2026-04-01 08:39:49 <phws> hi all, I'm still struggling to get data mode to function. I have a fat32 partition carrying boot files and apkvol file, then an ext4 parition mounted as /var. I'm getting errors at the start of boot about not being able to access cache and swclock does not properly restore the clock. My best guess is that this is due to /var being mounted too late.
2026-04-01 08:40:31 <phws> Now the exact same thing has happened for me on vms using both qemu and hyper-v, but now also on my raspberry pi with a physical ssd. This is my fstab: https://privatebin.net/?59a55aaba95a6260#4w24hHUhr1MSmXFRib2Ljz5LthzqWthNfLryjrERkEVn . Could anyone please confirm that this is actually how you're supposed to do "data" mode, because surely I must be doing something wrong?
2026-04-01 08:41:16 <phws> (the LBU is fine by the way, if I manually run "apk upgrade" it restores all the packages and with some fiddling I can fix the clock too, after which the system functions normally for a while)
2026-04-01 08:48:52 <phws> (thing's called apkovl, oops)
2026-04-01 09:57:46 <lopid> why does lnav seem to be losing logrotated files now :/
2026-04-01 11:48:27 <phws> regarding what I said earlier: I have now moved the .apkovl file to the ext4 partition and adjusted lbu.conf accordingly. So the fat32 part now only holds boot files, while the ext4 holds .apkovl, /var and /home. Still get a warning about cache at startup but so far everything seems to work, so perhaps that's one way to get it to mount that ext4 part early enough
2026-04-01 11:50:36 <ikke> I guess because nlplug-findfs tries to find the partition with the overlay file
2026-04-01 11:51:08 <phws> whatever works, I guess :)
2026-04-01 13:31:45 <seti_> having audio issues, trying to enable pipewire with the recommended openrc based method but it just returns 
2026-04-01 13:31:46 <seti_> ~ $ doas rc-service -U pipewire start
2026-04-01 13:31:46 <seti_> XDG_RUNTIME_DIR unset.
2026-04-01 13:31:46 <seti_> doas (leo@alpine) password: 
2026-04-01 13:31:46 <seti_> ~ $ doas rc-service -U pipewire status
2026-04-01 13:31:47 <seti_> ~ $ 
2026-04-01 13:31:47 <seti_> XDG_RUNTIME_DIR unset.
2026-04-01 13:31:58 <seti_> im guessing its because the dir is only set for my user right
2026-04-01 13:34:28 <visone[m]> seti_: You can add some like this to your `~/.profile` or shell config 
2026-04-01 13:34:28 <visone[m]> ```
2026-04-01 13:34:28 <visone[m]> ```
2026-04-01 13:34:28 <visone[m]> XDG_RUNTIME_DIR=$(mktemp -d /tmp/$(id -u)-runtime-dir.XXX)
2026-04-01 13:37:55 <invoked> they need coreutils for mktemp, which is not in the default install
2026-04-01 13:38:03 <invoked> (i think?)
2026-04-01 13:38:22 <seti_> busybox has a pared down one apparently
2026-04-01 13:38:27 <invoked> oh nvm me
2026-04-01 13:38:55 <seti_> just needed 6 X:s instead of 3 in the arg
2026-04-01 13:44:05 <visone[m]> invoked: So, try this way ...... (full message at <https://matrix.org/oftc/media/v1/media/download/AUgwFKmF8opkOiYeFxDPBoRqBCkVL4nECfyCTzQUm02Lh_n16oZLmmy5miD-MDVJ2ORq0ETwKIq2nXkweJMMWzxCedkWKDuwAG1hdHJpeC5vcmcvaVJNVnVRc1p6Tm9ZSW1wTkhuS3JJZXVk>)
2026-04-01 13:47:31 <invoked> that's how i've been doing it for years now :)
2026-04-01 13:50:39 <seti_> now i have issue instead with doas not applying keepenv properly so it still complains of no runtime dir
2026-04-01 13:50:59 <seti_> i dunno why but doas is really picky about how the doas.conf file should be
2026-04-01 13:59:46 <invoked> mktemp would be safer from hypothetical attacks, or at least the block should remove the existing dir if it finds it. /tmp gets cleared between boots in my case, but mktemp would still be better.
2026-04-01 14:00:44 <invoked> i probably pasted that in originally without thinking much about it.
2026-04-01 14:05:17 <invoked> seti_: what are you trying to do with doas?
2026-04-01 14:05:32 <invoked> what commands i mean
2026-04-01 14:05:56 <invoked> oh... you don't need doas to start user services
2026-04-01 14:07:44 <invoked> seti_: https://wiki.alpinelinux.org/wiki/OpenRC#User_services  might help you
2026-04-01 14:08:19 <invoked> 
ACTION is still waking up

2026-04-01 14:33:27 <seti_> invoked: trying to enable keepenv for my user and no password
2026-04-01 14:34:15 <seti_> even after adding that mktmp for the XDG_RUNTIME_DIR, i cant start the service, beause my env doesnt carry over with doas
2026-04-01 14:34:41 <seti_> the service complains that XDG_RUNTIME_DIR is unset when i try to start it
2026-04-01 14:34:45 <invoked> this doesn't make any sense
2026-04-01 14:35:05 <seti_> what doesnt?
2026-04-01 14:35:15 <invoked> you don't need doas. and even if you were trying to run doas as your user you'd 'doas -u <user>'
2026-04-01 14:35:46 <seti_> huh i thought openrc would only work as super user
2026-04-01 14:36:10 <invoked> for system services it does
2026-04-01 14:36:19 <invoked> for user services you just run them as your user directly
2026-04-01 14:36:34 <invoked> look at the wiki link i gave you for the quick overview
2026-04-01 14:36:51 <seti_> its getting even weirder, sorry kinda big log
2026-04-01 14:36:52 <seti_> ~ $ rc-service -U pipewire start
2026-04-01 14:36:52 <seti_> ~ $ rc-service -U wireplumber status
2026-04-01 14:36:52 <seti_>  * WARNING: pipewire is already starting
2026-04-01 14:36:52 <seti_>  * WARNING: wireplumber is already starting
2026-04-01 14:36:52 <seti_> ~ $ rc-service -U wireplumber start
2026-04-01 14:36:54 <seti_>  * You are attempting to run an openrc service on a
2026-04-01 14:36:54 <seti_>  * system which openrc did not boot.
2026-04-01 14:36:56 <seti_>  * another initialization system to boot this system.
2026-04-01 14:36:56 <seti_>  * You may be inside a chroot or you may have used
2026-04-01 14:36:58 <seti_>  * In this situation, you will get unpredictable results!
2026-04-01 14:36:58 <seti_>  * If you really want to do this, issue the following command:
2026-04-01 14:37:00 <seti_>  * touch /tmp/1000-runtime-dir.edaOAp/openrc/softlevel
2026-04-01 14:37:00 <seti_> ~ $ 
2026-04-01 14:37:10 <seti_> im absolutely on alpine and not in a chroot
2026-04-01 14:37:30 <seti_> this is after adding the export XDG_RUNTIME_DIR=$(mktemp -d /tmp/$(id -u)-runtime-dir.XXXXXX) to my .profile
2026-04-01 14:39:01 <invoked> your running system may be in a weird state with all your doas stuff. i'd reboot then follow the wiki example
2026-04-01 14:39:18 <invoked> stop mashing keys
2026-04-01 14:56:10 <courge_musque[m]> [@_oftc_seti_:matrix.org](https://matrix.to/#/@_oftc_seti_:matrix.org) most likely you started your DE with startx
2026-04-01 14:56:20 <courge_musque[m]> Or another way, but manually
2026-04-01 20:02:52 <WhyNotHugo> seti_: please use a pastebin next time; don't post long logs in chat.
2026-04-01 20:14:03 <Renan[m]> error: Failed to install org.gnome.Platform: Could not unmount revokefs-fuse filesystem at /var/tmp/flatpak-cache-2UFTM3/org.gnome.Platform-D6CZM3: Child process exited with code 1
2026-04-01 20:17:52 <Renan[m]> Renan[m]: how i can fix that?
2026-04-02 00:11:31 <Renan[m]> 
ACTION uploaded an image: (4KiB) < https://matrix.org/oftc/media/v1/media/download/AaB2qeqsaYavM8EHN71BstsV_SEyMxYLm6Pf6c_iwVYAWyLLRD_mIuS5pFEHsiBlgPW4WNp6en52QFigpqxBp_tCedk6DwqwAHVucmVkYWN0ZWQub3JnL1BTUmtZTWp2T1RSQkZIdmxHaFV3dmZncQ >

2026-04-02 00:11:40 <Renan[m]> i cant use bluetoothctl
2026-04-02 00:11:49 <Renan[m]> infinity waiting
2026-04-02 00:12:56 <Renan[m]> 
ACTION uploaded an image: (16KiB) < https://matrix.org/oftc/media/v1/media/download/ATTKNVPEGKaY_hmerkN-GwLIjiipkEZQ2mv4smAACHByAFnW_KTLkitsYEGfXuphfFieLBPl-nelCO6NjZvk_QJCedk6I8ggAHVucmVkYWN0ZWQub3JnL3hlT2RwUXVmU2drRlVqa0NCelFhbnpGYg >

2026-04-02 00:13:10 <Renan[m]> Renan[m]: another error when acessing other partition
2026-04-02 00:15:55 <Renan[m]> Renan[m]: i fixed this searching ddg haha
2026-04-02 00:25:10 <Renan[m]> Renan[m]: i am trying to mount an encrypted popos partition
2026-04-02 00:25:21 <Renan[m]> please, someone help me
2026-04-02 03:28:25 <rnkn> the NetworkManager wiki page is a good wiki page. good work Prabuanand
2026-04-02 03:30:10 <rnkn> seti_: I had this same problem and it's poorly documents. you need to symlink /etc/init.d/user.$USER -> /etc/init.d/user
2026-04-02 03:30:29 <rnkn> seti_: you then need to doas rc-update add user.$USER
2026-04-02 03:30:48 <rnkn> n.b. this is to the system rc, not the user
2026-04-02 03:30:58 <rnkn> *poorly documented
2026-04-02 03:32:17 <rnkn> you then need to mkdir ~/.config/rc/runlevels/{sysinit,gui}
2026-04-02 03:32:32 <rnkn> you then need to echo rc_env_allow="WAYLAND_DISPLAY XDG_RUNTIME_DIR" > ~/.config/rc/rc.conf
2026-04-02 03:32:46 <rnkn> you then need to rc-update -U <user service> gui
2026-04-02 03:32:57 <rnkn> then finally openrc -U
2026-04-02 03:33:08 <rnkn> it's a huge lot of trial/error/guessing
2026-04-02 05:58:09 <rnkn> how can I suspend the system without root?
2026-04-02 06:03:37 <WhyNotHugo> rnkn: powerctl lets you grant group-based permissions
2026-04-02 06:04:21 <amrfti> Could also do it via a doas rule
2026-04-02 06:11:45 <rnkn> amrfti: great idea! I just made a "echo mem > /sys/power/state" script called suspend and added that to doas.conf
2026-04-02 06:18:41 <rnkn> okay, roadblock. *I* can run this suspend script but rc-service can't
2026-04-02 06:22:47 <rnkn> doas: a tty is required
2026-04-02 06:23:26 <rnkn> WhyNotHugo: powerctl doesn't seem to be packaged?
2026-04-02 06:24:26 <WhyNotHugo> rnkn: it's in testing.
2026-04-02 06:24:53 <WhyNotHugo> how are you using doas? It requires a tty if you need a password. If you're doing some automation, you can configure it to allow a specific command for a specific user/group without password
2026-04-02 06:25:57 <rnkn> my doas.conf line is: permit nopass :wheel cmd /usr/local/bin/suspend
2026-04-02 06:31:06 <rnkn> okay I replaced my little script with zzz and it works
2026-04-02 06:32:48 <rnkn> ty for the help WhyNotHugo, amrfti
2026-04-02 07:56:08 <rnkn> can I mount external media as an unprivileged user without polkit?
2026-04-02 07:56:33 <rnkn> automount*
2026-04-02 08:12:27 <grawity> something like pmount could do that
2026-04-02 08:25:40 <rnkn> pmount does not seem to be packaged
2026-04-02 08:25:53 <rnkn> I think I will just accept that I need polkit
2026-04-02 08:26:05 <rnkn> at least I still don't have elogind
2026-04-02 09:39:21 <frag> is there a way to list or delete all shell functions defined?
2026-04-02 09:40:44 <quinq> rnkn, there's autofs
2026-04-02 09:41:54 <lopid> technically, `set` does list shell functions
2026-04-02 09:42:43 <frag> not for busybox ash it seems
2026-04-02 09:42:48 <lopid> o
2026-04-02 12:07:49 <ty3r0x> greetings, isn't dovecot on alpine ship with PAM plugin? If not, do you use password files or pgsql as alternative?
2026-04-02 12:09:39 <lopid> i have few users, so passwd-file
2026-04-02 12:28:30 <WhyNotHugo> rnkn: I use an mdev rule to change ownership of USB drives to my main user, and mount with fuse.
2026-04-02 12:28:46 <WhyNotHugo> It sucks tho, I intend to write a daemon for auto-mount via unprivileged users at some point.
2026-04-02 13:12:42 <invoked> ty3r0x: pam should be in there by default. also you can do multiple. eg where it checks pam, then falls back to /etc/passwd
2026-04-02 13:13:25 <invoked> you'll have to add dovecot-gssapi if you want to hook it up to AD or something
2026-04-02 13:19:52 <socksinspace> does anyone here have experience with virtio-pmem and dax (or an easier idea for getting VMs to leave the file caching to the host)
2026-04-02 13:43:19 <quinq> Hi
2026-04-02 13:43:31 <quinq> When trying to install meson, I get an error (edge/amd64)
2026-04-02 13:43:34 <quinq> ERROR: unable to select packages:
2026-04-02 13:43:34 <quinq>     breaks: meson-1.10.2-r1[python3~3.14]
2026-04-02 13:43:34 <quinq>   python3-3.12.12-r0:
2026-04-02 13:43:54 <ikke> quinq: what is trying to select pyton 3.12?
2026-04-02 13:43:58 <quinq> Even after an apk upgrade
2026-04-02 13:44:13 <quinq> oops sorry, next line to that is:
2026-04-02 13:44:14 <quinq>     satisfies: android-tools-35.0.2-r19[python3]
2026-04-02 13:44:20 <quinq> Does that answer the question?
2026-04-02 13:44:22 <bl4ckb0ne> is the python3.14 rebuild done on riscv64?
2026-04-02 13:44:48 <quinq> Then there's more than 400 lines of things like:
2026-04-02 13:44:48 <quinq>                boost1.84-python3-1.84.0-r5[so:libpython3.12.so.1.0]
2026-04-02 13:44:48 <quinq>                calibre-8.16.2-r1[so:libpython3.12.so.1.0]
2026-04-02 13:45:18 <ikke> quinq: did you try apk ugprade --available?
2026-04-02 13:45:34 <quinq> Actually, let-me paste the output instead of filtering: https://paste.nurd.space/7uc7Jslc7s16QGWTSlrSihMHdY77DuyoX3wxzvbDhBM
2026-04-02 13:47:16 <quinq> It wanted to remove a bunch of -pyc packages, which I was told to wait for it to resolve a few days ago
2026-04-02 13:47:48 <quinq> Well, I get the same after apk upgrade -a: https://paste.nurd.space/Z981a9pf5cQcOR8-tQdUh2iEavCwNAnqReY7y9YoWYM
2026-04-02 13:48:44 <invoked> quinq: what mirror are you using
2026-04-02 13:49:22 <quinq> http://dl-cdn.alpinelinux.org/alpine/edge/main
2026-04-02 13:49:34 <invoked> huh
2026-04-02 13:50:41 <quinq> I ran apk upgrade --prune too
2026-04-02 13:50:51 <quinq> wait
2026-04-02 13:50:52 <quinq> (24/26) Purging mupdf (1.26.12-r0)
2026-04-02 13:50:53 <quinq> wat
2026-04-02 13:50:56 <ikke> apk list -q --orphjaned
2026-04-02 13:51:10 <invoked> python3-dbg -dev packages hanging it up, maybe
2026-04-02 13:51:17 <quinq> mupdf has been dumped out of alpine?!
2026-04-02 13:51:42 <ikke> quinq: it has been temporarily disabled due to build failures
2026-04-02 13:52:06 <quinq> I use it daily for work
2026-04-02 13:52:08 <quinq> damn
2026-04-02 13:52:16 <quinq> $ apk list -q --orphaned
2026-04-02 13:52:16 <quinq> mupdf-libs
2026-04-02 13:52:37 <quinq> $ doas apk del mupdf-libs
2026-04-02 13:52:37 <quinq> World updated, but the following packages are not removed due to: mupdf-libs: zathura-pdf-mupdf
2026-04-02 13:52:38 <ikke> !95288
2026-04-02 13:53:00 <invoked> i have the same mupdf thing but it's not hanging me up
2026-04-02 13:53:23 <invoked> not advice, but if it were me, i'd try removing python3-dev python3-dbg and see what that gets me
2026-04-02 13:53:37 <invoked> (mashing keys)
2026-04-02 13:54:12 <quinq> Thanks iike, that doesn't look nice (the weird mujs integration into mupdf)
2026-04-02 13:54:17 <quinq> Sorry, ikke :)
2026-04-02 13:55:38 <quinq> invoked, after doing that, it wanted me to upgrade python3-gdbm, which I did
2026-04-02 13:55:43 <quinq> Same error then trying to install meson
2026-04-02 13:56:01 <quinq> That looks like a certain bug in the python upgrade, is it finished?
2026-04-02 13:57:11 <ikke> except for riscv64, the builders have finished
2026-04-02 13:57:39 <invoked> lol, what does python3-gdbm have to do with removing -dev and -dbg
2026-04-02 13:57:52 <invoked> this makes no sense
2026-04-02 13:58:06 <quinq> exactly
2026-04-02 13:58:44 <invoked> whatever you built against python3-dev just broke though, probably
2026-04-02 14:03:35 <invoked> i'd try --prune next (assuming you can recover from apk-cache if you need something), maybe --latest and see what happens, go through that list of things that depend on python
2026-04-02 14:04:40 <quinq> AH
2026-04-02 14:05:00 <quinq> Looks like it was py3-lib_users that was blocking
2026-04-02 14:05:14 <quinq> Yep, it's the broken package
2026-04-02 14:09:04 <quinq> Need to download 405.2 MiB of packages.
2026-04-02 14:09:04 <quinq> After this operation, 131.0 MiB of additional disk space will be used.
2026-04-02 14:09:12 <quinq> Python bloating
2026-04-02 14:11:41 <invoked> what's the meme?  the average claude commit is ~2000 lines :)
2026-04-02 14:12:57 <quinq> ikke, it seems that (some) packages in testing/ were not rebuilt with the python upgrade
2026-04-02 14:13:50 <ikke> Yes, that's correct 
2026-04-02 14:16:21 <cousin_luigi> Greetings.
2026-04-02 14:24:22 <mio> py3-lib_users package will be back shortly, !99975
2026-04-02 14:24:45 <mio> there was an issue with the source url, and so was not immediately rebuilt
2026-04-02 14:26:54 <cousin_luigi> I'm kind of new to alpine, do you think there's margin of improvement here https://paste.rs/yaz9r ? (perhaps the question is not for -devel)
2026-04-02 14:32:20 <quinq> Thank you for notifying, mio :)
2026-04-02 14:32:28 <quinq> So the python3 upgrade is not finished
2026-04-02 14:39:31 <mio> quinq: you're welcome. python3 upgrade is essentially completed in main/ and community/, aside from mupdf and py3-simframe awaiting a fix and riscv64 still processing packages in community/
2026-04-02 14:40:54 <mio> testing/ is in progress, a considerable portion of aports have been rebuilt, but there are still ones with issues that are gradually being resolved
2026-04-02 14:41:33 <mio> #18025 has a partial list if you'd like to check progress
2026-04-02 14:46:35 <quinq> Thanks again, that's pretty useful
2026-04-02 14:47:04 <quinq> Maybe those kind of big shifts could be published on the website news?
2026-04-02 14:47:36 <quinq> So that people have a direct way of keeping informed without having to browse the issue tracker and hopefuly encountering those
2026-04-02 15:22:39 <cousin_luigi> Can I build a git snapshot without uploading it to dev.alpinelinux.org ?
2026-04-02 15:48:24 <henrix> Hi! I was trying to install alpine in a laptop using the -extended iso
2026-04-02 15:48:25 <henrix> However, I can't use wifi (iwlwifi) because it can't load the fw
2026-04-02 15:48:25 <henrix> Installing linux-firmware-intel fails, saying it's trying to write to a read-only filesystem
2026-04-02 15:48:28 <henrix> Is this a know issue? Is there a workaround?
