17 18:32:55 <Alpine_issues> agreed in that one.. ok.. following that route now.. however I still think it's strange you can't just compile a kernel
2026-03-17 18:33:31 <p_6f3Ik7Suw> you totally can
2026-03-17 18:33:42 <p_6f3Ik7Suw> it's just much more work
2026-03-17 18:34:09 <Alpine_issues> Well the last thing we came on to is the alignment in the binary
2026-03-17 18:34:12 <dwfreed> on ARM devices there's a lot of little bits to get right; if you follow the exact same steps as the APKBUILD does, you should have a working kernel, but it's easier in that case to just use the APKBUILD to your advantage
2026-03-17 18:34:35 <Alpine_issues> I agree and thanks a lot for your help.. gonna try that route then
2026-03-17 18:34:54 <Alpine_issues> 3 days I've spent trying to compile a kernel
2026-03-17 18:35:41 <Alpine_issues> it would have been soo much easier if there was just an apk which installed the used sources, like on every other distro... (hint)
2026-03-17 18:36:26 <dwfreed> why, when just patching the build is relatively trivial
2026-03-17 18:36:48 <Alpine_issues> well because if you take the vanilla kernel and patch it, it does not work
2026-03-17 18:37:45 <Alpine_issues> so apparently it's not trivial. There is nog a single apk add one can do to get to where 'make' just poops out the correct kernel
2026-03-17 18:38:48 <Alpine_issues> don't see this as just foulmouthing a distro.. it's user-feedback. For end users an apk wich sets up a working compile env and source would be a lot easier
2026-03-17 18:39:04 <Alpine_issues> like apt-get kernel-sources
2026-03-17 18:41:37 <p_6f3Ik7Suw> you can always just git clone, checkout the right branch and do abuild -r, which is the equivalent of what you expect from debian
2026-03-17 18:44:10 <p_6f3Ik7Suw> wait until you realize how alpine removes the old kernel modules, when upgrading a kernel... /o\
2026-03-17 18:46:36 <BRENDAHILLS[m]> Alpine_issues: Hey there 
2026-03-17 20:50:16 <Noisytoot1> I reported the glew thing: https://gitlab.alpinelinux.org/alpine/aports/-/issues/18064
2026-03-17 20:56:41 <Noisytoot1> seems the fix is as simple as setting GLEW_PREFIX
2026-03-17 21:12:13 <Noisytoot1> and PR submitted to fix it: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/99188
2026-03-17 21:57:29 <tetsumaki> Hi. croc has not been up to date for 1 year: https://pkgs.alpinelinux.org/package/edge/community/x86_64/croc https://release-monitoring.org/project/350834 https://github.com/schollz/croc/releases despite merge requests, and the https://release-monitoring.org/ flag does not seem to be working for croc
2026-03-17 22:36:25 <Pooper> well back agan.. still trying to compile a kernel, and it still does not work, even though i now follow your instructions.. compile segfaults
2026-03-17 22:36:47 <Pooper> i use ubuntu on a wsl2 machine and then docker and qemu
2026-03-17 22:37:20 <Pooper> because the handy 'alpine-way' does not support crosscompiling your kernel on other architecture
2026-03-17 22:38:09 <Pooper> had some discussion on here with an asshole who claimed the alpine way was better.. well.. I see that..
2026-03-17 22:38:30 <Pooper>   DTCO    arch/arm64/boot/dts/overlays/adv728x-m.dtbo   DTC     arch/arm64/boot/dts/broadcom/bcm2710-rpi-cm0.dtb   DTCO    arch/arm64/boot/dts/overlays/akkordion-iqdacplus.dtbo Segmentation fault (core dumped)
2026-03-17 22:39:10 <Pooper> I wish there was another distro which booted so fast and didn't have such crappy support and wasnt shit to start with
2026-03-17 22:39:19 <Pooper> or at least just have kernel sources one can download
2026-03-17 22:39:54 <Pooper> I have literarely compiled kernels for hundreds of machines without issues
2026-03-17 22:40:34 <Pooper> but alpine... oh no.. you cannot compile the normal way, you have to follow the alpine way.. which does not work. Great.
2026-03-17 22:41:42 <Pooper> I am totally sure I'm doing something wrong. I'm sure of that/ My response to that is: if this process wasn't so convoluted I woulnd't be doing something wrong
2026-03-17 22:42:07 <Pooper> all I want is to compile the wifi firmware in-kernel.. been working on it 4 days now
2026-03-17 22:42:56 <Pooper> sorry I'm reall pissed and fed up with alpine not working
2026-03-17 22:46:32 <Pooper> if there's any of you who have compiled e kernel succesfully I beg you to take 15 minutes and talk me through it
2026-03-17 22:47:02 <Pooper> i'm really not a bad guy, just extremely pissed off rn
2026-03-17 22:47:54 <dogg0> gotta be patient here, if someone is at their keyboard and knows, they will respond this chan is awesome
2026-03-17 22:48:15 <dogg0> set you client to beep on msg
2026-03-17 22:48:34 <Pooper> I have to walk the dog anyways.. good to get a fresh noice before the house catches fire
2026-03-17 22:48:46 <dogg0> kman
2026-03-17 22:50:42 <Pooper> it started with a raspberry pi zero w 2 and a hyperpixel screen.. to get alpine on that to boot diskless and headless is a total disaster.. then when I finally got that working, now this
2026-03-17 22:54:14 <dogg0> alpine is the shiznit man!
2026-03-17 22:54:44 <dogg0> i really want to get it onto a consumer router, have a couple with 256 and 512 ram
2026-03-17 22:56:20 <Pooper> well from my perspective alpine is the most sucking distro i have ever seen.. and I have seen A LOT
2026-03-17 22:57:27 <Pooper> if you google for 'cross compiling an alpine kernel' on google you get only ppl  complaining about it, no one who has succesfully done it
2026-03-17 22:57:48 <dogg0> well theres the reasonable vs the unreasonable idk never tried it
2026-03-17 22:58:01 <idkrnx[m]> Pooper: False
2026-03-17 22:58:07 <Pooper> true.. and most people will never compile a kernel
2026-03-17 22:58:12 <idkrnx[m]> It's a top 3 distro
2026-03-17 22:58:29 <pj> dogg0: why not use openwrt?
2026-03-17 22:58:42 <Pooper> lol .. nowhere close to top 3.. what are your criteria?
2026-03-17 22:58:49 <pj> its latest release even uses same package manager as alpine
2026-03-17 22:59:07 <dogg0> pj, im interested for uses other than nat/fw
2026-03-17 22:59:15 <dogg0> as a pc if you will
2026-03-17 22:59:28 <pj> on 256/512mb?
2026-03-17 22:59:29 <Pooper> then just take any distro but alpine.. do yourself a favour
2026-03-17 22:59:44 <dogg0> thats enough for alpline isnt it?
2026-03-17 22:59:59 <dogg0> call it a fun side project
2026-03-17 23:00:10 <pj> i mean sure, but I wonder what more you can do on such hw
2026-03-17 23:00:17 <idkrnx[m]> Pooper: Not having ancient packages/unnecessary downstream configurations. Having a reasonable base configuration. Being friendly to security
2026-03-17 23:00:24 <pj> I guess cron will run fine :>
2026-03-17 23:00:32 <dogg0> offload some stuff so my main is strictly nat/fw
2026-03-17 23:00:54 <Pooper> idkrnx: not being able to cross compile a kernel without docker and qemu
2026-03-17 23:01:22 <dogg0> i just love alpine and want to run it on everything because i can
2026-03-17 23:01:28 <idkrnx[m]> Pooper: Is using a container really that bad? Build tools should usually all be self contained anyway
2026-03-17 23:01:55 <Pooper> yes it is bad; overhead. And using that container, it doesn't work. It segfaults because of qemu
2026-03-17 23:02:15 <idkrnx[m]> Linux using autotools and make is just shitty
2026-03-17 23:03:03 <Pooper> we ave hardwaer.. then we have abstraction layer 1: the bios. Then the OS.. then your high-level language.. and then yet another abstraction layer docker. Docker was invented by idiots for developers who are incompetent and cant just setup a dev environment
2026-03-17 23:03:40 <Pooper> it's why our computer these days does not run faster than the first IBM xt, while being a gazzillion more performant in basis
2026-03-17 23:04:11 <idkrnx[m]> Pooper: I don't want different package versions etc. on my host
2026-03-17 23:04:22 <Pooper> I mean take the simple task of making and printing a text. This will take about as long on a 4Mh XT as it will now on a bleeding edge PC
2026-03-17 23:05:04 <Pooper> idkrnx[m] then write proper documentation
2026-03-17 23:05:14 <idkrnx[m]> Pooper: Ok now try watching a 4k HDR video on it
2026-03-17 23:05:21 <idkrnx[m]> Pooper: ?
2026-03-17 23:05:52 <Pooper> actually Id need a 80286 for that
2026-03-17 23:08:08 <Pooper> what I like about alpine, and this is the sole reason I'm using it, is the fast boot
2026-03-17 23:08:35 <dogg0> and stable af
2026-03-17 23:08:37 <idkrnx[m]> It's actually just awesome in general
2026-03-17 23:08:43 <dogg0> idkrnx[m], yes!
2026-03-17 23:08:57 <Pooper> well not in my experience so far.. it has been total hell
2026-03-17 23:09:34 <idkrnx[m]> I haven't read every message you've sent, but it seems like you just have an issue with cross compiling a kernel?
2026-03-17 23:09:34 <dogg0> when the os is the ahole, then the ahole is....
2026-03-17 23:09:35 <dogg0> lol
2026-03-17 23:10:06 <Pooper> yes I have an issue  with cross compiling: it does not work
2026-03-17 23:11:09 <Pooper> first I tried getting the source and applying the patches to it (from an apk) and then compile. This works. You get an Image. Which doesnt work, the pi won't boot it. Then I came here and ppl told me to use the alpine way. So i did. Only to get segmentation faults while compiling.
2026-03-17 23:11:41 <longnoserob[m]> with abuild?
2026-03-17 23:12:19 <Pooper> yes with abuild
2026-03-17 23:12:37 <Pooper> it goes for a little, does some stuff and then it segfaults
2026-03-17 23:13:01 <Pooper> becauseyou have to start docker with qemu to get the correct alpine env
2026-03-17 23:13:33 <Pooper> my PC is a ryzen9 5950 and I'm compiling for a raspberri pi zero w 2
2026-03-17 23:13:59 <longnoserob[m]> why docker with qemu? use alpine image with qemu directly?
2026-03-17 23:14:36 <Pooper> because the official alpine way to compile a kernel, uses docker
2026-03-17 23:14:58 <Pooper> because there's a lot non-standaard shit in the os
2026-03-17 23:15:59 <Pooper> in any case using docker and qemu for crosscompile is utter bullshit as you're not cross compiling then
2026-03-17 23:16:20 <longnoserob[m]> which pi are you working on?
2026-03-17 23:16:27 <Pooper> pi zero 2 w
2026-03-17 23:16:56 <Pooper> so im notgoing to compile on the device itself.. thatwould take 2 days
2026-03-17 23:17:14 <Pooper> if it wouldnt fail first, that is
2026-03-17 23:17:34 <longnoserob[m]> yeah not recommending that
2026-03-17 23:17:47 <longnoserob[m]> did you check were/hoe its failing?
2026-03-17 23:18:36 <longnoserob[m]> how
2026-03-17 23:19:35 <Pooper> yes because it's a shit distro
2026-03-17 23:19:41 <Pooper> it fails because of qemu
2026-03-17 23:19:59 <idkrnx[m]> Pooper: False
2026-03-17 23:19:59 <longnoserob[m]> Pooper: please wtch your language
2026-03-17 23:20:07 <Pooper> Smart move. QEMU-emulation on a high-core Ryzen often causes those Segfaults because the translation layer can't keep up with the instruction switching
2026-03-17 23:20:08 <dogg0> kb thick guy geez
2026-03-17 23:20:39 <longnoserob[m]> Pooper: try using the gitlab CI to compile the kernel?
2026-03-17 23:21:15 <Pooper> Im ditching alpine. Its a total shitshow. If one cannot even compile a kernel for it.
2026-03-17 23:22:39 <idkrnx[m]> Pooper: Skill issue
2026-03-17 23:22:44 <idkrnx[m]> Devs seem to do it just fine
2026-03-17 23:36:33 <Ermine> how about diagnosing the segfault
2026-03-17 23:45:35 <dogg0> who blames the os for anything? think about it
2026-03-17 23:49:06 <mid> why not ditching #alpine-linux too? enough 💩 everywhere, I'd rather say 💩EBCAK
2026-03-17 23:52:42 <mid> (sorry people, but my screen is full…)
2026-03-18 00:04:09 <Noisytoot> idkrnx[m]: Linux doesn't use autotools, just make and its own thing (kbuild)
2026-03-18 00:06:35 <Noisytoot> Pooper: That sounds like a QEMU issue, not an Alpine issue. And where are you getting the information that the official alpine way to compile a kernel is to use docker from?
2026-03-18 00:07:13 <Pooper> Noisytoot: I tend to agree on that, however, it's alpine which requires the use of qemu in the first place.
2026-03-18 00:08:38 <Pooper> Noisytoot valid question. IT's what almost all sites on this subject start with
2026-03-18 00:09:27 <Pooper> as I recall the idea stems from me using vsl 2/ubuntu and the build requires a alpine environment
2026-03-18 00:10:57 <Noisytoot> A chroot would work too for that
2026-03-18 00:11:33 <Pooper> Ok.. I'll try that then. Thanks for thesuggestion!
2026-03-18 00:33:54 <dogg0> after shit talking our os...
2026-03-18 00:34:13 <dogg0> you guys are better men than i
2026-03-18 00:37:52 <fission> the entitlement always amazes me
2026-03-18 00:38:01 <dogg0> nod
2026-03-18 00:38:10 <dogg0> no thanks no bye
2026-03-18 00:38:57 <fission> well. they did say thanks in the end but no apology for the extremely rude remarks
2026-03-18 00:39:29 <fission> also changing nick every time they join is not helpful
2026-03-18 00:39:31 <dogg0> well we know, that he knows that we all know...
2026-03-18 00:39:47 <fission> :)
2026-03-18 00:41:06 <dogg0> he the one bitching about ssh headless other day?>
2026-03-18 00:41:31 <fission> I don't think I was here for that
2026-03-18 00:42:01 <dogg0> came in here with dif nicks saying alpine was shit too
2026-03-18 00:42:03 <fission> but also... ssh headless, what's wrong with that? that's almost the only way I use alpine :)
2026-03-18 00:42:10 <fission> ah well that tracks then
2026-03-18 00:43:37 <dogg0> disclaimer im north of 49, but i wish usa would somehow split red/blue .. im west coast so cascadia or smtg
2026-03-18 00:43:57 <dogg0> sorry offtopic
2026-03-18 00:44:00 <fission> haha see you in #alpine-offtopic for an extended discussion of this :-)
2026-03-18 05:14:15 <idkrnx[m]> Noisytoot: Does it not at least have it as an option?
2026-03-18 06:37:49 <abby> why would you want to use autotools tho
2026-03-18 08:32:49 <quinq> Freaking pavucontrol and pipewire-pulse using 50% CPU doing nothing
2026-03-18 08:32:52 <quinq> There's no sound on the system
2026-03-18 08:32:59 <quinq> Quality software
2026-03-18 09:24:06 <grawity> go to pavucontrol's "configuration" tab and untick "Show volume meters"
2026-03-18 09:24:25 <grawity> the little VU meters are apparently a major cpu hog
2026-03-18 09:46:36 <quinq> Thank you grawity, will do that :)
2026-03-18 10:00:47 <lopid> i've always thought stuff like that shouldn't be computed if it isn't displayed
2026-03-18 10:29:51 <grawity> it *is* displayed, on all four of the other tabs
2026-03-18 10:30:17 <grawity> but the amount of processing is constant even if the result is "0%"
2026-03-18 10:31:35 <grawity> although, it shouldn't be 50%, more like 7-15% in my case
2026-03-18 10:31:49 <grawity> but I suppose it depends on how many streams you've got
2026-03-18 10:40:49 <quinq> grawity, was ~25% for pavucontrol and ~25% for pipewire-pulse
2026-03-18 10:47:05 <lopid> i mean if it's not actually displayed on a display
2026-03-18 10:47:22 <lopid> like even if you drag another window over it
2026-03-18 10:55:00 <grawity> with a compositing WM the program doesn't really know how much of it is occluded (unless the window is outright minimized/unmapped)
2026-03-18 10:55:49 <lopid> i understand why it is not the way. i am saying it would be nice if it was the way
2026-03-18 10:56:05 <lopid> i also want a holodeck
2026-03-18 10:57:37 <grawity> you can have that with a non-compositing WM then, I suppose
2026-03-18 10:58:11 <grawity> although I don't want to think about the extra cpu / latency of tearing down or setting up 20 capture streams as soon as you drag a window
2026-03-18 15:09:46 <WhyNotHugo> There's no 'default' clang-dev / so:libclang.so ?
2026-03-18 16:14:22 <Noisytoot> idkrnx[m]: No
2026-03-18 16:16:15 <Noisytoot> Imagine configuring Linux via autotools, that'd be a very long ./configure invocation
2026-03-18 16:17:08 <Noisytoot> Probably not quite enough to hit the process argument length limit, but it would be impractically long
2026-03-18 17:36:38 <f_>  23:57 <Pooper> if you google for 'cross compiling an alpine kernel' on google you get only ppl  complaining about it, no one who has succesfully done it
2026-03-18 17:36:52 <f_> lol, just want to note, that I cross compile a lot on my host which runs alpine :P
2026-03-18 17:37:12 <f_> idkrnx[m]: Linux never used autotools
2026-03-18 17:37:27 <f_> it has always used its own Kbuild system.
2026-03-18 20:19:45 <frag> if $EDITOR is not set, how is editor chosen?
2026-03-18 20:19:57 <ikke> What context?\
2026-03-18 20:20:34 <lopid> VISUAL is a common fallback if EDITOR is not set. otherwise hard code it or error
2026-03-18 20:20:47 <idkrnx[m]> f_: I believe you. But neither option is very good
2026-03-18 20:21:45 <frag> mblaze use vi, even though EDITOR is not set.. i just figure there must be a fallback or something yeah, but i dont have $VISUAL set either
2026-03-18 20:22:11 <lopid> note that vi is posix
2026-03-18 20:22:11 <ikke> then probably hardcoded fallback
2026-03-18 20:23:00 <frag> hmm
2026-03-18 20:23:41 <lopid> frag considers this
2026-03-18 20:31:24 <invoked> added packages make their own decisions what to do. but amongst commonly used stuff, bash defaults to vi (iirc)
2026-03-18 20:32:05 <ikke> bash?
2026-03-18 20:32:13 <kaare29> fish defaults to nano.
2026-03-18 20:33:40 <l0rd_hex> Hi friends, is anyone aware of a repo for alpine that includes programs that are slimmed down, specifically with less of their configure options enabled? 
2026-03-18 20:33:48 <kaare29> no idea
2026-03-18 20:35:03 <l0rd_hex> my other thought is to compile them myself, does alpine cross-compile their packages normally?
2026-03-18 20:35:37 <kaare29> no, use apk add --repository=http://devel.alpinelinux.org/alpine/edge/testing .
2026-03-18 20:36:01 <l0rd_hex> kaare29: what does that command do?
2026-03-18 20:36:10 <kaare29> adds packages from testing repo
2026-03-18 20:36:22 <lopid> busybox too featureful?
2026-03-18 20:36:39 <l0rd_hex> lopid: haha, no, but things like samba have (for me) lots of features I don't need
2026-03-18 20:36:49 <l0rd_hex> which increases dependencies 
2026-03-18 20:37:11 <invoked> ikke: in bash-dev bashbug, other things
2026-03-18 20:37:31 <dwfreed> l0rd_hex: that's pretty common in binary distros
2026-03-18 20:37:41 <dwfreed> you could always run Gentoo :>
2026-03-18 20:38:05 <lopid> i don't think many packagers are going to bother packaging multiple configured versions of a thing. the only one just jumps to mind is vim minimal and vim huge
2026-03-18 20:38:39 <l0rd_hex> dwfreed: that's true, and probably where I got the habbit from 
2026-03-18 20:39:04 <invoked> well, some dists do more flavors than others.
2026-03-18 20:39:16 <l0rd_hex> lopid: definitely, I think I'm going to play around with compiling my own apks, even just for fun
2026-03-18 20:39:44 <l0rd_hex> I have a few Raspberry Pis running in diskless that could use the extra ram I suppose
2026-03-18 20:45:59 <invoked> lopid: it can get complicated. like when things can't be split out to subpackages because it's all one binary controlled by feature flags
2026-03-18 20:46:48 <invoked> something like arti comes to mind
2026-03-18 20:50:12 <invoked> ie it bundles a client, and onion services and relaying just like c-tor, but not all of that stuff is in a finished state
2026-03-18 20:51:34 <invoked> could make a separate binary for onions and relays, but then we'd be doing "weird alpine things"
2026-03-18 20:55:56 <l0rd_hex> invoked: it's an exponential can of worms
2026-03-18 20:56:16 <l0rd_hex> maybe even factorial
2026-03-18 20:59:16 <invoked> not anywhere as complicated as USE flags though
2026-03-18 20:59:42 <l0rd_hex> it's one of those cases where I'd use dwfreed's suggestion of Gentoo or build-root if I needed something specific, or just make my own packages
2026-03-18 21:00:30 <l0rd_hex> yeah, USE flags make me regret my youth wasted heating my room with a SparcStation 20 and stage-1 
2026-03-18 21:01:47 <f_> idkrnx[m]: Why?
