2025-04-01 12:08:19 <Soni> we would like to make an init system
2025-04-01 12:09:10 <Soni> where do we start?
2025-04-01 12:31:17 <pramodvu_> s6
2025-04-01 12:31:45 <pramodvu_> OR s6-rc
2025-04-01 12:31:47 <pramodvu_> OR 66
2025-04-01 15:35:09 <invoked> "good luck" (see you in 10 years)
2025-04-01 16:58:20 <mayli> So, anyone using `lbu` with a large system and limited memory?
2025-04-01 16:59:08 <frag> i always use unlimited memory
2025-04-01 16:59:19 <mayli> Something like a squashfs base system instead of loading everything into ram?
2025-04-01 17:11:05 <mayli> I am thinking about instead of unpacking .apkovl.tar.gz to ram, how about convert it to squashfs with tar2sqfs, then mount it with overlayfs, something similar with livecd, that avoids wasting ram on old/small devices.
2025-04-01 17:33:14 <mayli> Hmm, turns out I can just use the official iso + apkvol
2025-04-01 17:46:17 <voidastro[m]> alpine edge updates seem to not be working for a few days.. ERROR: unable to select packages:
2025-04-01 17:46:17 <voidastro[m]>   so:libxml2.so.16 (no such package):
2025-04-01 17:46:17 <voidastro[m]>     required by: clang16-extra-tools-16.0.6-r8[so:libxml2.so.16]
2025-04-01 17:51:46 <ikke> voidastro[m]: the builders are rebuilding packages to fix it
2025-04-01 17:52:02 <ikke> https://build.alpinelinux.org/
2025-04-01 18:55:56 <rhizoome> but if all packages are cached, alpine will install them before fstab is mounted. So after a reboot, everything is back in memory. But it turns out there is not that much. So I never tried to fix it.
2025-04-01 18:55:56 <rhizoome> overlay		/usr		overlay  lowerdir=/usr,upperdir=/zfs/grafter/usr/upper,workdir=/zfs/grafter/usr/work 0 0
2025-04-01 18:55:56 <rhizoome> mayli: I did 
2025-04-01 19:59:19 <realroot[m]> i disable the ui from chroot. i can use the tty
2025-04-01 19:59:27 <realroot[m]> chaingin to mdev maybe?
2025-04-01 19:59:27 <realroot[m]> nmcli shows interface unmanaged. how fix no keyboard and no mouse in the ui and all? setup-devd i think ruined it
2025-04-01 20:02:21 <ikke> Do you use something like seatd?
2025-04-01 20:06:47 <realroot[m]> not 100% sure i think yes
2025-04-01 20:07:02 <realroot[m]> can i check/enable services from chroot?
2025-04-01 20:07:52 <ikke> `find /etc/runlevels`
2025-04-01 20:08:50 <realroot[m]> when i ran setup-devd i answered y to "populate ..." i guess domething ran that already
2025-04-01 20:09:45 <realroot[m]> etc/runlevels/default/seatd
2025-04-01 20:10:19 <realroot[m]> i am running pinnacle, pre alpha now it was working fine
2025-04-01 20:11:36 <ikke> seems like the question to populate only matters for mdev
2025-04-01 20:11:55 <ikke> sorry, mdevd
2025-04-01 20:11:55 <realroot[m]> do iremove etc/runlevels/sysinit/mdev?
2025-04-01 20:12:12 <ikke> Do you want to switch to udev?
2025-04-01 20:12:22 <ikke> Would make sense with graphical environments
2025-04-01 20:12:49 <ikke> setup-devd udev would do that
2025-04-01 20:13:06 <realroot[m]> then yes do I need one for bluetooth?
2025-04-01 20:14:30 <ikke> one what?
2025-04-01 20:14:58 <SamuelMarks> I'm working on Alpine Linux builds for VMware, VirtualBox, Vagrant, Parallels, qemu, and Hyper-V; for x86_64 and aarch64.
2025-04-01 20:15:05 <realroot[m]> are mdev and mdevs for CLI?