2026-04-02 15:51:31 <courge_musque[m]> I put the fw for my ax210 in the extended image before flashing it on a usb drive
2026-04-02 15:52:19 <courge_musque[m]> There may be a way to ask iwlwifi to load it, but I don't remember well how
2026-04-02 15:53:04 <henrix> hmm... ok. that's a suggestion, but I should be able to install it on the -extended iso -- the firmware *is* there already
2026-04-02 15:53:54 <henrix> I mean, the .apk is there in the iso; but it fails to install it because it tries to write into /.modloop, which is mounted read-only
2026-04-02 15:54:18 <henrix> and I can't be the first person hitting this :-)
2026-04-02 15:55:20 <henrix> (oh, and this is a re-install! this laptop had alpine installed in the past, running edge. probably I used 3.22 or 3.21 to install it)
2026-04-02 16:30:13 <henrix> uff! looks like I found an ugly workaround using overlayfs :-/
2026-04-02 16:30:51 <henrix> I'll see if I can create an issue for this. I tried to find one, but couldn't find it
2026-04-02 17:56:20 <GeoVM> linuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinux
2026-04-02 17:56:30 <GeoVM> we love linux am i rightlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlin
2026-04-02 17:56:30 <GeoVM> uxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinux
2026-04-02 17:56:31 <GeoVM> linuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxli
2026-04-02 17:56:36 <GeoVM> nuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinu
2026-04-02 17:56:43 <GeoVM> xlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinux
2026-04-02 17:56:45 <GeoVM> linuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxli
2026-04-02 17:56:50 <GeoVM> nuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinu
2026-04-02 17:56:55 <GeoVM> xlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxl
2026-04-02 17:57:00 <GeoVM> inuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlin
2026-04-02 17:57:05 <GeoVM> uxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlinuxlin
2026-04-02 17:57:15 <GeoVM> oops sorry
2026-04-02 17:57:29 <dwfreed> O.o
2026-04-02 17:57:31 <GeoVM> this is a public vm like collabvm and whoever did that is getting banned now
2026-04-02 17:57:53 <dwfreed> the VM is getting banned now
2026-04-02 18:00:35 <Lilith> must've been a big fan of linux
2026-04-02 18:06:22 <quinq> jaja
2026-04-02 18:07:48 <quinq> Just ban the entire 104.16.0.0/12
2026-04-02 18:10:50 <KleeBoomBoomBakudan> hey
2026-04-02 18:10:59 <KleeBoomBoomBakudan> \"now download minecraft.exe free 10000TB size virus" on people XChat
2026-04-02 18:11:10 <KleeBoomBoomBakudan> oops
2026-04-02 18:11:18 <KleeBoomBoomBakudan> join me in genshin impact
2026-04-02 18:11:23 <KleeBoomBoomBakudan> my uid is 760438548
2026-04-02 18:11:27 <KleeBoomBoomBakudan> europe server
2026-04-02 18:11:29 <KleeBoomBoomBakudan> or kys
2026-04-02 18:12:12 <KleeBoomBoomBakudan> https://computernewb.com/collab-vm/user-vm/#amazingvm-w10
2026-04-02 18:12:14 <KleeBoomBoomBakudan> polecam
2026-04-02 18:12:16 <quinq> See what I meant
2026-04-02 18:12:18 <KleeBoomBoomBakudan> twoj stary
2026-04-02 18:12:20 <Lilith> id rather kms than join a klee player
2026-04-02 18:12:26 <KleeBoomBoomBakudan> hey
2026-04-02 18:12:28 <KleeBoomBoomBakudan> kys 
2026-04-02 18:12:45 <KleeBoomBoomBakudan> i
2026-04-02 18:12:50 <KleeBoomBoomBakudan> i use windows 95
2026-04-02 18:13:02 <quinq> Best OS, you got your street cred
2026-04-02 18:13:18 <KleeBoomBoomBakudan> can we talk pls
2026-04-02 18:13:29 <KleeBoomBoomBakudan> join amazingvm https://computernewb.com/collab-vm/user-vm/#amazingvm-w10
2026-04-02 18:13:44 <KleeBoomBoomBakudan> join now
2026-04-02 18:13:49 <aron> pls ban 104.16.0.0/12
2026-04-02 18:13:57 <KleeBoomBoomBakudan> why
2026-04-02 18:14:01 <KleeBoomBoomBakudan> im just talking
2026-04-02 18:14:04 <KleeBoomBoomBakudan> :sob:
2026-04-02 18:14:08 <KleeBoomBoomBakudan> im on a vpn
2026-04-02 18:14:10 <KleeBoomBoomBakudan> and this is collabvm
2026-04-02 18:14:11 <KleeBoomBoomBakudan> uservm
2026-04-02 18:14:26 <KleeBoomBoomBakudan> other people sent the bad stuff here
2026-04-02 18:14:30 <KleeBoomBoomBakudan> it's shared vm
2026-04-02 18:14:45 <KleeBoomBoomBakudan> w proton vpn
2026-04-02 18:14:52 <KleeBoomBoomBakudan> proton vpn so sexy
2026-04-02 18:15:05 <KleeBoomBoomBakudan> giga chad
2026-04-02 18:15:05 <penguinz_rule[m]> this is off topic so please shut it
2026-04-02 18:15:16 <KleeBoomBoomBakudan> if you want to stop this
2026-04-02 18:15:19 <KleeBoomBoomBakudan> join my vm and close the irc client
2026-04-02 18:15:27 <KleeBoomBoomBakudan> https://computernewb.com/collab-vm/user-vm/#amazingvm-w10
2026-04-02 18:15:29 <AmazingVMLover> YOU'RE off topic!
2026-04-02 18:16:01 <KleeBoomBoomBakudan> boom i destroyed the su
2026-04-02 18:16:06 <KleeBoomBoomBakudan> ahhj my balls
2026-04-02 18:16:09 <KleeBoomBoomBakudan> countdown
2026-04-02 18:16:11 <KleeBoomBoomBakudan> control my pc
2026-04-02 18:16:21 <KleeBoomBoomBakudan> love linux
2026-04-02 18:16:26 <KleeBoomBoomBakudan> arch better
2026-04-02 18:16:30 <KleeBoomBoomBakudan> i use arch btw uwu
2026-04-02 18:16:35 <quinq> Linux 95
2026-04-02 18:16:45 <dwfreed> KleeBoomBoomBakudan: enough
2026-04-02 18:16:48 <KleeBoomBoomBakudan> hell yeah linux 95
2026-04-02 18:16:57 <AmazingVMLover> you know what is better than windows 95 and linux 95 combined
2026-04-02 18:16:58 <AmazingVMLover> Barney OS
2026-04-02 18:17:06 <KleeBoomBoomBakudan> computernewb.com/collab-vm/user-vm#amazingvm-w10
2026-04-02 18:17:16 <KleeBoomBoomBakudan> 256mb ram on linux cool
2026-04-02 18:17:19 <KleeBoomBoomBakudan> ohio
2026-04-02 18:17:29 <KleeBoomBoomBakudan> 256mb ram cool 
2026-04-02 18:17:36 <KleeBoomBoomBakudan> https://cafewebs.com
2026-04-02 18:18:32 <dwfreed> ikke: perhaps +Mz would be good for a bit, if you're around
2026-04-02 18:24:19 <Blabi> hello
2026-04-02 18:24:28 <Blabi> https://computernewb.com/collab-vm/user-vm#amazingvm-w10
2026-04-02 18:24:29 <Blabi> join
2026-04-02 18:24:37 <Blabi> i heard about linux 95
2026-04-02 18:24:39 <Blabi> dzxf
2026-04-02 18:24:39 <Blabi> sdf
2026-04-02 18:24:39 <Blabi> d
2026-04-02 18:24:39 <Blabi> gd
2026-04-02 18:24:39 <Blabi> fg
2026-04-02 18:24:43 <Blabi> allah
2026-04-02 18:25:44 <Lilith> i wonder what do these sad existences have out of spamming an irc channel
2026-04-02 18:33:08 <AmazingVMLover> grrrrrrrr how dare u accuse me of being a spambot for breathing
2026-04-02 18:33:09 <AmazingVMLover> thats it
2026-04-02 18:33:12 <AmazingVMLover> u will install barney os now https://friends.citrons.xyz/~viba/os/barneyos/
2026-04-02 18:33:38 <aron> how ironic...
2026-04-02 18:33:56 <aron> I think they earn some money if someone clicks on those shady links
2026-04-02 18:34:23 <kaboooooooooooo> spam my site issueplate.netlify.app!
2026-04-02 18:34:23 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:24 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:24 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:24 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:25 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:25 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:27 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:27 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:34:29 <kaboooooooooooo> y site issueplate.netlify.app!
2026-04-02 18:37:31 <AmazingVMHater> O Brave people of the world of justice! O people of peace around the world! O people of goodness! May God bless you and your soul! Do Not Be Misled by the Voices of Falsehood, Let Truth Prevail Over the Lies of Our Enemies, No matter how painfull the truth is, Stand United Against Those Who Seek Our Harm, Hold Firm to the Truth of AmazingVM.
2026-04-02 18:37:32 <AmazingVMHater> https://amazingvm.geoegii555.eu.org/#amazingvm-w10
2026-04-02 18:46:11 <AmazingVMHater> O Brave people of the world of justice! O people of peace around the world! O people of goodness! May God bless you and your soul! Do Not Be Misled by the Voices of Falsehood, Let Truth Prevail Over the Lies of Our Enemies, No matter how painfull the truth is, Stand United Against Those Who Seek Our Harm, Hold Firm to the Truth of AmazingVM.
2026-04-02 18:46:12 <AmazingVMHater> https://amazingvm.geoegii555.eu.org/#amazingvm-w10
2026-04-02 18:50:38 <Reanimator> hi all
2026-04-02 18:58:56 <f_> reported to collabvm
2026-04-02 20:12:07 <cousin_luigi> this place is odd
2026-04-02 21:15:30 <efahl> Is this "bring out your weirdos week"?  :)
2026-04-02 22:09:22 <quinq> efahl, how many of those do you have? :)
2026-04-03 04:59:08 <ikke> Foxy[m]: I haven't noticed anything
2026-04-03 05:02:21 <Foxy[m]> damn gitlab. Not sure how to report my bug now
2026-04-03 05:03:48 <ikke> Foxy[m]: One option would be to send an email. If you are logged in, on the issue page, there a 3-dot button. There's a "email work item to this project"
2026-04-03 05:07:03 <Foxy[m]> worked like a charm. Ty!
2026-04-03 05:09:05 <knuxify[m]> oh wow, aports just hit 100 000 merge requests: aports!100000
2026-04-03 05:09:38 <knuxify[m]> congrats to the lucky 100000 :D
2026-04-03 06:28:25 <cousin_luigi> Greetings.
2026-04-03 06:28:41 <cousin_luigi> So, about my quaint little package, any suggestions?
2026-04-03 08:58:07 <ikke> cousin_luigi: for a personal package, it's fine. This cannot be submitted to aports though, since it's packaging proprietary binaries.
2026-04-03 11:06:30 <cousin_luigi> ikke: Not even community?
2026-04-03 11:06:38 <ikke> No
2026-04-03 11:13:50 <realroot[m]> i'm testing xonotic looks good
2026-04-03 11:22:18 <courge_musque[m]> Xonotic and Unvanquished are really great, that's a shame they're not more played
2026-04-03 11:23:27 <frag> oh no, theres a game on alpine?
2026-04-03 11:24:28 <frag> its like quake?
2026-04-03 11:25:47 <courge_musque[m]> They are fork from a quake engine iirc
2026-04-03 11:31:00 <courge_musque[m]> I checked, and Xonotic uses DarkPlaces derived from Quake, and Unvanquished uses Daemon, derived from ioQuake3
2026-04-03 11:31:26 <courge_musque[m]> But they both seem to focus on a Quake3-like gameplay
2026-04-03 11:31:28 <courge_musque[m]> https://en.wikipedia.org/wiki/File:Quake_-_family_tree.svg
2026-04-03 11:31:38 <frag> i played urban terror before, never really got into quake
2026-04-03 11:32:48 <lopid> xonotic not on that graph :/
2026-04-03 11:33:47 <frag> cool searchable mindmap tho
2026-04-03 11:34:08 <frag> kinda playing around with postscript right now
2026-04-03 12:08:50 <socksinspace> as it turns out, the easiest way to get a VM to rely on the host page cache is to use zfs, because that exposes caching controls, without having to use virtio-pmem as the disk
2026-04-03 12:09:24 <socksinspace> *smh my head*
2026-04-03 12:17:12 <cousin_luigi> ikke: Is there an alpine equivalent of AUR? For packages built locally?
2026-04-03 12:17:58 <cousin_luigi> Or perhaps a package containing a script that will download and install said files? I saw something like that packaging for rpm, not sure how to replicate it with abuild.
2026-04-03 12:18:13 <ikke> cousin_luigi: No, we do not have anything equivalent
2026-04-03 12:19:06 <ikke> But if you have the APKBUILD, doing abuild -r, and then installing the resulting package is reasonably easy
2026-04-03 12:19:27 <cousin_luigi> ikke: But where could said APKBUILD be hosted? 
2026-04-03 12:19:28 <Sertonix[m]> You could publish it on sone git forge. I use https://codeberg.org/sertonix/rports to push APKBUILDs which are not in aports
2026-04-03 12:19:31 <cousin_luigi> ok
2026-04-03 12:19:52 <cousin_luigi> good to know
2026-04-03 12:20:17 <ikke> Sertonix[m]: what does r in this case stand for?
2026-04-03 12:20:52 <Sertonix[m]> short for "random ports" (naming things is hard)
2026-04-03 12:21:07 <cousin_luigi> I thought -r was for autoinstalling dependencies?
2026-04-03 12:21:28 <cousin_luigi> anyway, couldn't one play with post-install scripts to that effect?
2026-04-03 12:21:52 <cousin_luigi> say, let's call it panasonic-firmware-installer?
2026-04-03 12:22:26 <cousin_luigi> or would extra untracked files damage referential integrity?
2026-04-03 12:22:37 <ikke> it would
2026-04-03 12:22:49 <ikke> You bypass apk and only use it to get you the install script
2026-04-03 12:23:06 <ikke> apk would not know about these firmware files or anything
2026-04-03 12:23:23 <cousin_luigi> a  post-uninst would get rid of them
2026-04-03 12:23:29 <cousin_luigi> too dodgy?
2026-04-03 12:25:50 <cousin_luigi> anyway I would like to make a feature request for a %ghost like function.
2026-04-03 12:26:27 <Sertonix[m]> It would for example not respect the --no-networking apk-tools flag
2026-04-03 12:27:33 <Sertonix[m]> If at all it should be a script manually invoked by the user
2026-04-03 12:28:35 <ikke> cousin_luigi: what would that feature do?
2026-04-03 12:29:30 <cousin_luigi> ikke: on rpm it allows to track files that are created after the installation.
2026-04-03 12:29:54 <cousin_luigi> Adopted by the package or something. I don't know the formal definition.
2026-04-03 12:30:17 <cousin_luigi> Sertonix[m]: hmm, --no-networking you say?
2026-04-03 12:30:24 <cousin_luigi> that's a massive bust
2026-04-03 14:41:31 <cousin_luigi> mercenary: Yeah, that's pretty much what I had in mind. Plus a post-deinstall script to clean up everything, but it's still not too clean.
2026-04-03 14:50:29 <realroot[m]> frag: i want to MR xonotic
2026-04-03 14:51:05 <realroot[m]> if somebody wants to do it I have files
2026-04-03 15:05:18 <courge_musque[m]> What, xonotic isn't already packaged?
2026-04-03 15:05:55 <courge_musque[m]> https://pkgs.alpinelinux.org/package/edge/community/x86_64/xonotic
2026-04-03 15:06:08 <realroot[m]> is that recent?
2026-04-03 15:06:39 <realroot[m]> no , well i thought it wasn't
2026-04-03 15:09:13 <courge_musque[m]> Well, Unvanquished isn't packaged, so it would be cool if you want to do it :)
2026-04-03 15:13:45 <realroot[m]> maybe
2026-04-03 15:33:00 <phws> hey all, to get forwarding within my wireguard net to work, I need to "iptables -I FORWARD 1 -i wg0 -o wg0 -j ACCEPT". awall policy "{ "in": "vpn", "out": "vpn", "action": "accept" }" simply gets ignored (zones are right). What would be the right way to specify it?
2026-04-03 15:36:03 <ikke> phws: good question, something I have been struggling myself
2026-04-03 15:36:14 <ikke> phws: For now, I use the zone for one, and iprange for the other direction
2026-04-03 15:37:38 <phws> thanks, that'll do for me too
2026-04-03 15:40:41 <invoked> for nftables, for passing wg0 specifically, i have this in /etc/nftables.d: https://tpaste.us/meyd
2026-04-03 15:42:10 <invoked> i sorely miss openbsd pf when i'm on linux.
2026-04-03 15:43:23 <ikke> I have never used it, so nothing to miss :P
2026-04-03 15:45:42 <invoked> :) it's likely to make you angry at linux, at least a little bit, so perhaps
2026-04-03 16:31:11 <Saijin_Naib[m]> I tried packaging unvanquished but got stuck
2026-04-03 16:31:43 <Saijin_Naib[m]> In i think nacl
2026-04-03 19:48:44 <WhyNotHugo> invoked: same
2026-04-04 00:02:31 <RavFX[m]> Hi there, question:
2026-04-04 00:11:26 <quinq> answer.
2026-04-04 00:12:28 <RavFX[m]> sorry for the delay, the matrix took forever to sync lol
2026-04-04 00:12:46 <RavFX[m]> ok, Normally when I do pull requestest for community package, A. Klitzing get assigned automatically (for stuff that already exist)
2026-04-04 00:13:10 <RavFX[m]> But I created a pull request to ad a new package and there is no one assigned, should I manually add him or something?
2026-04-04 00:17:48 <RavFX[m]> merge request in question is that one : https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/99676
2026-04-04 00:22:48 <idkrnx[m]> invoked: I didn't realize there was anything that openbsd actually did well
2026-04-04 00:23:12 <invoked> ok
2026-04-04 00:24:16 <idkrnx[m]> What did you like about it
2026-04-04 03:07:01 <runxiyu> how do I use tcp bbr with -virt?
2026-04-04 03:07:08 <runxiyu> seems like the module just doesn't exist
2026-04-04 03:25:11 <fission> runxiyu: which version of alpine are you using? (and on which architecture?)
2026-04-04 03:25:24 <dwfreed> runxiyu: I see it in file lists for all versions of alpine
2026-04-04 03:25:29 <fission> same
2026-04-04 03:25:47 <dwfreed> tcp_bbr.ko.gz
2026-04-04 03:30:26 <runxiyu> i'm on edge, VERSION_ID=3.24.0_alpha20260127, x86_64, 6.18.7-0-virt
2026-04-04 03:30:26 <runxiyu> hm, strange
2026-04-04 03:30:44 <runxiyu> yeah i should update that system...
2026-04-04 03:36:34 <dwfreed> linux-{lts,virt} in edge was updated to 6.18.9 almost 2 months ago
2026-04-04 03:38:37 <dwfreed> it perhaps would also be prudent to point out that BBR only helps for sending, not receiving
2026-04-04 03:52:40 <invoked> runxiyu: might help to give them more context about your problem
2026-04-04 03:53:08 <runxiyu> dwfreed: yeah, i understand that in general congestion control algorithms only pertain to sending
2026-04-04 03:53:28 <runxiyu> invoked, dwfreed: yeah i was stupid, i forgot to reboot after kernel upgrades and the module was not found as expected. sorry for the noise :(
2026-04-04 03:54:44 <icedmirna> i have problem with firmware kit :( 
2026-04-04 03:55:12 <invoked> runxiyu: np. just like dwfreed said though, not sure it will help with gfw & et al
2026-04-04 03:56:06 <runxiyu> yeah that would be completely different (and i've probed this a bit; gfw doesn't cause the packet losses i experience, it's just incompetent peering)
2026-04-04 03:56:25 <invoked> roger that
2026-04-04 03:56:38 <runxiyu> though for my isp's upstream packet loss issues, trying to mess with freebsd netgraph for a solution :P
2026-04-04 04:05:54 <cousin_luigi> ikke: wrt my previous question about post-install file creation, would it help if I created empty ones with the same names during installation?
2026-04-04 08:03:13 <runxiyu> has anyone got screen sharing usable on niri before
2026-04-04 09:27:28 <f_> is there a way to hold a package version in apk?