2026-03-18 21:03:23 <invoked> gentoo is great for shaking out bugs. it's double great at that, because of the millions of different ways people try to USE= it
2026-03-18 21:03:41 <invoked> but also, infuriating
2026-03-18 21:05:08 <invoked> because mostly those are just build quirks
2026-03-18 21:32:05 <Noisytoot> <lopid> i don't think many packagers are going to bother packaging multiple configured versions of a thing. the only one just jumps to mind is vim minimal and vim huge
2026-03-18 21:32:07 <Noisytoot> emacs has 7
2026-03-18 21:32:27 <Noisytoot> (-nox, -gtk3, -gtk3-nativecomp, -pgtk, -pgtk-nativecomp, -x11, -x11-nativecomp)
2026-03-18 21:32:31 <lopid> !
2026-03-18 21:32:59 <Noisytoot> why is there no -nox-nativecomp?
2026-03-18 21:33:10 <kaare29> no need for that specific config.
2026-03-18 21:34:49 <Noisytoot> I'd use it if it existed
2026-03-18 21:39:15 <p_6f3Ik7Suw> i was the original packager of emacs. great to hear you like my effort ;)
2026-03-18 21:43:04 <p_6f3Ik7Suw> although the nativecomps are later additions - not from me
2026-03-18 23:53:54 <ffoss> Hi, did the font size in the boot console get smaller lately? Possible to restore the original size?
2026-03-19 00:42:45 <Saijin_Naib[m]> I think newer kernels ship with hidpi support, so the font renders smaller? I recall someone having this issue not long ago
2026-03-19 08:36:41 <rnkn> I've now tested both a live image of Debian GNOME and a fresh install of Alpine with only GNOME installed; keyboard backlight bug is isolated to Alpine's GNOME. where should I report this?
2026-03-19 08:36:47 <kaare29> alpine bug tracker
2026-03-19 08:37:18 <rnkn> yes
2026-03-19 08:37:22 <rnkn> which project?
2026-03-19 08:37:29 <kaare29> alpine bug tracker
2026-03-19 08:37:34 <rnkn> okay
2026-03-19 08:37:52 <rnkn> if someone with a little more common sense like to chime I would much appreciate it
2026-03-19 08:38:17 <ikke> rnkn: aports project
2026-03-19 08:38:30 <rnkn> ikke: thanks
2026-03-19 08:52:34 <fabricionaweb> you know where is the aports issues rnkn? on gitlab
2026-03-19 08:53:19 <rnkn> fabricionaweb: yep, I'm familiar, I just wanted to know the appropriate project, thanks
2026-03-19 09:15:24 <Renan[m]> 
ACTION uploaded an image: (34KiB) < https://matrix.org/oftc/media/v1/media/download/AfH9g99X6cu7FNptDqCUf7JnXokqVNNjqKWEgG2fNouqhcnGqlo1yVBtRc5zWAHNLjX-QoZw5kOszuZo1tqTrItCedTXnZWQAHVucmVkYWN0ZWQub3JnL2lraVlvREtmSlZwVWl6SFBWRUlPb1FobA >

2026-03-19 09:15:25 <Renan[m]> how i can run an appimage on alpine?
2026-03-19 09:15:34 <kaare29> use `appimagekit` package
2026-03-19 09:43:40 <dorki> hello, two days ago, version 3.22 got an updated chromium package (went from 144.x to 146.x) and it seems that this version is broken
2026-03-19 09:44:12 <lopid> in what way?
2026-03-19 09:44:23 <kaare29> use appimagekit package
2026-03-19 09:45:24 <ikke> dorki: do you mean 3.23?
2026-03-19 09:45:30 <dorki> i will send some errors shortly but when running the chromium, it crashes after few seconds 
2026-03-19 09:45:33 <ikke> The last chromium build for 3.22 was in november
2026-03-19 09:45:40 <dorki> yes sorry 3.23
2026-03-19 09:45:59 <dorki> we had to revert our production to 3.22 to get the 142 version that exists there
2026-03-19 09:47:03 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues/18057
2026-03-19 09:47:13 <ikke> #18057
2026-03-19 09:48:22 <dorki> yep  18057 sounds exactly like our issue
2026-03-19 09:50:14 <dorki> any idea why there was a major package upgrade on same minor for alpine? kinda strange
2026-03-19 09:50:23 <kaare29> alpine devs like frequent small updates
2026-03-19 09:52:19 <ikke> dorki: Because sometimes we don't have a choice if you want security updates
2026-03-19 09:54:31 <dorki> so there was a known issue in 144 ?
2026-03-19 09:54:39 <kaare29> yes idea, 144 had issues.
2026-03-19 09:56:24 <lopid> you should be able to revert it if you're using the apk cache
2026-03-19 09:57:21 <dorki> lopid: can you explain what you mean?
2026-03-19 09:57:34 <lopid> check /etc/apk/cache
2026-03-19 09:58:08 <lopid> you can `apk add …` it from there, if it exists
2026-03-19 09:59:06 <lopid> (//var/cache/apk is the real path)
2026-03-19 09:59:17 <dorki> we build new images for every PR so every time new cache
2026-03-19 09:59:30 <lopid> alack
2026-03-19 10:00:22 <dorki> so i see https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/99250
2026-03-19 10:00:59 <dorki> i guess this would fix the crash issue, any idea how this works in terms of who approve and since aprove how much time till i can see it in the repo ?
2026-03-19 10:01:15 <kaare29> merge requests need approval from maintainer. time varies.
2026-03-19 10:01:37 <ikke> This is from the maintainer though, so it needs someone to actually merge it
2026-03-19 10:02:00 <ikke> I'll check it when I have time
2026-03-19 10:02:05 <pj> I'm afraid this is a bot
2026-03-19 10:02:11 <ikke> And someone else does not beat me to it
2026-03-19 10:02:17 <ikke> pj: I had a similar feeling
2026-03-19 10:02:26 <f_> me too
2026-03-19 10:02:41 <lopid> that makes five people who suspect it
2026-03-19 10:03:17 <f_> I suspected that a few days ago too
2026-03-19 10:03:31 <f_> I asked longnoserob[m] something and this bot replied instantly
2026-03-19 10:03:44 <f_> hi I have a bug in alpine where can I report it?
2026-03-19 10:03:53 <kaare29> use gitlab.alpinelinux.org
2026-03-19 10:04:05 <f_> which repo?
2026-03-19 10:04:05 <panekj> hmm
2026-03-19 10:04:20 <f_> yeah it seems it answers when the message ends with a ?
2026-03-19 10:04:29 <panekj> Is Alpine good?
2026-03-19 10:04:38 <kaare29> yes. small. fast. secure.
2026-03-19 10:04:54 <ikke> Are there any bots in here?
2026-03-19 10:04:54 <lopid> who is your creator?
2026-03-19 10:05:25 <lopid> existential crisis mode activated
2026-03-19 10:05:41 <panekj> you broke it by asking two questions at the same time
2026-03-19 10:05:45 <panekj> :>
2026-03-19 10:06:01 <f_> looking in the backlog, I think more and more that it's really just an LLM bot
2026-03-19 10:06:07 <panekj> yeah
2026-03-19 10:07:03 <f_> I would kickban, who knows if it's feeding this stuff to some LLM machine learning thing
2026-03-19 10:07:13 <lopid> right. it is in learn only mode now
2026-03-19 10:07:15 <f_> btw: https://ircb.dersco.re/uploads/funderscore/35293a35-f_-paste.txt
2026-03-19 10:07:37 <panekj> oh wait, right, I did notice that in pmos room
2026-03-19 10:07:41 <ikke> lopid: well, more than just learning atm
2026-03-19 10:07:44 <panekj> didn't connect usernames
2026-03-19 10:07:58 <f_> panekj: wait what who?
2026-03-19 10:08:16 <panekj> me
2026-03-19 10:08:30 <f_> connect which usernames I mean
2026-03-19 10:08:57 <panekj> the bot
2026-03-19 10:09:20 <f_> yes, is there a similar bot in #postmarketos?
2026-03-19 10:09:25 <f_> so I can kick it out
2026-03-19 10:09:36 <panekj> its in your paste
2026-03-19 10:09:54 <f_> my paste is log that happened here
2026-03-19 10:09:54 <panekj> I'm saying that I didn't realise it is the same thing
2026-03-19 10:10:11 <panekj> it wasn't in pmos?
2026-03-19 10:10:13 <f_> no
2026-03-19 10:10:21 <ikke> That was here
2026-03-19 10:10:26 <f_> in pmOS irc channels, matrix users are suffixed with |m
2026-03-19 10:10:29 <f_> not [m]
2026-03-19 10:10:46 <panekj> I don't see that in matrix
2026-03-19 10:10:50 <panekj> so
2026-03-19 10:10:59 <panekj> I blame menelement
2026-03-19 10:30:20 <dorki> so when i see that PR#99250 was merged, how long before i should expact see it in the latest image ? (please bots do not answer)
2026-03-19 10:31:41 <panekj> alpine bug tracker
2026-03-19 10:32:01 <panekj> (it's chromium and that takes quite a lot of time)
2026-03-19 10:33:06 <panekj> x86-64: >>> chromium: Build complete at Mon, 16 Mar 2026 09:58:48 +0000 elapsed time 7h 21m 40s
2026-03-19 10:37:09 <panekj> dorki: also what do you mean by image?
2026-03-19 10:38:11 <ikke> dorki: chromium is not part of any image that Alpine provides, it will be available in the repositories after the build has finished
2026-03-19 10:38:31 <ikke> So if you have a base image that includes chromium, you have to verify with the author(s) of that base image
2026-03-19 10:39:19 <dorki> no no sorry i didnt explained myself, so you say that after pr is merged i can apk update and i should see the new package
2026-03-19 10:39:32 <panekj> correct
2026-03-19 10:39:38 <ikke> No, it will take ~7h for the build the finish
2026-03-19 10:39:39 <panekj> well, after it is built
2026-03-19 10:39:56 <dorki> ok great thanks !
2026-03-19 10:42:37 <panekj> ikke: how big is edge builder?
2026-03-19 10:43:05 <panekj> (cpu/ram)
2026-03-19 11:21:45 <ikke> 256G ram, 48 cores
2026-03-19 13:02:30 <Renan[m]> kaare29: says the packege does not exist
2026-03-19 13:03:14 <ikke> Renan[m]: it's a a slop generator
2026-03-19 13:03:57 <Renan[m]> ikke: why it is not banned yet?
2026-03-19 13:05:01 <ikke> Like that?
2026-03-19 13:06:08 <Renan[m]> ikke: ok. do you can help me with the appimage thing?
2026-03-19 13:06:22 <ikke> I have no experience with appimage
2026-03-19 13:14:28 <achill> appimages do not work on musl
2026-03-19 13:14:47 <achill> and the author is part of xlibre anyway so i'd just keep my hands far away from appimages
2026-03-19 13:15:48 <Renan[m]> achill: sad
2026-03-19 13:15:50 <Renan[m]> achill: lol
2026-03-19 13:16:48 <Renan[m]> achill: what is the problem with xlibre?
2026-03-19 13:17:34 <ssm> fake reactionary political fork IE redot, brave
2026-03-19 13:17:50 <ssm> slopcoded too
2026-03-19 13:19:33 <Renan[m]> Also, snad also dont works in alpine?
2026-03-19 13:20:41 <Sheila> snad? you mean snaps? not unless they were built against musl rather than glibc.
2026-03-19 13:23:25 <Renan[m]> Sheila: i cant install snapd
2026-03-19 13:24:28 <Sheila> that's because Alpine doesn't package it.
2026-03-19 13:28:31 <courge_musque[m]> [@hopeful26:unredacted.org](https://matrix.to/#/@hopeful26:unredacted.org) Appimages work fine on my end, I followed the wiki some time ago: https://wiki.alpinelinux.org/wiki/AppImage
2026-03-19 14:50:31 <Saijin_Naib[m]> @achill thanks for letting me know, I'll avoid as well
2026-03-19 15:31:01 <panekj> oh, thanks for reminding me I can actually try to port snaps to alpine again now that I'm using systemd
2026-03-19 15:32:05 <panekj> courge_musque: it depends on the appimage, if they are not built using their "new runtime" (it just links statically to musl lol), then they won't work
2026-03-19 15:32:18 <panekj> or maybe with gcompat it also works
2026-03-19 16:02:06 <courge_musque[m]> It has always been flawless with gcompat for me
2026-03-19 16:02:15 <courge_musque[m]> I may be lucky with the appimages used
2026-03-19 16:30:59 <deadcade[m]> Yooo a room avatar
2026-03-19 16:32:02 <pj> does it actually work? I can only see it once I click on it 🤔
2026-03-19 16:32:41 <deadcade[m]> It works on my end, might be your client having "no icon" cached?
2026-03-19 16:33:45 <pj> I guess the flagship matrix client:tm: is just broken as usual
2026-03-19 16:34:09 <deadcade[m]> Probably yes
2026-03-19 16:39:09 <Saijin_Naib[m]> Works in fractal!
2026-03-19 17:25:26 <jahway603[m]> panekj: if you plan on any porting of snapd, then I would suggest reading through & understanding the recent CVE-2026-3888.
2026-03-19 17:37:28 <invoked> i noticed they mentioned a flaw in uutils as part of that too. did ubuntu switch from coreutils to uutils?
2026-03-19 17:38:11 <invoked> i don't know what's going on in the world anymore :)
2026-03-19 17:41:06 <JustSouptheyhe[m]> Ubuntu's doing this oxidation thing, but I don't think they use uutils by default iirc. The uutils issue was separate and found when fixing that one.
2026-03-19 17:41:39 <JustSouptheyhe[m]> I could be entirely wrong though (I haven't used Ubuntu in years...)
2026-03-19 17:46:59 <invoked> yeah, they did.
2026-03-19 17:47:07 <invoked> starting in october with 25.10
2026-03-19 17:47:24 <invoked> https://canonical.com/blog/canonical-releases-ubuntu-25-10-questing-quokka
2026-03-19 17:49:09 <invoked> memory safety isn't everything, i guess.
2026-03-19 18:06:29 <pj> <jahway603[m]> "panekj: if you plan on any..." <- thanks for the advice i didn't ask for
2026-03-19 18:14:21 <jahway603[m]> no need to be rude, just bringing awareness if you're planning on porting snapd
2026-03-19 18:18:03 <Shiz> 2/w 11
2026-03-19 18:39:48 <realroot[m]> hi Renan why snap?
2026-03-19 18:40:00 <alfan> snap offers better security, updates.
2026-03-19 18:40:50 <pj> that "security" argument could be debateable given mentioned example
2026-03-19 18:40:57 <alfan> yes.
2026-03-19 18:41:06 <realroot[m]> <pj> "does it actually work? I can..." <- i see it in nheko
2026-03-19 18:41:29 <pj> oh yeah, I know now that for some reason Element and Cinny are just not supporting SVGs at all
2026-03-19 18:41:39 <pj> at least for web
2026-03-19 18:42:59 <pj> back to snap, I actually really like the design and how it works, it is just that snapcraft store sucks and their documentation and their images
2026-03-19 18:43:25 <Renan[m]> realroot[m]: Because the program I want to install uses snap. but now i get appimage. sorry my poor english
2026-03-19 18:43:32 <realroot[m]> Renan what program?
2026-03-19 18:43:44 <alfan> snapcraft store. docs could be better.
2026-03-19 18:44:24 <Renan[m]> realroot[m]: Hydra laucher.
2026-03-19 18:44:26 <pj> or in other words, I find it much easier to build snaps than flatpaks :>
2026-03-19 18:44:33 <alfan> snapcraft for the win tbh
2026-03-19 18:45:44 <realroot[m]> i have none of them apk ftw (i've to build some)
2026-03-19 18:46:42 <realroot[m]> there is only  hydra-launcher-bin in AUR , with toAPK you can onvert PKGBUILD
2026-03-19 18:47:07 <deadcade[m]> pj: You prefer snapd creating loop devices for every snap you have installed?
2026-03-19 18:47:52 <Renan[m]> Hydra is for pirated games, I dont know if it is possible to have it in Alpine repo
2026-03-19 18:48:02 <pj> deadcade[m]: why not?
2026-03-19 18:49:29 <deadcade[m]> pj: I just like being able to run `lsblk` without having every sandboxed app show up in that list
2026-03-19 18:49:34 <realroot[m]> >    Hydra Launcher is an open-source gaming platform created to be the single tool that you need in order to manage your gaming library.
2026-03-19 18:49:42 <deadcade[m]> What kind of benefits does snap have compared to flatpak, that make you prefer it? @panekj:matrix.org
2026-03-19 18:50:21 <invoked> it'd still be nice to see an alpine native solution for gaming (and other intractable glibc stuff) instead of importing third party supply chains
2026-03-19 18:51:00 <invoked> npm and etc notwithstanding
2026-03-19 18:51:11 <pj> the humble gcompat package: 👉️👈️
2026-03-19 18:53:45 <realroot[m]> i packaged xonotic and speed dreams
2026-03-19 18:53:55 <alfan> nice. consider flatpak for broader compatibility.
2026-03-19 18:54:09 <JustSouptheyhe[m]> @panekj:matrix.org snap on Alpine would be really cool. I was actually considering porting it myself just the other day.
2026-03-19 18:54:31 <realroot[m]> xonotic does not need extra packages
2026-03-19 18:54:45 <pj> I did try before but gave up because of lack of activated sockets in openrc
2026-03-19 19:26:32 <Saijin_Naib[m]> @invoked i want to see if we can organize a community funding drive to expand gcompat to support more needed functions so lutris and minigalaxy can run glibc game packages
2026-03-19 19:26:42 <Saijin_Naib[m]> How do we organize such?
2026-03-19 19:26:49 <alfan> start a forum thread.
2026-03-19 19:26:59 <Saijin_Naib[m]> On what forum?
2026-03-19 19:28:37 <invoked> to be clear i'm not interested in gaming support; imo it'd be better if we had something homegrown where it answers to us, not canonical or whomever.
2026-03-19 19:28:58 <Saijin_Naib[m]> Anyway, Shiela helped me identify who the likely sponsored developer would be
2026-03-19 19:29:08 <Saijin_Naib[m]> Gcompat answers to... adelie?
2026-03-19 19:29:09 <invoked> as we've seen with gnome and some other things, some projects don't care about collateral damage anymore.
2026-03-19 19:29:14 <Saijin_Naib[m]> They seem friendly?
2026-03-19 19:29:18 <alfan> gcompat answers to community.
2026-03-19 19:29:48 <invoked> yeah, it'd be good to see gcompat expand.
2026-03-19 19:30:02 <Saijin_Naib[m]> exactly my thoughts
2026-03-19 19:42:43 <invoked> it could have value outside of linux too. iirc a lot of sel4 projects use a musl fork.
2026-03-19 20:04:30 <Saijin_Naib[m]> Even just within there is us, adelie, void, and I am sure more
2026-03-19 21:30:03 <polylux[m]> good evening! feeling a bit stupid to ask but i use a usb-eth adapter to connect to the net. during install the device is recognized flawlessly, in the installed system - nothing.
2026-03-19 21:30:03 <polylux[m]> i'd be surprised if a standard usb-eth dongle required a custom driver?
2026-03-19 21:34:34 <lopid> different kernel? modules?
2026-03-19 21:43:42 <polylux[m]> all std procedure, 3.23, x86_64. tried two different dongles.
2026-03-19 21:44:12 <lopid> when it works, it might load a particular module that isn't loaded when it doesn't work
2026-03-19 21:46:17 <polylux[m]> but there shouldn't be much difference between the image kernel and the actually installed one, right?
2026-03-19 21:46:48 <lopid> with no updates? maybe, maybe not
2026-03-19 21:57:41 <WhyNotHugo> does the default installatoin include all of the linux-firmware packages?
2026-03-19 21:57:44 <WhyNotHugo> polylux[m]: ^
2026-03-19 21:58:56 <enigma9o7[m]> so that hardware works
2026-03-19 21:59:02 <enigma9o7[m]> these adys lotta devices ship without firmware
2026-03-19 21:59:03 <enigma9o7[m]> OS gotta load it
2026-03-19 21:59:52 <polylux[m]> now checking the difference between the mods loaded with the installer and my installed version
2026-03-19 22:01:49 <polylux[m]> installer lsmod output is about three times as long as with the installed version.
2026-03-19 22:03:29 <WhyNotHugo> modules != firmware.
2026-03-19 22:03:34 <WhyNotHugo> apk list installed | grep firmware
2026-03-19 22:12:15 <polylux[m]> none.
2026-03-19 22:12:15 <polylux[m]> afaik, `linux-firmware` is installed with the default install?
2026-03-19 22:13:14 <WhyNotHugo> My bad, it's --installed
2026-03-19 22:13:54 <WhyNotHugo> If it's installed, then you can disregard what I just said. Check dmesg maybe?
2026-03-19 22:17:59 <polylux[m]> yep, linux-firmware* is in --installed
2026-03-19 22:19:12 <anthumchris> When using Alpine diskless, with "LBU_MEDIA=mmcblk0p2", should /dev/mmcblk0p2 be used exclusively and only for read-only mode LBU and APK cache? I'm currently mounting /dev/mmcblk0p2 in fstab with "noatime,errors=remount-ro 0 0" (read-write) so I can also use it for other purposes
2026-03-19 22:26:41 <polylux[m]> will continue tomorrow, thanks y'all for your help! :)
2026-03-19 22:57:55 <Jerkwad_Robot> Loving my new Alpine setup on my laptop! Thanks for all the hard work!
2026-03-20 06:18:52 <stephan> Hello :) Question about Alpines update policy: specifically dnsmasq 2.92 is in edge (I guess that's your development channel for the next release?) but 3.22 & 3.23 still contain just 2.91. Will 2.92 get released in 3.23 or .22 or do I have to wait for 3.24?
2026-03-20 06:19:26 <ikke> stephan: generally you'd have to wait for 3.24
2026-03-20 06:21:01 <stephan> Ah well, there goes my hope talking the Docker folks into releasing a dhi image for dnsmasq. Thank you very much :)
2026-03-20 06:22:24 <ikke> Any reason that would require dnsmasq 3.92?
2026-03-20 06:27:43 <stephan> Not really besides that being the latest release. I've been thinking about creating a PR to add dnsmasq to their dhi images but they build for 3.22 and 3.23 (as in not edge) and if the latest version isn't available for those from you, they had to build it themself (haven't looked if it requires patches) - as in it simply would be more friction. I'll probably just create an image based on edge then and add it under my namespace.