2025-04-01 20:15:11 <SamuelMarks> 
ACTION posted here and got no interest [no replies] https://lists.alpinelinux.org/~alpine/devel/%3CCAMfPbcYsT%2BM7oAwAhxXtYN-0vy%2BLSFwQKLJ8_jbv%2BuskTj%2BAxQ%40mail.gmail.com%3E

2025-04-01 20:15:30 <realroot[m]> one of mdev, mdev or udev
2025-04-01 20:15:41 <ikke> SamuelMarks: You have to have a bit more patience, we have been working on fixing the repositories the last few days
2025-04-01 20:16:05 <ikke> SamuelMarks: also, are you familiar with https://www.alpinelinux.org/cloud/
2025-04-01 20:16:20 <SamuelMarks> ikke - Even a "+1" in response to the email would probably go a long way!
2025-04-01 20:16:53 <SamuelMarks> ikke - No I hadn't seen that one; good to see local qcow2
2025-04-01 20:16:54 <ikke> You may want to join forces with the cloud team that already builds a lot of images for various clouds / noclouds
2025-04-01 20:17:00 <SamuelMarks> Very happy to
2025-04-01 20:17:17 <ikke> https://gitlab.alpinelinux.org/alpine/cloud/
2025-04-01 20:17:22 <ikke> and #alpine-cloud
2025-04-01 20:17:33 <ikke> (irc channel)
2025-04-01 20:18:28 <SamuelMarks> https://gitlab.alpinelinux.org/alpine/cloud/alpine-cloud-images just made an account no issues
2025-04-01 20:18:32 <SamuelMarks> ah cool I'll join their channel
2025-04-01 20:20:32 <SamuelMarks> I'm just happy to not create my own qcow2 like I was @ https://github.com/SamuelMarks/bento/blob/da04a6a/alpine-3.21-aarch64-virtualbox-build.sh
2025-04-01 23:55:10 <WhyNotHugo> Running `mmcli --scan-modems` I get: error: couldn't request to scan devices: 'GDBus.Error:org.freedesktop.ModemManager1.Error.Core.Failed: PolicyKit authorization error: 'authority not found''
2025-04-01 23:55:14 <WhyNotHugo> Any idea what this means?
2025-04-02 00:49:05 <broski[m]> From Trusted and Vouched Dealers... (full message at <https://matrix.org/oftc/media/v1/media/download/AeG3hdeFci2fgFqrqTnS9oZX5VG-6tb5kU2d5qvZPmz0m6AE5JWBWL9IkA_qtSMr_Ie-mCBrqU9dyYPbs0VK8qFCeWPBIpcgAG1hdHJpeC5vcmcvekVwbmN0Vk1kQlBNWnROZHNYSlRXRXZ6>)
2025-04-02 01:07:24 <WhyNotHugo> answering my own question: polkit needs to be running for modemmanager. See !!82215
2025-04-02 01:16:52 <voidastro[m]> <ikke> "voidastro: the builders are..." <- it's fixed.
2025-04-02 14:29:48 <andar1an> hello all. I have been using ifupdown-ng on alpine, and am doing some more advanced config today and am curious about configuration keywords. With ifupdown inet, and inet6 are often seen. However, in the admin guide for ifupdown-ng inet is not listed, though it seems to work. Does anyone know what the deal is with this keyword for ifupdown-ng? probably going to just leave off and see what happens
2025-04-02 14:58:59 <ikke> andar1an: they're not required for ifupdown-ng
2025-04-02 15:01:55 <andar1an> thank you @ikke, Many blogs tend to have it, and since ifupdown-ng is mostly compatible with many other configurations I think it just hasn't presented and issue, but great to know I can leave off 
2025-04-02 16:38:47 <frag> is there a way to reload soundcards, so if device 0 is a usb and you unplug it, you can get another card as nr 0?
2025-04-02 17:13:45 <lopid> i looked into a way to simulate replugging a usb device. there were a couple of examples but nothing worked
2025-04-02 18:09:47 <frag> lopid thanks for trying :)
2025-04-02 19:14:31 <andar1an[m]> Can udev rules work for sound cards?