2026-04-04 09:31:21 <quinq> f_, look for constraints in the man-pages
2026-04-04 09:47:17 <f_> ahh right
2026-04-04 09:47:19 <f_> ty
2026-04-04 10:18:07 <quinq> De nada f_ :)
2026-04-04 13:54:56 <pu> Hi, what process rotates /var/log/messages on a stock Alpine install? logrotate is not configured to, but I still get a messages.0 file.
2026-04-04 13:55:55 <ikke> pu: busybox syslogd
2026-04-04 13:56:14 <ikke> -b N            N rotated logs to keep (default 1, max 99, 0=purge)
2026-04-04 13:56:49 <pu> ikke: and when does rotation trigger?
2026-04-04 13:57:13 <ikke> -s SIZE         Max size (KB) before rotation (default 200KB, 0=off)
2026-04-04 14:00:41 <pu> ikke: Thanks, I was looking for configuration files, but it's basically just the defaults. Thanks.
2026-04-04 14:00:50 <ikke> yup
2026-04-04 15:15:23 <bdprom> hi all, the networking service slows down boot on my laptop as udhcpc searches for DHCP servers while not connected to any wifi connections, wpa_supplicant eventually connects and udhcpc succeeds, but i'd rather this happen in the background instead of holding up my boot.. is this possible?
2026-04-04 15:20:26 <p_6f3Ik7Suw> my preference is to not do wifi/dhcp on boot, but explicitly if needed after being dropped in DE
2026-04-04 15:20:34 <ikke> bdprom: try something like `dhcp-opts -A 0` in /etc/networking/interfaces
2026-04-04 15:25:50 <bdprom> ikke: thanks, i assume you meant udhcpc_opts
2026-04-04 15:26:07 <ikke> bdprom: Both should work
2026-04-04 15:26:14 <ikke> https://github.com/ifupdown-ng/ifupdown-ng/blob/main/doc/interfaces-dhcp.scd#dhcp-related-options
2026-04-04 15:26:25 <ikke> it would be udhcpc-opts, with a -
2026-04-04 15:27:18 <bdprom> awesome, thanks
2026-04-04 15:27:43 <bdprom> hmm, i was looking for a list of valid dhcp executor options, but the interfaces-dhcp man page is missing, weird
2026-04-04 15:27:57 <bdprom> (man pages for all other executors are there)
2026-04-04 15:29:27 <ikke> bdprom: I think it's because it's stil using the ifupdown scripts for that
2026-04-04 15:30:54 <bdprom> ahh i see
2026-04-04 15:31:35 <ikke> hmm
2026-04-04 15:32:15 <bdprom> `/usr/libexec/ifupdown-ng/dhcp is owned by ifupdown-ng-0.12.1-r7`
2026-04-04 15:32:27 <bdprom> seems like it's from ifupdown-ng
2026-04-04 15:33:30 <bdprom> maybe this is a bug? interfaces-dhcp man page not being packaged, i mean
2026-04-04 15:36:02 <ikke> bdprom: maybe dhcp should be added to https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/ifupdown-ng/APKBUILD#L44
2026-04-04 15:36:04 <ikke> not sure
2026-04-04 15:39:31 <bdprom> hmm yeah, i might take a look when i get chance, i'm not too familiar with ifupdown-ng's build scripts
2026-04-04 15:45:52 <ikke> Just tried it, that's not sufficient
2026-04-04 15:46:36 <ikke> It's just plain missing from the Makefile
2026-04-04 15:52:38 <bdprom> oh wow, i guess that should be fixed upstream then
2026-04-04 15:58:35 <ikke> At least reported upstream
2026-04-04 16:15:29 <bdprom> yup
2026-04-04 16:16:07 <bdprom> hmm, what are most people running, edge or stable (v3.23)?.. i like the idea of having less churn/instability, but it's also nice having the latest features and stuff..
2026-04-04 16:18:23 <bdprom> i guess it's relatively straightforward to run stable but compile cherry picked packages from aports to a local repo if needed
2026-04-04 17:02:21 <ikke> yeah, that's not too difficult
2026-04-04 17:02:39 <ikke> For my personal devices I use edge, but for servers I use a stable release
2026-04-04 17:14:59 <rx_> i uhh just run edge everywhere
2026-04-04 18:11:06 <pooper> hello folks. I was here earlier this week with a boot issue.. got into a fight with some assholes here.. their conclusion was it was a skill issue.. anyone remember this?
2026-04-04 18:12:32 <pooper> well this dude with skill issues wrote his own kernel driver for the hyperpixel square screen, in the process fixing a kernel bug and submitted a kernel fix, wrote his own .init file, and now has the setup so far it goes from plugging in the wallsocket to lvgl app usable on screen (includinf wifi, bt) in 5.5s
2026-04-04 18:12:44 <pooper> I'd like to see any of you do the same
2026-04-04 18:13:57 <pooper> oh it's a monolytic kernel, fully stripped, with everything built in, including wifi firmware and the hyperpixel driver i wrote myself. I wrote it myself as the standard hyperpixel driver isn't capable of that. Also that doesn't show you the kernel splash, as it starts halfway the boot. Mine shows the kernel splash.
2026-04-04 18:14:20 <pooper> not bad eh.. for a dude with skill issues..
2026-04-04 18:41:31 <cousin_luigi> pooper: Shouldn't you make such remonstration to the person who actually told you that?
2026-04-04 18:41:59 <pooper> Thats what I'mn doing
2026-04-04 18:42:37 <pooper> see here the result. I'm not making it up :P https://x.com/Dr__Snuggels/status/2040494688566693979
2026-04-04 19:05:31 <n2j3> preach 
2026-04-04 19:06:43 <n2j3> pooper: vibecoded?
2026-04-04 19:07:16 <n2j3> where would you utilise a square screen? fridge door?
2026-04-04 19:08:43 <pooper> anywhere you want a square screen.. in this case this will be a smart home system for in a caravan
2026-04-04 19:14:53 <invoked> i don't recall any such fight
2026-04-04 19:15:10 <invoked> but congrats
2026-04-04 19:38:35 <f_> oh that fight
2026-04-04 19:39:22 <f_> where you claimed that Alpine couldn't crosscompile linux kernels
2026-04-04 19:40:26 <f_> (which is false, I crosscompile kernels and u-boot on alpine natively all the time, no docker or qemu or anything like that)
2026-04-04 19:41:17 <f_> watch your behavio
2026-04-04 19:41:18 <f_> watch your behavior*
2026-04-04 20:06:28 <pooper> f__nothing wrong with by behaviour, that was triggered by the ppl in this chat who couldn't aswer my question and started attacking me
2026-04-04 20:07:35 <pooper> and I never stated one cannot crosscompile linux kernels. I blamed alpine for not just having a kernel source package
2026-04-04 20:08:49 <pooper> what happend was I came here, pissed off, and asked a question too complex for the ppl in this group.. so they started calling me dumb and having skill issues
2026-04-04 20:08:59 <f_> EOD
2026-04-04 20:09:10 <pooper> well.. I've prooved them wrong. NON of those fuckers could have done what I now have done
2026-04-04 20:09:25 <f_> oh I can assure you we know how to write kernel modules
2026-04-04 20:09:29 <pooper> writing your own kernel driving, finding a bug in the kernel, submitting a bugpatch
2026-04-04 20:09:39 <f_> some of us here do that all the time
2026-04-04 20:09:44 <f_> anyway. Offtopic here
2026-04-04 20:09:49 <Lilith> inb4 "i am GOD"
2026-04-04 20:09:52 <pooper> well do it. Make a own driver for the hyperpixel square screen
2026-04-04 20:10:04 <f_> I won't because I have no use for it and you're not paying me to do it
2026-04-04 20:10:18 <quinq> pooper, in any case, I'd advise taking down a notch your tone
2026-04-04 20:10:21 <pooper> no I am not god. I am proud of what i did, yes. But the fact of the matter is you assholes were calling ME a dumbass
2026-04-04 20:10:28 <f_> tone down
2026-04-04 20:10:33 <pooper> why?
2026-04-04 20:10:36 <pooper> or what?
2026-04-04 20:10:45 <quinq> Being “triggered” by whatever doesn't give you the right to start ranting out on this channel insulting people collectively like a child
2026-04-04 20:10:53 <pooper> why wasn't no-one told to tone down when you guys told me I was a dumbass?
2026-04-04 20:11:06 <f_> anyway, I'm upgrading -a my install and I see a bunch of -pyc packages getting removed. Is that normal?
2026-04-04 20:11:29 <pooper> skill issue
2026-04-04 20:11:30 <f_> the non-pyc ones are not removed, e.g. pmbootstrap isn't removed, only pmbootstrap-pyc
2026-04-04 20:11:43 <f_> pooper: I ask you to please stop acting like a child
2026-04-04 20:11:48 <quinq> f_, apparently yes
2026-04-04 20:11:50 <pooper> so you realize now how annoying this is?>
2026-04-04 20:12:02 <ikke> Please calm down
2026-04-04 20:12:10 <quinq> They'll be replaced by the 3.14 pyc then
2026-04-04 20:13:38 <ikke> pooper: I'm going to remove the +q, but please do not continue this discussion
2026-04-04 20:14:07 <f_> quinq: ok, started the upgrade -a, we'll see what happens :)
2026-04-04 20:14:15 <pooper> ikke go eat shit moron.. like I care.. I tried to teach you fuckers something but apparently you have your head to far up your arses to understand
2026-04-04 20:14:17 <f_> (I need to upgrade my pmbootstrap)
2026-04-04 20:14:42 <quinq> f_, I can guarantee that nothing too bad happened to me ;)
2026-04-04 20:15:02 <quinq> Thanks ikke :)
2026-04-04 20:15:17 <f_> oh I have all night to troubleshoot if something goes wrong
2026-04-04 20:15:24 <quinq> jeje
2026-04-04 20:15:52 <f_> I learned to live on the bleeding edge
2026-04-04 20:16:06 <f_> though arch breaks slightly more often
2026-04-04 20:16:15 <f_> due to AUR and such things
2026-04-04 20:39:59 <invoked> "you assholes" ... just an average $spring_equinox_holiday on internet chat rooms, i guess
2026-04-04 20:42:49 <f_> ok turns out I have py3.12 packages still installed
2026-04-04 20:42:53 <f_> too bad
2026-04-04 20:43:34 <quinq> Just wait a bit :)
2026-04-04 20:47:39 <f_> I cannot so I just git cloned for now :)
2026-04-04 21:09:22 <omni> f_: I assume you have some of the python aports in testing/ that still need to be rebuilt installed?
2026-04-04 21:43:37 <WhyNotHugo> f_: `doas apk add --simulate python3~3.14` should help find packages which are missing a rebuild :)
2026-04-04 21:44:26 <WhyNotHugo> (I use doas to ensure the implicit apk-update)
2026-04-04 22:28:34 <f_> omni: yep
2026-04-04 22:55:50 <bdprom> what are people using for local mail delivery, e.g. cron output/sendmail?
2026-04-04 22:55:56 <bdprom> dma + s-nail seem decent
2026-04-05 00:09:10 <omni> I've been a happy user of dma in the past
2026-04-05 04:00:14 <WhyNotHugo> bdprom: I used opensmtpd in the past, but it's an overkill for _just_ that.
2026-04-05 05:14:47 <orcaa> yo im tryna make a user service for emacs daemon its based on the gentoo service (gitweb.gentoo.org/emacs-tools.git/tree/emacs.conf?h=emacs-daemon) i got it to kinda work by removing "unset XDG_RUNTIME_DIR" but openrc complains that the pidfile specified isnt writeable none of the other user services i have (pipewire etc) have pidfiles, so is it unnecessary?
2026-04-05 05:15:19 <orcaa> sorry for bad grammar
2026-04-05 05:16:36 <orcaa> also should i have asked in #alpine-devel?
2026-04-05 06:03:30 <orcaa> did more research and found out i need one, just tried setting it to /run/usr/$(id -u "${command_user}")/${RC_SERVICENAME}.pid but openrc start stop daemon says that setgroups() failed
2026-04-05 06:06:24 <orcaa> could have set it like that with xdg runtime dir oops
2026-04-05 06:26:24 <runxiyu> no idea how openrc user daeons work yet :(
2026-04-05 06:46:06 <orcaa> aww thanks anyway runxiyu
2026-04-05 10:13:57 <bdprom> omni: thanks, good to get a vote for it, it seems its dma sendmail isn't sendmail-compatible though, it doesnt understand the -s flag :/
2026-04-05 10:14:22 <bdprom> WhyNotHugo: yeah i looked into opensmtpd, but i need something to literally just take messages from cron, and deliver to my user somewhere locally
2026-04-05 10:15:34 <bdprom> i wish udhcpc would stop holding up my boot to connect to a wireless network and acquire an ip address too :(
2026-04-05 10:24:02 <p_6f3Ik7Suw> bdprom: what's wrong with not setting up networking during boot, but only manually when already in a DE?
2026-04-05 10:24:29 <p_6f3Ik7Suw> or the solution that ikke gave with the param you can set in the config?
2026-04-05 10:45:09 <bdprom> p_6f3Ik7Suw: the param didnt help, and what do you mean by manual?
2026-04-05 10:45:16 <bdprom> starting the networking service manually on each boot?
2026-04-05 11:40:41 <p_6f3Ik7Suw> bdprom: yes, but no. well, not all networking only the wifi/dhcp. after loging in you can just pop a terminal, and from command line history start wifi and if needed dhcp. i always thought of this automatic wifi on thing as an anti-feature, either most of the time i'm at home or the office, and i have a static ip and i really only need wifi when i'm off the cable - like to a
2026-04-05 11:40:43 <p_6f3Ik7Suw> meeting room or the garden. but most of the time, having wifi on is completely unneccessary, and dhcp even less. if i go somewhere else i also don't want my device to be wifi on, if i do, i want to explicitly and with-a-human-in-the-loop connect to whatever i connect to.
2026-04-05 11:41:35 <p_6f3Ik7Suw> and i boot to a DE very quickly, and if i really need wifi, it's not a big deal to ^R in the shell history to the appropriate setup and start it.
2026-04-05 11:48:48 <bdprom> p_6f3Ik7Suw: makes sense, though i'm always on wifi, i don't use ethernet unless something needs fixing.. i might just comment out the 'auto wlan0' directive in interfaces then use 'ifup wlan0' when i'm booted into Sway
2026-04-05 11:49:09 <bdprom> it's a bit more manual than i'd like but maybe i can automate it with an 'exec' line in sway config
2026-04-05 12:04:39 <p_6f3Ik7Suw> but if you're at home, why don't you use static ips for your own devices in your own network?
2026-04-05 12:28:08 <bdprom> what's assigning these static ips?
2026-04-05 13:45:45 <bdprom> what are people using to search contents of all/remote packages? i.e. `pacman -F` from Arch, or apt-file
2026-04-05 13:58:02 <quinq> There was apk-file, but it doesn't work anymore and seems to be abandoned
2026-04-05 13:58:14 <quinq> Otherwise there's https://pkgs.alpinelinux.org/ :/
2026-04-05 13:58:26 <quinq> Maybe there's a better option I don't know about
2026-04-05 14:09:14 <invoked> apk-file would probably work with a patch. it's failing on the url it's trying to use
2026-04-05 14:09:49 <invoked> pkgs.alpinelinux.org won't take an empty branch=
2026-04-05 14:10:06 <invoked> also arch list has changed, some things
2026-04-05 14:10:23 <invoked> but i'm guessing it would still work
2026-04-05 15:28:27 <MrAdvert[m]> Humble greetings to all, here for a brief advert and not to take much of your time 🚨 ❄️ 
2026-04-05 15:28:27 <MrAdvert[m]> We got products available and many more. Check out the shop below to see products available and remember you owe yourself happiness.  join to get promo updates 👇 👇👇👇👇
2026-04-05 15:28:27 <MrAdvert[m]> Are you going through depression, anxiety and stress or a lover of cannabinoids and psychedelics or physique Enhancement products to be in shape?
2026-04-05 15:28:27 <MrAdvert[m]> https://t.me/+c-LC9ed_hBgwYmZh
2026-04-05 15:30:14 <quinq> How is weed going to improve my physique?
2026-04-05 16:13:28 <Shiz> fucking drugs peddlers
2026-04-05 16:15:56 <quinq> All those questions left unsanswered :(
2026-04-05 16:16:03 <quinq> *unanswered
2026-04-05 17:32:00 <maria[m]> advertising a british telegram channel in an international irc channel is the best way to get customers, or so i've heard
2026-04-05 19:36:26 <WhyNotHugo> I'm low-key bothered by the fact that having openrc installed also pulled openrc user services. I wish we had separate `-openrc` and `-openrc-user` packages.
2026-04-05 19:37:09 <achill> why
2026-04-05 19:44:44 <WhyNotHugo> it just pads the list of installed packages with noise
2026-04-05 19:44:55 <WhyNotHugo> and often times I think "oh, cool, this now has an init script", but it only has a _user_ init script.
2026-04-05 19:46:09 <achill> yeah so what
2026-04-05 19:46:13 <f_> WhyNotHugo: what package are you thinking of specifically
2026-04-05 19:46:30 <f_> like, it doesn't make sense for, e.g. pipewire to have a system init script
2026-04-05 19:46:40 <achill> every new complexity makes it hrader for people especially new people
2026-04-05 19:46:41 <f_> pipewire is supposed to run as your user
2026-04-05 19:47:00 <achill> pipewire actually would work as a system service
2026-04-05 19:47:09 <f_> would != should
2026-04-05 19:47:49 <WhyNotHugo> f_: great example actually, pipewire support running as a system service, and that even makes more sense in boxes where you run things like mpd as a system service.
2026-04-05 19:48:05 <WhyNotHugo> (mpd and say, kodi)
2026-04-05 19:49:34 <f_> ok, another example would be xdg-desktop-portal :p
2026-04-05 19:49:45 <WhyNotHugo> I did say "it low key bothers me", not "i'm absolute appaled by"
2026-04-05 19:49:55 <f_> that's true
2026-04-05 19:50:09 <WhyNotHugo> podman-openrc only has a system service, and no user one
2026-04-05 19:50:33 <f_> it probably should have a user one
2026-04-05 19:50:57 <WhyNotHugo> Docker has a separate package for it tho: docker-rootless-extras-openrc
2026-04-05 19:51:02 <WhyNotHugo> But mostly because it has extra depends.
2026-04-06 23:58:09 <popefucker2> I modified the elogind service so it starts before network connectivity is established
2026-04-07 13:17:23 <runxiyu> hi, im looking to package a updated version of community/drawterm.
2026-04-07 13:17:28 <runxiyu> they have separate x11 and wayland builds
2026-04-07 13:17:35 <runxiyu> what would be an appropriate way to deal with this?
2026-04-07 13:17:53 <runxiyu> i'm inclined to keep the x11 version at drawterm, but create a subpackage for the wl one
2026-04-07 13:17:59 <longnoserob[m]> subpackages?
2026-04-07 13:18:36 <runxiyu> err, forgot what those were called
2026-04-07 13:18:44 <runxiyu> i remember theyre claled subpackages though?
2026-04-07 13:26:37 <ikke> Yes, subpackages
2026-04-07 13:37:06 <runxiyu> gitlab down
2026-04-07 13:37:10 <runxiyu> gitlab down? *
2026-04-07 13:37:12 <runxiyu> or is it just me
2026-04-07 13:38:01 <runxiyu> seems like just me, nvm.
2026-04-07 13:38:35 <runxiyu> can't seem to be able to create a MR though
2026-04-07 13:44:42 <ikke> Do you get an 500 error when opening the page?
2026-04-07 13:51:43 <rx_> ikke: yes
2026-04-07 13:57:18 <ikke> Then you need to make sure the target branch (master) on your remote is up-to-date
2026-04-07 13:57:34 <ikke> Or use the `New Merge` request button instead of following the link
2026-04-07 13:57:38 <rx_> ikke: right, it isn't, thanks
2026-04-07 13:58:19 <rx_>  i assume the merge base traversal was too resource consuming to handle or something?
2026-04-07 14:00:28 <ikke> Yeah
2026-04-07 14:01:00 <ikke> It times out, resulting in a 500
2026-04-07 14:03:07 <runxiyu> do you use commit-graph?
2026-04-07 14:53:04 <ikke> runxiyu: gitlab manages the repos, and it does not enable it by default afaik
2026-04-07 14:53:28 <ikke> I did run the command to generate them on alpine/aports, but I don't think it's maintained, or helps with forks
2026-04-07 15:36:36 <runxiyu> right, i forgot how forks work on gitlab
2026-04-07 15:36:55 <runxiyu> idk how to even implement cross-repo merge base with upstream git...