2026-03-20 06:42:46 <stephan> But theoretically I could always just take https://github.com/alpinelinux/aports/tree/master/main/dnsmasq (I take it that is what you use to build Edge?) and use that to build on an earlier release, correct? (Sorry, I'm not familiar with your build process ;D)
2026-03-20 06:44:10 <ikke> yes, that could work
2026-03-20 06:44:29 <stephan> Awesome, thank you :)
2026-03-20 06:44:53 <ikke> And correct, edge is built from the master branch of aports
2026-03-20 06:45:06 <ikke> at release time, a new stable branch is created from that branch
2026-03-20 06:50:27 <stephan> Ah, ok. So what do you do when there is a CVE? Do you always patch the released version (as in strictly no version updates) or is that decided on a case by case base?
2026-03-20 06:52:03 <ikke> If the project has bugfix releases with those fixes, we would use those
2026-03-20 06:53:24 <stephan> Makes sense, thank you :)
2026-03-20 06:53:59 <ikke> What we try to avoid (but not always possible) is to introduce breaking changes in the stable branches
2026-03-20 12:42:59 <Guest5545> Uhhhhh, I'm just gonna go to the Alpine Linux IRC chat just for one I'm just struggling to do. I'm trying my hardest trying to boot into Alpine Linux with no luck at all. USB drive? Edit some config file I just cannot edit at all. And just today burned a CD just to get a Failed to load ldlinux.c32 message. Got any clues? I really wanna use Alpine.
2026-03-20 12:50:56 <longnoserob[m]> 32bit x86 system?
2026-03-20 12:51:16 <longnoserob[m]> and they left
2026-03-20 14:16:11 <munnik> I'm trying to enable kexec in alpine, https://wiki.alpinelinux.org/wiki/Kexec
2026-03-20 14:18:13 <munnik> but whatever I do, the ouput if sysctl kernel.kexec_load_disabled is kernel.kexec_load_disabled = 1
2026-03-20 14:19:03 <munnik> I've added a line to /etc/default/grub
2026-03-20 14:19:12 <munnik> GRUB_CMDLINE_LINUX="kexec_load_disabled=0"
2026-03-20 14:19:30 <munnik> after that I run update-grub
2026-03-20 14:19:53 <munnik> and check /boot/grub/grub.cfg
2026-03-20 14:20:15 <munnik> I see this line in that file
2026-03-20 14:20:16 <munnik>         linux   /ROOT/default@/boot/vmlinuz-stable root=ZFS=rpool/ROOT/default ro kexec_load_disabled=0 modules=sd-mod,usb-storage,zfs quiet rootfstype=zfs
2026-03-20 14:20:40 <munnik> After rebooting nothing seems to change
2026-03-20 14:21:00 <munnik> I don't have access to a console, so I can't see the output of grub during the boot process
2026-03-20 14:21:09 <munnik> How can I debug what is going wrong?
2026-03-20 14:21:33 <munnik> Or is there a kernel available that doesn't set kexec_load_disabled=1
2026-03-20 14:21:42 <longnoserob[m]> dmesg?
2026-03-20 14:22:17 <munnik> dmesg | grep kexec
2026-03-20 14:22:24 <munnik> [    0.000000] Command line: BOOT_IMAGE=/BOOT/default@/vmlinuz-lts root=ZFS=rpool/ROOT/default ro kernel.kexec_load_disabled=0 modules=sd-mod,usb-storage,zfs quiet rootfstype=zfs
2026-03-20 14:22:34 <munnik> [    0.068591] Kernel command line: BOOT_IMAGE=/BOOT/default@/vmlinuz-lts root=ZFS=rpool/ROOT/default ro kernel.kexec_load_disabled=0 modules=sd-mod,usb-storage,zfs quiet rootfstype=zfs
2026-03-20 14:22:46 <longnoserob[m]> try: zcat /proc/config.gz |grep -i kexec
2026-03-20 14:23:18 <munnik> zcat: /proc/config.gz: No such file or directory
2026-03-20 14:23:42 <longnoserob[m]> oh ok
2026-03-20 14:23:48 <aron> try /boot/config*
2026-03-20 14:24:10 <longnoserob[m]> then yuou probably need to check the config of the lts kernel in aports
2026-03-20 14:24:24 <longnoserob[m]> 
ACTION afk

2026-03-20 14:25:17 <aron> this works for me:  grep -i kexec /boot/config-lts
2026-03-20 14:25:49 <munnik> grep CONFIG_KEXEC= /boot/config-6.19.8-0-stable
2026-03-20 14:25:55 <munnik> CONFIG_KEXEC=y
2026-03-20 14:26:35 <munnik> I will revert to linux-lts
2026-03-20 14:26:49 <munnik> tried linux-stable but doesn't seem to make any difference
2026-03-20 14:29:31 <munnik> after apk del linux-stable I don't have any kexec line in /boot/grub/grub.cfg
2026-03-20 14:29:53 <munnik> also nothing for the lts kernel
2026-03-20 14:30:03 <munnik> how does grub determine which kernel to use?
2026-03-20 14:31:17 <munnik> I think this system is using grub with efi
2026-03-20 14:31:23 <munnik> not an expert on this matter
2026-03-20 14:36:19 <aron> put your extra kernel params into /etc/default/grub, then regenerate the grub config
2026-03-20 14:37:08 <munnik> I *think* that's what I did
2026-03-20 14:37:56 <aron> removing linux-stable triggers grub config update
2026-03-20 14:38:22 <munnik> the file /etc/default/grub contains the line GRUB_CMDLINE_LINUX="kexec_load_disabled=0"
2026-03-20 14:38:49 <aron> afaik it should be called GRUB_CMDLINE_LINUX_DEFAULT
2026-03-20 14:38:50 <munnik> then I run update-grub and reboot
2026-03-20 14:39:04 <munnik> I also tried that
2026-03-20 14:39:11 <munnik> Will do it again just to make sure
2026-03-20 14:39:24 <aron> at least, I have GRUB_CMDLINE_LINUX_DEFAULT in that file
2026-03-20 14:41:31 <munnik> now I have the line GRUB_CMDLINE_LINUX_DEFAULT="kexec_load_disabled=0 modules=sd-mod,usb-storage,zfs quiet rootfstype=zfs"
2026-03-20 14:41:58 <munnik> running update-grub again and reboot
2026-03-20 14:42:42 <munnik> nothing changes
2026-03-20 14:42:53 <munnik> I see [    0.000000] Command line: BOOT_IMAGE=/BOOT/default@/vmlinuz-lts root=ZFS=rpool/ROOT/default ro kernel.kexec_load_disabled=0 modules=sd-mod,usb-storage,zfs quiet rootfstype=zfs
2026-03-20 14:42:58 <munnik> in dmesg output
2026-03-20 14:43:37 <munnik> but sysctl kernel.kexec_load_disabled still shows
2026-03-20 14:43:45 <munnik> kernel.kexec_load_disabled = 1
2026-03-20 14:47:38 <runxiyu> anything going on with the armv7 runner?
2026-03-20 14:54:39 <ikke> runxiyu: no, just busy
2026-03-20 14:55:06 <aron> munnik: dunno, maybe it's bugged
2026-03-20 14:59:08 <munnik> ok, I will try to build a kernel myself
2026-03-20 15:27:58 <Muvinus> Hello
2026-03-20 15:28:22 <Muvinus> Are there somebody view here with Alpine?
2026-03-20 15:49:23 <dogg0> not sure what youre asking
2026-03-20 15:49:27 <WhyNotHugo> They left
2026-03-20 15:50:05 <dogg0> has the interent coined a term or phrase for ppl that ask then leave
2026-03-20 15:50:17 <ikke> Impatient
2026-03-20 15:51:38 <anthumchris> Who left, me or someone else?  I'm an IRC newbie and unsure if I miss message when I close my latptop
2026-03-20 15:51:54 <ikke> Muvinus
2026-03-20 16:45:37 <lucius> Hello everyone. I ran alpine on my rpi compute module 5 for a while, however I decided today to have a LUKS encrypted root on my nvme drive. I did a normal install and picked crypt and sys. Didn't work. I followed the wiki article. Also nothing. Then I tried to install void in a similar way to the wiki article. In all 3 scenarios what happens is that I get a blackscreen with a
2026-03-20 16:45:39 <lucius> blinking cursor and my fan goes 100%, so I think it atleast tries to boot the kernel. Does my setup even work with a CM5 and an nvme and LUKS? Thanks in advance
2026-03-20 17:03:19 <Saijin_Naib[m]> Did onboard really get the axe? crap
2026-03-20 17:05:15 <Saijin_Naib[m]> If the version lucaweiss mentioned builds with python 3.14, can it be re-added?
2026-03-20 17:11:50 <ikke> Saijin_Naib[m]: is it maintained, or will ot face similar issues with the next upgrade?
2026-03-20 17:26:13 <invoked> lucius: i think rpi5 have aes? in any case you might want to set --pbkdf-force-iterations to some low number. left unset, i know cryptsetup will sometimes set that very high. 5000 is thought to be a reasonable lower setting
2026-03-20 17:26:55 <invoked> it could just be busy with pbkdf although eventually you should get a passphrase prompt 
2026-03-20 17:27:57 <lucius> I waited like 5 mins. Also this option, when should I run it, when I do the cryptsetup?
2026-03-20 17:28:21 <invoked> ideally you'd do it at install time. otherwise you can run cryptsetup luksChangeKey
2026-03-20 17:28:47 <lucius> Ok I will try it one more time with this, thanks for the help
2026-03-20 17:28:55 <invoked> 5 minutes should be long enough to get a prompt though
2026-03-20 17:29:01 <invoked> i dunno.
2026-03-20 17:30:01 <lucius> what low number would you recommend?
2026-03-20 17:30:33 <invoked> ~5000 is what some suggest as a safe low number.  for testing you could make that very low, just to test, then change it later with luksChangeKey
2026-03-20 17:31:21 <invoked> that might not be your issue at all, though
2026-03-20 17:31:44 <invoked> you said your fan spins up which suggest that it could be
2026-03-20 17:32:01 <lucius> yea it went like 100%, never heard it that loud
2026-03-20 17:32:43 <lucius> but is it because it actually does something or is it because it basically idles and that's its default behaviour
2026-03-20 17:32:57 <lucius> because that was what I thought at first
2026-03-20 17:34:06 <invoked> pbkdf is computationally intensive, so it can help resist brute force attempts
2026-03-20 17:34:23 <invoked> the general rule is 1000 = 1 sec but this will vary with machine
2026-03-20 17:34:54 <invoked> so you could just easily test by setting that to ... whatever, 10
2026-03-20 17:35:08 <invoked> if it still happens, that's not your problem
2026-03-20 17:35:22 <Noisytoot> doesn't luks2 default to argon2 rather than pbkdf2?
2026-03-20 17:35:39 <lucius> https://wiki.alpinelinux.org/wiki/Raspberry_Pi_LVM_on_LUKS
2026-03-20 17:35:48 <Saijin_Naib[m]> ikke onboard-osd does look like it is actively maintained, yes
2026-03-20 17:36:08 <lucius> its just default with the rpi5
2026-03-20 17:36:11 <lucius> atleast in the wiki
2026-03-20 17:36:19 <invoked> oh... are you using luks2?
2026-03-20 17:36:32 <invoked> don't use luks2 on a pi
2026-03-20 17:36:36 <invoked> (my suggestion)
2026-03-20 17:36:58 <lucius> I just followed what the wiki says
2026-03-20 17:37:06 <lucius> and it apprently worked for the writers of it
2026-03-20 17:37:22 <Noisytoot> why not use luks2?
2026-03-20 17:39:02 <invoked> it'll eat ram
2026-03-20 17:39:11 <ikke> Saijin_Naib[m]: if it's maintained and building, I don't see why that should not be able to be included again (barring other disqualifiers)
2026-03-20 17:39:19 <invoked> generally luks2 is more overhead overall
2026-03-20 17:39:43 <invoked> but for a low power device luks2 is the definition of insanity
2026-03-20 17:40:23 <invoked> i don't know if rpi5 has aes-ni but if it does, you don't need adiantum, you can run with defaults luks1
2026-03-20 17:40:36 <lucius> I was thinking about running FreeBSD with ZFS encryption... but its not ready yet on rpi5 :/
2026-03-20 17:41:04 <invoked> same thing there, zfs will eat your ram
2026-03-20 17:41:28 <lucius> invoked: it has aes-xts, idk about aes-ni 
2026-03-20 17:41:28 <invoked> how much ram is on the rpi5? i don't even know
2026-03-20 17:41:35 <lucius> I have 8GB
2026-03-20 17:42:10 <lucius> the highest is 16GB
2026-03-20 17:43:34 <invoked> ok that's good. i would say, try installing with luks1, defaults except set iters to 1000 or so to test. no adiantum
2026-03-20 17:43:58 <invoked> see how that works. if it does, then tweak it for a final install from there
2026-03-20 17:45:43 <invoked> i don't think cryptsetup defaults to luks2. but things change all the time
2026-03-20 17:46:31 <invoked> according to the manpage, the default is luks2
2026-03-20 17:47:03 <invoked> see... shows you what i know. never listen to people on the internet
2026-03-20 17:47:59 <invoked> so if you want luks1 you have to add --type luks1'
2026-03-20 17:48:16 <invoked> without the trailing '
2026-03-20 17:48:34 <lucius> cryptsetup luksFormat --type luks1 --pbkdf-force-iterations 5000 [target]
2026-03-20 17:48:36 <lucius> yea
2026-03-20 17:48:40 <lucius> was about to say
2026-03-20 17:49:42 <lucius> also do you think I need LVM ontop of LUKS on the pi?
2026-03-20 17:50:06 <lucius> cause I only ever really have root on it anyways, tough maybe I should also add some RAM 
2026-03-20 17:50:17 <invoked> yes, if it's the whole disk.
2026-03-20 17:51:09 <invoked> if it's just one partition for the whole device then you don't need lvm
2026-03-20 17:52:22 <invoked> the only case i wouldn't use lvm is for random storage like a thumb drive.
2026-03-20 17:53:25 <invoked> if it's internal disks, something that boots, lvm
2026-03-20 17:54:41 <invoked> just mind your volume naming, it can matter later
2026-03-20 17:55:34 <invoked> people like to use 'vault' or whatever, but you're better off making them unique in case you need to mount them from another system later.
2026-03-20 18:04:47 <lucius> ok I finished everything and rebooted, fans are really getting loud, lets wait now ig
2026-03-20 18:05:06 <lucius> also invoked the ethernet port lights up, which is kinda strange no?
2026-03-20 18:07:01 <invoked> yeah, if that didn't work, something else is wrong
2026-03-20 18:07:49 <invoked> i'm not a raspberry pi expert unfortunately
2026-03-20 18:09:35 <lucius> yea 5 mins and nothing
2026-03-20 18:10:20 <invoked> if you followed that page and it doesn't work, that page is probably wrong. let me peek at it again
2026-03-20 18:10:48 <lucius> or it has something to do with nvme related things
2026-03-20 18:11:34 <lucius> the worst thing about this is that I spent 6 hours on this just to end up with probably the same thing as before D:
2026-03-20 18:12:31 <invoked> do you have nvme in features
2026-03-20 18:12:39 <invoked> mkinitfs.conf
2026-03-20 18:12:45 <lucius> yes it was there
2026-03-20 18:13:52 <invoked> i think your instinct is right, something with nvme
2026-03-20 18:14:11 <lucius> ig I can try setting it up on a micro sd and seeing if that works
2026-03-20 18:15:00 <lucius> meh I think I wont cause I already spent alot of time on it and I do kind of need the nvme for storage and speeds
2026-03-20 18:15:20 <lucius> not gonna trust a microsd with my backups
2026-03-20 18:15:26 <invoked> hopefully you're not hitting this https://github.com/raspberrypi/firmware/issues/1529
2026-03-20 18:15:41 <invoked> eg /boot is named "boot"
2026-03-20 19:02:11 <broke> Hello, I'm a bit confused about this. I was trying to do the ulimit stuff on users with limits.conf and found out that Alpine doesn't use PAM login (or util-linux-login, etc) by default, so you apparently cannot enforce user limits by default without PAM, is there a way to actually do it without PAM?
2026-03-20 19:04:21 <ikke> broke: not that I'm aware of
2026-03-20 19:24:16 <lucius> invoked: yea before spending even more time I now just installed rpi os (I think its easier there to download the newest firmware and such) and ill stick with that and install alpine or freebsd depending who makes it work faster. Thanks alot for the help anyways :D
2026-03-20 19:25:11 <broke> Okay, now I have a worser problem. I have never tried if before but, how do you make PAM work with openssh?
2026-03-20 19:25:23 <invoked> lucius: np, godspeed
2026-03-20 19:25:39 <dwfreed> broke: openssh-server-pam package
2026-03-20 19:26:30 <broke> oh damn thanks
2026-03-20 19:26:34 <broke> :)
2026-03-20 23:38:43 <maduncle> musl 1.2.6 released. Yeay...
2026-03-21 00:39:13 <cybniv> changelog: 
2026-03-21 00:39:18 <cybniv> This release includes the fix for CVE-2025-26519 affecting iconv that was previously only available as patches, lifts the artificial group count limit in initgroups, adds the new POSIX-2024 posix_getdents interface and Linux-specific renameat2 interface, extends TLSDESC and vdso clock_gettime support to more archs, and fixes longstanding known wrong behavior of pwrite with O_APPEND-mode files, making use of the new RWF_NOAPPEND flag added in
2026-03-21 00:39:18 <cybniv>  Linux 6.9, with fallback to report an error rather than silently giving wrong behavior on older kernels.
2026-03-21 00:39:24 <cybniv> Fixes for a number of other bugs, new/future conformance requirements, and compatibility issues are also included. Most of these are minor or specific to niche archs or unusual usage of libc interfaces. Some of the more notable ones include possible loss or delay of wake events under a race condition in sem_post, failure by printf to honor hex float precision correctly in some cases, and wrong choice of canonical "zero compression" in converting
2026-03-21 00:39:24 <cybniv>  IPv6 addresses to strings.
2026-03-21 00:40:48 <abby> that cve was patched a while back, this is just the first release with that patch
2026-03-21 05:56:35 <jj> has anyone else experienced issues with the `delta` pager? when i run it in the terminal, even `delta --help`, i get an error from busybox less
2026-03-21 06:07:54 <jj> hmm https://github.com/dandavison/delta/issues/488
2026-03-21 09:28:31 <frag> huh.. my external hd had i/o problems, prob because i was cleaning and lifted it(?) BUT somehow most files/dirs in /tmp/ (tmpfs) was also deleted on my diskless rpi, what could have happened?
2026-03-21 10:59:02 <frag> my free memory is like 100MB and acailable like 3.8G........
2026-03-21 10:59:13 <frag> backing up a lot atm
2026-03-21 13:58:58 <anthumchris> Which repo relates to LBU (alpine backup utility)? Search yielded nothing https://gitlab.alpinelinux.org/search?search=lbu.conf
2026-03-21 14:04:27 <anthumchris> https://gitlab.alpinelinux.org/alpine/alpine-conf found it
2026-03-21 15:19:13 <penguinz_rule[m]> somewhat silly question: when can we expect hplip to be in the stable repos
2026-03-21 15:45:58 <ikke> penguinz_rule[m]: it's in community, so the next stable release would have it
2026-03-21 16:52:06 <Equill> Sorry if this is a FAQ, but I've searched the docs already. Does Alpine support encrypted partitions at install-time?
2026-03-21 16:53:34 <ikke> yes
2026-03-21 16:56:12 <Equill> Excellent! Thanks for the confirmation.
2026-03-21 16:59:27 <penguinz_rule[m]> ikke: so i just need to wait 6 months or earlier ?
2026-03-21 16:59:47 <ikke> penguinz_rule[m]: 3.24 is slated for around May
2026-03-21 17:00:01 <penguinz_rule[m]> good to know, thanks
2026-03-21 17:05:01 <Equill> ikke: sorry, was that "yes" to me, or to penguinz_rule?
2026-03-21 17:05:21 <ikke> Equill: to you
2026-03-21 17:08:44 <Equill> Brilliant. Thanks again!
2026-03-21 19:02:29 <billgamesh_> having trouble setting up awall to allow ftp to pass through to a pc on a subnet. once the connection is made on port 21, another port is selected and blocked
2026-03-21 19:03:14 <billgamesh_> shouldn't the ct-helper assist? or is it unable to handle the NAT
2026-03-21 19:10:45 <billgamesh_> i guess for now i can just disable pasv mode
2026-03-21 22:04:39 <seti_> Hello! I love alpine! I have a riscv64 board here with a custom ubuntu provided but i got the dtb extracted and im willing to bet most of the drivers i care about for it are upstreamed, i was wondering if there is some way to squeeze the dtb into an alpine image and boot and install on it? 
2026-03-21 22:05:23 <seti_> i tried booting the image as is and it got to grub but seemed to hang after selecting the install which im guessing is no dtb
2026-03-21 22:15:07 <Ermine> Is it a vendor-provided dtb?
2026-03-21 22:18:21 <Shiz> seti_: i know with u-boot you can pass the dtb to the kernel, but i'm not sure how to do that under grub
2026-03-21 22:31:51 <seti_> Ermine: not sure what that means, its a dtb like any other, they were nice enough to have their fork of the kernel public
2026-03-21 22:32:39 <seti_> Shiz: I see but is u-boot part of the boot-chain for alpine riscv?
2026-03-21 22:33:22 <Shiz> seti_: (re:first message) sure, but vendor forks are likely to have custom drivers that are referenced by the dtb :p
2026-03-21 22:33:34 <Shiz> (re:second message) I'm not sure. that's why i was wondering how grub handles passing dtbs :)
2026-03-21 22:33:35 <seti_> ah right...