2025-04-02 19:16:01 <andar1an[m]> Also, for iwd, is it by design that connecting a station to a password protected network doesn't prompt for password if connected by Ethernet? Or could that be a vulnerability? I noticed this today and wonder if anyone else has
2025-04-02 19:17:50 <AledCuda[m]> Got it, thanks
2025-04-02 19:18:32 <ZombieChicken> Hello. Anyone here running Alpine on a Pinebook Pro? I can't seem to get it to boot.
2025-04-02 19:18:52 <grawity> andar1an[m]: depends, do what exactly with sound cards
2025-04-02 19:19:04 <f_> ZombieChicken: What issues are you having?
2025-04-02 19:19:22 <andar1an[m]> Similar tobl network card naming rules via udev
2025-04-02 19:19:37 <ZombieChicken> Bad Linux ARM64 Image magic! was the issue. I'm also looking at the extlinux.conf and trying to figure out where it's trying to guess where to boot from.
2025-04-02 19:19:48 <andar1an[m]> Was reference to what Frag said earlier
2025-04-02 19:20:03 <f_> ZombieChicken: What kernel image are you trying to boot?
2025-04-02 19:20:25 <ZombieChicken> default for the aarch64 image; vmlinuz-lts
2025-04-02 19:20:34 <f_> mmhmm
2025-04-02 19:20:53 <grawity> andar1an[m]: in theory they can create symlinks to /dev/snd/card**, but in practice I don't know if that's of any use for ALSA
2025-04-02 19:21:04 <ZombieChicken> Does Alpine's u-boot now know how to decompress the kernel? Last I checked, it didn't like compressed kernels...
2025-04-02 19:21:31 <grawity> network interfaces are kinda special in that they don't exist in /dev so renaming them is a thing
2025-04-02 19:21:56 <ZombieChicken> f_: I literally /just/ dd'ed u-boot from the same tarball onto an SD card, formatted the rest (starting block at 62000, just to be sure I didn't clobber anything) ext4, and ran with it.
2025-04-02 19:22:09 <andar1an[m]> @grawity maybe they could also reference device Id and type in proc
2025-04-02 19:22:26 <ZombieChicken> and, of course, tar xpvf'ed the tarball onto the resulting partition
2025-04-02 19:22:27 <andar1an[m]> For Attr
2025-04-02 19:22:36 <grawity> missed my point I think
2025-04-02 19:23:14 <andar1an[m]> Ah, sorry, small screen. Didn't see message about /dev
2025-04-02 19:23:38 <andar1an[m]> I didn't know network cards were unique like that, cool
2025-04-02 19:23:44 <andar1an[m]> Ty
2025-04-02 19:24:07 <ZombieChicken> f_: Any clue what might be wrong?
2025-04-02 19:24:34 <f_> is your vmlinuz compressed?
2025-04-02 19:25:25 <ZombieChicken> file says it's a PE32+ exec
2025-04-02 19:25:27 <ssm> the z in linuz implies compression I think?
2025-04-02 19:25:45 <ZombieChicken> ssm: that is what I was thinking.
2025-04-02 19:26:15 <ZombieChicken> but I havn't messed with alpine much at all. I tossed it on a rpi02w before, and that's been that so far.
2025-04-02 19:26:24 <ava_mateo21[m]> Hello everyone 
2025-04-02 19:26:28 <f_> ssm Sometimes apparently not.
2025-04-02 19:26:49 <ZombieChicken> so I'm unfamiliar with the various failure modes for it's boot process
2025-04-02 19:27:17 <ZombieChicken> f_: The kernel I normally run is ~25MB. The vmlinuz-lts is 9.9MB. I think it's probably compressed.
2025-04-02 19:28:40 <ZombieChicken> I'm going to reformat the partition vfat and try again. Not like it takes a long time to reextract that tarball.
2025-04-02 19:28:50 <ZombieChicken> and the wiki suggest using it anyways.
2025-04-02 19:30:34 <ZombieChicken> okay, rebooting. BBL
2025-04-02 20:09:24 <user0> ssm: yes I thought so, like in zImage
2025-04-02 21:31:05 <bent_fingers> i found that busybox less fails to render a text file correctly if the lines are too long
2025-04-02 21:31:27 <bent_fingers> any reason to curtail line lengths generally?