2026-04-07 16:07:32 <infinitywisdom[m]> what filesystem does alpine linux use by default?
2026-04-07 16:09:22 <ikke> ext4
2026-04-07 16:24:35 <bdprom> is anyone using a relatively full-featured desktop with mdev/mdevd? it seems like pipewire, pulse, and a bunch of other stuff require udev, and libudev-zero is insufficient
2026-04-07 16:24:51 <bdprom> it seems like eudev is unmaintained at this point?
2026-04-07 16:32:32 <dwfreed> one of the main real motivators for eudev was musl support; now that systemd officially supposrts musl, there's not really much reason for eudev to continue to exist
2026-04-07 16:39:08 <invoked> iirc not really a musl thing. it started with gentoo and gentoo has since started pulling udev from systemd
2026-04-07 16:39:48 <dwfreed> it started with Gentoo because Gentoo wanted musl support
2026-04-07 16:40:55 <invoked> i thought it was more to do with openrc.
2026-04-07 16:41:22 <dwfreed> No, udev has always been available independently, even after merging with systemd
2026-04-07 16:44:05 <Guest6626> oh, I didn't know official musl support was being worked on, let alone merged! out of curiosity, is there work to enable systemd as an option in alpine? 
2026-04-07 16:47:00 <invoked> no
2026-04-07 16:47:32 <invoked> but, see postmarketos, which is based on alpine
2026-04-07 16:50:39 <invoked> fwiw i was at gentoo when udev was introduced, so i'm aware of what happened. i just don't recall how well musl even worked back then. we had enough problems just getting arch flags stable
2026-04-07 17:04:43 <bdprom> dwfreed: ah cool, so it just needs to be packaged
2026-04-07 17:05:51 <bdprom> i was hoping there'd be a non-systemd option but it seems like there isnt anymore
2026-04-07 17:07:11 <frag> alpine is getting systemd?
2026-04-07 17:08:19 <Sertonix[m]> No
2026-04-07 17:08:22 <bdprom> frag: read again, you made several leaps there :p
2026-04-07 17:13:17 <frag> ah, phew.. kinda worried about the police state digital id stuff..
2026-04-07 17:14:33 <invoked> bdprom: eudev is still udev. it's not some other code
2026-04-07 17:15:43 <bdprom> invoked: i know, it was forked off a long time ago, but it seems to be unmaintained these days, no activity in the repo for 2 years
2026-04-07 17:16:13 <bdprom> i know it still works, but for an important piece of my system(s), i'd rather use something that was actively maintained
2026-04-07 17:16:46 <bdprom> and like frag said, there are objections to be had for using systemd, so i was looking for alternatives
2026-04-07 17:18:58 <invoked> mdev is mainly it but, it will never try to do everything
2026-04-07 17:20:06 <invoked> thus, udev is required for some/many use cases 
2026-04-07 17:22:20 <bdprom> indeed, not ideal :(
2026-04-07 17:24:04 <invoked> it's possible the uu-people are oxidizing something, but i haven't bothered to look
2026-04-07 17:25:00 <frag> you said it was "cool", your word
2026-04-07 17:25:11 <bdprom> uu people? uutils?
2026-04-07 17:25:23 <invoked> yeah
2026-04-07 17:25:32 <invoked> but i think they leave systemd alone
2026-04-07 17:25:42 <frag> udev is mostly needed for hotplugging? like unplug/replug the mouse?
2026-04-07 17:25:47 <bdprom> ah, interesting
2026-04-07 17:26:15 <bdprom> frag: i was referring to potentially packaging systemd-udevd, not the whole of systemd
2026-04-07 17:26:37 <Sertonix[m]> frag: Then it would be easy to replace it with mdev. But it also has a device database with metadata an such
2026-04-07 17:27:46 <frag> ah ok..
2026-04-07 17:28:24 <frag> tried it a little bit once, but couldnt get it to work and just stuck with eudev
2026-04-08 00:40:10 <RavFX[m]> guys
2026-04-08 00:40:10 <RavFX[m]> Need help with something. I have a machine in PROD that is stuck with 3.21. When I try to update it, I get that error
2026-04-08 00:40:10 <RavFX[m]> ERROR: unable to select packages:
2026-04-08 00:40:10 <RavFX[m]>   vlan-2.3-r1:
2026-04-08 00:40:10 <RavFX[m]>     breaks: ifupdown-ng-0.12.1-r7[!vlan]
2026-04-08 00:40:11 <RavFX[m]> I have my vlan stuff in the /etc/network/interface. That machine need to be upgraded and I keep delaying, Any ideal solution? I need to have minimum downtime.
2026-04-08 00:40:11 <RavFX[m]>     satisfies: world[vlan]
2026-04-08 00:41:31 <RavFX[m]> I have a bunch of line in my /etc/network/interface like, iface eth0.99 inet manual
2026-04-08 00:41:31 <RavFX[m]> my local clanker tell me to shove all vlan configuration in iface eth0 inet manuel with like post-up ip link add link eth0 name eth0.99 type vlan id 99, etc, etc, etc.
2026-04-08 00:41:31 <RavFX[m]> It is the way to go?
2026-04-08 00:41:31 <RavFX[m]>   blablabla
2026-04-08 00:46:10 <RavFX[m]> I guess I can just not upgrade, at the end it's just a machine in my homelab that exclusively run VMs
2026-04-08 01:05:14 <fission> hmm how can one even install vlan ordinarily
2026-04-08 01:05:32 <fission> alpine:~# apk add vlan
2026-04-08 01:05:33 <fission> (3/4) Installing busybox-ifupdown (1.37.0-r30)
2026-04-08 01:05:33 <fission> (4/4) Installing vlan (2.3-r1)
2026-04-08 01:05:33 <fission> (2/4) Purging bridge (1.5-r5)
2026-04-08 01:05:33 <fission> (1/4) Purging ifupdown-ng (0.12.1-r7)
2026-04-08 01:05:35 <fission> Executing busybox-1.37.0-r30.trigger
2026-04-08 01:05:41 <fission> never mind, answered my own question :-)
2026-04-08 01:14:49 <fission> RavFX[m]: https://wiki.alpinelinux.org/wiki/VLAN   says that ifupdown-ng can do it (presumably without the vlan package) in the way described there
2026-04-08 01:14:56 <fission> maybe a different syntax than you're currently using though
2026-04-08 01:15:14 <RavFX[m]> I take a look
2026-04-08 05:25:21 <cousin_luigi> RavFX[m]: I prefer more descriptive names, but that's me.
2026-04-08 05:31:49 <rnkn> I'm currently using alsamixer to change the sound volume but is there something that can live in my xfce4-panel system tray? I'm using pipewire
2026-04-08 05:47:38 <seti_> There is currently no alternative gpu drivers for nvidia users on alpine apart from using nouveau right?
2026-04-08 05:48:07 <seti_> I have a 3060ti and get seemingly random crashes in sway
2026-04-08 05:58:54 <hazardaj_nombroj> rnkn: i think there's an xfce4 panel applet for volume, you don't need a system tray thing specifically
2026-04-08 05:59:23 <hazardaj_nombroj> there is `volumeicon`, however, but it's in the testing repo
2026-04-08 06:00:08 <hazardaj_nombroj> `xfce4-pulseaudio-plugin` should work for you
2026-04-08 06:00:21 <hazardaj_nombroj> if you have `pipewire-pulse`, it will work flawlessly
2026-04-08 06:05:08 <seti_> Im considering the nvk drivers, i dont do anything heavy and just want something that wont crash randomly
2026-04-08 06:05:28 <seti_> are they in some repo in alpine or do i need to build them myself
2026-04-08 06:12:26 <hazardaj_nombroj> think nouveau is the only option right now? https://wiki.alpinelinux.org/wiki/NVIDIA
2026-04-08 06:20:55 <dwfreed> seti_: according to https://docs.mesa3d.org/drivers/nvk.html so long as you're on 3.22 or newer, you're already using NVK
2026-04-08 06:20:56 <rnkn> hazardaj_nombroj: awesome, that works
2026-04-08 06:21:21 <rnkn> hazardaj_nombroj: is this the only option? I have an irrational disdain for pulseaudio
2026-04-08 06:21:43 <hazardaj_nombroj> pipewire provides a pulseaudio interface, you're not installing any actual pulseaudio
2026-04-08 06:21:47 <dwfreed> ^
2026-04-08 06:22:02 <dwfreed> seti_: but note that NVK is the userspace graphics interface to the GPU; it is not the kernel driver.  The kernel driver is still Nouveau
2026-04-08 06:22:23 <seti_> ah
2026-04-08 06:22:38 <seti_> :'(
2026-04-08 07:06:03 <frag> why doesnt keyboard/mouse work with default mdev setup in xorg?
2026-04-08 07:08:08 <ikke> Perhaps because it does not have any rules for those kinds of devices?
2026-04-08 07:10:24 <frag> phaps
2026-04-08 08:47:42 <Sertonix[m]> frag: Did you try adding the user to the input group and which ownership the input devices have?
2026-04-08 09:48:28 <rnkn> I see the syncthing-openrc is a system-level service; is this the right way to use syncthing instead of starting it at the user level?
2026-04-08 09:58:26 <hazardaj_nombroj> huh, apparently flatpak needs an update to 1.16.4, there's a vulnerability that allows for sandbox escape
2026-04-08 09:58:39 <phws> rnkn: all I can say is I'm using it that way and I can confirm it actually spawns a new syncthing process running as the syncthing user
2026-04-08 09:59:11 <phws> but I'm not qualified to properly assess any security concerns compared to running it as a user directly
2026-04-08 10:22:04 <ikke> rnkn: user services are relatively new and still experimental. So most packages do not provide them yet
2026-04-08 10:36:26 <ivy-> if i want to run a few KVM VMs on Alpine for a test/dev environment, what should i be looking at?  i'd prefer command-line management that doesn't involve XML, if possible...
2026-04-08 11:09:31 <lopid> https://wiki.alpinelinux.org/wiki/KVM might be a start
2026-04-08 11:11:09 <Sertonix[m]> Depending on your needs you could write a shell script that runs qemu with various arguments. I use this script to quickly setup temporary VMs: https://codeberg.org/sertonix/alpine-scripts/src/branch/main/alpine-virt
2026-04-08 11:21:41 <Shiz> I wrote some simple scripts to manage qemu vms a while ago https://codeberg.org/shiz/vm-utils
2026-04-08 11:21:48 <Shiz> still using them in my own production
2026-04-08 11:24:57 <Shiz> example of the node i'm ircing for, using a microvm with direct kernel boot https://txt.shiz.me/NmIzZDc1MG
2026-04-08 11:39:34 <rnkn> phws: you don't get any permission errors when the syncthing user tries to access/change files in your $HOME?
2026-04-08 11:41:15 <phws> ah, I keep all its files in its user folder
2026-04-08 11:41:49 <phws> but I guess it makes sense that it would not be able to do that by default, no
2026-04-08 11:42:53 <rnkn> I have it syncing ~/Documents and ~/Pictures
2026-04-08 11:48:00 <dopice> RavFX[m]: fission: If you check https://wiki.alpinelinux.org/wiki/VLAN you'll see that the vlan package is no longer needed and conflicts with ifupdown-ng. The vlan fuctionallity is now incorporated into ifupdown-ng. It was a while back now but once I migrated I removed the vlan package before the upgrade. Upgraded, rebooted and I did not have to touch the interface config.
2026-04-08 11:48:24 <phws> rnkn: I think you should be able to change the user it runs as by defining SYNCTHING_USER in /etc/conf.d/syncthing
2026-04-08 12:00:52 <ivy-> thanks.  i'm trying to avoid libvirt... maybe i'll just write my own scripts
2026-04-08 12:08:14 <dopice> ivy-: Is it important that it's KVM? I have a machine running Alpine and Xen. It works outstanding. You don't have to fight with XML and I personally prefer Xen over KVM. The only thing that could be better regarding Xen is the documentation...
2026-04-08 12:08:45 <dopice> But if the machine is used for other things as well, maybe Xen isn't the best way to go...
2026-04-08 12:10:17 <socksinspace> i think, if you don't want libvirt, then using qemu more directly, either with shell scripts, or just from it's own cli the cli is your best option, years ago i used quickemu to save myself from fiddling with the raw qemu cli, not sure how/what that project is doing these days 
2026-04-08 12:11:12 <ivy-> dopice: this is ppc64le, i don't think Xen works there
2026-04-08 12:15:36 <socksinspace> ok, i would assume that disqualifies quickemu, afaik the focussed on aarch64 and x86_64
2026-04-08 12:16:10 <socksinspace> s/the/they/
2026-04-08 12:17:11 <ivy-> quickemu looks interesting but a bit more magical than i'm looking for, i was hoping for something like vm-bhyve on freebsd, if anyone knows that.  i'll probably just write something myself though
2026-04-08 12:18:28 <dopice> ivy-: Ah, okay! I got no Linux experice with ppc64le except AIX and IBMi.
2026-04-08 12:18:53 <ivy-> i'm actually not 100% sure kvm works here since i haven't tested it yet :D but someone told me it does
2026-04-08 12:19:27 <socksinspace> i'm 99.9% sure it should work
2026-04-08 12:23:02 <Sertonix[m]> You can check qemu-system-ppc64 -accel help
2026-04-08 12:25:59 <socksinspace> i was going by my foggy recollection of https://www.qemu.org/docs/master/about/build-platforms.html#supported-host-architectures
2026-04-08 12:30:40 <ivy-> yeah, that lists kvm, so we should be good
2026-04-08 12:34:12 <ctrl> Hello, I have a question about creating a custom ISO with mkimg. I need the image to autolog into a regular user. My profile uses a genapkovl script, that sets up a startup script with adduser and passwd. Is there a better way to do this?
2026-04-08 13:49:35 <ivy-> on FreeBSD, we have a device driver called nmdm(4) which is a two-ended virtual serial line (null modem), so you can connect a VM to one side, then connect to the other side using tip(1) or similar... does qemu on Linux have anything similar to that?
2026-04-08 13:50:07 <longnoserob[m]> pty?
2026-04-08 13:51:18 <ivy-> i'm not sure pty(4) is the same thing, doesn't that require specific ioctls to set up correctly?  nmdm can be used by anything expecting a serial line
2026-04-08 13:52:55 <longnoserob[m]> yeah
2026-04-08 13:53:02 <longnoserob[m]> maybe socat something?
2026-04-08 13:54:08 <ivy-> is that how people normally do VM serial consoles with qemu? 
2026-04-08 13:55:05 <longnoserob[m]> maybe qcom already has some option?
2026-04-08 13:55:31 <ivy-> what is qcom?  assuming not Qualcomm :-)
2026-04-08 13:55:51 <longnoserob[m]> erm qemu
2026-04-08 13:55:58 <ivy-> ah
2026-04-08 13:56:15 <longnoserob[m]> 
ACTION is working on sdm845 kernel testing

2026-04-08 13:56:35 <ivy-> i'll have a fiddle, although maybe -serial mon:stdio in tmux is the easiest way
2026-04-08 14:04:28 <Sertonix[m]> I think you can have qemu create a socket and then connect to it with something like netcat
2026-04-08 14:05:06 <ivy-> i was trying to avoid that since i don't like random unauthenticated listening sockets
2026-04-08 14:06:22 <Sertonix[m]> With unix domain sockets you kind of have authentication through access
2026-04-08 14:13:43 <Sertonix[m]> I have this in my notes as example: -serial unix:qemu-serial-socket,server,nowait
2026-04-08 14:17:12 <Sertonix[m]> If it should be a good out of the box experience then I guess you are right with tmux being the easiest way
2026-04-08 14:17:32 <socksinspace> the "character device options" section of the qemu docs list a whole boatload of possibilities
2026-04-08 15:09:39 <Shiz> ivy-: 
2026-04-08 15:09:41 <Shiz> chardev stdio,id=log0,signal=on
2026-04-08 15:09:43 <Shiz> chardev socket,id=tty0,mux=on,path=$consolefile,server=on,wait=off
2026-04-08 15:09:45 <Shiz> mon chardev=tty0,mode=readline
2026-04-08 15:09:47 <Shiz> is what I do
2026-04-08 15:09:50 <Shiz> and then the consolefile itself can be subject to normal unix permissions/acl :p
2026-04-08 15:10:16 <Shiz> ah, together with:
2026-04-08 15:10:18 <Shiz> device virtio-serial-device
2026-04-08 15:10:20 <Shiz> device virtconsole,chardev=tty0
2026-04-08 15:10:22 <Shiz> device virtconsole,chardev=log0
2026-04-08 16:11:00 <Renan[m]> 
ACTION uploaded an image: (16KiB) < https://matrix.org/oftc/media/v1/media/download/ARgyseHDiiR-fCFQ2FSxMp6uo0r8FaiOdbkeilimDrT04RRwa0ssU2zE64xyal7UGnR6JccyZMSzT-_a-6f0QFtCedtfWH-AAHVucmVkYWN0ZWQub3JnL2Z0bUNHS0JpWXpnaHFPVHVsTUljc1pOWg >

2026-04-08 16:11:01 <Renan[m]> what i need to do?
2026-04-08 16:11:13 <lopid> briefly describe the problem
2026-04-08 16:15:06 <Renan[m]> lopid: its happen when i try to see an encrypted hd
2026-04-08 16:15:21 <lopid> what is?
2026-04-08 16:15:26 <Renan[m]> i put the password and this happen
2026-04-08 16:16:00 <Renan[m]> lopid: cant you see the image?
2026-04-08 16:16:11 <lopid> can't you describe it?
2026-04-08 16:16:48 <dwfreed> an image of text, how nice
2026-04-08 16:17:02 <lopid> 
ACTION , justified

2026-04-08 16:17:13 <dwfreed> anyway, the message is saying you don't have udisks installed
2026-04-08 16:17:17 <Renan[m]> lopid: i cant acess the hd
2026-04-08 16:17:34 <dwfreed> err, wait, misread
2026-04-08 16:18:04 <Renan[m]> https://upload.disroot.org/r/LT8LrGJs#jgn9CqIBbGDF/wzcob2JhVUwmgGCYL3EkeCxzMObdCo=
2026-04-08 16:18:04 <Renan[m]> if someone cant see the image
2026-04-08 16:18:20 <Renan[m]> An error occurred while accessing '438.1 GiB Encrypted Drive', the system responded: An unspecified error has occurred: No such interface “org.freedesktop.UDisks2.Filesystem” on object at path /org/freedesktop/UDisks2/block_devices/dm_2d3
2026-04-08 16:18:25 <Renan[m]> lol, i can copy
2026-04-08 16:18:35 <lopid> hooray
2026-04-08 16:18:51 <Renan[m]> my english is poor
2026-04-08 16:19:14 <lopid> does lsblk see it?
2026-04-08 16:19:59 <Renan[m]> lopid: yes. i can see the disk but i cant acess it
2026-04-08 16:20:11 <dwfreed> pastebin the lsblk output
2026-04-08 16:23:09 <Renan[m]> dwfreed: localhost:~$ lsblk... (full message at <https://matrix.org/oftc/media/v1/media/download/AQUaqBb1_PNCq5v-c7JwkCS2pBkd4PNXXwlOugUhFTGP4pwRbro-dJbtSLnqSj5h5QJwlQOvAgY45Fvb124nPz5CedtgCmkgAG1hdHJpeC5vcmcvVEtjcVl2RnNkblpuaVFmZGVNRnVUcnRD>)
2026-04-08 16:26:54 <qbit> hola
2026-04-08 16:30:25 <penguinz_rule[m]> apk info | grep udisks
2026-04-08 16:30:43 <penguinz_rule[m]> if that returns nothing, theres your error
2026-04-08 16:30:54 <dwfreed> they have udisks, I misread the error
2026-04-08 16:31:12 <dwfreed> the error is saying there is not a filesystem object at the disk path
2026-04-08 16:31:21 <penguinz_rule[m]> ah
2026-04-08 16:31:28 <dwfreed> (would not have gotten the disk path if udisks isn't installed)
2026-04-08 16:32:47 <Renan[m]> penguinz_rule[m]: localhost:~$ apk info | grep udisks
2026-04-08 16:32:47 <Renan[m]> udisks2
2026-04-08 16:32:47 <Renan[m]> udisks2-libs
2026-04-08 16:32:47 <Renan[m]> udisks2-bash-completion
2026-04-08 16:32:47 <Renan[m]> udisks2-openrc
2026-04-08 16:32:57 <Renan[m]> i have udisks installed
2026-04-08 16:33:15 <qbit> I am trying to bootstrap alpine on a google wifi device ( https://openwrt.org/toh/google/wifi ) - i don't yet have a suzyq cable.. anyone have experience doing something similar?