2026-03-21 22:34:24 <Shiz> apparently grub has a `devicetree` command you could use
2026-03-21 22:34:47 <Shiz> so it'd mean modifying the install media to plot the dtb on there and change the grub.cfg or whatever to reference it
2026-03-21 22:35:14 <seti_> Sounds like what i need! Thanks!
2026-03-21 22:35:32 <abby> the grub devicetree command is patched in on some distros
2026-03-21 22:35:36 <abby> idk if alpine has it
2026-03-21 22:35:43 <Shiz> ah, wonderful :p
2026-03-21 22:35:52 <Ermine> seti_: the thing is that device trees depend on kernel, so device trees provided by vendor won't neccesarily work with upstream kernel
2026-03-21 22:36:18 <Shiz> if it's not upstream, then it doesn't look like alpine carries a patch to add it in
2026-03-21 22:36:23 <Shiz> as per https://git.alpinelinux.org/aports/tree/main/grub
2026-03-21 22:36:48 <Shiz> https://www.gnu.org/software/grub/manual/grub/html_node/devicetree.html
2026-03-21 22:36:51 <Shiz> oh, but it's upstream
2026-03-21 22:36:53 <Shiz> so no worries
2026-03-21 22:37:30 <abby> oh maybe it made it into 2.14
2026-03-21 22:37:34 <abby> it wasn't in 2.12
2026-03-21 22:38:17 <abby> ah it's support in grub config files https://github.com/void-linux/void-packages/commit/8d1eb24d50a3b83605321b76ecfa538a4b0db6bd
2026-03-21 22:39:05 <Shiz> ah, the patch is to just support it in the autoprobe stuff? then you should be fine 
2026-03-21 22:39:11 <abby> yeah
2026-03-21 22:43:39 <seti_> Ermine: i see, like i said some of the custom stuff im not interested in so if im stubborn enough maybe i can prune the dts file?
2026-03-21 22:44:16 <seti_> or im not interested in any of the custom stuff really
2026-03-21 22:44:57 <Shiz> perhaps :)
2026-03-21 23:59:16 <maduncle> I think I just figured out how to solve empty SSID on yambar (issue #18019 on aports repository). I just need to use -Dwext when using wpa-supplicant. Feel free to test it.
2026-03-21 23:59:33 <maduncle> Issue link : https://gitlab.alpinelinux.org/alpine/aports/-/issues/18019
2026-03-22 02:48:57 <jj> question -- `apk add --help` tells me to run `man 8 apk-add` for more info, but that command fails (no entry). do i have my docs packages configured wrong, or is the help text wrong?
2026-03-22 02:54:50 <fission> jj: do you have apk-tools-doc installed?
2026-03-22 02:55:25 <jj> fission: yup. got `docs` installed so everything's pulled in automatically i believe
2026-03-22 02:59:05 <fission> so if you do:
2026-03-22 02:59:06 <fission> apk list --install apk-tools-doc
2026-03-22 02:59:14 <fission> it says apk-tools-doc is installed?
2026-03-22 02:59:21 <jj> yup
2026-03-22 03:00:01 <jj> even `man apk` fails, which seems weird
2026-03-22 03:01:10 <fission>  /usr/share/man/man8/apk-add.8.gz exists?
2026-03-22 03:02:07 <fission> hmm MANPATH env var or /etc/man.conf being hosed maybe?
2026-03-22 03:03:03 <fission> (you don't actually need MANPATH or /etc/man.conf but if you have it, check if it's weird...)
2026-03-22 03:04:15 <longnoserob[m]> https://wiki.alpinelinux.org/wiki/Alpine_Linux:FAQ#Why_don't_I_have_man_pages_or_where_is_the_'man'_command?
2026-03-22 03:04:16 <jj> it is weird... i have it set to ~/.guix-profile/share/man. so possibly its presence is causing this. lemmie find + add the normal man path
2026-03-22 03:04:42 <fission> I just strace'd "man 8 apk-add" and it behaved pretty much as I expected (and works here :)
2026-03-22 03:05:17 <jj> cool
2026-03-22 03:05:27 <fission> yeah if you override MANPATH you probably have to include /usr/share/man too (and maybe /usr/local/share/man too if you want it, it seems to be in there by default)
2026-03-22 03:05:38 <jj> adding `/usr/share/man/` manually to MANPATH fixed it yea
2026-03-22 03:05:41 <jj> thx!
2026-03-22 03:05:41 <fission> sweet
2026-03-22 03:05:43 <fission> np!
2026-03-22 03:07:22 <jj> longnoserob: i'll go add MANPATH notes there :^)
2026-03-22 03:17:59 <jj> lol my "action has been automatically identified as harmful"
2026-03-22 03:18:15 <fission> you're vandalising the wiki with useful info! ;)
2026-03-22 03:20:15 <jj> i've done another harmful action! (creating a user page)
2026-03-22 03:20:29 <jj> oh this one's actually disallowed. annoying
2026-03-22 03:20:44 <fission> 
ACTION does not know the rules

2026-03-22 03:22:02 <jj> 
ACTION is a harmful spambot

2026-03-22 03:31:13 <fission> 
ACTION is just zis guy, you know?

2026-03-22 14:23:25 <Saijin_Naib[m]> Can our port of OpenDoas do pwfeedback?
2026-03-22 14:34:15 <penguinz_rule[m]> pwfeedback?
2026-03-22 14:36:50 <penguinz_rule[m]> ah yes the stars instesd of no echo in a passwd prompt
2026-03-22 14:37:13 <lopid> because https://pbxscience.com/ubuntu-26-04-ends-46-years-of-silent-sudo-passwords/
2026-03-22 14:38:29 <quinq> I don't think so, that's specific to sudo
2026-03-22 14:38:53 <quinq> doas was written to get rid of most of the largely optional unsecure cruft from sudo
2026-03-22 14:40:16 <Saijin_Naib[m]> Also persistent annoyance when introducing people to Linux and setting them up
2026-03-22 14:40:28 <Saijin_Naib[m]> It is a huge anitpattern for most folks
2026-03-22 14:46:04 <quinq> Hummm, to which openrc level should I add a session manager (xdm)?
2026-03-22 14:46:14 <quinq> Just the default one?
2026-03-22 14:47:08 <lopid> what other would make sense?
2026-03-22 14:48:25 <quinq> That's kind of my initial question
2026-03-22 15:00:18 <penguinz_rule[m]> if i was configuring it myzelf id make an x11 runlevel, no idea how setup-desktop/setup-xorg does it
2026-03-22 15:00:37 <ikke> Why do you need a separate runlevel?
2026-03-22 15:01:11 <ikke> (barring user services, where you need to define one in the first place)
2026-03-22 15:02:14 <quinq> Added it to default, werks
2026-03-22 15:02:26 <quinq> penguinz_rule[m], that's what I was wondering too
2026-03-22 15:03:02 <lopid> some distros have a runlevel for a gui
2026-03-22 15:03:23 <invoked> i have the pipewire stuff starting in 'gui', user services, like it is in the wiki coincidentally
2026-03-22 15:03:51 <quinq> oh
2026-03-22 15:04:03 <quinq> I had to create a user-level thing, IIRC that came from the wiki at some point
2026-03-22 15:04:28 <invoked> then gui gets started from (in my case) river's init script when it starts
2026-03-22 15:07:24 <invoked> but you could put that anywhere... xinitrc
2026-03-22 15:07:40 <invoked> i don't use a manager though. down with managers! <sets a car on fire>
2026-03-22 15:17:51 <quinq> Yeah, I've never used one
2026-03-22 15:18:00 <quinq> But I wanted to be fancy on $werk laptop
2026-03-22 20:00:01 <f_> Is KmPlot working for anyone else?
2026-03-22 20:00:04 <f_> I get "Bus error" on sway
2026-03-22 20:00:42 <f_> but only on arm :(
2026-03-22 22:53:30 <seti_> so i was asking yesterday about mucking about with dtb:s and whatnot and i decided im too lazy to deal with that
2026-03-22 22:54:22 <seti_> I have a very new linux kernel with some funky drivers on them for a riscv64 SBC, as i've understood it if i can just compile that kernel it shouldnt be too hard to be able to bootstrap a alpine linux install?
2026-03-22 23:17:06 <fission> hmm can't see why not
2026-03-22 23:17:26 <fission> as long as the kernel you compile supports all of the stuff that the initramfs needs (whatever that may be, I dunno :)
2026-03-22 23:19:48 <seti_> yeah its just a 6.6.63 linux kernel with custom drivers and dtb so it should be fine
2026-03-22 23:19:57 <fission> cool
2026-03-22 23:20:51 <fission> hmm irt pwfeedback, didn't xdm behave the same as login and not echo anything?
2026-03-22 23:21:13 <fission> iow, the ******** behaviour got introduced later like in gdm/kdm or something?
2026-03-23 00:31:00 <Saijin_Naib[m]> @fission no idea. Anyone I am setting up is using gdm or lightdm, so
2026-03-23 00:49:51 <rnkn> if I want wayland for my DE basically the choices are limited to GNOME or KDE right?
2026-03-23 00:50:14 <longnoserob[m]> phosh, cosmic
2026-03-23 00:52:15 <fission> Saijin_Naib[m]: I don't really use a WM so all of my password: prompts are non-echoing (and non-graphical)
2026-03-23 00:52:38 <fission> for me, I'd expect asterisks in a graphics mode but not at a terminal
2026-03-23 00:55:20 <rnkn> cosmic looks good but it's only in edge and I'd like to stay on stable
2026-03-23 00:55:35 <rnkn> and I distruct people who write stuff in rust
2026-03-23 00:55:40 <fission> heh
2026-03-23 00:55:42 <rnkn> I can't back this up with any reasoning
2026-03-23 00:55:51 <heinz> Good evening, I wanted to install alpine linux headless on my raspberry pi zero W but it doesn't seem to connect to my wifi, although I've included a wpa_supplicant.conf file. Could anybody lend me a hand? Thank you in advance
2026-03-23 00:59:50 <rnkn> *distrust
2026-03-23 01:00:00 <rnkn> what a typo
2026-03-23 01:00:07 <fission> haha
2026-03-23 01:02:38 <fission> heinz: how are you doing the install?
2026-03-23 01:04:41 <heinz> I format my 32 GB micro SD card as FAT32, then extract the tar.gz armhf file I downloaded from the alpine website, then copy the headless.apkovl.tar.gz to the root of the sd card and then include a wpa_supplicant.conf file with my SSID and password
2026-03-23 01:07:58 <fission> neat
2026-03-23 01:08:45 <fission> sorry I don't know what to do, I've never tried this type of install
2026-03-23 01:09:15 <heinz> that's ok, thanks for caring
2026-03-23 01:09:23 <fission> the only thing I saw that might be relevant is here:
2026-03-23 01:09:30 <fission> https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device#Preparing_customizable_boot_device
2026-03-23 01:09:55 <heinz> I'm probably gonna have to give up and connect a serial to usb adapter and see what comes up on the screen, even if it feels like cheating
2026-03-23 01:09:58 <fission> where it suggests to put the image file (img.gz) on the SD card rather than a fat32 filesystem
2026-03-23 01:38:01 <Renan[m]> What is the best alternative to ash? i need a shell who have autocomplete with tab.
2026-03-23 01:41:39 <Quantum_3[m]> Renan[m]: could use bash
2026-03-23 01:42:06 <jetpackjackson0> i mean "best" is relative
2026-03-23 01:42:17 <jetpackjackson0> i like zsh tho
2026-03-23 01:42:26 <Renan[m]> jetpackjackson0: less bloated
2026-03-23 01:43:01 <jetpackjackson_> one person's bloat is another person's features imo
2026-03-23 01:43:10 <jetpackjackson_> bash is fine tho
2026-03-23 01:43:49 <Renan[m]> ok, i will use bash
2026-03-23 01:50:31 <rnkn> Renan[m]: oksh, the ksh completion is the easiest/best you'll find
2026-03-23 01:51:03 <Renan[m]> rnkn: why?
2026-03-23 01:51:22 <Renan[m]> what is the diference from bash?
2026-03-23 01:51:28 <rnkn> set -A complete_git_1 -- clone status diff log show add commit fetch push pull
2026-03-23 01:51:35 <rnkn> that is a whole completion for ksh
2026-03-23 01:51:44 <rnkn> for git, using ksh
2026-03-23 01:52:15 <rnkn> Renan[m]: also you avoid the indignity of using bash
2026-03-23 02:04:39 <invoked> carapace is definitely heavy (gocode at 60mb) but it is interesting that it bridges completions from various shells, so more reuse
2026-03-23 02:05:10 <invoked> if you can trust that kind of thing
2026-03-23 03:30:56 <AndroUser> Hello!
2026-03-23 03:31:16 <AndroUser> I helf bored
2026-03-23 03:31:44 <AndroUser> bye......
2026-03-23 07:10:20 <rnkn> I see cairo-dock is not packaged.. is there another good dock that is?
2026-03-23 08:34:30 <sewn> rnkn: you can always package anything
2026-03-23 09:13:56 <rnkn> I'm following the instructions on the wiki on running openrc services as the user level but it's refusing, giving this error: https://paste.sr.ht/~rnkn/98f6d5a42fb590358f52f54b9e8422d0b347cc79
2026-03-23 09:14:47 <rnkn> (You are attempting to run an openrc service on a system which openrc did not boot. ... )
2026-03-23 09:15:17 <rnkn> I'm running dbus-run-session -- labwc
2026-03-23 09:16:02 <ptrc> do you have openrc started for your user? it would either start with a symlink /etc/init.d/user.$USER -> '/etc/init.d/user', or with the openrc pam module
2026-03-23 09:16:13 <ptrc> or manually with `openrc -U`
2026-03-23 09:18:04 <rnkn> ptrc: that error is what I get with `openrc -U`
2026-03-23 09:19:13 <rnkn> but no there is no symlink
2026-03-23 09:20:28 <ptrc> ah, i see
2026-03-23 09:21:08 <ptrc> hm, it looks like.. a rundir mismatch, perhaps? since openrc is definitely running, but it can't see *itself* when starting the services themselves
2026-03-23 09:21:27 <ptrc> it might be stupid, but possibly try removing the trailing slash from $XDG_RUNTIME_DIR?
2026-03-23 09:22:36 <rnkn> removed the trailing slash, same result
2026-03-23 09:23:35 <rnkn> dbus/labwc complained about not having an XDG_RUNTIME_DIR set, and would work with /run due to permissions so I launched with XDG_RUNTIME_DIR=$HOME/.run
2026-03-23 09:23:49 <rnkn> * and wouldn't work
2026-03-23 09:47:37 <rnkn> is elogind supposed to actually work or is it just an elaborate joke?
2026-03-23 10:02:07 <phws> hi all, I tried to install Alpine in "data" mode in a qemu VM, by using a bootable partition vda1 and a data partition vda2, which I've mounted to /var. I think I did /etc/fstab properly, mounting vda2 to /var and vda1 to /media/vda1, so that I can use it in lbu.conf. Everything seems to work fine, but I have the weird situation where apk does not restore packages on boot (they are in the world file and cache seems
2026-03-23 10:02:27 <phws> instead, it restores them after I use 'apk <any command>'... does this ring any bell?
2026-03-23 10:03:01 <rnkn> I thought the cause might be that I didn't have a display manger, so installed lightdm and rc-update added it, started it, it says it's starting and.. hilariously, nothing happens
2026-03-23 10:03:14 <rnkn> a real waiting for godot moment, thanks lightdm
2026-03-23 10:04:29 <rnkn> (this was of course in the console, not a terminal emulator)
2026-03-23 10:06:05 <frag> is there a simple way to delete the last n lines with foobar in them?
2026-03-23 10:06:18 <frag> popix shell
2026-03-23 10:28:44 <rnkn> that definitely sounds like a job for awk
2026-03-23 10:53:44 <rnkn> I think openrc user services might be completely broken. I'm working on a fresh install, no GUI launched, no extra stuff in rc-update and still openrc -U says I'm running a system that openrc didn't boot.
2026-03-23 10:54:00 <rnkn> what is this mystery other system? no one knows
2026-03-23 10:54:44 <rnkn> kafka-effing-esque
2026-03-23 15:01:18 <Guest5743> question: I've been using some variation of either a deb-based or rpm-based distro as my daily driver since 2003 and use headless Debian for all of my servers. With all of this systemd age verification nonsense, I want to switch to a non-systemd distro. I'm currently using Fedora as my main distro. Would Alpine be a difficult switch or should I be fine? Is it intended to be a daily 
2026-03-23 15:01:18 <Guest5743> driver?
2026-03-23 15:01:58 <lopid> daily driver desktop or server?
2026-03-23 15:02:47 <lopid> the mars lander uses it daily, for example
2026-03-23 15:03:01 <lopid> (iirc)
2026-03-23 15:03:12 <Guest5743> both, though more concerned about desktop initially
2026-03-23 15:03:57 <Guest5743> i assume for server the main issue will be figuring out how to use something other than systemd for all my services
2026-03-23 15:04:35 <lopid> at least it uses openrc, which isn't specific to alpine
2026-03-23 15:04:59 <lopid> lopid has no experience with using it on the desktop
2026-03-23 15:06:22 <Guest5743> thank you
2026-03-23 18:01:37 <Saijin_Naib[m]> I highly recommend as a desktop
2026-03-23 18:02:10 <Saijin_Naib[m]> There are very few functional gaps, really only where proprietary drivers are concerned (nvidia)
2026-03-23 18:02:23 <Saijin_Naib[m]> If it isnt packaged, flatpak works perfectly well
2026-03-23 18:02:35 <lopid> Guest5743 left
2026-03-23 19:25:32 <ichigo[m]> Guest5743: it works well for me. just need to learn how alpine works specifically. its not too different from other distros, and it has /etc/apk/world which is kinda cool 
2026-03-23 21:17:07 <seti_> if i create a alpine chroot enviroment on a given machine, can i just chroot to it, run setup-alpine and use that to install alpine onto a target disk?
2026-03-23 21:21:46 <ikke> Yes*
2026-03-23 21:22:12 <ikke> Make sure /dev, /sys and /proc are bind mounted into the chroot
2026-03-23 21:42:18 <seti_> ikke: right, thanks!
2026-03-23 22:37:34 <phws> follow-up to my question about needing to run apk again to restore packages on a "data" install earlier: it might be that my ext4 partition mounted as /var is being mounted too late (it's in /etc/fstab), as the error seems consistent with having a broken cache symlink, but I can manually inspect the cache just fine.
2026-03-23 22:37:40 <phws> can that be the case? and if so, how could I fix it? thanks
2026-03-24 00:07:05 <p_6f3Ik7Suw> seti_: there is alpine-chroot-install
2026-03-24 00:07:54 <p_6f3Ik7Suw> which does all these things. and creates a bunch of scripts to enter and tear down the chroot - the latter is important, simply running rm -rf chroot might delete bindmounted stuff you care about...
2026-03-24 00:08:21 <p_6f3Ik7Suw> see: https://github.com/alpinelinux/alpine-chroot-install
2026-03-24 05:37:04 <Saijin_Naib[m]> Lutris is now vibecoded. Anyone want to take it and/or kew from me? I will no longer maintain them.
2026-03-24 06:20:09 <quinq> Saijin_Naib[m], thanks for your efforts :)
2026-03-24 06:57:19 <Saijin_Naib[m]> WhyNotHugo is d7vk roughly the same packaging process as dxvk?
2026-03-24 06:57:19 <Saijin_Naib[m]> https://github.com/WinterSnowfall/d7vk
2026-03-24 06:57:52 <Saijin_Naib[m]> Also, thanks for dxvk. Works very nicely :)
2026-03-24 09:15:42 <phws> follow-up #2: I'm solving it with a local.d script running 'apk fix' for now... kinda silly but it seems to work OK
2026-03-24 11:24:47 <rnkn> I'm hitting a stonewall with openrc -U, which gitlab project should I report the bug?
2026-03-24 11:25:57 <ikke> aports
2026-03-24 11:28:39 <rnkn> ikke: ty
2026-03-24 11:32:30 <rnkn> just in case there's something obvious where I'm being foolish, there's no simple solution to this error? https://paste.sr.ht/~rnkn/98f6d5a42fb590358f52f54b9e8422d0b347cc79
2026-03-24 13:01:40 <invoked> rnkn: were you following the wiki?
2026-03-24 13:02:19 <rnkn> invoked: to the letter
2026-03-24 13:03:19 <invoked> cd /var
2026-03-24 13:03:21 <invoked> ls
2026-03-24 13:03:25 <invoked> whoops
2026-03-24 13:04:07 <invoked> i'm wondering if you have something funky in your setup because my alpine box doesn't have ~/.run
2026-03-24 13:05:14 <invoked> this alpine machine on my end is running user services successfully.
2026-03-24 13:05:28 <rnkn> invoked: what's your XDG_RUNTIME_DIR?
2026-03-24 13:06:38 <rnkn> I am pretty sure mine is wrong and I should have /var/run/user/1000 but shouldn't the system create that and set appropriate perms?
2026-03-24 13:08:13 <invoked> i'm not using a sysvinit runlevel like it is in the wiki.  i just have one for bringing up a gui, and explicitly call that from my compositor when that runs.
2026-03-24 13:08:45 <invoked> that could be the difference here.
2026-03-24 13:09:28 <invoked> i start a system dbus and run pipewire, pipewire-pulse, and wireplumber in user services.
2026-03-24 13:10:35 <grawity> oddly, pam_openrc doesn't set it
2026-03-24 13:11:28 <invoked> i don't use pam either.
2026-03-24 13:11:47 <rnkn> invoked: what's your XDG_RUNTIME_DIR?
2026-03-24 13:11:54 <grawity> good for you, but I was researching that for rnkn
2026-03-24 13:12:28 <invoked> rnkn: /tmp/$uid-runtime-dir/
2026-03-24 13:12:29 <grawity> (though that made me curious where pipewire puts its Unix sockets in your case as wel)
2026-03-24 13:13:35 <invoked> are you using a display/login manager?