2025-04-02 21:32:17 <bent_fingers> i would rather not have adventitious line endings sporadically interposed throughout the documents 
2025-04-02 22:37:43 <fjdzqoFsDUV2> you can just install a proper less
2025-04-02 22:40:05 <anelli> hi frends. anyone got factorio running? i `apk add gcompat' but it still no worky. i'm trying out factorio 2.0.42 demo. it gives some `symbol not found' errors
2025-04-02 22:40:07 <anelli> thx
2025-04-02 22:52:05 <dwfreed> anelli: you'd be much better off using flatpak than gcompat
2025-04-02 23:14:33 <candrew> The testing mirrors for Alpine edge x86_64 don't seem to have been updated this month
2025-04-02 23:14:55 <candrew> On dl-cdn.alpinelinux.org as well, not just my mirror
2025-04-02 23:17:06 <candrew> I'm getting stuck in a dependency conflict with the icu upgrade (the electron rebuild against icu 76 doesn't seem to be on any of the mirrors)
2025-04-03 00:21:13 <candrew> actually I have seen some updated mirrors now. bit odd though.
2025-04-03 06:06:13 <rhizoome> candrew: I received electron electron-34.4.1-r2.apk via my mirror.
2025-04-03 06:10:48 <ikke> candrew: the builders have been stuck for a while due to various upgrades 
2025-04-03 07:35:26 <candrew> ikke: ah right, thanks
2025-04-03 09:09:34 <rhizoome> what is the fastest way to get patch if I fixed bugs in src/ of package? I usually copy src/ to src.patch and unpack again.
2025-04-03 09:13:28 <ikke> I init a git repo before making the change 
2025-04-03 09:13:41 <ikke> In $builddir
2025-04-03 09:14:01 <rhizoome> @ikke nice! thanks!
2025-04-03 09:15:20 <ikke> git init -q && git add -q . && git commit -m.
2025-04-03 09:15:26 <ikke> Something like that 
2025-04-03 12:00:00 <danci1973|2> Hello.
2025-04-03 12:03:15 <danci1973|2> I recently deployed a new router / firewall based on Alpine Linux 3.21.2 ... I'm running net-snmp snmpd for collecting network traffic data with mrtg (remotely) and for some inexplicable reason, smtpd crashes with 'segfault' pretty much exaclty after 10 minutes of running.
2025-04-03 12:04:52 <danci1973|2> Apr  3 11:47:56 router /etc/init.d/snmpd[8984]: start-stop-daemon: no matching processes found
2025-04-03 12:04:52 <danci1973|2> Apr  3 11:47:56 router start-stop-daemon[8984]: Will stop /usr/sbin/snmpd
2025-04-03 12:04:52 <danci1973|2> Apr  3 11:47:56 router /etc/init.d/snmpd[8984]: start-stop-daemon: no matching processes found
2025-04-03 12:04:52 <danci1973|2> Apr  3 11:47:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:52 <danci1973|2> Apr  3 11:47:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:54 <danci1973|2> Apr  3 11:48:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:54 <danci1973|2> Apr  3 11:48:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:56 <danci1973|2> Apr  3 11:49:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:56 <danci1973|2> Apr  3 11:49:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:58 <danci1973|2> Apr  3 11:50:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:04:58 <danci1973|2> Apr  3 11:50:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:00 <danci1973|2> Apr  3 11:51:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:00 <danci1973|2> Apr  3 11:51:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:02 <danci1973|2> Apr  3 11:52:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:02 <danci1973|2> Apr  3 11:52:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:04 <danci1973|2> Apr  3 11:53:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:04 <danci1973|2> Apr  3 11:53:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:05 <ghodawalaaman> danci1973|2 use pastebin
2025-04-03 12:05:06 <danci1973|2> Apr  3 11:54:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:06 <danci1973|2> Apr  3 11:54:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:08 <danci1973|2> Apr  3 11:55:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:08 <danci1973|2> Apr  3 11:55:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:10 <danci1973|2> Apr  3 11:56:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:10 <danci1973|2> Apr  3 11:56:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:12 <danci1973|2> Apr  3 11:57:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:12 <danci1973|2> Apr  3 11:57:56 router snmpd[9009]: systemstats_linux: unexpected header length in /proc/net/snmp. 237 != 224
2025-04-03 12:05:14 <danci1973|2> Apr  3 11:57:56 router kernel: [2321383.211307] snmpd[9009]: segfault at 7f841048a280 ip 00007f8410c51bac sp 00007fffff1270a0 error 4 in libnetsnmpmibs.so.40.2.1[3bbac,7f8410c43000+a8000] likely on CPU 0 (core 0, socket 0)
2025-04-03 12:05:14 <danci1973|2> Apr  3 11:57:56 router kernel: [2321383.211307] snmpd[9009]: segfault at 7f841048a280 ip 00007f8410c51bac sp 00007fffff1270a0 error 4 in libnetsnmpmibs.so.40.2.1[3bbac,7f8410c43000+a8000] likely on CPU 0 (core 0, socket 0)
2025-04-03 12:05:16 <danci1973|2> Ouch, sorry about that.