2026-04-08 16:34:10 <Renan[m]> 
ACTION uploaded an image: (87KiB) < https://matrix.org/oftc/media/v1/media/download/AeaVFS6Nqor8hMxoziU8VONlQJgSN6HG_kJ6Se7SR1pdNKmr0ftDRlkU3q-5ONFTZ-VFZJzPpIKg1hiiXjX9-45Cedtgq8CAAHVucmVkYWN0ZWQub3JnL3pCY0lOZEhGVVFrR2Zhbm1iUlpFc1lrTA >

2026-04-08 16:34:14 <Renan[m]> i have this error in flatpak too
2026-04-08 21:06:45 <amadaluzia> Is alpine rootfs statically linked?
2026-04-08 21:07:31 <amadaluzia> I'm pretty sure everything is dynamically linked but I want to be certain
2026-04-08 21:51:44 <quinq> $ file /bin/busybox
2026-04-08 21:51:44 <quinq> /bin/busybox: ELF 64-bit LSB pie executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib/ld-musl-x86_64.so.1, BuildID[sha1]=660ba2ec89871a1e0e6b5c9e6a57fbf9250a9925, stripped
2026-04-09 00:31:24 <amadaluzia> quinq: I see. I assume you would need to rebuild the rootfs in order to have a statically built Alpine Linux system, or has someone else already done that?
2026-04-09 00:41:12 <WhyNotHugo> amadaluzia: I suspect you'd need to build it yourself. Why do you want it to be statically linked?
2026-04-09 00:49:26 <amadaluzia> LD_PRELOAD attack mitigations
2026-04-09 00:51:05 <dwfreed> if they can LD_PRELOAD, they can just run a different binary
2026-04-09 00:56:31 <longnoserob[m]> stere0
2026-04-09 01:04:33 <mlody> Hi, is there typescript language server in Alpine repositories>
2026-04-09 01:04:34 <mlody> ?
2026-04-09 01:07:06 <mlody> It's in npm, but I don't trust npm.
2026-04-09 01:11:54 <dwfreed> I mean, Alpine would basically be pulling from NPM
2026-04-09 01:17:54 <comk> dwfreed: well, if packaged in alpine it would pulll the sources from github and do its own packaging. all microsoft but can still make a difference with supply chain trust 
2026-04-09 01:18:28 <comk> mostly because git sources are more visbible and easy to review
2026-04-09 01:19:43 <WhyNotHugo> Nobody's reviewing the entire source code for things like tsserver. I think it's more sensible to run in inside docker or podman (that's what I do).
2026-04-09 01:20:00 <WhyNotHugo> Note that you'll still be pulling dependencies from npm anyway
2026-04-09 01:20:08 <comk> maybe not from scratch but many eyes on the full diffs
2026-04-09 01:20:25 <dwfreed> comk: so they just commit the vuln to the git repo
2026-04-09 01:20:35 <dwfreed> most of the recent exploits have been committed
2026-04-09 01:23:28 <WhyNotHugo> You can review tsserver, but will yuo also review its 37 build dependencies, which are pulled as binaries from npm?
2026-04-09 01:23:39 <comk> WhyNotHugo actually, yes
2026-04-09 01:24:20 <dwfreed> WhyNotHugo: is that recursive, or not? because if not, those 37 dependencies probably have another 37 dependencies each :>
2026-04-09 01:24:45 <comk> well it does force ne to be more conservative with their stack
2026-04-09 01:24:46 <dwfreed> before you know it, you've spent a year reading half the JS ecosystem, and you still haven't built tsserver
2026-04-09 01:25:02 <comk> i had ripped out axios from the last place over a year ago already
2026-04-09 01:25:14 <WhyNotHugo> dwfreed: it does not include recursive dependencies.
2026-04-09 01:25:20 <dwfreed> :)
2026-04-09 01:25:30 <dwfreed> and in that year, half those dependencies have gone through half a dozen updates
2026-04-09 01:25:56 <WhyNotHugo> and the build system from tsserver unconditionally pulls and uses binaries from npm. You'd have to rewrite the build system to "truly" build from source.
2026-04-09 01:26:11 <WhyNotHugo> I hear it's being rewritten in go :P
2026-04-09 01:26:40 <dwfreed> So one nightmare to another :D
2026-04-09 01:27:02 <WhyNotHugo> Go actually builds from source, I don't think it ever pulls binaries.
2026-04-09 01:27:15 <comk> WhyNotHugo: There is an option: build deps from scratch too and substitute private npm registry. 
2026-04-09 01:27:35 <dwfreed> WhyNotHugo: Sure, but it's still a nightmare depgraph
2026-04-09 01:27:39 <comk> then curse about the packages that are not reproducble and t mismatching hashes and deal with those individually :p
2026-04-09 01:28:40 <WhyNotHugo> comk: it's defienitely an option
2026-04-09 01:29:23 <comk> ...i guess this sums up why nobody took the effort to package typescript-language-server for alpine yet (:
2026-04-09 09:20:42 <fossman[m]> Hi. Where do I report issues with the setup-desktop scripts?
2026-04-09 09:21:38 <fossman[m]> After plasma setup it does not boot into the login manager as it used to before...
2026-04-09 09:27:51 <puretryout[m]1> I'm not sure that would be a bug with the setup-desktop script. I changed SDDM yesterday to fix a race condition, perhaps that broke it instead for you
2026-04-09 10:27:46 <fossman[m]> I remember seeing an error message in red saying something like sddm not found towards the end of the setup
2026-04-09 10:27:54 <fossman[m]> * 
2026-04-09 11:11:28 <realroot[m]> does make prefix=/usr  go in build() only? to avoid /usr/local
2026-04-09 11:28:03 <ikke> realroot[m]: Depends on the makefile, but if it's provided to ./configure, it's persistent and not necessary to repeat
2026-04-09 11:28:16 <monotux[m]> why do you want to avoid /usr/local?
2026-04-09 11:28:59 <realroot[m]> in only build it did not work i'll try both then
2026-04-09 11:32:37 <ikke> monotux[m]: packages would not install anything in /usr/local
2026-04-09 11:33:51 <monotux[m]> ah I thought he just wanted to install some custom software by compiling it himself, maybe I missed the backlog
2026-04-09 11:34:12 <monotux[m]> when I install non-packaged stuff I put it into /usr/local to avoid issues with future updates
2026-04-09 11:35:52 <ikke> monotux[m]: the `in build()` inplies APKBUILD, which implies packaging it
2026-04-09 11:36:26 <monotux[m]> my mistake, sorry!
2026-04-09 11:48:53 <frag> is it worth it to use mdev for a dwm setup? got it working with this, except if i unplug/replug mouse it no longer works, also when computer falls asleep mouse/kb seem to stop working... https://x0.at/7WaF.txt
2026-04-09 12:02:19 <Sertonix[m]> The replug issue is probably due to the hardcoded paths. I suspect that they will be event5 and larger when.
2026-04-09 12:03:26 <frag> some old documentation suggested using /dev/input/mice, but didnt seem to work for me
2026-04-09 12:03:30 <Sertonix[m]> Is that kind of specific device configurarion not avoidable?
2026-04-09 12:04:59 <frag> maybe sertonix, im just trying random tips from the internet
2026-04-09 12:06:16 <Sertonix[m]> If you really need something like /dev/input/mice you could add a rule for that to /etc/mdev.conf
2026-04-09 12:08:19 <Sertonix[m]> Removing or commenting out the "Device" lines might be worth a try
2026-04-09 12:09:17 <frag> in /usr/share/X11/xorg.conf.d/40-libinput.conf ??
2026-04-09 12:10:48 <Sertonix[m]> Yes, I think
2026-04-09 12:12:58 <frag> okok, will try later, ty!
2026-04-09 12:13:30 <Sertonix[m]> The evdev config section here might help as well https://xorg-team.pages.debian.net/xorg/howto/configure-input.html
2026-04-09 12:21:29 <frag> dont understand it, but looks promising!
2026-04-09 12:32:38 <Sertonix[m]> I think I found the problem. The generic device identification uses MatchIsPointer which according to some page is implemented using udev
2026-04-09 12:36:51 <frag> oh, so it wont work with mdev?
2026-04-09 12:37:44 <ivy-> i'm not sure if this is on-topic here, but does anyone know how to make btop show usage graph for all CPU threads?  currently it's only showing 0-164, but the system has 176 in total: https://people.freebsd.org/~ivy/tmp/btop.png
2026-04-09 12:40:07 <Sertonix[m]> frag: There is no good udev device db for mdev+libudev-zero at the moment. For a specific system it is probably possible to setup mdev.conf rules for the necessary devices.
2026-04-09 12:43:27 <Sertonix[m]> There seems to be a few generic checks in libudev-zero. But I don't know how to check if and why they fail
2026-04-09 12:44:09 <bdprom> frag: you could get away with mdev(d) for most things, but some apps will require libudev, e.g. pipewire.. there's libudev-zero, which helps with many other apps (not pipewire unfortunately :()
2026-04-09 12:45:31 <frag> my system is quite simple, so i think it might work, but will see how hard it is to get working, maybe eudev is better after all..
2026-04-09 12:46:24 <bdprom> postmarketos has packaged systemd-udevd, i'm working up the courage to try and bring it over to alpine
2026-04-09 12:47:01 <bdprom> unfortunately, if you want a well supported + maintained devd implementation that works on a "modern" desktop system, systemd-udevd is pretty much the only option afaik :(
2026-04-09 12:48:16 <frag> what to choose, freedom or convenience..
2026-04-09 12:49:40 <bdprom> indeed.. though tbf, at least we *have* options
2026-04-09 12:50:41 <ikke> ivy-: not sure, do you have an unusual cpu layout?
2026-04-09 12:51:33 <ivy-> ikke: i don't think it's very unusual, 2 sockets x 22 cores x SMT4
2026-04-09 12:52:37 <Sertonix[m]> It would be much simpler if X11/wayland would accept a generic human input device which may or may not have any number of buttons and pointers. But it seems like people want to group everything in "clear" categories such as keyboard or mice...
2026-04-09 12:55:51 <ikke> ivy-: what does `cat /sys/devices/system/cpu/online` return?
2026-04-09 12:57:39 <ivy-> hyacinth# cat /sys/devices/system/cpu/online
2026-04-09 12:57:40 <ivy-> 0-175
2026-04-09 13:00:21 <ikke> ivy-: and `grep processor /proc/cpuinfo | wc -l`?
2026-04-09 13:00:33 <ivy-> hyacinth# grep processor /proc/cpuinfo | wc -l
2026-04-09 13:00:34 <ivy-> 176
2026-04-09 13:00:40 <ikke> No idea then
2026-04-09 13:02:42 <ivy-> so apparently a workaround is to just make the terminal taller... but i'd rather it used more space for cpu and less space for everything else
2026-04-09 13:03:04 <ikke> Ok, so it's just a UI thing
2026-04-09 13:51:31 <socksinspace> the cpu layout may not be too outlandish, but the amount of threads i would qualify as "above average", to put it lightly :D
2026-04-09 14:52:46 <ivy-> socksinspace: probably, although... can't you get way more threads than that in a single socket on x86 nowadays?
2026-04-09 14:58:55 <dwfreed> ivy-: I've never used btop, can you manually change the size of the sections?
2026-04-09 14:59:54 <ivy-> dwfreed: that's what i was trying to work out, i don't see anything obvious in the help
2026-04-09 14:59:57 <longnoserob[m]> check the nice menu?
2026-04-09 15:07:49 <dwfreed> I don't see anything either; it's probably an off by 1 bug
2026-04-09 15:07:59 <dwfreed> because 1 more row would show all the CPUs
2026-04-09 15:08:25 <dwfreed> (btw, in 5th gen epyc, the largest single socket thread count is 384)
2026-04-09 15:08:47 <ivy-> if i make the window smaller, it only shows 131 CPUs, so i think this is "expected" behaviour, just not very useful
2026-04-09 15:09:08 <ivy-> i wonder if the size of the memory/network panels is simply fixed
2026-04-09 15:10:02 <dwfreed> or at least has a minimum
2026-04-09 15:40:43 <socksinspace> ivy-: *i* can't get that many threads, because i'm broke af, but sure, big corpos can get crazy amounts of threads, not sure they use btop to monitor their sytems tho :P
2026-04-09 15:41:16 <ivy-> trust me, i couldn't afford this either, the vendor donated it for testing :D
2026-04-09 15:43:25 <socksinspace> i did wonder how a regular person found room for a computer that i estimate to be >20K$ in their budget
2026-04-09 16:50:12 <roselandgoose> it seems that simply installing vulkan-tools or mesa-vulkan-nouveau breaks himitsud/hiprompt-gtk ...
2026-04-09 16:50:52 <roselandgoose> can anyone who understands the graphics stack point me towards where I should be looking to understand/debug that?
2026-04-09 16:51:25 <roselandgoose> I'm not used to the mere presence of a package breaking running programs/preventing them from launching successfully
2026-04-09 16:52:06 <roselandgoose> there's nothing in /var/log/messages that I register as relevant (which includes sway's output via logger)
2026-04-09 16:52:37 <roselandgoose> I have one NVIDIA GPU enabled, running on nouveau
2026-04-09 16:52:44 <roselandgoose> edge, upgraded earlier today
2026-04-09 16:53:52 <roselandgoose> (it was much less surprising to see that adding WLR_RENDERER=vulkan to my sway launch script caused it to fail, although also without helpful logs yet)
2026-04-09 18:48:23 <frag> is it a problem to have this at the end of the $ENV file?  "$0", or generally source a file with functions and this on the end?
2026-04-09 19:20:07 <lopid> it is if you don't want to run another shell instance
2026-04-09 19:36:09 <frag> ah, thats why i had such a complex code.. dont need it anyway
2026-04-09 22:15:08 <Noisytoot> <socksinspace> i did wonder how a regular person found room for a computer that i estimate to be >20K$ in their budget
2026-04-09 22:15:12 <Noisytoot> You can get second-hand POWER9 systems for <$2k sometimes
2026-04-09 22:15:59 <Noisytoot> or <$3.5k at least (I know someone who has one and won't say how much it cost exactly, but less than the minimum usable raptor blackbird system, which is $3.5k)
2026-04-09 22:22:03 <socksinspace> aye, so it's only 3k outside my budget ;P
2026-04-09 22:26:09 <penguinz_rule[m]> i am assuming radare2/iaito became slopware since it has been updated in edge
2026-04-09 22:27:01 <penguinz_rule[m]> also i noticed a endResetModel error on some objects being called on v3.23 iaito. unsure if this reproduces. i just have an alpine binary called and whenever i select a function i get that error printed and then a crash
2026-04-09 22:28:32 <penguinz_rule[m]> s/(binary) called/\1 opened with it
2026-04-09 23:34:18 <histopson[m]> MASTER THE ART OF HACKING 🕹️... (full message at <https://matrix.org/oftc/media/v1/media/download/AQSwXtOFZ061ignaPXXAnW9ZBpxJVg1-YGjo4iqfTdTZk-qmofbdxAUwzf0kf0ySjVdBTG-H8mMDE8pl1hUm6sdCedvLG-GAAG1hdHJpeC5vcmcvZmJDeGZOYmV2Z3BiSHN0d1VxcXZiV2hN>)
2026-04-10 06:34:52 <frag> think i asked this before long ago and the answer is no, but is there a way to specify running a function, just like you can specify running a command with 'command <command>'
2026-04-10 07:24:26 <rnkn> I know this is a longshot, irrational even, but has anyone got Davinci Resolve to work on Alpine?
2026-04-10 07:28:05 <rnkn> ah I'll just use Kdenlive
2026-04-10 09:37:04 <bdprom> frag: you mean in the shell? yeah, just define it in one of your startup files using 'my_function() { ... }', then call it from the prompt using 'my_function arg1 arg2...'
2026-04-10 09:44:18 <ikke> bdprom: From my understanding of their question is that, imagine you have an alias defined called my_function as well, how do you force it to use the function instead of alias, similar like how `command my_function` would force it to use an executable instead of function or alias
2026-04-10 09:44:26 <ikke> And I'm not aware of any way to do that
2026-04-10 09:48:36 <fission> $ foo() { echo 'foo'; }
2026-04-10 09:48:36 <fission> $ foo
2026-04-10 09:48:36 <fission> bar
2026-04-10 09:48:36 <fission> $ \foo
2026-04-10 09:48:36 <fission> $ alias foo='echo bar'
2026-04-10 09:48:38 <fission> foo
2026-04-10 09:49:00 <ikke> Ah, \ works with functions as well
2026-04-10 09:49:12 <fission> I didn't know it but just tried it now
2026-04-10 09:50:22 <bdprom> oh i didn't know that either, i wonder how portable that is
2026-04-10 09:50:58 <fission> I only tested with /bin/ash
2026-04-10 09:51:23 <fission> but just tested /bin/sh on my macOS box and behaves the same
2026-04-10 09:51:29 <fission> so possibly posixy
2026-04-10 10:16:07 <frag> yes i like functions with short one char names, so would be easier to search for them in script if i did something like 'runfunction c', but eh... works either way
2026-04-10 12:59:41 <Sertonix[m]> You could do runfunction() { "$@"; } and just have that as a convention in your code.
2026-04-10 14:03:34 <Alice_> Hello there! I've been reading the initramfs init script [https://github.com/alpinelinux/mkinitfs/blob/master/initramfs-init.in] and noticed that a lot of variables like $ROOT and $MOCK are used without ever being defined. Since /init is the first process and the environment is empty, I assume unset variables just expand to an empty string. Is that the intended behavior? And if that's the case, why do these variables even exist? Would appreciate any
2026-04-10 14:09:24 <quinq> Yes I suppose
2026-04-10 14:09:41 <quinq> They exist because then you'd be able to run the same script in a different environment, by defining those variables
2026-04-10 14:10:06 <quinq> For example you could define ROOT and then the script would treat that path as root, instead of directly '/'
2026-04-10 14:10:45 <quinq> And MOCK, I assume, would be to simulate commands instead of actually calling them and modifying the current environment
2026-04-10 14:10:57 <quinq> (for example MOCK=echo)
2026-04-10 14:12:19 <Alice_> Alright, thanks for the clarification! It's my first time using Linux, so I'm trying to understand how this whole thing works. The MOCK=echo idea is honestly really smart, I hadn't even thought of that...
2026-04-10 14:15:14 <quinq> Don't worry, that's what real free software is about, sharing information, helping others, etc. :)
2026-04-10 14:25:52 <lopid> aw
2026-04-10 15:55:26 <edsonwolf> Good afternoon. How do I enable history in an LXC container on Alpine Linux?
2026-04-10 15:58:04 <ikke> What kind of history?
2026-04-10 21:31:13 <edsonwolf> There are only two commands left in .ash_history. https://pastebin.com/QER9agBU
2026-04-11 07:24:53 <popefucker2> if I want to compile a slightly modified kernel, I just do 'make olddefconfig' with the changed config option in .config right?
2026-04-11 07:59:04 <frag> has anyone gotten mdev to work with replugging mouse/kb in xorg? think ill stick with eudev, but i guess my mind has not totally given up.. even if the mouse get the same /dev/input/event number it wont work...
2026-04-11 11:40:38 <andrewm> hello guys, anybody has broken sddm and plasma after updating to the latest qt?
2026-04-11 11:48:57 <courge_musque[m]> sddm is broken for months for me
2026-04-11 11:49:17 <courge_musque[m]> What's happening in your case?
2026-04-11 11:58:59 <andrewm> courge_musque[m]: in most cases on system boot it just doesn't start when trying to launch it manually it hangs and I can't switch to another tty, sddm.log looks fine tho I don't see any errors, plasma fails to start with errors like "Not connected to D-Bus server" but I'm starting it with dbus-run-session and qdbus6 has output, busctl shows that there are connections, then it tries to do 
2026-04-11 11:58:59 <andrewm> shell integration but fails as well, I even tried to set QT_WAYLAND_SHELL_INTEGRATION manually but that didn't help too
2026-04-11 12:01:55 <courge_musque[m]> I had a very similar issue, but it was really random and sometimes I got it on a computer and not on another
2026-04-11 12:01:58 <courge_musque[m]> Good luck :(
2026-04-11 12:24:09 <andrewm> that's sad I can't use my linux installation anymore because of that, switched from windows only to use alpine :(
2026-04-11 12:34:35 <courge_musque[m]> Maybe try to install a DE that uses lightdm, like xfce
2026-04-11 12:34:56 <courge_musque[m]> Disable the sddm service at boot, and replace it by lightdm
2026-04-11 12:52:29 <andrewm> well I checked output with plugins debug and I see that alpine plasma-integration package is outdated and it always fails with version mismatch 396032 395776
2026-04-11 12:52:57 <andrewm> alpine package is 6.6.4 but upstream is 6.6.8
2026-04-11 12:53:12 <andrewm> maybe it's the real reason
2026-04-11 13:12:12 <andrewm> YES I fixed it
2026-04-11 13:12:26 <andrewm> compiled newer package version and it works now
2026-04-11 13:28:12 <andrewm> well with notable exception that keybinds don't work at all
2026-04-11 13:37:58 <broycraft> I'm trying to run a diskless image loading zfs and zfs-lts, but I'm noticing a mismatch between the kernel verion of linux-lts in the 3.23 repository and the most current .iso build for 3.23.3. Is it possible to update the kernel in a diskless install without having an .iso for it?