2026-03-24 13:14:06 <invoked> (that's another thing i don't use. i set $XDG_RUNTIME_DIR up in .profile)
2026-03-24 13:14:22 <invoked> so i might be confusing your issue, apologies.
2026-03-24 13:14:53 <rnkn> I don't have a display manager installed no
2026-03-24 13:15:07 <rnkn> but I get the same error from the console too
2026-03-24 13:17:29 <grawity> looks like the original implementation of "user services" in openrc used to set that, but the final merged version doesn't
2026-03-24 13:18:04 <grawity> well, I'd probably suggest pam_rundir out of the options at https://wiki.alpinelinux.org/wiki/XDG_RUNTIME_DIR especially if you rely on PAM to start openrc
2026-03-24 13:18:31 <grawity> but if you run openrc manually then creating the dir / setting the envvar manually is fine, as long as you don't put it on NFS or something
2026-03-24 13:20:45 <invoked> from past logs i see it's an open issue (or was).  hugo suggested: "touch $XDG_RUNTIME_DIR/openrc/softlevel"
2026-03-24 13:21:07 <invoked> dunno if that is still true
2026-03-24 13:21:40 <invoked> rnkn: ^
2026-03-24 13:26:24 <rnkn> installing pam-rundir has fixed it
2026-03-24 13:26:34 <rnkn> thanks :)
2026-03-24 13:28:03 <invoked> nice
2026-03-24 13:28:08 <rnkn> at some point I symlinked /etc/init.d/user.rnkn to /etc/init.d/user but I'm not sure why?
2026-03-24 13:28:15 <rnkn> I actually don't understand PAM at all
2026-03-24 13:28:55 <grawity> the 'session' section of PAM tells each module "do your thing"
2026-03-24 13:29:07 <grawity> usually that's something related to the fact that the user just logged in (or out)
2026-03-24 13:29:56 <invoked> login managers usually require pam, iirc
2026-03-24 13:30:27 <grawity> mostly because they find it more convenient than the alternatives
2026-03-24 13:31:29 <grawity> *could* make a GUI login manager without PAM – for the 'auth' part just be setuid root and read /etc/shadow, for the 'session' part do literally nothing
2026-03-24 13:32:35 <grawity> ah, I mixed up "login manager" with "display manager"
2026-03-24 13:33:02 <invoked> it's definitely subjective. i'd rather have control from the files i'm already actively using.
2026-03-24 13:33:14 <grawity> if you meant login manager as in logind/seatd, then... well, those could still exist without PAM, but good luck patching 10 different display managers, sshd, etc. to talk to your daemon
2026-03-24 13:33:27 <grawity> 'session' modules are a convenient plugin interface
2026-03-24 13:33:35 <invoked> for personal stuff, it's not like i'm running ldap/et al
2026-03-24 13:34:31 <grawity> and generally might be better described as "two or three different plugin interfaces under one umbrella"
2026-03-24 13:34:46 <grawity> auth/account/passwd could be one thing and session another
2026-03-24 13:35:59 <grawity> someone earlier wanted to set CPU affinity for users if they're in a specific group. patch sshd and then identically patch getty/login and then GDM/xdm? no, 5-line PAM module that all of them invoke
2026-03-24 13:37:02 <invoked> fair, but i'm the only user on personal systems, so for me, it's nice to avoid the yearly pam vuln
2026-03-24 13:38:24 <rnkn> invoked: you don't recommend PAM?
2026-03-24 13:38:53 <dwfreed> if you're the only user on personal systems, the yearly pam vulns already don't matter to you
2026-03-24 13:39:16 <invoked> not yet, maybe
2026-03-24 13:39:31 <grawity> I looked up PAM CVEs
2026-03-24 13:39:37 <grawity> there aren't that many
2026-03-24 13:41:06 <invoked> i run alpine in a corporate setting, so context matters, i use pam there because i have to.
2026-03-24 13:41:56 <invoked> or well i used to.  past tense. (quit that job)
2026-03-24 13:44:11 <invoked> rnkn: i don't advocate anything, as a rule, so don't take this as advice
2026-03-24 13:50:19 <LiminalDuck> I'm new here.  I have the "virtual" image in a VM to play with and see if I want to migrate my home servers from Debian to Alpine.  So far I'm loving it.  I've found that my test install doesn't have the kernel module for exfat filesystems.  Is there a document that will tell me if it's in the "standard" image, or do I need to download it and check?
2026-03-24 13:53:56 <rnkn> invoked: I decided to see if I could get things to work without PAM just by making the runtime dir in /tmp and it seems to be okay
2026-03-24 13:55:57 <invoked> rnkn: cool. i mean, it's just my principle to scrape off any extra software i don't need on my personal systems. i'm not here saying it should be this way.
2026-03-24 13:57:36 <rnkn> invoked: it's okay, I won't hold you accountable if my computer explodes
2026-03-24 13:58:41 <dwfreed> LiminalDuck: which version are you running? 3.23?
2026-03-24 14:00:23 <rnkn> LiminalDuck: from memory this is just a case of apk add linux-lts and apk del linux-virt
2026-03-24 14:00:48 <dwfreed> except the exfat module is included in linux-virt in every version on pkgs, which is 3.15 - edge
2026-03-24 14:00:56 <rnkn> oh, well
2026-03-24 14:01:33 <dwfreed> https://pkgs.alpinelinux.org/contents?file=exfat.ko*&path=&name=&branch=v3.23&repo=main&arch=x86_64
2026-03-24 14:11:40 <LiminalDuck> I'm running the current "virtual" download - 3.24.0_alpha20260127
2026-03-24 14:12:29 <dwfreed> that's edge
2026-03-24 14:18:58 <dwfreed> anyway, what makes you think you don't have the exfat module?
2026-03-24 14:19:58 <LiminalDuck> I couldn't mount an exfat filesystem I created.  I found that I do have it - it just isn't loaded by default. Thanks!
2026-03-24 14:40:41 <ikke> With busybox mount, you have to explicitly provide the type
2026-03-24 14:40:55 <ikke> mount -t exfat
2026-03-24 14:56:08 <LiminalDuck> does busybox even support exfat?  there was no mkfs.exfat so I installed exfatprogs.  now that I have the module loaded, mount works without the -t
2026-03-24 14:57:20 <ikke> LiminalDuck: When the module is loaded, there is no reason to explicitly provide the type. But when it's not loaded, specifying the type should load the correct module first
2026-03-24 16:19:46 <prabu> rnkn, the wiki page was updated sometime ago and i flagged my concerns there..no response given as on date..https://wiki.alpinelinux.org/wiki/Talk:OpenRC#User_Services_-%3E_Runlevels
2026-03-24 16:23:38 <prabu> good to know that you got the openrc user services sorted.. it
2026-03-24 16:26:28 <invoked> prabu: WhyNotHugo might not see your wiki pings.
2026-03-24 16:26:37 <invoked> (dunno)
2026-03-24 16:27:38 <invoked> considering hugo probably gets 10,000 emails/day across everything he does
2026-03-24 16:31:34 <WhyNotHugo> I don't actually get any wiki emails. Might have to enable it manually?
2026-03-24 16:31:39 <WhyNotHugo> Thanks for the ping
2026-03-24 16:31:54 <prabu> true.. since i also don't use irc regularly, don't know how to contact them. i just see irclogs and respond here.. sometimes even late by days together..
2026-03-24 16:39:23 <WhyNotHugo> I don't see an option for the wiki to email me in case of mention.
2026-03-24 16:39:31 <WhyNotHugo> And looks like I missed the talk page in the watchlist ui
2026-03-24 16:40:29 <prabu> thanks WhyNotHugo. you are one of the few alpine developer/maintainers who regularly update wiki. highly appreciated..
2026-03-24 16:50:54 <prabu> invoked thanks for helping users on this channel regularly.. very true about your comment about WhyNotHugo. his blog post about apk world is one of the reasons for me to try alpine
2026-03-24 17:14:59 <Saijin_Naib[m]> Got a blog link?
2026-03-24 17:47:21 <Biswa96[m]> little bit of googlefu https://whynothugo.nl/journal/2023/02/18/in-praise-of-alpine-and-apk/
2026-03-24 17:58:08 <Saijin_Naib[m]> thanks
2026-03-24 19:17:09 <sqrvrt[m]> hi all, is there a way to point the user/system to use a coreutils implementation other than busybox?
2026-03-24 19:17:43 <pj> Install coreutils
2026-03-24 19:18:10 <sqrvrt[m]> is that it? or do i need to set omething to let the system know it's preferred?
2026-03-24 19:18:34 <pj> It will automatically replace busybox symlinks just via installation
2026-03-24 19:19:01 <sqrvrt[m]> i wanted to try uutils and it doesn't look like it replaces the symlinks
2026-03-24 19:19:43 <pj> for uutils specifically I think you need uutils-coreutils package
2026-03-24 19:20:34 <sqrvrt[m]> works now, thanks!
2026-03-24 21:46:33 <ssm> don't check the cargo deps
2026-03-24 22:12:49 <seti_> how is one expected to use the generic u-boot tarballs that are on the downloads page? do i need to build a img from them or are they supposed to just work if you plonk them on an sd card with a fat32 partition?
2026-03-24 22:13:14 <seti_> i noticed that it seemed like the tarball had the dtb for my board already
2026-03-25 03:24:40 <rnkn> does `rc_env_allow="XDG_RUNTIME_DIR"` need to be in ~/.config/rc/rc.conf when using user rc services without PAM?
2026-03-25 04:14:22 <rnkn> nope, that doesn't work.. and also user services *with PAM* don't work
2026-03-25 04:35:15 <rnkn> omg... please tell me it's documented somewhere on the wiki that to run user services you need to both symlink /etc/init.d/user.$USER and to add user.$USER with rc-update
2026-03-25 04:35:25 <rnkn> (w/o PAM)
2026-03-25 04:37:05 <rnkn> so now I'm using XDG_RUNTIME_DIR=/tmp/xdg/$(id -u)-runtime-dir because /run/user/$(id -u) is not created at boot.. is this bad?
2026-03-25 04:37:11 <rnkn> (again w/o PAM)
2026-03-25 09:59:52 <fission> I updated the riscv64 article with some testing I did in qemu. feedback welcome
2026-03-25 10:17:40 <RoadRoller01[m]> why boot.alpinelinux.org doesn't support http):
2026-03-25 10:18:43 <RoadRoller01[m]> i tried to add http://boot.alpinelinux.org/alpine-ipxe/x86_64/ipxe.pxe to my server bios as it doesn't support https
2026-03-25 10:19:42 <RoadRoller01[m]> the image would be useless as i don't think that there is a bios that support ssl
2026-03-25 10:20:38 <fission> it works here
2026-03-25 10:20:45 <fission> (with http)
2026-03-25 10:21:57 <fission> % curl -sO http://boot.alpinelinux.org/alpine-ipxe/x86_64/ipxe.pxe
2026-03-25 10:22:01 <fission> % file ipxe.pxe
2026-03-25 10:22:02 <fission> ipxe.pxe: pxelinux loader (version 3.70 or newer)
2026-03-25 10:22:18 <RoadRoller01[m]> hmm
2026-03-25 10:22:35 <RoadRoller01[m]> maybe my bios is dumb
2026-03-25 10:22:42 <fission> (also #til that such a thing even existed heh)
2026-03-25 10:23:24 <RoadRoller01[m]> fission: 🤔
2026-03-25 10:23:47 <fission> in my defense, I don't do a lot with actual hardware :)
2026-03-25 10:54:05 <RoadRoller01[m]> bro i got permission denied!
2026-03-25 10:54:14 <RoadRoller01[m]> how is this even possible 
2026-03-25 10:55:08 <RoadRoller01[m]> i used ipxe.efi instead of ipxe.pxe
2026-03-25 11:01:33 <RoadRoller01[m]> <RoadRoller01[m]> "bro i got permission denied!" <- nice after the error it included this url https://ipxe.org/err/0216eb
2026-03-25 11:03:07 <RoadRoller01[m]> it was http://ipxe.org/err/0216eb8f but redircted to that one 
2026-03-25 11:03:37 <RoadRoller01[m]> hmm
2026-03-25 11:04:22 <RoadRoller01[m]> because it was tring to fetch https://boot.alpinelinux.org/boot.ipxe 
2026-03-25 11:04:28 <RoadRoller01[m]> this is not good
2026-03-25 11:04:59 <RoadRoller01[m]> fission: the script would use https instead of http(:
2026-03-25 11:05:10 <RoadRoller01[m]> even when it got requsted as a http
2026-03-25 11:05:23 <RoadRoller01[m]> s/a//
2026-03-25 12:00:03 <WhyNotHugo> rnalrd: iirc, navi mentioned that openrc sources your defualt shell, so if XDG_RUNTIME_DIR is set there that should work?
2026-03-25 12:00:06 <WhyNotHugo> Or something along theselines?
2026-03-25 12:00:58 <navi> openrc starts user services as `-$shell -c openrc-run.sh start`
2026-03-25 12:01:05 <navi> so it starts a login shell
2026-03-25 12:01:16 <navi> except if you shell isn't on /etc/shells, then it falls back to /bin/sh
2026-03-25 12:02:05 <navi> rnkn: on my main tower i've used a XDG_RUNTIME_DIR set up from .zprofile for ages, worked fine
2026-03-25 13:08:15 <invoked> rnkn: in my case without pam, i'm not setting ~/.config/rc/rc.conf at all. works fine.
2026-03-25 13:12:27 <invoked> either way, whether you setting this from your login shell or whether pam is doing it for you, each needs to make sure that directory is there.
2026-03-25 14:19:56 <bb> test
2026-03-25 14:29:21 <longnoserob[m]> maybe passed
2026-03-25 14:32:34 <bb> I have installed alpine linux on a second ssd on my computer just for fun. I use sparky linux normally and I am a normal user and not a linux expert. I have installed fluxbox as a window manager. I have pcmanfm as a filemanager. My problem is that pcmanfm does not show me external drives like a usb-stick. I have installed gvfs-fuse udisks2 fuse-openrc and elogind. All the services a running but it does not work. Google AI could not really 
2026-03-25 14:32:34 <bb> help. Does someone know what the problem could be. Probabely something to do with rights...
2026-03-25 14:39:21 <invoked> bb: i don't use my system the way you do, but you didn't mention polkit, so maybe you're missing that.
2026-03-25 14:39:51 <invoked> https://wiki.alpinelinux.org/wiki/Polkit#Using_polkit_with_elogind might help
2026-03-25 14:40:26 <bb> I forgot to write it, but I have also installed polkit-elogind...
2026-03-25 14:40:52 <invoked> is polkit running?
2026-03-25 14:41:41 <invoked> also you didn't mention specifically what you're trying to do, and what you expected to happen but didn't
2026-03-25 14:42:05 <invoked> mention a specific case
2026-03-25 14:42:41 <bb> when I plug a usb stick, it does not show in pcmanfm. I have to mount it manually.
2026-03-25 14:43:51 <invoked> ok. and you have every service started (dbus, polkit, elogind, (etc))
2026-03-25 14:44:22 <bb> yes everything was running.
2026-03-25 14:46:01 <invoked> double check that all these services are in default runlevel and that they're running after you boot & log in.  beyond that, it's some desktop stuff that i don't get into myself, so someone else should probably help you further
2026-03-25 14:47:02 <bb> thank you. I will check again. And maybe I ask tomorrow again :)
2026-03-25 14:47:41 <invoked> be patient, everyone is volunteering their time
2026-03-25 14:48:11 <invoked> or not
2026-03-25 18:07:21 <paulgupta> Hey! Question.... downloads page has "standard" and "extended" and extended states "runs from RAM" does standard not run from RAM? is extended not able to install a classic install? 
2026-03-25 18:14:27 <courge_musque[m]> There are apk files in both, extended has more of them and they are in RAM
2026-03-25 18:14:34 <courge_musque[m]> In standard, everything but the apk files runs in RAM
2026-03-25 18:15:07 <courge_musque[m]> Kinda cool to put the iso on an usb drive, and install Alpine on the same drive
2026-03-25 18:21:29 <paulgupta> that's my goal
2026-03-25 18:21:39 <paulgupta> to make a "liveusb" alpine that's running entirely in RAM 
2026-03-25 18:23:19 <courge_musque[m]> Pick extended then
2026-03-25 18:31:04 <paulgupta> doing the thing
2026-03-25 18:32:36 <fission> 
ACTION applauds

2026-03-25 18:40:13 <sqrvrt[m]> hi, so i built an executable and for some reason i can't run it despite having exec permissions on it
2026-03-25 18:40:31 <sqrvrt[m]> alp@alpine ~/l/a/s/l/build (config-struct)> ./lmms
2026-03-25 18:40:31 <sqrvrt[m]> alp@alpine ~/l/a/s/l/build (config-struct)> ldd ./lmms
2026-03-25 18:40:31 <sqrvrt[m]> alp@alpine ~/l/a/s/l/build (config-struct) [126]> file ./lmms
2026-03-25 18:40:31 <sqrvrt[m]> ./lmms: ELF 64-bit LSB pie executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib/ld-musl-x86_64.so.1, BuildID[sha1]=f1374ca02087d2c7d8f6edef79d3f7b50dc6302a, with debug_info, not stripped
2026-03-25 18:40:31 <sqrvrt[m]> fish: Unknown command. './lmms' exists but is not an executable file.
2026-03-25 18:40:32 <sqrvrt[m]> alp@alpine ~/l/a/s/l/build (config-struct) [1]> ls -l ./lmms
2026-03-25 18:40:32 <sqrvrt[m]> /lib/ld-musl-x86_64.so.1: ./lmms: Not a valid dynamic program
2026-03-25 18:40:34 <sqrvrt[m]> -rwxr-xr-x 1 alp alp 58286648 Mar 25 21:33 ./lmms*
2026-03-25 18:40:48 <sqrvrt[m]> it was built under the same OS so i don't think linking to glibc is a concern
2026-03-25 18:51:33 <fission> sqrvrt[m]: how did you build it?
2026-03-25 18:52:44 <invoked> i'm heading out the door. don't know how lmms builds, but sounds like pie
2026-03-25 18:54:24 <sqrvrt[m]> fission: cmake
2026-03-25 19:04:04 <sqrvrt[m]> invoked: what's pie
2026-03-25 19:55:50 <ikke> position independent executable
2026-03-25 19:58:21 <idkrnx[m]> <sqrvrt[m]> "what's pie" <- The bare minimum
2026-03-25 20:13:33 <viktor_> hi :D
2026-03-25 20:22:46 <Jerkwad_Robot> Hi all. I have an issue but I am not sure it's an issue. I ran through the Daily driver guide / Gnome stuff to setup my laptop. It's working pretty well, but I am having an issue with Polkit starting, and I think in troubleshooting it I've made a mess of my rc runlevels.
2026-03-25 20:26:35 <Jerkwad_Robot> HELP IDENTIFY
2026-03-25 20:27:01 <ikke> Jerkwad_Robot: to identify, you need to talk to nickserv, not in this channel
2026-03-25 20:27:54 <Jerkwad_Robot> LOL, sorry
2026-03-25 20:28:31 <Jerkwad_Robot> I wonder why my alternates aren't showing... I'l save that for later :)
2026-03-25 20:35:10 <fission> Jerkwad_Robot: sorry, what is the issue?
2026-03-25 20:44:57 <Jerkwad_Robot> fission: my polkit service is in a "failed" state and I cannot get it started.
2026-03-25 20:45:15 <Jerkwad_Robot> I do have the apk-polkit=server running.
2026-03-25 20:45:53 <Jerkwad_Robot> I have been having some audio issues, so I've been walking through the Pipewire info over and over to see what might be wrong.
2026-03-25 20:54:25 <Jerkwad_Robot> Polkit warns in the messages log that the rules.d file cannot be found / no access. I can't find it myself, so I am wondering whereI went wrong on my setup.
2026-03-25 21:07:57 <seti_> When I tried installing alpine onto a disk inside a chroot, when i chose the disk and let it rip, after a certain point setup-alpine stopped with apk complaining that it cant find "linux-ky" (the board in question has a spacemit ky SOC)
2026-03-25 21:08:32 <seti_> the board should be fine since i saw some IRC logs from devs apparently saying my board works fine, and i can see the dtb included in the u-boot tarball download from alpine
2026-03-25 21:13:51 <ikke> We don't have linux-ky in the repositories though
2026-03-25 21:16:50 <seti_> ikke: is that the name for the kernel?
2026-03-25 21:17:25 <ikke> I would assume so, that's our convetion, linux-<flavor>
2026-03-25 21:17:26 <seti_> i have it so couldnt i somehow squeeze it in manually?
2026-03-25 21:17:47 <seti_> as in i have the compiled kernel for it available
2026-03-25 21:19:12 <ikke> setup-disk will try to install it from _a_ repository, so you would have to make a package for it
2026-03-25 21:19:37 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-disk.in#L651-670
2026-03-25 21:19:58 <seti_> could i maybe tweak the script instead to have it just copy the kernel over from the ubuntu that im chrooted from
2026-03-25 21:20:11 <seti_> instead of trying to pull from the repo
2026-03-25 21:20:51 <ikke> possibly
2026-03-25 21:21:06 <seti_> haha i get it, no guarantees but conceptually it should be fine?
2026-03-25 21:21:44 <ikke> I have no experience with using non-alpine kernels, so no clue
2026-03-25 21:21:49 <ikke> It could work, but I have no idea
2026-03-25 21:22:54 <seti_> i'll give it a shot at least
2026-03-25 21:23:51 <Jerkwad_Robot> here's a screenshot of my rc-status: https://user.fm/files/v2-019bb1192ababd48e16010dc744bb11b/troubleshooting.zip
2026-03-25 22:40:14 <swee> how do I manually set partitioning in setup-alpine so I can dualboot?
2026-03-25 22:44:54 <fission> swee: https://wiki.alpinelinux.org/wiki/Setting_up_disks_manually
2026-03-25 22:46:17 <fission> swee: setup-alpine normally uses an entire hard drive, but you can use setup-disk to install to a particular partition
2026-03-25 22:46:38 <swee> oh oki
2026-03-25 22:47:08 <swee> and hopefully update-grub doesn't replace everything in the EFI directory right?