2025-04-03 12:05:26 <danci1973|2> Thought I had the pastebin in my buffer. :(
2025-04-03 12:06:08 <danci1973|2> https://pastebin.com/raw/LgwBVPhW
2025-04-03 12:08:59 <danci1973|2> Any idea what might be going on or how to debug this (on a remote router)?
2025-04-03 12:10:30 <ikke> danci1973|2: did you search for any known issues with smtpd?
2025-04-03 12:11:00 <ikke> Seeing https://github.com/net-snmp/net-snmp/issues/829
2025-04-03 12:12:22 <fabricionaweb> that could be post on tpaste.us :(
2025-04-03 12:12:56 <danci1973|2> I did - the 'unexpected header lenght' is apparently due to incompatibility with kernel v6.x and isn't causing segfaults... And apparently there is a patch for this - https://github.com/net-snmp/net-snmp/pull/785 .
2025-04-03 12:13:38 <danci1973|2> I couldn't find an issue that would match my case - segfaulting after 10 minutes...
2025-04-03 12:15:12 <ikke> danci1973|2: can you try to set ulimit -c unlimited, and then run snmpd manually and see if you get the segfault?
2025-04-03 12:15:33 <danci1973|2> ikke: Will do right now.
2025-04-03 12:38:01 <danci1973|2> ikke: Sorry, a phone call distracted me - but unfortunately, ulimit -c didn't help (https://pastebin.com/raw/EBv2yM2k)
2025-04-03 12:38:35 <ikke> danci1973|2: It was not a solution, but a way to get a core dump
2025-04-03 12:38:47 <ikke> do you have a file called `core` in your current directory?
2025-04-03 12:38:48 <danci1973|2> ikke: I see, that it did.
2025-04-03 12:39:06 <danci1973|2> I do.
2025-04-03 12:39:22 <ikke> ok, now `apk add gdb net-snmp-dbg`
2025-04-03 12:39:43 <ikke> and then gdb /usr/sbin/snmpd core`
2025-04-03 12:41:14 <danci1973|2> Done... https://pastebin.com/raw/tQ5bKDgJ
2025-04-03 12:41:37 <ikke> Can you run `bt`
2025-04-03 12:42:12 <ikke> And then create an issue in the aports project with that information
2025-04-03 12:42:43 <ikke> oh: https://github.com/net-snmp/net-snmp/issues/293
2025-04-03 12:43:09 <ikke> ncopa already filed an issue regarding this
2025-04-03 12:43:32 <danci1973|2> Maybe if I disable disk monitoring, it will stop crashing.
2025-04-03 12:44:41 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues/12309#note_171617
2025-04-03 12:46:08 <ikke> ftr: commenting out `includeAllDisks 10%`
2025-04-03 12:46:10 <danci1973|2> ikke: Thank you, it seems like the issue is solved for me as I don't really need disk monitoring...