2026-04-11 13:43:09 <broycraft> I was able to get my wifi working by unpacking/repacking the modloop squashfs container to add linux-firmware modules. But zfs-lts are kernel modules, and the repo version doesn't match the .iso version for a diskless install.
2026-04-11 14:37:53 <broycraft> Or, alternatively, am I better off just building an .iso image myself including the required modules and firmware?
2026-04-11 15:24:25 <Sertonix[m]> Depending on what you want you can update kernel/modloop/firmware with update-kernel or you can generate a custom iso.
2026-04-11 15:28:04 <Sertonix[m]> But using the official iso without the kernel from the iso does not seem right.
2026-04-12 05:24:36 <rnkn> anyone know of linux running on a macbook neo?
2026-04-12 06:55:59 <jpp1> rnkn: Your only option with the Neo would be a VM
2026-04-12 11:23:41 <penguinz_rule[m]> speaking of packaging, is putting stuff into /opt ok or no. i am thinking of packaging opensmalltalk-vm for my own potential downstream aports and idk how else i would package it
2026-04-12 11:24:17 <rnkn> jpp1: drat, I'm looking to reduce complexity 
2026-04-12 11:24:26 <ikke> penguinz_rule[m]: No, packages should not put anything in /opt, abuild by default even does not allow it
2026-04-12 11:25:05 <ikke> penguinz_rule[m]: for your own repo, ofcourse you are free to do whatever you want
2026-04-12 11:31:17 <penguinz_rule[m]> i ask about /opt because by default the squeak executable builds in its /building/linux64x64/squeak.cog.spur/build with a bunch of it s own libraries stored there
2026-04-12 11:31:44 <penguinz_rule[m]> so if i just copied that build directory to /opt it would be simple enough, the user would just have to add to path
2026-04-12 11:33:24 <penguinz_rule[m]> i assume the course of action then is to copy build to /usr/share/squeak and then symlink squeak to /usr/bin or so? since ik it works when you symlink $0
2026-04-12 12:30:37 <f_> jpp1 rnkn: there might be asahi support
2026-04-12 12:33:55 <f_> if there isn't now there may be later I guess
2026-04-12 12:50:05 <jpp1> f_ rnkn https://asahilinux.org/
2026-04-12 12:51:12 <f_> yes I know what asahi is :)
2026-04-12 12:51:49 <f_> there isn't support right now, maybe don't ask Neo-related questions in their channels right now
2026-04-12 13:26:47 <andar1an> is anyone else having issues with pid file creation for init services recently?
2026-04-12 13:27:59 <andar1an> I see the file created, but get fopen errors when trying to init service. I can get it to work here and there, but there is no seeming reason why it doesn't work, even if not using a command_user
2026-04-12 13:28:39 <invoked> is this for a package you created, or all installed packages?
2026-04-12 13:28:49 <andar1an> it is specifically for anubis
2026-04-12 13:29:03 <andar1an> but curious it it affects any others
2026-04-12 13:29:27 <invoked> i don't run anubis but i'm not seeing any pid file issues elsewhere
2026-04-12 13:29:31 <andar1an> I can create an issue if it is specific, I just don't know what is going on yet
2026-04-12 13:29:47 <andar1an> the defauly config will start, but that is it
2026-04-12 13:30:33 <andar1an> and it doesn't appear to be an anubis error, anubis will start with same config without openrc
2026-04-12 13:31:03 <andar1an> one can use a command_user, check path on a directory and set correct permissions, and specifiy pid location, but still issue. 
2026-04-12 13:31:22 <quinq> Maybe anubis is DOS'ing your PID files
2026-04-12 13:31:58 <andar1an> I have tried command_background=true because it seems poorly behaved
2026-04-12 13:32:03 <andar1an> didn't really help
2026-04-12 13:33:09 <invoked> if you can cat the init.d file > tpaste  (or something) that might tell us more
2026-04-12 13:33:43 <andar1an> yes, I can share, 1 sec. Would you like original or modified?
2026-04-12 13:34:13 <invoked> the original, at least
2026-04-12 13:35:02 <andar1an> original: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/anubis/anubis.initd
2026-04-12 13:35:34 <andar1an> current state of tinkering: 
2026-04-12 13:35:58 <andar1an> (is it possible to do multiline in irc?)
2026-04-12 13:36:05 <invoked> it is, but don't
2026-04-12 13:36:10 <invoked> use a paste service
2026-04-12 13:36:21 <andar1an> ok, is that tpaste you mentioned?
2026-04-12 13:36:46 <invoked> yes, you can install it.  apk add tpaste
2026-04-12 13:39:38 <invoked> but anyway, based on the packaged init, it should work. something is strange on your system
2026-04-12 13:39:57 <andar1an> the default does work
2026-04-12 13:40:23 <andar1an> it is the moment you cnange even something like a port that it starts to misbehave
2026-04-12 13:42:08 <andar1an> my current tinker state: https://tpaste.us/QKpQ
2026-04-12 13:46:45 <invoked> name=$RC_SVCNAME  why are you doing this?
2026-04-12 13:47:29 <invoked> i don't think that's the problem, but it's weird :)
2026-04-12 13:47:37 <andar1an> because anubis will require separate instances per subdomain behind a reverse proxy, so it is easy to just create a new init file based on name
2026-04-12 13:47:59 <andar1an> just copy pasta and change a few details
2026-04-12 13:48:39 <invoked> ok. i don't see anything obviously wrong there.
2026-04-12 13:48:48 <andar1an> me neither, that
2026-04-12 13:48:54 <invoked> at least with the init script
2026-04-12 13:48:56 <andar1an> 's why I am pulling out hair lol
2026-04-12 13:49:00 <invoked> i don't know what anubis expects
2026-04-12 13:49:17 <andar1an> it isn't anubis erring which is weird
2026-04-12 13:49:24 <andar1an> I can run with same config on cli
2026-04-12 13:49:52 <andar1an> and verbose and debug openrc init indicates a pid file is missing, but I can look right at it haha
2026-04-12 13:50:01 <invoked> yeah but in the init script you are feeding it  : ${plcyfile:="$cfgdir/media.botPolicies.yaml"}
2026-04-12 13:50:10 <invoked> is that a typo and it should be "policyfile"
2026-04-12 13:50:27 <andar1an> I just tried to keep it short
2026-04-12 13:50:41 <andar1an> it is plcyfile referenced in checkpath start_pre
2026-04-12 13:51:09 <andar1an> ends up being /etc/anubis/policyfilename.yaml
2026-04-12 13:51:12 <invoked> right
2026-04-12 13:52:09 <andar1an> I have tried via an env config in /etc/anubis, as well as env vars via conf.d. It just is weird not getting any explicit detail as to what is failing
2026-04-12 13:52:16 <andar1an> only thing I get is PID error
2026-04-12 13:52:35 <invoked> i dunno. i'd suggest opening a gitlab about it
2026-04-12 13:52:43 <invoked> you're probably in a hurry though
2026-04-12 13:53:04 <andar1an> no, not really. I can create an issue. 
2026-04-12 13:53:42 <andar1an> just haven't had time to get back into mr's yet. This is all trying to get my forgejo public for job apps lol
2026-04-12 13:54:14 <andar1an> It was cool to watch the insane amount of bot traffic handled by the default though.
2026-04-12 13:54:38 <andar1an> will create now, thank you @invoked
2026-04-12 13:54:58 <invoked> i didn't fix anything but you're welcome
2026-04-12 13:55:10 <andar1an> knowing I am not losing my mind fixed my brain haha
2026-04-12 13:55:14 <andar1an> :P
2026-04-12 13:56:02 <longnoserob[m]> stere0
2026-04-12 13:56:37 <andar1an> Do you think it would be worth suggesting in issue that it may be nice to include policy config as part of default? 
2026-04-12 13:57:12 <andar1an> currently it does everything via env variable, but anubis should also be referencing a policy file. There is a substantial default included in upstream repo
2026-04-12 13:57:48 <invoked> sure, why not
2026-04-12 14:10:58 <andar1an> could nftables cause an issue with fopen? I would think no, but now I am wondering
2026-04-12 14:14:28 <andar1an> I don't think that makes sense anyways, because same nftables config did allow a working instance, so ignore. My bad
2026-04-12 14:30:40 <andar1an> rough issue create, will refine as I dig in if anything comes up. Thanks again
2026-04-12 15:25:22 <f_|mo> andar1an: envvar is outdated now I think
2026-04-12 15:26:45 <f_> oh oop you left
2026-04-12 17:31:41 <inmary[m]> https://t.me/+pa-CiKYv9-5lNWM8
2026-04-12 17:31:41 <inmary[m]> Get in touch with this platform for greatness you’ll definitely thank me later
ℹ️❤️
👇🏻👇🏻👇🏻
2026-04-12 17:34:42 <quinq> How do we get in touch?
2026-04-12 17:42:32 <ssm> quinq: high frequency cosmic ray directed into scammer's storage device
2026-04-12 17:44:46 <quinq> 
ACTION starts polishing the mirror antennas

2026-04-13 01:55:16 <rnkn> jpp1: thanks but that's only for M series chips, the neo uses an A series (iPhone)
2026-04-13 02:04:40 <jpp1> ok
2026-04-13 02:34:05 <Advert[m]> https://t.me/+c-LC9ed_hBgwYmZh
2026-04-13 07:49:15 <Atque> Oh well, it's getting colder. I wonder if more people use alpine in the winter.
2026-04-13 08:10:28 <socksinspace> i feel like we have more activity from the northern hemisphere in here, so for many it's getting warmer now
2026-04-13 10:58:38 <courge_musque[m]> Or Alpine is just working so well that nobody has to ask question on IRC :(
2026-04-13 11:01:14 <cousin_luigi> socksinspace: Actually it's getting temporarily colder here.
2026-04-13 11:10:09 <socksinspace> cousin_luigi: yeah, bit of a rainy day here in germany as well
2026-04-13 12:09:37 <xolitude> my pc boots into TTY and gnome doesn't run after following wiki.alpinelinux.org/wiki/Securing_Alpine_Linux and rebooting
2026-04-13 12:22:24 <invoked> xolitude: sounds probable that you disabled a service you need. since the wiki page doesn't explicitly say which ones are "unnecessary"
2026-04-13 12:23:37 <invoked> that wiki page has some problems. it doesn't cover ipv6, assumes people have rsyslog installed, doesn't suggest which services could be unnecessary.
2026-04-13 12:24:21 <invoked> you should probably try backing out your changes 1 thing at a time until you figure out what broke things.
2026-04-13 15:48:30 <kaathewise> Q: is there a guide on setting up and using opencl on Alpine?
2026-04-13 15:49:42 <kaathewise> I tried installing mesa-rusticl, but this just caused clinfo to start segfaulting, so I'm not sure if I'm holding it wrong or if I should report it upstream
2026-04-13 16:09:29 <courge_musque[m]> Gpu model?
2026-04-13 16:12:14 <kaathewise> None, I'm on a laptop
2026-04-13 16:21:35 <courge_musque[m]> There's still a gpu
2026-04-13 16:22:03 <courge_musque[m]> Or you are cpu-rendering, and I don't think any laptop does that
2026-04-13 16:38:21 <kaathewise> AMD Radeon 890M, an iGPU.  But I'm only using the compute shaders
2026-04-13 16:39:22 <kaathewise> I was able to successfully launch Vulkan compute shaders on top of mesa-dri-gallium before, so theoretically it should be possible 
2026-04-13 17:17:46 <courge_musque[m]> 890m is recent and well supported, so maybe a package issue?
2026-04-13 17:18:04 <courge_musque[m]> Would be weird that it went unnoticed upstream
2026-04-13 17:42:58 <kaathewise> Yeah, I decided to look deeper into it after Your message, and turns out it is indeed rusticl
2026-04-13 17:43:59 <kaathewise> rusticl is still unstable, and it is the cause of the clinfo segfault
2026-04-13 17:51:16 <kaathewise> Thanks for the help!  I'll go see if I can make a proper bug report for the upstream
2026-04-13 17:56:08 <realroot[m]> anybody tried ladybird in alpine?
2026-04-13 17:56:24 <realroot[m]> Repository      : AUR Name            : ladybird Version         : 20260320-2
2026-04-13 18:02:32 <cousin_luigi> mail: Cannot create lockfile /var/mail/user.lock: Error creating temporary lockfile
2026-04-13 18:02:38 <cousin_luigi> Is this normal?
2026-04-13 18:11:50 <quinq> cousin_luigi, maybe /var/mail doesn't exist
2026-04-13 19:30:55 <cousin_luigi> quinq: It's there, along with the mailbox for my user
2026-04-13 19:31:06 <cousin_luigi> only I can't delete anything from it
2026-04-13 19:32:00 <cousin_luigi> I even created it by hand with the right permissions but mail still complains
2026-04-13 19:34:52 <quinq> ok
2026-04-14 01:28:08 <popefucker2> question about AppImages: if it links against glibc and gcompat isn't enough, do I just need to make my own AppImage with a different distro?
2026-04-14 07:25:36 <uocci> ciao
2026-04-14 07:25:41 <uocci> !list
2026-04-14 07:45:46 <jpp1> Personally, I don't use AppImages; I trust the official packages more.
2026-04-14 07:56:47 <PureTryOut> popefucker2: you'll need an appimage that links against Musl, so yes you'll have to make your own. But there are enough problems with Appimage that I too would recommend staying away from it and favor solutions like the distro packages or Flatpak
2026-04-14 08:20:23 <popefucker2> Yeah after poking around in appimage land I've come to the same conclusion
2026-04-14 10:23:50 <jammy> hallo
2026-04-14 10:39:14 <jammy> nobody up i see
2026-04-14 10:49:12 <longnoserob[m]> have some patience
2026-04-14 10:50:47 <jammy> ill wait i didnt sleep tho
2026-04-14 11:14:07 <realroot[m]> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1118055 i ahve this with offlineimap
2026-04-14 11:14:23 <realroot[m]>  module 'importlib' has no attribute 'machinery'
2026-04-14 11:16:42 <realroot[m]> https://github.com/OfflineIMAP/offlineimap3/pull/205
2026-04-14 11:17:53 <realroot[m]> they released https://github.com/OfflineIMAP/offlineimap3/releases/tag/v8.0.2
2026-04-14 11:17:53 <realroot[m]> achill ^
2026-04-14 11:18:16 <realroot[m]> OfflineIMAP v8.0.1 (2025-10-28)
2026-04-14 11:44:42 <quinq> Hi
2026-04-14 11:45:04 <quinq> I'm using pipewire (because it's the new fad and web browsers require it) and firefox for $werk meetings
2026-04-14 11:45:20 <quinq> Since a few weeks they tell me that my sound is horrible, it cuts, it's clipped, etc.
2026-04-14 11:45:41 <quinq> On two different computers (same Edge/amd64 though)
2026-04-14 11:46:03 <quinq> Am I the only one with that problem? Maybe I'm missing something to do on top of this?
2026-04-14 15:46:19 <bdprom> quinq: is it just a problem with your mic? playback is ok? try running `ps -Heo nice,args` and see if pipewire/wireplumber has a nice value of < 0 (-11 i think usually?)
2026-04-14 15:47:21 <bdprom> i'm not sure pipewire is actually required for firefox btw? alsa works fine for playback only, and my webcam worked fine without pipewire last time i checked (didnt test a call, or mic)
2026-04-14 15:49:48 <quinq> bdprom, I don't think it is, as I said it's the same on two different hardware, and it was working fine before
2026-04-14 15:50:19 <quinq> No, everything's running at nice 0
2026-04-14 15:51:10 <mid> quinq: that may help: I'm not using the mic on Alpine, though I have same kind of problem on another distro (using FF with KDE and pipewire), and I found that each time after a reboot I have to use alsamixer (!) to lower the capture level of my webcam which is (re)set to the max. I don't know why but without this tweak I have exactly the same sound problems in calls
2026-04-14 15:51:42 <quinq> bdprom, firefox works with apulse, but when there's the slightest load on the machine (which happens always as it's doing video in a web-browser), firefox decides to drop the line and it doesn't recover from that (sound just stops working)
2026-04-14 15:52:02 <bdprom> quinq: https://wiki.alpinelinux.org/wiki/PipeWire#Realtime_scheduling
2026-04-14 15:52:08 <bdprom> maybe try that first
2026-04-14 15:52:15 <quinq> mid, that might help, thanks!
2026-04-14 15:52:29 <bdprom> you seem to be missing realtime privileges, so if the cpu is loaded, pipewire will suffer, cut out, stutter, etc
2026-04-14 15:52:36 <bdprom> mid's idea is good for clipping too 
2026-04-14 15:52:49 <quinq> oh, is that new?
2026-04-14 15:52:50 <bdprom> i reduce my source volume in pipewire to 0.40
2026-04-14 15:52:57 <quinq> Because I did look at the wiki in order to find a solution
2026-04-14 15:53:01 <quinq> Must have missed that
2026-04-14 15:53:06 <quinq> Will try with the pipewire group
2026-04-14 15:53:16 <bdprom> good luck :)
2026-04-14 15:53:46 <quinq> Thank you very much too! :)
2026-04-14 15:54:32 <mid> on my setup it's independent of pipewire mic level setting, that is what is weird, I can see this capture volume to the max only in alsamixer… well, give it a try if you can, just to be sure
2026-04-14 15:54:38 <mid> well, hope that helps
2026-04-14 16:01:04 <quinq> okkk
2026-04-14 16:01:10 <quinq> I just tested with a coworker
2026-04-14 16:01:15 <quinq> Looks like the group thing worked
2026-04-14 16:20:07 <bdprom> nice! rtkit+polkit would also work afaik, but i prefer not running extra daemons unless necessary
2026-04-14 16:29:33 <quinq> same
2026-04-14 19:26:13 <popefucker2> Firefox works fine with just pulseaudio
2026-04-14 19:26:23 <popefucker2> On my machine :^)
2026-04-14 19:36:48 <mlody> popefucker2: Cool nickname :D
2026-04-14 19:37:14 <lopid> no, m'lody
2026-04-14 20:04:12 <swee> where is ldap support on cyrus-sasl?
2026-04-14 20:04:17 <swee> (saslauthd)
2026-04-14 20:09:55 <popefucker2> Thanks mlody, I think I probably registered popefucker ages ago and forgot the password. Unsure
2026-04-14 20:10:14 <popefucker2> Either that or there are other popefuckers out there
2026-04-14 20:27:38 <swee> oh and also, i saw that cyrus-sasl doesn't have a config file for saslauthd
2026-04-14 20:52:08 <mid> swee: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/41484
2026-04-14 20:58:21 <swee> mid: i need to compile that myself?
2026-04-14 21:02:37 <mid> I meant I made an MR for this but it was refused (for the reasons explained in the comments)
2026-04-14 23:13:37 <swee> spectacle still segfaults in alpine edge after I just ran apk upgrade
2026-04-15 10:47:49 <prabu> quinq, the need to create necessary groups has been added to the pipewire wiki page.
2026-04-15 11:46:56 <realroot[m]> modprobe mhi_pci_generic modprobe: FATAL: Module mhi_pci_generic not found in directory
2026-04-15 11:54:52 <f_> realroot[m]: reboot
2026-04-15 11:57:00 <realroot[m]> hi f_ it is not that
2026-04-15 11:57:22 <f_> I'm not sure, you just gave us this error without any context
2026-04-15 11:57:28 <f_> explain :)
2026-04-15 11:57:53 <realroot[m]> i am trying to run modprobe mhi_pci_generic
2026-04-15 11:58:07 <realroot[m]> I've read that I need that driver
2026-04-15 11:58:20 <f_> need it for what?
2026-04-15 11:59:06 <realroot[m]> i modprobed mhi_net
2026-04-15 12:00:51 <realroot[m]> is this one https://github.com/torvalds/linux/blob/master/drivers/bus/mhi/host/pci_generic.c?
2026-04-15 12:03:28 <jvvv> realroot[m]: what arch and kernel are you currently running?