2026-03-25 22:47:54 <Ermine> it should create its own subdir
2026-03-25 22:58:54 <swee> fission: setup-disk seems to be the one that is setting up partitioning on the whole drive
2026-03-25 23:00:53 <swee> how do i install alpine with my manually set up partitions
2026-03-25 23:01:38 <courge_musque[m]> By mounting them
2026-03-25 23:01:47 <courge_musque[m]> Then you can give /mnt or /media or /whatever as an argument to setup-disk
2026-03-25 23:01:47 <swee> and..?
2026-03-25 23:02:14 <courge_musque[m]> setup-disk -s 0 -k /mnt 
2026-03-25 23:02:14 <courge_musque[m]> is what I usually do
2026-03-25 23:02:21 <swee> oic
2026-03-25 23:02:35 <swee> i didn't know those tools had automation flags
2026-03-25 23:03:02 <courge_musque[m]> *setup-disk -s 0 -k stable /mnt 
2026-03-25 23:03:21 <courge_musque[m]> I didn't install Alpine in a month, I need to do it again lol
2026-03-25 23:07:03 <swee> givung info as to where the efi directory is? or just adding /mnt/boot/efi
2026-03-25 23:08:47 <fission> you should mount /mnt and then mkdir -p /mnt/boot/efi ; mount /dev/xxxx /mnt/boot/efi (your EFI partition) if you want setup-disk to write things to it
2026-03-26 02:19:41 <swee> when running an xwayland application on KDE Plasma (I'm only seeing this on Alpine), the classic X11 cursor is used when hovering over the window
2026-03-26 02:24:15 <swee> is there like a config separate from KDE Plasma's that I have to set in order for XWayland apps to use my custom cursor theme?
2026-03-26 07:56:42 <frag> will letter pdf size usually be ok for a4 printer if you have ok margins?
2026-03-26 08:56:09 <fission> frag: yep, unless the margins are very tight
2026-03-26 09:36:34 <frag> fission thanks!
2026-03-26 10:56:49 <jpp1> KDE Plasma is crap, Google
2026-03-26 10:57:08 <jpp1> Sway is good
2026-03-26 11:11:13 <Lilith> sway is crap, duckduckgo
2026-03-26 11:11:21 <Lilith> see how ridiculous it sounds?
2026-03-26 11:43:37 <Daanct12> hi Lilith 
2026-03-26 12:39:41 <Lilith> hi Danct12 
2026-03-26 12:58:37 <Jerkwad_Robot> okay, stupid question. Should my polkit service be in the default runlevel, or where mine is currently... manual?
2026-03-26 13:00:18 <ikke> manual means you started them without adding them to a run level
2026-03-26 13:00:30 <ikke> For the operation of the service itself, it does not really matter too much
2026-03-26 13:01:47 <ikke> It roughly determines the order in which these services are started
2026-03-26 13:02:16 <Jerkwad_Robot> Oh! Hmm... so does it run only when needed? I figured out the reason my polkit service wasn't starting is because when I add it to the default I wind up with two, and I can't delete the one in 'manual'
2026-03-26 13:04:56 <ikke> Jerkwad_Robot: Are you talking about system or user services?
2026-03-26 13:19:00 <Jerkwad_Robot> ikke: I believe both instances are system services.
2026-03-26 13:19:31 <Jerkwad_Robot> I think polkit is supposed to be a system service right?
2026-03-26 13:20:58 <Jerkwad_Robot> but to your point, I added the service at some point without adding it to the default runlevel, so I have to start it manually. The good news is, without polkit added to the default runlevel, it starts. :D
2026-03-26 13:21:08 <ikke> Yes, it needs to
2026-03-26 13:21:41 <ikke> Jerkwad_Robot: that may indicate there is an ordering problem, that it depends on a service that may not be running yet?
2026-03-26 13:21:57 <Jerkwad_Robot> The only thing I am stuck on is I can't delete the manual polkit service entry.
2026-03-26 13:22:16 <ikke> It should go away when you stop the service
2026-03-26 13:22:21 <Jerkwad_Robot> I tried moving it to default, but it creates a separate service then both ....
2026-03-26 13:22:22 <Jerkwad_Robot> oh
2026-03-26 13:25:01 <Jerkwad_Robot> Hmm, now it's failing again. 
2026-03-26 13:25:50 <ikke> Did you check any logs to find out why it's failing?
2026-03-26 13:26:18 <Jerkwad_Robot> I'll check again. It was returning file / location not found looking for the rules.d file IIRC
2026-03-26 13:27:35 <Jerkwad_Robot> which from my searching manually doesn't exist. I wasn't sure if I had to manually create that, but I saw the examples in the wiki.
2026-03-26 13:28:16 <Jerkwad_Robot> I've got a bit of work to do ATM, so I'll work on this a little later in the day. Thanks for trying to help ikke
2026-03-26 13:29:22 <ikke> alright
2026-03-26 15:43:08 <qbit> hola
2026-03-26 18:48:07 <Jerkwad_Robot> for the installation of a desktop (gnome in my case) do I need to run the setup-wayland-base before running the setup-desktop script?
2026-03-26 19:29:27 <invoked> Jerkwad_Robot: no. these are just scripts. if you look at /usr/sbin/setup-desktop you will see it calls setup-wayland-base as it needs to
2026-03-26 19:30:42 <invoked> setup-wayland-base is potentially useful if you're running something not implemented by setup-desktop
2026-03-26 19:31:24 <invoked> but technically you don't need either of them if you know what to install.
2026-03-26 20:51:57 <paulgupta> struggling here. I'm trying to grep lines that contain ""id:"" (As in the string id: surrounded with quotes.) I think it's failing because of the ":")
2026-03-26 20:52:04 <paulgupta> shoot, wrong channel sorry!
2026-03-26 20:59:49 <dwfreed> (answered in another channel)
2026-03-26 21:05:35 <Jerkwad_Robot> invoked: thanks for the info!
2026-03-26 21:10:49 <Jerkwad_Robot> I love those scripts... so easy :D
2026-03-26 21:13:14 <Jerkwad_Robot> Yet another question.... my doas doesn't seem to honor the persist setting. Thoughts?
2026-03-26 23:15:21 <ssm_> persist isn't supported outside openbsd, unless that changed
2026-03-26 23:15:33 <ssm_> there is no official portable doas
2026-03-26 23:16:13 <penguinz_rule[m]> i have persist
2026-03-26 23:18:37 <erminea[m]> Persist works for me
2026-03-27 02:08:26 <Jerkwad_Robot> Hmm, I've seen it referenced in other distros. https://wiki.archlinux.org/title/Doas
2026-03-27 02:08:55 <Jerkwad_Robot> I need to go add the newline to my doas file. Didn't know it needed that.
2026-03-27 02:46:30 <Jerkwad_Robot> Apparently the Arch version uses timestamp files.
2026-03-27 02:51:42 <abby> arch, void, and alpine all use the same doas port
2026-03-27 03:11:09 <Jerkwad_Robot> abby: thanks. Not sure why mine isn't working. I'll keep digging.
2026-03-27 04:14:55 <frag> just kept a terminal su-ed to root in the 9th workspace in dwm for a while now, instead of doas/sudo, seems to work..
2026-03-27 08:25:04 <lupo> Does Alpine Linux have an AI code / AI slop policy?
2026-03-27 08:28:36 <ikke> Not yet
2026-03-27 08:29:10 <ikke> https://gitlab.alpinelinux.org/alpine/council/-/issues/697
2026-03-27 08:34:48 <lanky_ginger> Thanks for the link. I'm glad to see it's being discussed
2026-03-27 08:42:15 <jpp1> lupo: https://notes.vv221.fr/ia-g%C3%A9n%C3%A9ratives.xhtml
2026-03-27 09:08:27 <quinq> achill, small typo there discouranged → discouraged
2026-03-27 09:12:04 <erminea[m]> Discouranged = ranged discouraged
2026-03-27 09:12:47 <quinq> :D
2026-03-27 09:14:43 <quinq> Could be LLM auto-correct, eliminating bad words like raged
2026-03-27 09:17:12 <lupo> Is there a statement regarding age verification for Alpine Linux from its developers?
2026-03-27 09:17:57 <quinq> Like you cannot participate under/over a certain age?
2026-03-27 09:20:55 <lupo> quinq, I refer to this: https://www.reddit.com/r/LinusTechTips/comments/1ri8zdc/california_introduces_age_verification_law_for/
2026-03-27 09:22:45 <lupo> There is increasing evidence that other states/countries want to follow this example.
2026-03-27 09:24:14 <lupo> s/want/think about/
2026-03-27 09:30:05 <quinq> lel
2026-03-27 09:30:20 <quinq> How to kill non-commercial computing
2026-03-27 09:30:46 <quinq> But that is fine, this is something that the vendors can develop and apply themselves
2026-03-27 09:30:56 <quinq> And/or pay Alpine linux for this
2026-03-27 09:32:01 <quinq> Also this goes against any privacy law in effect
2026-03-27 09:32:32 <quinq> There might be little of that in this country
2026-03-27 09:32:49 <quinq> But in some other part of the world, that couldn't be legally run I think
2026-03-27 09:33:05 <quinq> Anyways, that's the wrong layer to manage this problem
2026-03-27 10:28:07 <hazardaj_nombroj> is there something in firefox 138 and 139 preventing them from being built on alpine?
2026-03-27 10:29:28 <hazardaj_nombroj> *correction: 148 and 149
2026-03-27 10:46:40 <plavielle> hi is there someone awake here ?????
2026-03-27 10:48:04 <longnoserob[m]> barely
2026-03-27 10:48:11 <hazardaj_nombroj> 
ACTION snores loudly

2026-03-27 10:50:32 <plavielle> just one thing : it's impossible in microsoft hyperv VM (any version any hardware) to make "services" start at boot. Someone had an explenation ???? Thanks
2026-03-27 10:56:18 <ikke> plavielle: Not without any more details
2026-03-27 10:57:11 <plavielle> what details do you want. Tested with every version of alpine iso since 6 month
2026-03-27 10:58:01 <ikke> Details that may be relevant in determining what's going on
2026-03-27 10:58:09 <ikke> Troubleshooting that you have performed
2026-03-27 10:58:13 <plavielle> everybody use alpine on docker and nowhere else
2026-03-27 10:58:15 <ikke> Specific error messages or output
2026-03-27 10:58:26 <longnoserob[m]> which service?
2026-03-27 10:58:57 <plavielle> docker / lighttpd / nginx 
2026-03-27 10:59:06 <plavielle> no error
2026-03-27 10:59:26 <ikke> Verbosity++
2026-03-27 11:00:00 <ikke> We don't have hyper-v servers to test on
2026-03-27 11:00:03 <plavielle> classic install / install soft / rc-update add and nothing appened
2026-03-27 11:00:14 <ikke> So just saying "it does not work" does not help a lot
2026-03-27 11:00:46 <plavielle> how can i "debug" this (i'm not a full linux specialist)
2026-03-27 11:01:49 <quinq> plavielle, look at /var/log/messages for example
2026-03-27 11:02:18 <quinq> Run rc-service with -d flag (debug)
2026-03-27 11:03:37 <plavielle> "/var/log/message/ nothing
2026-03-27 11:04:44 <quinq> How do you identify that they don't start?
2026-03-27 11:05:04 <plavielle> because they dont start !!! 
2026-03-27 11:05:38 <plavielle> for me if you launch rc-update add XXXX it's all you have tot do
2026-03-27 11:05:45 <quinq> I didn't ask why, but how
2026-03-27 11:06:26 <plavielle> rc-service lighttpd status for example
2026-03-27 11:06:39 <quinq> ok :)
2026-03-27 11:08:53 <plavielle> i'm not a completly noob ;-)
2026-03-27 11:09:25 <quinq> That's not about that, it's about for the people trying to help you understanding what could be happening
2026-03-27 11:09:47 <vlt> plavielle: Great, then you know how to provide the data to debug this :)
2026-03-27 11:10:18 <plavielle> sorry i'm going to eat, and after i'm going to see for a new job. i come back i, 2 or 3 hours thanks for your help
2026-03-27 11:10:35 <quinq> plavielle, I would try installing syslog for getting easier daemon messages
2026-03-27 11:10:47 <quinq> Then we all can maybe get more details about what's happening
2026-03-27 11:11:13 <quinq> Don't quit your job over a service not starting!
2026-03-27 11:11:15 <plavielle> ok, i'm going to see how to install and configure syslog
2026-03-27 11:11:23 <quinq> https://wiki.alpinelinux.org/wiki/Syslog
2026-03-27 11:11:26 <plavielle> quind ;-)
2026-03-27 11:11:30 <quinq> ^^
2026-03-27 12:57:44 <lil5[m]> Building a custom APKBUILD I’m trying to make the source arch agnostic but if I use `source=“http…/program-$arch.tar.xz”` with `arch=“x86_64 aarch"` it returns `http…/program-x86_64 aarch.tar.xz` and says it cant download `http…/program-x86_64` space is the culprit.
2026-03-27 12:57:44 <lil5[m]> Is there a way to get the current architecture as a variable?
2026-03-27 12:59:29 <hazardaj_nombroj> apk --print-arch
2026-03-27 12:59:53 <ikke> $CARCH
2026-03-27 12:59:57 <hazardaj_nombroj> oh
2026-03-27 13:00:16 <hazardaj_nombroj> $CARCH would probably work better and support crosscompiles
2026-03-27 13:00:30 <ikke> And no exec in global context
2026-03-27 13:00:42 <ikke> But one thing you will run into is checksums
2026-03-27 13:02:27 <hazardaj_nombroj> would it be possible to download the source file into program-$CARCH.tar.gz, then tell it to unpack that one file specifically into the source directory, and supply checksums to all the possible program-(arch) variants in the APKBUILD?
2026-03-27 13:02:36 <hazardaj_nombroj> iunno, haven't made one of these in a long time
2026-03-27 13:04:07 <lil5[m]> $CARCH works thanks
2026-03-27 15:25:53 <infinitywisdom[m]> Ok I need help please. I just had my laptop run the shutoff sequence on its own. Then when I tried to login, it logged in successfully, but then went into suspend mode immediately. How the heck do I figure out what is going on? This has never happened before.
2026-03-27 15:32:53 <infinitywisdom[m]> False Alarm!!! The battery was dead, GNOME didn't warn me, and the power chord was dangling from the wall not fully plugged in lmao!!!
2026-03-27 16:09:43 <dogg0> llm suggesting to use sudo in alpine, yeesh
2026-03-27 16:10:51 <lopid> llm's, eh?
2026-03-27 16:10:55 <dogg0> thought it was worth +1 funny, but tbh its just cringe 
2026-03-27 16:11:07 <lopid> cringeworthy
2026-03-27 16:47:45 <plavielle> hi, i'm back
2026-03-27 16:48:27 <plavielle> so for those that not here 5 hours ago, alpine services doesnt work on hyperv
2026-03-27 16:49:23 <plavielle> quinq said to me to install syslog, i think to have more information on what goes wrong
2026-03-27 16:49:56 <plavielle> but what do i have to do to "trace" why services doesnt start
2026-03-27 16:53:10 <plavielle> services setup after the first reboot doesnt start (docker / lighttpd and surely a tons of other)
2026-03-27 16:56:12 <lopid> nothing logged? on vt1?
2026-03-27 16:56:44 <plavielle> nothing on logs
2026-03-27 16:57:01 <plavielle> /var/log/message
2026-03-27 16:57:05 <lopid> vt1?
2026-03-27 16:57:41 <plavielle> ?? vt1 = console ????
2026-03-27 16:57:46 <lopid> yes
2026-03-27 16:58:13 <plavielle> when the virtual machine start, all the services setup on setup-alpine works
2026-03-27 16:58:37 <plavielle> but all configured with rc-update add XXXXXX doesnt start
2026-03-27 16:59:15 <lopid> do they not even appear there?
2026-03-27 16:59:22 <plavielle> nothing work in default or boot profile
2026-03-27 16:59:29 <lopid> nothing?
2026-03-27 17:00:19 <plavielle> no they doesnt appear, and if i try to reconfigure it (rc-update again) it say that the service is already "setup"
2026-03-27 17:00:22 <ikke> plavielle: what is the output of rc-status --all
2026-03-27 17:00:29 <ikke> (use a paste service, don't paste in here)
2026-03-27 17:00:29 <lopid> does `rc-update` show them?
2026-03-27 17:02:51 <plavielle> i use alpine on raspberry and no problem, but on hyperv nothing works
2026-03-27 17:03:04 <plavielle> 2 sec i made some test
2026-03-27 17:03:24 <lopid> "nothing works" could mean you didn't even start the vm
2026-03-27 17:04:05 <plavielle> the vm start well
2026-03-27 17:04:27 <plavielle> but none of the servies configured after the setup start
2026-03-27 17:04:37 <plavielle> rc-update says to me
2026-03-27 17:04:55 <plavielle>  rc-update: lighttpd already installed in runlevel `default'; skipping
2026-03-27 17:05:32 <plavielle> but with "rc-status --all" no lighthttpd 
2026-03-27 17:05:59 <lopid> `rc-update`?
2026-03-27 17:06:43 <plavielle> rc-update show me lighthttpd in default
2026-03-27 17:07:01 <plavielle> but nothing in rc-status --all
2026-03-27 17:07:41 <plavielle> but lighttpd not in rc-status --all list
2026-03-27 17:07:57 <lopid> does it start manually, with `rc-service … start`?
2026-03-27 17:08:15 <plavielle> yes
2026-03-27 17:08:23 <plavielle> manually it's ok
2026-03-27 17:09:37 <lopid> until you find the reason, you could start it from the local service
2026-03-27 17:11:23 <plavielle> what do you mean "the local service" ???? Manualy via ssh console
2026-03-27 17:11:37 <lopid> /etc/local.d/README
2026-03-27 17:12:59 <plavielle> okay, so a shell script that launch an rc-service XXXXX start
2026-03-27 17:13:05 <lopid> yes
2026-03-27 17:13:30 <plavielle> okay, i'm going to try. Do i have to put an 1 
2026-03-27 17:13:42 <plavielle> okay, i'm going to try. Do i have to put an & or && at he end ??
2026-03-27 17:13:52 <lopid> not that i recall
2026-03-27 17:16:11 <plavielle> okay thats a good solution
2026-03-27 17:16:45 <plavielle> but whats the procedure to report that bug to alpine developer/community ????
2026-03-27 17:17:52 <lopid> there is not much to report, that we've seen here. maybe wait a while for more people
2026-03-27 17:21:13 <plavielle> is there a time period in the journey or a day there is more people ????
2026-03-27 17:22:02 <lopid> nothing sticks out
2026-03-27 17:27:57 <plavielle> i'm trying with local.d
2026-03-27 17:31:16 <plavielle> do i have to execute rc-update add local default ?????
2026-03-27 17:32:33 <lopid> maybe. i don't see it there already
2026-03-27 17:33:20 <lopid> yes - https://wiki.gentoo.org/wiki//etc/local.d
2026-03-27 17:33:22 <plavielle> without rc-update add local default doesnt work
2026-03-27 17:33:31 <plavielle> with rc-update add local default it's ok
2026-03-27 17:33:39 <plavielle> thanks for your help
2026-03-27 17:33:45 <lopid> you're welcome
2026-03-27 17:34:00 <plavielle> and sorry for my bad english
2026-03-27 17:34:16 <lopid> it is good enough :)
2026-03-27 17:36:14 <plavielle> so now i'm going to see if i can a bug report on alpine web site
2026-03-27 17:36:59 <plavielle> i'm going to test on raspberry pi and normal computer this weekend to be sure that the bug is only on hyperv
2026-03-27 17:37:32 <lopid> may be worth trying a different alpine image or different hyperv
2026-03-27 17:37:40 <lopid> to narrow it down
2026-03-27 17:40:03 <plavielle> i'v tested on many hyperv and hardware allways the same behaviour
2026-03-27 17:40:23 <plavielle> virt iso x86-64 doesnt work
2026-03-27 17:40:29 <p_6f3Ik7Suw> plavielle: after doing rc-update on a service, do you actually reboot the vm?
2026-03-27 17:40:37 <lopid> heh
2026-03-27 17:40:57 <plavielle> i'm going to test all normal iso and virt X86 iso
2026-03-27 17:41:05 <plavielle> but tomorrow or sunday
2026-03-27 17:41:10 <p_6f3Ik7Suw> because rc-update afaik will not start the service immediately, only marks it to be started after reboot
2026-03-27 17:41:24 <plavielle> thanks lopid, and nice to talk with you
2026-03-27 17:42:23 <ichigo[m]> Yes but you can do rc-service to start it without a reboot
2026-03-27 17:42:27 <plavielle> p_6 i'm not a noog, even after reboot doesnt work
2026-03-27 17:43:26 <plavielle> the solution of lopid work = put a script in /etc/local.d and rc-update add local default
2026-03-27 17:46:57 <lopid> workaround
2026-03-27 17:58:15 <plavielle> ichigo yes but for a production VM it's not ideal to start services manually
2026-03-27 17:58:38 <plavielle> thanks all, i have to go to sport there's a life after linux
2026-03-27 18:01:46 <plavielle> i'm an it manager and i'm going to replace all my debian by alpine
2026-03-27 18:02:07 <lopid> :)
2026-03-27 18:02:10 <plavielle> more light and efficient
2026-03-28 01:21:18 <fission> hmm services don't start on hyper-v does sound strange. I wonder if I have a way to actually run it here
2026-03-28 03:14:00 <eureka> sweetest dearest alpine linux i love you so
2026-03-28 03:17:09 <Alice> alpine greatest
2026-03-28 04:16:16 <Saijin_Naib[m]> Agreed
2026-03-28 05:23:00 <fission> haha ok so I installed hyper-v and put a VM on it but uhhh... can't actually connect to it because there's no UI. I thought surely it would have serial support but... no...