2026-04-15 12:04:00 <realroot[m]> x86_64 6.18.22-0-lts
2026-04-15 12:05:03 <realroot[m]> https://gist.github.com/clouedoc/14af89b947ddf7805e4b1765ba025cbf 
2026-04-15 12:05:03 <realroot[m]> >Driver 	mhi-pci-generic
2026-04-15 12:05:09 <jvvv> # MHI_BUS_PCI_GENERIC is not set
2026-04-15 12:05:51 <jvvv> for the linux-lts kernels, it is only set for aarch64
2026-04-15 12:07:04 <realroot[m]> do I have to recompile  the kernel right? it's been a lot
2026-04-15 12:10:38 <f_> yes and you can request for that to be enabled
2026-04-15 15:21:39 <WhyNotHugo> The old linux-edge kernel used to expose /dev/dmesg, but linux-stable does not. Was this intentionally removed, or was it just a side-effect of aligning with linux-lts?
2026-04-15 15:24:04 <achill> a side effect
2026-04-15 15:24:41 <WhyNotHugo> An MR to enable that would have to keep both aligned, right?
2026-04-15 15:28:50 <WhyNotHugo> Er, hang on, it's /dev/kmsg. 
2026-04-15 15:31:30 <Sertonix[m]> I vaguely remember linux-edge exposing dmesg to unpriviliged users. But might be wrong
2026-04-15 15:34:53 <achill> yeah, so the unrestricted dmesg is something i'd avoid, but i dont know any downsides of /dev/kmsg
2026-04-15 16:15:04 <nanon> 📯 https://alpinelinux.org/posts/Alpine-3.20.10-3.21.7-3.22.4-3.23.4-released.html
2026-04-15 16:23:04 <WhyNotHugo> i mixed the two up; the unrestricted dmesg is a kernel config and is still off. my kmsg simply had wrong permissions.
2026-04-15 17:09:12 <realroot[m]> still compiling with max priority how long
2026-04-15 17:10:02 <realroot[m]> does it takr for abuild and linux-lts?
2026-04-15 18:30:07 <realroot[m]> in aarch64 it was way quicker weird
2026-04-15 18:30:36 <realroot[m]> i should benchmark cpu
2026-04-15 19:25:21 <realroot[m]> 345m
2026-04-15 19:34:53 <popefucker2> > in aarch64 it was way quicker weird
2026-04-15 19:34:53 <popefucker2> That's not too unusual, x64 has many more drivers available usually
2026-04-15 19:36:43 <realroot[m]> does it take longer in alpine? I did it in x86_64 some time ago with a very crappy CPU
2026-04-15 19:46:53 <popefucker2> I compiled linux-virt the other day on my laptop, took about 30 minutes
2026-04-15 19:47:13 <popefucker2> Maybe a little more. I didn't time it
2026-04-15 19:57:24 <realroot[m]> that was something like 4.19 iirc
2026-04-15 20:16:41 <realroot[m]> i did compile that too
2026-04-15 20:16:42 <realroot[m]> not on purpose
2026-04-15 20:23:29 <realroot[m]> <realroot[m]> "https://gist.github.com/clouedoc..." <- does that work only with m.2 to pcie? and not m.2 to usb?
2026-04-15 22:58:05 <shadow> hi all, I'm submitting a request to Rust language team to promote RISC-V support musl target to (Tier 2, as it is now and addition of) "with Tools"
2026-04-15 22:59:09 <shadow> I think this is somewhat Alpine related, as Alpine is used frequently in CI/CD docker image workflows for musl-ness so those familiar with Alpine Linux distro and Rust language may want to get involved?
2026-04-15 22:59:53 <shadow> it's issue 982 at rust-lang/compiler-team on their github page if so :)
2026-04-15 23:03:19 <shadow> the downstream-upstream of that is rust-lang/docker-rust which publishes images only for "with tools" supported targets of Rust language, noticeably Alpine Linux is missing there although Debian (unknown-gnu, not musl) images are available.  So... for wider adoption of Alpine and RISC-V I'm hopeful that https://github.com/rust-lang/docker-rust/issues/272 will be possible if the request to compiler-team is successful
2026-04-15 23:51:26 <ivy-> i am slightly concerned about the status of my zfs pool: https://people.freebsd.org/~ivy/tmp/zfs.txt - why does it show the device as nvme1n1 instead of nvme1n1p4?  
2026-04-15 23:52:25 <shadow> how do you mean
2026-04-15 23:52:48 <kiki[m]> does anybody know the best dektop/windows manager for alpine?
2026-04-15 23:52:57 <kiki[m]> i was thinking of dwl but idk
2026-04-15 23:53:19 <shadow> nvme1n1 is the whole-block device and pN are the partitions
2026-04-15 23:53:21 <ivy-> shadow: in the zpool status output, it lists the device as nvme1n1, implying it's using the entire device (?) but it should only be using nvme1n1p4
2026-04-15 23:53:58 <shadow> oh I see it now
2026-04-15 23:53:59 <ivy-> maybe this is a quirk of linux zfs?  on freebsd it shows the partition:           vtbd0p3   ONLINE       0     0     0
2026-04-15 23:55:00 <popefucker2> > does anybody know the best dektop/windows manager for alpine?
2026-04-15 23:55:00 <popefucker2> I'm biased towards sway
2026-04-15 23:55:06 <popefucker2> It just werks
2026-04-15 23:55:21 <shadow> I've never used zfs userspace tools so I'm not sure either
2026-04-15 23:55:23 <kiki[m]> interesting
2026-04-15 23:55:44 <kiki[m]> is hyprland on alpine any good?
2026-04-15 23:55:55 <popefucker2> Idk
2026-04-15 23:56:52 <shadow> ivy-: does this happen to be a desktop machine with GNOME ? what does Gnome disks utility say about it
2026-04-15 23:57:04 <ivy-> shadow: no, it's a remote server without a GUI
2026-04-15 23:57:23 <shadow> oh okay given a moment I could probably conjure up a way to do this on commandline
2026-04-15 23:57:26 <ivy-> as far as i can tell, everything is working fine, but i was going to do some repartitioning and noticed this and now i'm a bit worried
2026-04-15 23:59:39 <shadow> ivy-: I don't know what to make of this, can you review http://blog.asiantuntijakaveri.fi/2024/01/fix-zfs-device-names.html and tell us if that sounds applicable?
2026-04-16 00:00:26 <ivy-> shadow: i don't think that's related... i imported the pool that was created on freebsd, it was never a whole-disk pool
2026-04-16 00:00:31 <shadow> also not recommended to follow those instructions just review and I'm trying to figure out something from this
2026-04-16 00:04:40 <shadow> I guess if you transition to WNN identifiers then it will be clearer what is going on
2026-04-16 00:05:13 <shadow> https://almonde.org/posts/zfs-setting-up-zpool/   and yes sorry if I'm posting links to web searches with content that are already familiar or not applicable
2026-04-16 00:05:26 <ivy-> well, the WWN refers to the LUN, i'm not sure whether that would affect whether it includes the partition number or not
2026-04-16 00:05:44 <shadow> that's probably why then
2026-04-16 00:05:57 <shadow> your concern makes sense from UX perspective
2026-04-16 00:07:31 <shadow> I don't know these commands, ivy- what is the output of zpool list  in addition to status which you shared with us?
2026-04-16 00:07:57 <ivy-> NAME   SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOT
2026-04-16 00:08:00 <ivy-> data  1.81T   306G  1.51T        -       48G    10%    16%  1.00x    ONLINE  -
2026-04-16 00:17:25 <shadow> @ivy- the terminology you are looking for is known as "expanded space" in ZFS
2026-04-16 00:17:56 <shadow> read through the man page for zpool-list 8
2026-04-16 00:18:44 <ivy-> hm, i know how zfs expansion works, but i don't think that's relevant to this?
2026-04-16 00:19:15 <shadow> or I totally misunderstood that teriminology, seems that is likely
2026-04-16 00:19:16 <shadow> there's three examples, one of them expands the pool listing to show the components of the pool thing that does the other stuff
2026-04-16 00:20:43 <ivy-> i'm not trying to expand the pool, though
2026-04-16 00:20:43 <shadow> the third example suggests a json output with all fields 
2026-04-16 00:21:23 <shadow> no I get that I meant how to find this in the man page, it's near that example. Also I misunderstood the context of "expansion" .... it's referring to that other thing you are familiar with and I am not aware of
2026-04-16 07:11:32 <burnix>  hi
2026-04-16 07:11:54 <burnix> is there a good tutorial to make alpine from scratch ??
2026-04-16 07:19:44 <longnoserob[m]> did yiu check the wiki?
2026-04-16 07:36:32 <Justin[m]> Does the Alpine ISO work with Ventoy USB booting?
2026-04-16 07:37:05 <Justin[m]> Nevermind, it does, just sat on Mounting boot media, please wait .... for a fair while.
2026-04-16 07:59:26 <burnix> not now but i'm going to see but i have to report bugs on gitlab
2026-04-16 08:00:21 <burnix> i try to replace debian with alpine, but alpine doesnt work on VM system.
2026-04-16 08:01:46 <burnix> The virt iso doesnt work on hyperv (all versions). only the standard 64 iso work on win1126h2-pro hyperv and vmware 25h2 
2026-04-16 08:20:12 <Justin[m]> Can I set Alpine to not require network on boot? 
2026-04-16 08:24:46 <Justin[m]> I think I got it figured, I moved wpa_supplicant from boot to default so they don't block boot.
2026-04-16 09:06:46 <kiki[m]> <Justin[m]> "Can I set Alpine to not require..." <- yes just leave blank or type none
2026-04-16 09:07:02 <kiki[m]> not install tho
2026-04-16 09:14:36 <omni> sounds like some boot order dependency issue?
2026-04-16 09:15:19 <Justin[m]> omni: It is a laptop and the WiFi on the system is quite flaky.
2026-04-16 09:15:35 <Justin[m]> So it tries to connect but if the adapter disappears it just sits and waits.
2026-04-16 09:25:46 <omni> oh, ok
2026-04-16 09:26:34 <omni> btw, for wifi I can recommend iwd over wpa_supplicant, perhaps that would also help in this case?
2026-04-16 10:48:27 <realroot[m]> once i have the kernel modules modem should work right?
2026-04-16 13:33:29 <qbit> hio
2026-04-16 13:33:43 <qbit> seems the gpg sig for x86_64 miniroot is missing: https://dl-cdn.alpinelinux.org/alpine/v3.23/releases/x86_64/alpine-minirootfs-3.23.4-x86_64.tar.gz.asc
2026-04-16 13:37:41 <ikke> ncopa: ^
2026-04-16 13:44:16 <ncopa> uploaded. thanks.
2026-04-16 13:44:40 <ncopa> the checksum for minirootfs s390x failed
2026-04-16 13:54:05 <qbit> yay, ty
2026-04-16 14:28:32 <qbit> is my understanding of aports' pkgrel correct?: if you bump pkgver, pkgrel should go to 0
2026-04-16 14:29:05 <Biswa96[m]> correct
2026-04-16 14:29:11 <qbit> (coming from OpenBSD we have REVISION, which follows the above pattern - but with removal entirely)
2026-04-16 14:29:12 <qbit> sweet, ty
2026-04-16 14:56:47 <qbit> bah, should have checked for existing stuff https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/99810
2026-04-16 14:57:59 <qbit> IANAL: but the distribution clause hasn't been changed at all.. so it should be fine to include? (we do in openbsd: https://codeberg.org/OpenBSD/ports/src/branch/master/net/py-rns/Makefile#L13 )
2026-04-16 19:05:29 <runxiyu> a
2026-04-16 20:32:41 <Saijin_Naib[m]> @whynothugo , is DE-specific stuff support in-scope for darkman? Yeah, right, since the user has to provide the scripts?
2026-04-16 20:33:16 <Saijin_Naib[m]> I'm thinking about switching off xfce4-night-mode genmon script, and learning more about darkman
2026-04-16 20:34:06 <Saijin_Naib[m]> It seems it only triggers scripts on day/night events, right? If so, already way better as genmon is a continual set interval
2026-04-16 21:30:09 <datakinds> hi team. running into a problem setting up an alpine container inside docker that I cannot find info about elsewhere. the dockerfile is super simple -- just uses apk to download `openssh` and runs `/usr/sbin/sshd -Dddd`. no change to the config. after failing auth on an initial connection, the sshd server will crash with the error "mm_reap: child exited with status 255". anyone seen behavior like this before? can upload f
2026-04-16 21:31:14 <datakinds> it's not a seccomp issue, that fails with a different status code (31). that's the closest behavior I could find elsewhere online
2026-04-16 21:40:41 <popefucker2> since I updated the other day, WPA_supplicant needs me to pass -C /run/WPA_supplicant or it won't create the runtime control socket
2026-04-16 21:40:59 <popefucker2> Sorry mobile wants me to capitalize WPA
2026-04-16 21:41:56 <popefucker2> I'll try and file a report of some kind but just thought I'd alert people here in case I'm not the only one
2026-04-16 22:00:42 <popefucker2> Apparently a closed issue since 2022. Don't know why it wasn't in my config tho
2026-04-17 00:01:18 <WhyNotHugo> Saijin_Naib[m]: you can manually toggle at any time with `darkman toggle`.
2026-04-17 00:02:12 <WhyNotHugo> There's also a desktop entry so you can "toggle darkman" via a launcher.
2026-04-17 00:02:13 <Saijin_Naib[m]> Sweet, thanks. So I could keep a shortcut for it for manual toggle, so keep my current UX
2026-04-17 00:02:25 <Saijin_Naib[m]> Ah, awesome
2026-04-17 00:02:39 <Saijin_Naib[m]> Can it set envvars?
2026-04-17 00:03:06 <Saijin_Naib[m]> So, it could set GTK_THEME and other apps would see that envvar?
2026-04-17 00:03:28 <WhyNotHugo> It can run any custom script as hooks. You'd need to write a script that maybe sets a value for this variable somewhere that apps read at launch?
2026-04-17 00:04:08 <WhyNotHugo> IIRC: gtk4 uses the xdg-dekstop-portal to read the dark/light mode (darkman has integration for it), and for gtk3 you can just change the current theme to a light/dark variant.
2026-04-17 00:04:56 <WhyNotHugo> Changing GTK_THEME for running apps is not really feasible.
2026-04-17 00:12:39 <mewt> Hi, I am trying to mount my EFI partition before doing an install in a chroot. Attempting to mount it gives "codepage cp437 not found" in dmesg
2026-04-17 00:13:23 <mewt> What do I need so that cp437 will work?
2026-04-17 00:14:05 <mewt> well, if I can find out what codepages are supported by the kernel on the installation media and select that one, that works too
2026-04-17 00:18:50 <Saijin_Naib[m]> Ah, okay, didnt know if being an openrc service meant setting envvars meant it would behave differently than say, doing so in a termimal as my user, which only affects apps launched from that terminal/session
2026-04-17 00:19:59 <Saijin_Naib[m]> I am playing with patched libadwaita (libadwaita-without-adwaita) and setting that manually does help gtk4 apps respect my Yaru theme
2026-04-17 00:23:59 <jvvv> mewt: What environment are you working from?
2026-04-17 00:25:17 <mewt> alpine-standard-3.23.3-x86_64.iso
2026-04-17 00:26:57 <jvvv> Just a moment, I'll boot a vm with that so I can follow along better.
2026-04-17 00:33:26 <WhyNotHugo> Saijin_Naib[m]: environment variables affect that process which sets them and their children/descendants.
2026-04-17 00:33:45 <WhyNotHugo> openrc has some magic where one service "exports" them to a file and other services "import" them from that file.
2026-04-17 00:33:54 <WhyNotHugo> But the root behaviour is the same.
2026-04-17 00:34:30 <Saijin_Naib[m]> Ah, okay, so that rough edge remains. 
2026-04-17 00:34:31 <Saijin_Naib[m]> It seems the DE itself would need to manage that, then
2026-04-17 00:37:12 <mewt> thanks, happy to provide info once you're ready, just lmk
2026-04-17 00:47:19 <jvvv> mewt: I've booted it up, just partitioning etc to catch up
2026-04-17 00:50:14 <mewt> ok i understand the problem now
2026-04-17 00:50:55 <mewt> /lib/modules is a symlink to /.modloop/modules
2026-04-17 00:51:03 <mewt> ./modloop should be the flashdrive's 2nd partition
2026-04-17 00:51:08 <mewt> this *also* fails, with the same problem
2026-04-17 00:51:27 <mewt> do I need to download some package to get the module to load, or something, here?
2026-04-17 00:53:10 <jvvv> Are you trying to mount it before or after you  chroot?
2026-04-17 00:53:16 <mewt> before
2026-04-17 00:53:51 <mewt> cp437 is needed for the partition on the flashdrive to mount too, the symlink goes nowhere because .modloop can't be mounted appropriately out of the gate afaict
2026-04-17 00:55:12 <mewt> is there something i need to install with apk to be able to deal with vfat properly
2026-04-17 00:55:48 <mewt> i already have dosfstools as you might have gleaned
2026-04-17 00:58:45 <mewt> does ls -lh /lib/modules give "no such file or directory" for you on boot
2026-04-17 01:01:25 <jvvv> Huh, weird, it mounted with the codepage no problem for me.
2026-04-17 01:02:56 <jvvv> I was hoping to reproduce what you are seeing.
2026-04-17 01:03:10 <mewt> i'll try one reboot
2026-04-17 01:04:12 <mewt> oh, now it's back...why
2026-04-17 01:07:03 <mewt> ok, and it stays up once apk can install things..hm
2026-04-17 01:08:42 <mewt> ok, and it mounts. sorry about that! I'm not sure how I caused modloop to die, nor the codepage support...
2026-04-17 01:09:03 <mewt> onward I guess, thanks for trying it
2026-04-17 01:10:19 <jvvv> no problem, transient issue or something in the system being flaky, hard to say what caused it.
2026-04-17 01:10:37 <mewt> maybe some piece of the setup script kills modloop and I ran that without remembering...
2026-04-17 01:11:13 <jvvv> seems plausible
2026-04-17 01:29:15 <mewt> ...oh, even stupider. I did umount -a
2026-04-17 02:05:16 <jvvv> yeah, I get my share of face-palms, too.
2026-04-17 02:08:41 <mewt> one other thing, how does mkinitfs select what volumes to unlock? It persistently goes for the wrong one, and doesn't seem to obey the same /etc/default/grub config that works for arch
2026-04-17 02:10:25 <mewt> (my standard setup, I guess I can just say, is to have luks2 on separate boot/"everything-else" drives, with pbkdf2 set as the pbkdf for the boot one since grub can't deal with argon2id yet
2026-04-17 02:11:00 <mewt> gets into early userspace fine, then tries to unlock boot again and gives up because it expects /dev/myvg/root to exist already when that's the other volume
2026-04-17 02:16:35 <jvvv> I've always just followed https://wiki.alpinelinux.org/wiki/Setting_up_encrypted_volumes_with_LUKS. I never setup encrypted drives before I started using alpine as my daily driver
2026-04-17 02:18:24 <jvvv> And I've never encrypted anything other than root and some removables, never any boot partitions
2026-04-17 02:37:37 <mewt> well, it's not hook order, I've ruled that out
2026-04-17 02:37:49 <mewt> i can get it to boot if I just unlock the volume and mount on /sysroot
2026-04-17 02:38:57 <mewt> that's not a very sustainable way to do things but it worked, I guess
2026-04-17 04:42:53 <Saijin_Naib[m]> Jesus, librewolf is not an easy bump
2026-04-17 04:43:05 <Saijin_Naib[m]> Sertonix , I tried, but I failed. Sorry.
2026-04-17 04:59:36 <prabu> Saijin_Naib, if your needs are simple, i.e occasional dark mode switching, i've a script that does the toggle and it works almost everywhere..https://git.sr.ht/~prabuanand/dotfiles/blob/master/.config/sway/scripts/toggle_theme
2026-04-17 05:04:06 <Saijin_Naib[m]> That looks great, but I am using xfce so I would need to adapt, I think
2026-04-17 05:08:05 <prabu> i use a minimal desktop with sway and mostly gtk applications..sure it will need some customizing for Xfce.. see if it helps.. i just did not want to add another application for this occassional theme switching...
2026-04-17 05:23:23 <rnkn> does Firefox use GTK settings to get its light/dark theme (when set to automatic)?
2026-04-17 05:31:06 <Saijin_Naib[m]> It should yes
2026-04-17 05:31:17 <Saijin_Naib[m]> It seems to here when using xfce4-night-mode via GenMon
2026-04-17 06:10:06 <Saijin_Naib[m]> Darkman is sweet!
2026-04-17 06:10:41 <Saijin_Naib[m]> Adapted the settings from xfce4-night-mode, and it works brilliantly. Would example scripts to cover all of xfce settings be welcome in the repo, WhyNotHugo?
2026-04-17 06:28:56 <Saijin_Naib[m]> Now just gotta figure how to change WINE theme and KDE/QT theme, haha
2026-04-17 07:36:00 <Sertonix[m]> [@saijin_naib:matrix.org](https://matrix.to/#/@saijin_naib:matrix.org) I think we need to drop the librewolf package
2026-04-17 07:50:49 <rnkn> Saijin_Naib[m]: are you running wayland?
2026-04-17 07:51:35 <rnkn> I'm trying `gsettings set org.gnome.desktop.interface color-scheme 'prefer-dark'` and it has no effect on Firefox
2026-04-17 08:00:26 <rnkn> I've now installed xdg-desktop-portal-wlr and added it to user rc runlevel gui
2026-04-17 08:00:30 <rnkn> still no dark mode
2026-04-17 08:04:55 <rnkn> I've installed Firefox with flatpak
2026-04-17 08:46:32 <rnkn> omg I did it
2026-04-17 08:46:48 <rnkn> 
ACTION cues "We Are the Champions" by Queen

2026-04-17 09:04:08 <hazardaj_nombroj> yeah, i think you also need the gnome settings daemon running or smth like that
2026-04-17 09:04:13 <hazardaj_nombroj> or gsd-desktop-schemas
2026-04-17 09:04:19 <hazardaj_nombroj> i had the same problems some time ago
2026-04-17 09:04:33 <hazardaj_nombroj> (though i used firefox installed directly, not through flatpak)
2026-04-17 09:09:07 <rnkn> xdg-desktop-portal-gtk
2026-04-17 09:09:23 <rnkn> and start the session with dbus-run-session
2026-04-17 12:44:31 <WhyNotHugo> Saijin_Naib[m]: thanks! yup, examples for these are always welcome.
2026-04-17 12:45:21 <WhyNotHugo> rnkn: iirc, firefox uses gtk3, so to change it, change the gtk theme to a -light or -dark variant
2026-04-17 14:18:03 <Saijin_Naib[m]> @Sertonix are you too overloaded? 
2026-04-17 14:21:41 <Saijin_Naib[m]> @WhyNotHugo sweet, I'll open an MR. They may need to be cleaned up, but I hope they are useful none-the-less. 
2026-04-17 14:22:17 <WhyNotHugo> Thanks. As long as they're clear enough for the next person to adopt them that's good enough. A lot of the examples hardcode a theme name, or whatnot. That's fine.
2026-04-17 16:08:29 <Sertonix[m]> <Saijin_Naib[m]> "@Sertonix are you too overloaded..." <- It is too much effort for too little use. Almost all changes can be done through firefox config files
2026-04-17 16:11:21 <Saijin_Naib[m]> Ahh... okay. Makes sense
2026-04-17 16:26:51 <cousin_luigi> bash won't use ~/.bashrc upon login. If I launch it again it will.
2026-04-17 16:26:53 <cousin_luigi> What gives?
2026-04-17 16:28:46 <cousin_luigi> oh nevermind, it was .bash_profile
2026-04-17 16:38:42 <Biswa96[m]> you can execute .bashrc in .bash_profile. it helps to keep stuff in one file.
2026-04-17 16:57:42 <jahway603[m]> RCvXQ!]^{xq'vu;JDn{L5mWyyF-c_ixD/c)Q2/5~=vrVEX:%1j
2026-04-17 17:27:22 <henrix> jahway603[m]: time to change password?
2026-04-17 17:28:05 <jahway603[m]> LOL
2026-04-17 17:28:23 <jahway603[m]> yep
2026-04-17 18:33:27 <omni> don't worry, I won't remember that
2026-04-17 21:12:45 <Saijin_Naib[m]> @Sertonix want to revive and maintain SeaMonkey so hexchat and thunderbird and liferea can be dropped since it does it all? 🤣
2026-04-17 21:13:12 <Saijin_Naib[m]> I really do want to revive that aport, though. I tried already, but could not get a build 😭
2026-04-17 21:20:32 <Sertonix[m]> (I removed hexchat already)
2026-04-17 21:59:37 <Noisytoot1> rip, I use hexchat
2026-04-17 21:59:46 <mid> +1
2026-04-17 22:03:35 <Sheila> https://zoitechat.zoite.net/ picks up where hexchat left off.
2026-04-17 22:50:17 <WhyNotHugo> Can I declare that a subpackage DOES support an architecture which the main pacakge does not?
2026-04-17 23:00:18 <ptrc> i think it's not gonna trigger the builders?
2026-04-17 23:01:06 <ptrc> you *could* try doing some weird hacks by overwriting arch= in package() maybe...?
2026-04-18 00:42:43 <omni> in a way the main package is a subpackage of the aport too..
2026-04-18 00:43:11 <omni> WhyNotHugo: for what aport is it you want to do this?
2026-04-18 00:44:07 <WhyNotHugo> omni: !97109. libpiper is a library, but piper-tts doesn't actually use it.
2026-04-18 01:08:32 <omni> WhyNotHugo: so you want libpiper available on more architectures than the python packages can be enabled on? (due to python dependencies not being available on all architectures)
2026-04-18 01:09:13 <omni> how about renaming the aport to libpiper and have the python parts be subpackages of that?
2026-04-18 01:09:48 <WhyNotHugo> That might be a good approach.
2026-04-18 01:10:05 <WhyNotHugo> Frankly, this looks like two separate projects in a single git repo, maybe they should just be two packages, despite using the same source?
2026-04-18 01:10:22 <WhyNotHugo> They only partially overlap in makedepends and have no mutual dependency.
2026-04-18 01:11:24 <omni> that happens, and if it is the same source it seems easier to maintain in a single aport
2026-04-18 01:24:26 <Saijin_Naib[m]> @Sertonix so it is decided, we bring back SeaMonkey 🤣
2026-04-18 01:45:07 <mid> Sheila: thanks!
2026-04-18 01:45:30 <Sheila> np
2026-04-18 04:13:48 <Saijin_Naib[m]> WhyNotHugo , just made a MR to add the new scripts I have that I _think_ cover everything I was able to do with xfce4-night-mode, minus some unmerged ones that toggled kvantum and qt6ct themes (never got those working correctly)
2026-04-18 04:14:16 <Saijin_Naib[m]> To darkman, sorry
2026-04-18 07:14:52 <realroot[m]> WARNING: gstreamer-ptp-helper-1.26.9-r0: failed to preserve usr/libexec/gstreamer-1.0/gst-ptp-helper: xattrs
2026-04-18 08:44:04 <rnkn> is alpine US-based?
2026-04-18 08:46:53 <rnkn> nvm, found the info on open collective
2026-04-18 08:48:40 <frag> based norwegian alpine distro
2026-04-18 08:58:29 <cousin_luigi> but the Alps are not in Norway...
2026-04-18 09:10:09 <frag> many ppl dont know what that means, me included before i watched the alpinist, nice movie btw, but not related to (alpine) linux
2026-04-18 09:41:16 <cousin_luigi> frag: Then why mention it? o.O?
2026-04-18 09:44:19 <frag> for fun and education, im so sorry
2026-04-18 09:45:50 <achill> rnkn: alpine isn't based on a location, there are volunteers all over the world, and as we dont have a legal entity, there is no location it cant be based on aswell
2026-04-18 09:58:13 <jpp1> For me, it's reminiscent of the Alpine Renault. :)
2026-04-18 10:36:49 <omni> https://en.wikipedia.org/wiki/Lyngen_Alps
2026-04-18 13:02:40 <cousin_luigi> jpp1: I remember that one.
2026-04-18 16:15:04 <realroot[m]> maybe it's like debian made by DEBra's boyfriend IAN, so ALPhea and ermINE
2026-04-18 16:15:45 <realroot[m]> <realroot[m]> "WARNING: gstreamer-ptp-helper-1..." <- `apk fix gstreamer-ptp-helper` ^
2026-04-18 18:16:11 <Ermine> ?
2026-04-18 18:49:49 <oleksiidn[m]> Hi does Alpine has Qt6Keychain ARM package? I can't find it.
2026-04-18 18:50:10 <oleksiidn[m]> s/has/have/
2026-04-18 19:01:07 <Biswa96[m]> the package name is qtkeychain
2026-04-18 19:40:24 <quinq> mupdf is back \o/
2026-04-18 19:40:36 <quinq> algitbot \o/
2026-04-18 20:56:07 <km> clear
2026-04-18 20:57:10 <quinq> roger
2026-04-18 21:15:27 <lopid> over
2026-04-18 21:32:26 <omni> sync
2026-04-18 22:44:16 <amorbts> hey guys good day
2026-04-18 22:48:11 <amorbts> sorry to bother you but my system is stuck at boot, last message is "* Starting local ..."
2026-04-19 05:13:28 <qaz_2025> Hello
2026-04-19 05:13:41 <qaz_2025> how to install amd-smi?
2026-04-19 05:23:31 <qaz_2025> aka rocm-smi
2026-04-19 06:00:06 <qaz_2025> there is a way to request the amd /rocm-smi package?
2026-04-19 06:00:18 <qaz_2025> tried to complie nut it stuck at 32%
2026-04-19 22:25:21 <wsenn> howdy, just installed alpine on my t430. the install went fine, no errors or anything, but on first boot, it failed on wlan supplicant. how to troubleshoot? 
2026-04-19 22:37:16 <longnoserob[m]> dmesg and logread -b 0 ?
2026-04-19 22:37:21 <longnoserob[m]> missing wlan-fw?
2026-04-19 23:14:30 <wsenn> looks like firmware missing - it's an intel hd4000 which is ancient and should be picked up as intel.
2026-04-19 23:16:40 <Ermine> longnoserob[m]: there's no logread on Alpine by default iirc
2026-04-19 23:23:56 <omni> wsenn: do you have linux-firmware-intel installed?
2026-04-19 23:25:20 <wsenn> don't think so, is this something i can do during the iso boot (before install is complete)? cuz right now, don't have internet :).
2026-04-19 23:25:46 <wsenn> on that machine, this is mint on my tuxedo laptop
2026-04-19 23:29:11 <longnoserob[m]> Ermine: oh i forgot
2026-04-19 23:56:32 <wsenn> hooked up eth0 and installed firmware, reboot and supplicant worked fine. off to trying it during install, cuz doing the find a cable, hook it up to download a file, disconnect dance is annoying.
2026-04-20 00:09:10 <wsenn> how to get write access for apk before reboot?
2026-04-20 02:54:23 <popefucker2> wsenn: you mean write access to the installed root fs? I think the best way to do that is mount the root partition on /mnt and do apk --root /mnt <commands>
2026-04-20 02:55:01 <popefucker2> or do you mean write access to the installation medium? I don't think that's possible
2026-04-20 09:31:10 <tobby> is there some handy way to run glibc depending software besides using its flatpak variant?
2026-04-20 09:34:26 <tobby> regarding laptop
2026-04-20 09:40:26 <courge_musque[m]> gcompat can work if you're lucky
2026-04-20 09:40:33 <courge_musque[m]> https://wiki.alpinelinux.org/wiki/Running_glibc_programs#Running_glibc_programs
2026-04-20 09:42:47 <tobby> oh Appimages work out of the box, sorry for being stoopid
2026-04-20 09:42:49 <tobby> courge_musque[m]: thanks
2026-04-20 09:52:26 <jpp1> Snap, flatpak, PPA, appimages…  Why is it better not to use them?
2026-04-20 09:53:39 <jpp1> Why not trust them? Because it's always possible to introduce malicious code into an application.
2026-04-20 10:06:22 <tobby> i still have to rely on some software that requires it, and find tweaks on how to at least reduce it's possible harm such as containerization. I haven't got enough of powerfull pc to host such stuff in dedicated vm
2026-04-20 11:20:39 <jpp1> ok
2026-04-20 11:20:48 <tobby> it's normal running alpine on main computer, no?
2026-04-20 11:22:34 <jpp1> I don't know
2026-04-20 11:23:39 <frag> yes, im normal
2026-04-20 11:35:21 <omni> me too, no one's said it to me but I'm sure
2026-04-20 11:50:18 <popefucker2> > Snap, flatpak, PPA, appimages…  Why is it better not to use them?
2026-04-20 11:50:18 <popefucker2> Having tried to use a lot of AppImage tools, I think the approach to software engineering in that community is very sloppy
2026-04-20 15:50:36 <specialk> is the apkovl method for headless recommended? 
2026-04-20 15:57:57 <strophy_> I had a pretty good experience using it last month to deploy an inventory of Pi Zero cameras on Alpine, it's well maintained but does require a moderate amount of reading and experimentation
2026-04-20 16:07:18 <specialk> ah it works :) 
2026-04-20 16:13:19 <specialk> strophy_: zerocams? 
2026-04-20 16:15:02 <strophy_> I was using Arducam SKU B0154 with imx219 sensor, I documented my steps here: https://github.com/macmpi/alpine-linux-headless-bootstrap/discussions/54
2026-04-20 16:18:37 <WhyNotHugo> How can I unlock FDE remotely via SSH? Is there some well-defined path to running SSH from an initfs?
2026-04-20 16:21:51 <WhyNotHugo> Ah, https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/86
2026-04-20 17:58:45 <specialk> this alpine uses 37MB ram 😊 
2026-04-21 04:25:00 <maduncle94[m]> Could you guys confirm that yambar 1.11.0 and its foreign-toplevel module works if it's built using GCC 15.x ? Thanks.
2026-04-21 04:30:18 <jpp1> hi
2026-04-21 12:02:44 <ivy-> is it intentional that linux-lts doesn't include dmraid?  and is there no way to get it aside from building my own kernel?
2026-04-21 14:00:38 <dwfreed> ivy-: is there some config other than DM_RAID ?
2026-04-21 14:02:11 <ivy-> i don't think so:
2026-04-21 14:02:12 <ivy-> 2!hyacinth ~# gzip -dc /proc/config.gz|grep DM_RAID
2026-04-21 14:02:13 <ivy-> # CONFIG_DM_RAID is not set
2026-04-21 14:08:11 <dwfreed> what kernel are you running? 'uname -a'
2026-04-21 14:08:39 <ivy-> Linux hyacinth.le-fay.org 6.18.22-0-lts #1-Alpine SMP 2026-04-11 15:53:25 ppc64le GNU/Linux
2026-04-21 14:10:06 <dwfreed> specifically unset in the ppc64le kernel; it *is* set in the x86_64 kernel
2026-04-21 14:10:22 <ivy-> odd.  should i file a bug, or was that done on purpose?
2026-04-21 14:10:29 <ivy-> it's rather annoying since lvm mirroring depends on dmraid
2026-04-21 14:10:53 <dwfreed> why not just use mdraid under the LVM?
2026-04-21 14:12:10 <ivy-> this is for the root/boot filesystem, i would rather only manage a single subsystem (lvm)
2026-04-21 17:17:19 <ivy-> so... should i file a bug?  (i think i will do, if no one has any other ideas)
2026-04-21 22:42:00 <halloy6401> Hi, I have a question.  I installed Alpine in a virtual machine to explore possibily switching to it (I currently use Debian on all my systems).  I managed to get everything I care about working except for one use case: the ability to download a pre-built proprietary game from itch.io and run it.  As an example, I'll use https://wizbane.itch.io/the-legend-of-lumina
2026-04-21 22:42:21 <halloy6401> When I extract the archive and run it from the terminal, I get the error message
2026-04-21 22:42:40 <halloy6401> >/bin/ash: The-Legend-of-Lumina-linux.x86_64: not found
2026-04-21 22:43:43 <halloy6401> Which I assume is due to it relying on glibc.  I tried installing both Debian and Void-Glibc in distroboxes, and launching the Legend of Lumina binary from within a distrobox.  I instead get another error.
2026-04-21 22:43:54 <longnoserob[m]> it is probably compiled against glibc and not musl
2026-04-21 22:44:08 <halloy6401> >Your video card driver does not support any of the supported OpenGL versions.
2026-04-21 22:44:59 <halloy6401> To compare, I tried installing the supertuxkart debian and void packages (inside the distroboxes) and launching them.  They gave the same error: it seems no distroboxes are able to use OpenGL.
2026-04-21 22:45:35 <halloy6401> My next thought was that the virtual machine I am running Alpine in doesn't have working GPU drivers, so I tried installing SuperTuxKart from the Alpine package repository, and *that* did work.
2026-04-21 22:45:38 <longnoserob[m]> suoertuxkart is in aport IIRC
2026-04-21 22:46:34 <halloy6401> Yeah, I understand STK is in aport: I was just trying to determine whether the lack of OpenGL drivers was system-wide or whether it was only happening inside a distrobox.
2026-04-21 22:50:25 <halloy6401> The issue seems to be that distrobox containers can't access my GPU, although they can still run 2D graphical applications (I tested the deb packages of supertux2 and adonthell, the latter of which is not in aports).
2026-04-21 23:35:45 <courge_musque[m]> Did you try to install gcompat?
2026-04-21 23:46:12 <halloy6401> Inside the distrobox or outside?
2026-04-21 23:46:53 <halloy6401> Oh wait it only makes sense to install it outside.  Lemme try.
2026-04-21 23:48:50 <unvFK9I3vgc> you need gpu-passthrough
2026-04-21 23:49:53 <unvFK9I3vgc> but if you run alpine natively, i can assure you steam via flatpack runs all games perfectly well
2026-04-21 23:50:29 <unvFK9I3vgc> messing around with gpu passtrhough is possible, but not trivial from the last time i tried that a couple years ago
2026-04-21 23:51:07 <halloy6401> Ah, that makes sense.
2026-04-21 23:51:55 <halloy6401> I tried installing gcompat and then running the Legend of Lumina binary (in the VM but outside the distrobox) and I got the same error about my GPU not supporting required OpenGL versions.  
2026-04-21 23:53:31 <halloy6401> I figured the Steam flatpak would work, I was more worried about Itch games that are distributed as tarballs.  I'll try dual-booting Alpine on my laptop at some point to test.
2026-04-21 23:54:12 <unvFK9I3vgc> i recommend running those itch games in a debian chroot
2026-04-21 23:54:25 <unvFK9I3vgc> or ubuntu, or whatever
2026-04-21 23:54:37 <unvFK9I3vgc> chroots have access to the gpu
2026-04-21 23:54:52 <unvFK9I3vgc> (if bind mounted correctly)
2026-04-21 23:55:12 <unvFK9I3vgc> i'
2026-04-21 23:55:45 <unvFK9I3vgc> m runnig alpine for a decade or so as my daily driver, and am a happy (occasional) gamer too
2026-04-21 23:57:20 <halloy6401> Does chroot offer an advantage over distrobox?
2026-04-22 00:00:35 <unvFK9I3vgc> isn't distrobox qemubased?
2026-04-22 00:00:41 <unvFK9I3vgc> like in a proper vm?
2026-04-22 00:00:52 <halloy6401> I don't think so?  Distrobox is podman with different default settings.
2026-04-22 00:01:24 <unvFK9I3vgc> sorry i don't know this newfangled stuff. podman?
2026-04-22 00:01:44 <halloy6401> In the same family as Docker (but I don't know the differences).
2026-04-22 00:01:53 <unvFK9I3vgc> ah. ok, so not a proper vm.
2026-04-22 00:01:56 <halloy6401> I currently mainly use it to run a void container with Sagemath.  
2026-04-22 00:02:44 <unvFK9I3vgc> well, in that case i'm pretty sure gpu stuff should work, if /dev and possibly /sys is properly bind mounted inside the chroot (docker and podman seem to be more complex chroots really)
2026-04-22 00:03:44 <courge_musque[m]> halloy6401 no, gcompat must be installed inside the alpine container
2026-04-22 00:03:51 <courge_musque[m]> Not in the host
2026-04-22 00:04:57 <halloy6401> Ah, I was not clear: I have nested containers.  I have Debian on bare metal, with Alpine in a virtual machine (Gnome Boxes/QEMU), with a debian distrobox container inside the Alpine VM.  
2026-04-22 00:05:18 <halloy6401> It seems like it only makes sense to install gcompat on Alpine, not on either of the Debian levels.
2026-04-22 00:05:52 <unvFK9I3vgc> yes, gcompat needs to be in the alpine system
2026-04-22 00:06:52 <unvFK9I3vgc> ok, so you need gpu-passthrough from the bare metal debian to the qemu alpine virtual machine
2026-04-22 00:07:22 <halloy6401> I think I'm just going to try dual-booting instead of gpu-passthrough.
2026-04-22 00:07:24 <unvFK9I3vgc> and then in the debian distrobox container you need access to the /dev of the alpine host, which has access to the root debian host
2026-04-22 00:38:26 <courge_musque[m]> debian in alpine in debian is quite something ngl