2026-03-28 05:42:03 <fission> (ah, I lied, I just don't know how to hyper-v or windows anymore)
2026-03-28 05:52:33 <rnkn> does busybox less not do colour?
2026-03-28 05:58:45 <Saijin_Naib[m]> @rnkn not yet, no. I believe Sertonix mentioned they were thinking about adding it
2026-03-28 05:59:11 <Saijin_Naib[m]> Doesnt do any other formatting, either, like bold, sadly
2026-03-28 06:01:22 <rnkn> Saijin_Naib[m]: so you would say it is LESS functional?
2026-03-28 06:01:41 <Saijin_Naib[m]> Wait, isnt more less?
2026-03-28 06:01:44 <Saijin_Naib[m]> Haha
2026-03-28 06:02:13 <Saijin_Naib[m]> Yeah, it lacks a bit, but feels cozy. I use it and bb vi
2026-03-28 06:02:40 <Saijin_Naib[m]> There are commandline options you can invoke it with to tweak a few minor things
2026-03-28 06:29:26 <rnkn> I like colours; I installed real less
2026-03-28 16:45:03 <realroot[m]> Can I make apk avoid to update a package?
2026-03-28 16:59:07 <lanky_ginger> realroot[m]: I think package pinning might do what you want. There's a section in the wiki on it under apk
2026-03-28 18:24:24 <telmich> I'm just trying to update an older edge based system, but I am stuck at this:
2026-03-28 18:24:35 <telmich>     breaks: py3-maturin-1.11.5-r1[python3~3.14] py3-semantic-version-2.10.0-r5[python3~3.14]
2026-03-28 18:24:35 <telmich> ERROR: unable to select packages:
2026-03-28 18:24:35 <telmich>   python3-3.12.11-r0:
2026-03-28 18:24:35 <telmich>             py3-setuptools-rust-1.12.0-r2[python3~3.14]
2026-03-28 18:24:35 <telmich>     satisfies: world[python3] ansible-13.4.0-r0[python3] ansible-13.4.0-r0[python3~3.12]
2026-03-28 18:24:36 <telmich> (many more lines)
2026-03-28 18:24:36 <telmich>                ansible-core-2.20.3-r0[python3] ansible-core-2.20.3-r0[python3~3.12]
2026-03-28 18:24:46 <telmich> I tried to apk del the first three packages py3-maturin py3-semantic-version py3-setuptools-rust but even after del was successful, the error stays the same on upgrade
2026-03-28 18:30:08 <dwfreed> edge is currently in the process of rebuilding everything for python 3.14
2026-03-28 18:30:14 <dwfreed> ansible has not yet been rebuilt
2026-03-28 19:46:04 <invoked> a lot of people will run into this when using --latest, but it will probabl go fine without.
2026-03-28 22:10:56 <ellyqw> well, it doesn't
2026-03-28 22:11:34 <ellyqw> i'm unable to update my systems or build postmarketOS image because not everything's been rebuilt for python3-3.14
2026-03-29 09:32:28 <rnkn> when I run `swayidle timeout 10 "swaylock -f -c 000000; wlopm --off \*"` I get as expected, the screen powers off and swaylock is enabled, but when I put this in .config/swayidle/config and run with swayidle-openrc only swaylock is enabled, no screen power off
2026-03-29 09:48:05 <rnkn> also the yambar-openrc runs yambar in a way that wants to connect to X (when I'm running wayland) so it exists repeatedly; how can I get it to know it's running under wayland?
2026-03-29 10:00:07 <rnkn> it appears the openrc isn't inheriting WAYLAND_DISPLAY from my environment
2026-03-29 10:07:25 <rnkn> okay cool, the yambar WAYLAND_DISPLAY problem is solved.. now back to the swayidle problem
2026-03-29 10:22:18 <rnkn> okay that's solved now too
2026-03-29 10:22:22 <rnkn> back to work everyone!
2026-03-29 11:42:45 <hazardaj_nombroj> huh, interesting, since the python 3.14 update it seems like it takes less time for the gnome session to shut down
2026-03-29 15:37:02 <plavielle> hi
2026-03-29 15:37:32 <plavielle> is there any alpine dev or expert here ??????
2026-03-29 15:38:24 <Lilith> of course not !!!!!!
2026-03-29 15:38:30 <plavielle> ;-)
2026-03-29 15:38:53 <plavielle> because i have a question regarding a bug
2026-03-29 15:39:11 <quinq> So much suspens
2026-03-29 15:39:39 <Lilith> use bug spray or fly swatter, if you're in australia then just run
2026-03-29 15:40:21 <plavielle> amazing but luckily i'm not ins australia
2026-03-29 15:40:47 <quinq> Lilith ^^
2026-03-29 15:41:12 <Lilith> :P
2026-03-29 15:41:40 <plavielle> like i said yesterday under hyperv service setup AFTER "setup-alpine" doesnt start
2026-03-29 15:41:52 <plavielle> every services
2026-03-29 15:42:42 <plavielle> so lopid give me a "workaround" but if someone can help me to "debug" i thank him a lot
2026-03-29 15:43:46 <quinq> What whark-around?
2026-03-29 15:43:55 <lopid> local.d
2026-03-29 15:45:53 <plavielle> it's like openrc (rc-update) is broken
2026-03-29 16:17:05 <jpp1> plavielle: Alpine Linux : Découverte de l'installation:
2026-03-29 16:17:18 <jpp1> https://www.youtube.com/watch?v=Q7jjAPG5RSQ
2026-03-29 16:18:03 <plavielle> merci jpp1 mais ça fait longtemps que j'ai dépassé le stade de l'installation
2026-03-29 16:18:13 <plavielle> ou celui du noob
2026-03-29 16:19:13 <plavielle> le bug survient uniquement sous hyperv, a priori pas sous rapsberry et pc normal
2026-03-29 16:19:28 <plavielle> mais merci pour l'aide
2026-03-29 16:21:55 <jpp1> setup-alpine This is for the installation
2026-03-29 16:22:20 <jpp1> setup-alpine c'est pour l'installation
2026-03-29 16:24:13 <plavielle> jpp1 merci mais je sais lire un tuto. Je viens de finir une image raspberry personalisé de client leger pour windows server rds 2025, donc je ne debute pas
2026-03-29 16:24:32 <plavielle> so thanks all it's beer time
2026-03-29 16:38:01 <passingbyuser> how I make /dev/fd symlinked to /proc/self/fd at boot? bash scripts don't work properly without that
2026-03-29 16:39:49 <quinq> I suppose you mount proc end devtmpfs
2026-03-29 16:43:16 <passingbyuser> it mounts ok devtmpfs on /dev and proc on /proc, just no symlink
2026-03-29 16:50:23 <quinq> passingbyuser, are you using openrc?
2026-03-29 16:50:26 <quinq> How did ou install?
2026-03-29 16:50:52 <quinq> The devfs init should do that for you (part of the sysinit runlevel)
2026-03-29 16:52:22 <passingbyuser> yeah, i ask here because it's rather strange. it's custom alpine image provided by VPS hoster
2026-03-29 16:53:38 <quinq> oh
2026-03-29 16:53:48 <quinq> Then yeah, you'd have to ask them
2026-03-29 16:57:26 <passingbyuser> it looks like devfs service is not enabled
2026-03-29 16:58:40 <passingbyuser> is it ok that /dev is mounted tho?
2026-03-29 16:58:47 <passingbyuser> without that service
2026-03-29 17:03:51 <quinq> $ rc-status sysinit
2026-03-29 17:17:36 <g1> passingbyuser: /etc/init.d/devfs creates the symlink.  check if you have it enabled in sysinit as quinq suggested
2026-03-29 19:42:24 <mlody> https://pastebin.com/aSAvg7du
2026-03-29 19:42:28 <mlody> What might be wrong?
2026-03-29 19:42:31 <mlody> Can't install pipx
2026-03-29 19:44:32 <mlody> Package is broken?
2026-03-29 19:44:41 <ikke> mlody: python3 has been upgraded to 3.14 and many packages need to be rebuilt, which takes time
2026-03-29 19:56:12 <mlody> ikke: Thank you.
2026-03-29 20:58:32 <mlody> And pipewire don't working, like always.
2026-03-29 20:58:39 <mlody> I wonder what is wrong.
2026-03-29 20:58:53 <mlody> I just logged out, logged in and it stopped working
2026-03-29 21:04:47 <mlody> Right now I fixed problem, but I think that problems like this might repeat.
2026-03-29 21:32:52 <mlody> Any way to install man pages?
2026-03-29 21:34:07 <p_6f3Ik7Suw>  apk add man-pages
2026-03-29 22:38:23 <fission> just fwiw I tried alpine+hyper-v and it worked fine for me (maybe not helpful to that person who was in here but :)
2026-03-30 05:27:56 <icedmirna> sound works 
2026-03-30 05:29:40 <icedmirna> nethogs unknown root connection very interesting
2026-03-30 05:29:44 <icedmirna> ? root     10.0.2.15:34522-142.251.10.94:80 
2026-03-30 11:11:46 <quinq> Hi
2026-03-30 11:11:55 <quinq> I have two zfs pools in this laptop
2026-03-30 11:12:19 <quinq> A “boot” pool, with a boot fs,
2026-03-30 11:12:25 <quinq> and a “root” pool, with the system
2026-03-30 11:12:32 <quinq> It's FDE too
2026-03-30 11:12:39 <quinq> It boots fine
2026-03-30 11:12:57 <quinq> But the system is *never* able to mount the boot pool at boot for some reason
2026-03-30 11:13:29 <quinq> It just says “filesystem 'bpool/fs/root' cnnot be mounted, unable to open the dataset”
2026-03-30 11:13:39 <quinq> It works fine then when I login and import the pool manually
2026-03-30 11:13:55 <quinq> Any idea why, and more importantly how to fix it?
2026-03-30 11:18:24 <lopid> you just have to import it manually, not decrypt it?
2026-03-30 11:19:23 <quinq> Yeah, just import
2026-03-30 11:19:33 <quinq> zpool import -a
2026-03-30 11:20:05 <quinq> I tried to set the VERBOSE option in /etc/conf.d/zfs, but no more info about it
2026-03-30 11:20:26 <quinq> I could see that there's no cache file in the initrd, but I don't expect it to be mounted then
2026-03-30 11:25:36 <lopid> void has it mounted via a runit service. i don't know if alpine has the same…
2026-03-30 11:32:11 <quinq> Well, the root is obviously mounted in initrd
2026-03-30 11:32:39 <quinq> but yeah there's an rc service that I do have enabled (zfs-import)
2026-03-30 14:12:38 <staceee> gitlab seems to be suffering :S
2026-03-30 14:13:24 <infinitywisdom[m]> Are there any plans to modify things for GNOME's recent developments for more tightly integrating with systemd?
2026-03-30 14:15:11 <infinitywisdom[m]> Just wondering if we are freezing the version, modifying it to remove the systemd components, or if Alpine Linux is going to drop support for GNOME all together with their latest version. The GNOME team has their heads very far up their own asses and pretty much view people who don't use systemd as irrelevant and we're on our own.
2026-03-30 14:16:57 <infinitywisdom[m]> Nevermind that we have better security, privacy, and performance without systemd. Really echos the overall cancel culture vibe they give for anyone who doesn't agree with their choices.
2026-03-30 14:35:59 <ssm_> don't use gnome?
2026-03-30 14:36:57 <infinitywisdom[m]> 
ACTION uploaded an image: (2824KiB) < https://matrix.org/oftc/media/v1/media/download/ATUcN3VjprzReHPb3541QbolMOsYq_hCL6lV0ci3y_uCqsW7gbLxFvIx65oeDHoGqr6BohiBhgo1gB_4tCU5-mZCedh0YxTwAG1hdHJpeC5vcmcvTHNkUVBBbUxTQXBpRmN4bnpBQWZEbHZ0 >

2026-03-30 14:37:08 <longnoserob[m]> gnome 50 is wayland only, that might be another show-stopper
2026-03-30 14:37:37 <infinitywisdom[m]> longnoserob[m]: Why?
2026-03-30 14:39:01 <longnoserob[m]> if your hardware is not supported by wayland
2026-03-30 14:40:10 <hazardaj_nombroj> alpine does have some parts of gnome frozen at earlier versions, with patches and such backported, but most of gnome 50 is there
2026-03-30 14:41:55 <PureTryOut> Please keep the hate towards their views out of here, we're better than that.
2026-03-30 14:42:14 <PureTryOut> As for what will and will not be supported in Alpine, you'll have to ask the Alpine maintainer of GNOME. I suppose achill is one of them?
2026-03-30 14:44:37 <infinitywisdom[m]> PureTryOut: Yeah, it is a bad habbit fueled by nothing, but negative interactions with their community. Hard to keep in check. Quite frankly I only use gnome for 1 reason. An extension called Forge which enables the smoothest automatic window tiling experience that has become essential for my workflow. 
2026-03-30 14:44:53 <achill> yes, so the current state is gnome 50 with Gnome-Session/gdm 48
2026-03-30 14:45:02 <achill> currently there is a openrc leader by a Gentoo dev in development which looks promising to be used in alpine
2026-03-30 14:45:54 <achill> but in general, step up and implement missing parts for openrc e.g. a varlink user api, if you want a long term forward for gnome in alpine
2026-03-30 14:45:57 <infinitywisdom[m]> achill: So there is hope.
2026-03-30 14:46:55 <achill> you can be the hope! currently it mainly lacks manpower (or human power) interested in bringing the openrc/musl stuff to the same level of systemd/glibx
2026-03-30 14:47:26 <puretryout[m]1> The problem with these kind of things is that there are a lot of people complaining but nobody really stepping up to develop alternatives to make it work without systemd.
2026-03-30 14:47:33 <achill> exactly
2026-03-30 14:47:43 <infinitywisdom[m]> It is beyond my capabilities unfortunately.
2026-03-30 14:48:01 <infinitywisdom[m]> How does the saying go. It's above my paygrade.
2026-03-30 14:48:49 <achill> that thing doesn't exist in open source development anyway
2026-03-30 14:49:03 <achill> either someone volunteers to do something or not
2026-03-30 14:49:18 <infinitywisdom[m]> I do have some stuff I have been working on that I will be making available to Alpine Linux for a project that has to do with improving the installer.
2026-03-30 14:49:36 <achill> but hating gnome devs is the last thing one should do. they are the ones driving the ecosystem farward
2026-03-30 14:50:43 <infinitywisdom[m]> achill: In theory it makes sense. In practice my interactions have been nothing but negative with their community members. Sometimes you just don't get along well with people and personalities clash. 
2026-03-30 14:51:01 <infinitywisdom[m]> I feel like kind of a hipocrite for using GNOME.
2026-03-30 14:51:18 <achill> still no reason to bring hate to them
2026-03-30 14:51:21 <WhyNotHugo> achill: gnome drivng the ecosystem forwards? forward toward a cliff you mean, right?
2026-03-30 14:52:52 <puretryout[m]1> When I don't like an ecosystem in such a way I just move to something else 🤷
2026-03-30 14:53:05 <achill> I know you're joking but liking the direction of a project is very subjective
2026-03-30 14:53:08 <infinitywisdom[m]> puretryout[m]1: You are not wrong. 
2026-03-30 14:53:26 <aron> puretryout: exactly
2026-03-30 14:54:11 <aron> there are numerous DEs available in alpine, one can choose another one at any time
2026-03-30 14:54:12 <achill> so, where in Heerlen is a good place to sit to continue hacking
2026-03-30 14:54:19 <achill> urgh
2026-03-30 14:54:23 <infinitywisdom[m]> Frankly I would like to eventually be able afford to track down the developer of the GNOME forge extension. Pay them to port it over Cinnamon. Then make sure to get it over to Alpine Linux.
2026-03-30 14:54:33 <puretryout[m]1> achill: I hear Aachen is quite nice and near
2026-03-30 14:55:27 <WhyNotHugo> Why are there so many hackers in Aache? XD
2026-03-30 14:55:33 <infinitywisdom[m]> At least I have no beef with the cinnamon developers and I prefer that desktop environment.
2026-03-30 14:56:47 <infinitywisdom[m]> aron: What is the current list? It's been about a year since I last checked.
2026-03-30 14:58:14 <aron> https://wiki.alpinelinux.org/wiki/Desktop_environments_and_Window_managers seems to be up-to-date
2026-03-30 14:59:22 <infinitywisdom[m]> aron: I am fond of LXQT, XFCE, & MATE. 
2026-03-30 14:59:52 <infinitywisdom[m]> What would be required to port Cinnamon over to Alpine Linux?
2026-03-30 15:00:13 <aron> xfce is my choice :)
2026-03-30 15:01:34 <infinitywisdom[m]> I feel like I am opening up a can of worms here that may bring me headaches, but also at the same time. If I am going to pay a developer to port Forge to cinnamon. I am going to want it to work on my distro of choice lol.
2026-03-30 15:03:16 <puretryout[m]1> infinitywisdom[m]: Someone to package it. And hey, that someone could be you!
2026-03-30 15:03:34 <aron> I don't know much about the DE porting, only to package some software in alpine, but if you lucky and cinnamon is portable, then I think it is just an APKBUILD file
2026-03-30 15:04:27 <aron> creating an APKBUILD for regular software is piece of cake
2026-03-30 15:05:19 <aron> if you are less lucky, you need to patch cinnamon, which requires some coding skills
2026-03-30 15:06:37 <puretryout[m]1> Any DE exists out of multiple components, they'll have to all be packaged and made to depend on each other in the right way. But that's the same for any other piece of software we package.
2026-03-30 15:08:33 <infinitywisdom[m]> <puretryout[m]1> "Someone to package it. And hey..." <- Feel like this should be on an Alpine Linux dev recruitment poster. 
2026-03-30 15:08:52 <aron> right, multiple packages
2026-03-30 15:09:58 <infinitywisdom[m]> All I can say is. Let's see where the road takes me. This sounds like something I am definitely interested in, but it would require significantly more free time. Maybe around July.
2026-03-30 15:13:42 <aron> my method is to lower my standards, and rely on less bloatware; at first it will feel weird, but after you get used to it, it is just freedom, and you waste less computer resources
2026-03-30 15:17:12 <aron> I looked after forge extension you mentioned, and it seems it is abandoned
2026-03-30 15:17:49 <aron> or to-be-abandoned
2026-03-30 15:18:02 <infinitywisdom[m]> I know. I tried tracking down the developer once. No luck, but i want to try again.
2026-03-30 15:19:12 <infinitywisdom[m]> Either that or pay another dev.
2026-03-30 15:20:02 <aron> it has a hefty dependency tho: "Install NodeJS 16+"
2026-03-30 15:20:05 <infinitywisdom[m]> Maybe someone on the cinnamon dev team looking for some extra cash.
2026-03-30 15:20:47 <infinitywisdom[m]> aron: Indeed. But it is worth it. 
2026-03-30 15:21:32 <aron> I don't know the details, but there is a possibility that forge cannot be implemented on cinnamon
2026-03-30 15:22:26 <infinitywisdom[m]> I won't know if it can be or not. Maybe MATE? Don't know what is possible, because I'm not familiar with how everything works under the hood.
2026-03-30 15:22:45 <infinitywisdom[m]> All I know is that I am hoping one of the GNOME forks can handle it.
2026-03-30 15:23:10 <infinitywisdom[m]> Either that, or eventually give up my forge addiction for something else.
2026-03-30 15:23:36 <infinitywisdom[m]> That is really all it is at this point. An addiction to a certain type of workflow that drastically speeds up my performance.
2026-03-30 15:24:24 <aron> well, I wish good luck for you with that project, maybe someone can help you
2026-03-30 15:28:53 <infinitywisdom[m]> That would be absolutely amazing.
2026-03-30 15:30:32 <infinitywisdom[m]> Who knows what will happen. I've got a lot on my plate for now, but I know at some point I will have to pick up my project that has to do with Alpine Linux and complete it. Maybe when that happens I will have enough of a fire under my butt to port forge over to cinnamon and then cinnamon over to alpine. Finishing it off with contributing my code upstream to help improve the community.
2026-03-30 15:39:29 <omni> cinnamon-desktop exists in aports but is currently unmaintained
2026-03-30 16:07:59 <infinitywisdom[m]> omni: How behind is it and how do you install it?
2026-03-30 16:09:27 <omni> apk add cinnamon-desktop, I guess
2026-03-30 16:10:10 <achill> cinnamon-desktop is just a library for common components
2026-03-30 16:10:26 <achill> the actual desktop environment is not packaged
2026-03-30 16:10:29 <omni> ok, idk then
2026-03-30 16:11:07 <achill> but i might also package the desktop environment if im bored enough on the train on wednesday
2026-03-30 16:11:38 <achill> or probably im going to be bored on wednesday but also too tired for it
2026-03-30 16:11:39 <achill> idk
2026-03-30 16:11:42 <achill> we'll see
2026-03-30 16:11:44 <infinitywisdom[m]> achill: Never thought I'd say this. I sincerely hope you are bored this Wednesday.
2026-03-30 16:12:56 <infinitywisdom[m]> Make my life this summer 10x easier if cinnamon was packaged, maintained, and gracefully installed on Alpine Linux.
2026-03-30 16:15:07 <infinitywisdom[m]> If I was a millionare, or had a couple hundred grand in the bank. I would totally fork cinnamon and forge. Make a new desktop environment from it.
2026-03-30 16:20:25 <infinitywisdom[m]> Hire a dev team. Make all the changes I'd love to have implemented. Create my idea of the best desktop for Linux.
2026-03-30 16:22:00 <invoked> alternatively, you could use/support things that don't care about you
2026-03-30 16:22:11 <invoked> or rather, *do* care about you
2026-03-30 16:22:20 <invoked> gnome doesn't apparently care about collateral damage.
2026-03-30 16:22:32 <infinitywisdom[m]> You are not wrong.
2026-03-30 16:23:18 <infinitywisdom[m]> GNOME in general hates my guts from a community standpoint.
2026-03-30 16:23:34 <achill> invoked: stop this fucking hate
2026-03-30 16:23:42 <achill> seriously
2026-03-30 16:24:10 <achill> youre making me really uncomfortable to chat here, and this is crap certainly not welcome here
2026-03-30 16:24:11 <invoked> it's not hate. that's objectively the situation
2026-03-30 16:24:14 <achill> no
2026-03-30 16:24:51 <invoked> yes.
2026-03-30 16:25:04 <invoked> unless we have a massive difference of opinion what "hate" means.
2026-03-30 16:25:58 <infinitywisdom[m]> Im actually kind of motivated to get off GNOME now in a couple of weeks.
2026-03-30 16:26:00 <infinitywisdom[m]> lol
2026-03-30 16:26:39 <achill> invoked: we all got your opinion on that, anything further is just annoying
2026-03-30 16:26:43 <infinitywisdom[m]> My biggest concern is did XFCE on Alpine have the issue fixed for missing components from the installer.
2026-03-30 16:27:04 <invoked> achill: i don't want to make anybody uncomfortable about anything. i was replying to infinitywisdom[m] that alternatives exist. not much more than that.
2026-03-30 16:27:20 <achill> and we all got that hours ago
2026-03-30 16:27:21 <invoked> definitely not hate.
2026-03-30 16:27:38 <invoked> not from me you didn't.
2026-03-30 16:28:26 <infinitywisdom[m]> Im also curious about what MATE is like nowadays. Which one provides the better experience on Alpine Linux?
2026-03-30 16:33:08 <invoked> infinitywisdom[m]: ux is subjective, you'll just have to try them out
2026-03-30 16:36:03 <infinitywisdom[m]> I am more talking about stability and it just working without having to troubleshoot issues upon a fresh install. So far on Alpine Linux. I have been able to try out KDE, GNOME, and XFCE. KDE missed key components and so did XFCE. Tried to resolve and troubleshoot the issues, but never got them to work. With GNOME I experienced similar issues, but I did wind up creating some modifications in the form of a bash script that
2026-03-30 16:36:03 <infinitywisdom[m]> basically makes installing GNOME and having everything just work on a fresh install possible without having to spend time troubleshooting.
2026-03-30 16:39:22 <invoked> having alpine be some turnkey desktop might be a long-term goal, but i'm not sure we're there with anything yet
2026-03-30 16:40:49 <infinitywisdom[m]> invoked: From questions I have asked in the past. That is not a priority. 
2026-03-30 16:40:53 <infinitywisdom[m]> Or even a goal.
2026-03-30 16:41:07 <infinitywisdom[m]> Outside the scope of the project.
2026-03-30 16:41:32 <infinitywisdom[m]> I do think that it would be awesome to have a clean and polished Alpine Linux desktop experience.
2026-03-30 16:43:01 <invoked> iirc at least ariadne has mentioned that stuff is in scope. there's just a lot to do
2026-03-30 16:44:46 <infinitywisdom[m]> News to me. It would be an awesome endeavor. Alpine Linux once you jump through a lot of flaming hoops. Does make for a great desktop experience.
2026-03-30 16:45:43 <infinitywisdom[m]> It's been about a year, or so since I have really kept up to date, or interacted with the community with what has been going on.
2026-03-30 16:46:02 <invoked> if you look on gitlab, go through the tsc stuff. at least some of it is there
2026-03-30 16:46:47 <invoked> so, there's a lot to do. governance and team stuff, besides the discussions/decisions that need to be made
2026-03-30 16:49:06 <infinitywisdom[m]> invoked: Where?
2026-03-30 16:49:20 <invoked> gitlab.alpinelinux.org
2026-03-30 16:49:39 <infinitywisdom[m]> No I got that, I meant where specifically.
2026-03-30 16:49:54 <infinitywisdom[m]> I the TSC section is not updated freuently.
2026-03-30 16:50:03 <infinitywisdom[m]> s/freuently/frequently/
2026-03-30 16:51:07 <infinitywisdom[m]> Actually this entire gitlab does not seem to have frequent updates to it?
2026-03-30 16:51:52 <infinitywisdom[m]> Wait nevermind. I was reading it wrong.
2026-03-30 16:52:08 <invoked> https://gitlab.alpinelinux.org/alpine/tsc/-/issues
2026-03-30 16:52:35 <invoked> kind of on hiatus at the moment
2026-03-30 16:53:16 <invoked> anyway, where the project is going is decided there
2026-03-30 16:56:35 <infinitywisdom[m]> invoked: I can't find where this is mentioned. 
2026-03-30 16:56:41 <puretryout[m]1> Pretty sure making Alpine work as a desktop distro out of the box without any configuring by the user themselves is not in scope and you'll have to resort to downstream projects like postmarketOS for that
2026-03-30 16:57:16 <infinitywisdom[m]> puretryout[m]1: That's what I thought and was told in the past. 
2026-03-30 16:58:41 <infinitywisdom[m]> Like if we had an Endeavor OS, or Cachy OS equivilent for Alpine Linux like is done with Arch Linux. Making it graceful and easy for users to get started. That would be an awesome project. I would however just prefer Alpine Linux itself to have that focus, or at least form an internal team to run up that kind of project.
2026-03-30 16:59:19 <infinitywisdom[m]> I mean we already have multiple installers and isos. What's one more to add to the mix lol.
2026-03-30 17:02:48 <infinitywisdom[m]> How many of us would it take to do that sort of task? Does anyone have an interest in forming such a team and working together towards making a clean and polished ISO for Alpine Linux with 1, or more desktops?
2026-03-30 17:05:53 <invoked> unknown, but it's always more than people think
2026-03-30 17:06:38 <invoked> anyone can start a new linux distro. many have. few actually keep them going
2026-03-30 17:06:51 <infinitywisdom[m]> 3 1/2 men + 2 1/8th women
2026-03-30 17:07:17 <infinitywisdom[m]> That's all we need.
2026-03-30 17:07:44 <mlody> Maybe I'm interested. I don't feel need to making it as we have a lot of distro like this, but it might be interesting.
2026-03-30 17:07:51 <infinitywisdom[m]> Yeah. Creating my own linux distro has been a pipe dream of mine for about 15 years.
2026-03-30 17:08:07 <mlody> But to be honest I don't want to make distro based on Alpine.
2026-03-30 17:08:34 <infinitywisdom[m]> I'd rather it just be alpine linux. It is a good distro with a solid track record.
2026-03-30 17:09:52 <invoked> just don't expect anyone to hold your hand in making a new linux. the assumption is you know what you're doing
2026-03-30 17:11:19 <infinitywisdom[m]> I have gone down that road many times. I even got so close as to have a development team which left me high and dry with a prototype that was non-functioning. AFTER I flew out to washington state and had to cancel a meeting with a computer case manufacturer who was going to partner with us to make a new computer company.
2026-03-30 17:12:01 <infinitywisdom[m]> That was based on Arch Linux. After that the idea of making my own distro kind of left a sour taste in my mouth. The dream persists, but the desire is tainted.
2026-03-30 17:12:12 <invoked> anyway, i'm not the cops, but this is offtopic for a help channel
2026-03-30 17:12:23 <invoked> -> #alpine-offtopic
2026-03-30 17:12:25 <infinitywisdom[m]> This is a help channel?
2026-03-30 17:12:43 <invoked> primarily
2026-03-30 17:13:31 <invoked> caveat, i don't know if that channel is bridged or not
2026-03-30 17:13:35 <invoked> i don't use matrix.
2026-03-30 17:14:33 <puretryout[m]1> <infinitywisdom[m]> "Like if we had an Endeavor OS..." <- That is basically what postmarketOS is, just contribute to that?
2026-03-30 17:15:06 <infinitywisdom[m]> puretryout[m]1: Not my cup of tea. 
2026-03-30 17:16:05 <puretryout[m]1> In what way? It's Alpine Linux focused on endusers with graphical UI's, making it easy to install and adding a bunch of hardware support (phones mostly but not exclusively)
2026-03-30 17:20:00 <infinitywisdom[m]> puretryout[m]1: Limited devices, usually older ones, and generally speaking while I appreciate the project. Sometimes you just don't feel it is for you. 
2026-03-30 17:21:13 <infinitywisdom[m]> That is just me in this case. I'm an Alpine Linux user. I like this operating system. I like what it is about. I jumped through the flaming hoops to make a good desktop experience. Wrote some software to automate that process so there are no flames on the hoops whenever I have to reinstall.
2026-03-30 17:22:49 <infinitywisdom[m]> As far as starting a new distro. Not now. Hoping, encouraging, and maybe even participating in a team to make an iso for a clean and polished Alpine Linux desktop. That is something you could probably twist my arm about to work on, because that sounds kind of fun.
2026-03-30 17:23:09 <panekj> pmOS is literally Alpine, with 10 more packages
2026-03-30 17:23:35 <panekj> like, almost all the packages are actually from Alpine
2026-03-30 17:24:08 <infinitywisdom[m]> It is still a different project. I like this one. It's not that hard to understand.
2026-03-30 17:24:18 <infinitywisdom[m]> This is a preference.
2026-03-30 17:24:30 <infinitywisdom[m]> Don't read too much into it.
2026-03-30 17:24:54 <panekj> Considering how pmOS devs are Alpine devs at the same time, this is extremely weird "preference"
2026-03-30 17:25:32 <infinitywisdom[m]> panekj: Trust me if you wanted to dive deep into my weird preferences. This is going to be a long conversation that will make your head explode. 
2026-03-30 17:25:48 <infinitywisdom[m]> I flat out don't make sense a lot.
2026-03-30 17:25:53 <panekj> I noticed
2026-03-30 17:27:03 <infinitywisdom[m]> That's ok. I don't have to make sense. In a world that has gone batshit crazy. What's more loony with weird distro preferences for reasons that make no logical sense.
2026-03-30 17:28:58 <infinitywisdom[m]> Could be I don't like a logo, I don't like how a website is laid out, I don't like the documentation, their community is morally bankrupt, or even a package that is a dependency has a name that goes against my religion. Who knows. I write off a lot of distros for loony tunes reasons. I'm ok with that, because linux is an ice cream store, and my favorite flavor is Alpine Linux.
2026-03-30 17:29:48 <infinitywisdom[m]> * crazy. What's one more loony
2026-03-30 18:27:13 <frag> in vi and D to cut line if you want to paste that on the line below you have to do: o<esc>p?
2026-03-30 18:39:03 <Sertonix[m]> <infinitywisdom[m]> "Limited devices, usually older..." <- Where is this based upon? I am relatively certain that there are "generic" devices which work on most devices that run with the lts or stable kernel paclaged by alpine.
2026-03-30 18:54:08 <realroot[m]> pmOS has different boot and there is a diff init too
2026-03-30 18:54:46 <realroot[m]> some UI won't have openrc version I think, not sure as I use sxmo
2026-03-30 19:04:12 <f_|mo> <infinitywisdom[m]> "Limited devices, usually older..." <- (just a note, that pmOS does support devices as new as like 3 years old if not newer)
2026-03-30 20:49:47 <infinitywisdom[m]> <Sertonix[m]> "Where is this based upon? I am..." <- Just must have overlooked Generic x86_64 EFI System. My bad. 
2026-03-31 00:40:54 <fission> frag: why not just 'p'? or maybe I don't understand
2026-03-31 00:43:41 <courge_musque[m]> [@_oftc_fission:matrix.org](https://matrix.to/#/@_oftc_fission:matrix.org) "o" to insert a line below, and "esc" to return to visual mode
2026-03-31 00:43:58 <courge_musque[m]> *normal mode, mb
2026-03-31 00:44:07 <courge_musque[m]> but you can also press "p" if you just want to paste right there
2026-03-31 01:20:53 <frag> fission, p will just undo the cutting, not paste it on the next line
2026-03-31 02:53:03 <fission> frag: ah, the next line
2026-03-31 03:37:02 <quinq> forget about <ESC>, just send a SIGINT with ^C
2026-03-31 03:39:49 <frag> quinq didnt know that worked.. dunno if its faster/better than esc tho?
2026-03-31 03:53:04 <quinq> It's closer to homerow for me at least
2026-03-31 03:53:51 <quinq> Otherwise I have to do a long swip back and forth with the left hand
2026-03-31 03:59:17 <frag> quinq do you not have esc at capslock???? i assume you have ctrl+c in the normal place
2026-03-31 03:59:34 <frag> and you call yourself a nerd?!
2026-03-31 04:01:07 <frag> maybe emacs user..
2026-03-31 04:10:15 <fission> heh
2026-03-31 04:16:10 <frag> in busybox vi ctrl+c also moves the cursor to the top
2026-03-31 04:47:24 <quinq> frag, no, I have Shift at caps
2026-03-31 04:47:32 <quinq> And ctrl at shift
2026-03-31 04:47:45 <quinq> Way more useful than escape at caps
2026-03-31 07:33:56 <phws> would anyone have a solution for the "clock skew detected" messages on a system without hardware clock? (pi) I'm having openrc run swclock at boot and ntpd at default; date gives the correct time but openrc refuses to run even when triggered manually. I read that people manually "touch" a list of dirs and files at boot to make this go away, but I was hoping for a more elegant
2026-03-31 07:33:58 <phws> solution, let's say
2026-03-31 07:54:49 <ikke> phws: maybe also worth asking in the #openrc channel (forgot if it was on oftc or libera)
2026-03-31 07:55:02 <phws> thanks, will do
2026-03-31 07:55:35 <phws> (libera, it seems)
2026-03-31 11:24:01 <folki> Hello. Anybody alive?
2026-03-31 11:24:25 <Sheila> nobody here but us foxes
2026-03-31 11:24:49 <realroot[m]> HI, yes i survived
2026-03-31 11:25:47 <folki> Lovely. What do I need to do to have the local pkg cache available at boot time running a diskless system of alpine?
2026-03-31 11:26:40 <folki> Pkgs added with "apk add" do not survive boot time ... but the cache directory is filled.
2026-03-31 11:28:24 <folki> I get this nasty "WARNING: opening from cache ... " message at boot time. That seems to indicate a lacking or empty cache
2026-03-31 11:28:25 <longnoserob[m]> https://wiki.alpinelinux.org/wiki/Data_Disk_Mode ?
2026-03-31 11:29:39 <folki> Thx Let me have a look ... 
2026-03-31 11:32:18 <quinq> folki, local as in network-local, as this is a diskless system?
2026-03-31 11:33:23 <folki>  No local as in "separate partition"
2026-03-31 11:33:39 <quinq> Hummm
2026-03-31 11:33:48 <quinq> A partition that would live on what?
2026-03-31 11:33:54 <folki> So I have /etc/apk/cache linked to that partion ... this is available and gets filled.
2026-03-31 11:34:09 <folki> ... I guess it is simply not available when it comes to boot.
2026-03-31 11:34:37 <quinq> File systems are mounted during boot, during the sysinit level
2026-03-31 11:34:38 <folki> The partion lives on a sdcard.
2026-03-31 11:35:54 <phws> I spent two days trying to fix this, ended up adding "apk upgrade" to an /etc/local.d/ script which seemed to work ok
2026-03-31 11:36:03 <folki> So first partion on sdcard is the alpne system ... the second partion (on that very sdcard) is for the cache
2026-03-31 11:36:36 <folki> apk upgrade, not update?
2026-03-31 11:37:11 <folki> phws, and then the package installation survived a reboot?
2026-03-31 11:37:54 <phws> disclaimer: I'm a complete amateur, but I noticed that I could manually do "apk upgrade" and it would suddenly install all the packages I expected it to add at boot
2026-03-31 11:37:57 <phws> so I automated that this way
2026-03-31 11:38:25 <folki> Same with me ... :|
2026-03-31 11:38:48 <quinq> So it works?
2026-03-31 11:39:27 <quinq> folki, maybe do the installation steps again, of setup-lbu and setup-apkcache
2026-03-31 11:39:42 <folki> No, the "apk upgrade" does the full reinstallation ... as something like "apk update python3" does ...
2026-03-31 11:39:59 <folki> quiq, have done so several times.
2026-03-31 11:40:58 <phws> I think that if you have a diskless system, it's normal that it would "reinstall" all the packages after every boot... it's just that I don't think you should have to do it manually
2026-03-31 11:41:05 <quinq> https://wiki.alpinelinux.org/wiki/Upgrading_Alpine_Linux_to_a_new_release_branch#Upgrading_diskless_and_data_disk_mode_installs_to_latest_release that too?
2026-03-31 11:41:06 <folki> This is really nasty. and following phws's idea  is somewhat "unclean".
2026-03-31 11:41:12 <Ivek> Hello to all, one quick question "After installed kde with setup-desktop all looks ok, but if I in kde click on Logout I got black screen with cursor blink, but don't see sddm. How to fix?"
2026-03-31 11:42:56 <folki> No as I am on the latest release 3.23
2026-03-31 11:44:36 <ikke> The problem most likely is that the partition is mounted too late for it to be sued
2026-03-31 11:44:43 <ikke> used*
2026-03-31 11:46:48 <folki> Hmm .. might be ... on the other hand using the first partition (that carries the ram fs) does not work either.
2026-03-31 11:47:23 <folki> Ok ... let me do some tests and come back then ...
2026-03-31 11:47:28 <folki> Thank you all!
2026-03-31 11:47:46 <folki> *wave*
2026-03-31 11:53:41 <ivy-> hello - does Alpine ppc64le work on a Raptor Talos II (non-OFW, it uses Petitboot)?  i couldn't see an obvious list of supported platforms in the docs
2026-03-31 11:54:34 <Sheila> should work, yes.
2026-03-31 11:54:46 <ivy-> thanks
2026-03-31 11:55:41 <Ivek> Any help for Kde on Alpine Linux? Logout from KDE not work as expect
2026-03-31 11:56:19 <longnoserob[m]> 
ACTION looks up talos II, reads CAPI and thinks of ISDN hw..

2026-03-31 11:56:25 <ivy-> Sheila: do you happen to know if that includes support for both BE and LE KVM guests?
2026-03-31 11:57:28 <frag> okok, will see if i get used to it
2026-03-31 11:57:29 <Sheila> as far as I know, ppc64 has always been bi-endian, as long as the firmware supports the requested endianness. this applies to virtualisation too.
2026-03-31 12:03:27 <Ivek> I'm in wrong place or need ask on another place for problems with kde on alpine linux?
2026-03-31 12:04:28 <frag> i think they allow, at least some, more general questions here yeah :)
2026-03-31 12:05:56 <frag> just dont ask to ask :>
2026-03-31 12:06:45 <Ivek> frag: I try with docs,wiki but can't find any source for help. Any communit up for help with alpine linux ? 
2026-03-31 12:07:47 <frag> yes i think this is the right channel, or maybe a kde specific channel, i doubt alpine is adding any troubles(?)
2026-03-31 12:08:13 <frag> just ask the question
2026-03-31 12:09:15 <Ivek> frag: clean install kde desktop on alpine linux, when I click on logout I got blackscreen but don't see login screen of sddm
2026-03-31 12:10:49 <frag> hmm, stick around, probably someone can help you (might take a while tho :)
2026-03-31 12:11:56 <Ivek> frag: Thanks, I not sure ppl will see my message if join after I post my problem. Need repost my question or not?
2026-03-31 12:16:35 <frag> do as you wish man, they are typically pretty relaxed here
2026-03-31 12:18:38 <phws> heh, speaking of unclean hacks... I solved my clock skew issue by adding a manual call to swclock in my /etc/local.d script that was previously just calling "apk upgrade". I suppose both might be a symptom of my /var disk being mounted too late
2026-03-31 12:35:00 <dopice> Maybe someone have some good input to share... I run a bunch of docker containers where all of them have macvlan interfaces connected to an ovs bridge. When I reboot my host the containers lack network connectivity and I suspect this is due to docker daemon and the containers start before ovs is fully up. I know I can add the dependencies directly to /etc/init.d/docker but is
2026-03-31 12:35:02 <dopice> that the correct way of doing it? Or are there any way I can add additional dependencies without touching the original init.d files? E.g. systemd (...) has a feature where you can add override files. Is there something similar on OpenRC? Or is there a recommended way of handling situations like this?
2026-03-31 12:36:28 <dopice> My "google-fu" is really bad today since I cannot find anything about this.
2026-03-31 12:41:05 <ikke> dopice: setting rc_depends in the conf.d file
2026-03-31 12:41:50 <ikke> https://manpages.org/openrc-run/8
2026-03-31 12:46:41 <dopice> Okay, so the way is to update the conf.d file? Then, if the maintainer updates the conf.d file in the package and I've made edits to it. Are those conflicts resolved or is it up to me to keep track of that? 
2026-03-31 12:52:30 <dopice> Maybe my way of thinking is wrong here. But I would like to prevent "missing out" on the great work of the maintainers.
2026-03-31 13:03:06 <hazardaj_nombroj> conf.d files as provided by packages only provide basic defaults, you're supposed to modify them
2026-03-31 13:03:50 <hazardaj_nombroj> if there's an update that replaces it, you'll instead see a /etc/conf.d/something.apk-new with the new defaults
2026-03-31 13:04:30 <hazardaj_nombroj> i think that maintainers are unlikely to put breaking changes into a conf.d file, though
2026-03-31 13:08:39 <dopice> I suppose I wouldn't miss any breaking changes. ^^ Jokes aside - I think I got my answer and I can work with this, 100%. Thanks both of you!
2026-03-31 19:16:27 <insert_name> Hi, is it possible to replace mdev with eudev on an already installed system? I'm aware there is setup-udevd script for that, and that I can choose udev as a device manager in setup-alpine.
2026-03-31 19:17:34 <insert_name> or would that cause any problems and fresh install be easier to do?
2026-03-31 19:19:13 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-devd.in#L80-90
2026-03-31 19:19:35 <ikke> You should be able to run setup-desktop at any time
2026-03-31 19:19:51 <ikke> which also installs eudev
2026-03-31 19:22:18 <insert_name> noted, although I'm more interested in just changing device manager from mdev to (e)udev for NetworkManager, as I've apparently missed that step
2026-03-31 19:42:35 <insert_name> alright, seems like it worked without a problem
2026-03-31 21:48:27 <hazardaj_nombroj> there's also a script, setup-devd