2025-04-03 12:46:21 <ikke> glad to hear
2025-04-03 13:22:36 <invoked> sertonix[m]: librewolf update changed the config path to ~/.mozilla away from ~/.librewolf
2025-04-03 13:25:26 <ikke> https://codeberg.org/librewolf/issues/issues/2400
2025-04-03 13:27:51 <invoked> thanks
2025-04-03 13:30:02 <invoked> i'm going to roll back, but this is significant breakage, even for edge
2025-04-03 18:04:04 <ssm> invoked: finally imperialism I can support
2025-04-03 19:38:25 <feuerteufel> N8 @ all
2025-04-03 19:46:01 <ikke> o/
2025-04-03 20:16:41 <f_> \o
2025-04-03 20:18:39 <user0> o/
2025-04-04 01:08:16 <zayd_webirc> How do I get Qualcomm Wi-Fi cards working on Alpine? Mine is a NCM825, it doesn't show up in `ip addr`  but it appears under `lspci`
2025-04-04 01:11:20 <fjdzqoFsDUV2> same as with any other distro/kernel i suppose? loading the correct kernel module?
2025-04-04 01:16:59 <zayd_webirc> fjdzqoFsDUV2: I don't think I see anything relevant in /lib/modules for this
2025-04-04 01:22:31 <fjdzqoFsDUV2> are the modules in mainline?
2025-04-04 01:24:14 <Ermine> Seems like it should be ath12k driver
2025-04-04 01:31:49 <zayd_webirc> Ermine: does Alpine's kernel package not have that enabled or did I forget to do something because I have that installed and it seems to not be in /lib/modules, the only thing I could find for it on my system is /lib/firmware/ath12k
2025-04-04 01:37:06 <Ermine> zayd_webirc: seems like the module is not built
2025-04-04 01:39:12 <zayd_webirc> Ermine: about how much of a pain would it be building it myself
2025-04-04 01:43:22 <Ermine> zayd_webirc: not much, since it's mainline. Grab kernel sources, pick a config from https://git.alpinelinux.org/aports/tree/community/linux-edge for your arch, copy it to kernel tree, make oldconfig, then set CONFIG_ATH12K to M, and build the kernel
2025-04-04 01:56:06 <zayd_webirc> Ermine: oh should I be on linux-edge? the installer put me on -lts
2025-04-04 01:56:46 <zayd_webirc> sorry for the slow responses, my desktop with the IRC client open and the laptop with the Wi-Fi problem plugged into ethernet are on opposite sides of my home
2025-04-04 04:39:59 <Saijin_Naib[m]> @zayd_webirc edge and lts do not have a common baseline config. I used to modify lts for my machines that needed tweaks before I submitted MRs to enable that hardware
2025-04-04 10:33:41 <Ermine> zayd_webirc: in case of -lts, use config for main/linux-lts . But seems like ath12k is a new driver, and you'd better have a newest kernel version because it takes a few versions for driver to become mature and stable enough
2025-04-04 14:02:57 <muusemuuse> I'm trying to run alpine linux diskless with lbu throwing data back onto the SD card occasionally. I'm using a nanopi R4s.  I have u-boot working, I have the alpine linux files on the correct partition, GRUB loads, but I cant get past GRUB without the error 
2025-04-04 14:03:07 <muusemuuse> I'm trying to run alpine linux diskless with lbu throwing data back onto the SD card occasionally. I'm using a nanopi R4s.  I have u-boot working, I have the alpine linux files on the correct partition, GRUB loads, but I cant get past GRUB without the error "Mounting boot media failed"
2025-04-04 14:04:46 <muusemuuse> it will happily boot and run from a USB flash drive, but the SD card will only reach GRUB
2025-04-04 15:22:34 <DaKnig[m]> hey
2025-04-04 15:23:45 <DaKnig[m]> upon installing postmarketos in my phone (alpine based), I got an ugly font for my local language. would it be a welcome contribution to add a package for Rubic font?
2025-04-04 15:23:49 <DaKnig[m]> the font from Google
2025-04-04 15:24:22 <ikke> if the font is not too big
2025-04-04 15:24:32 <DaKnig[m]> what is big
2025-04-04 15:24:34 <DaKnig[m]> define
2025-04-04 15:25:00 <DaKnig[m]> 348kB
2025-04-04 15:25:07 <DaKnig[m]> (not including italic)
2025-04-04 15:25:10 <ikke> that's fine
2025-04-04 15:25:17 <ikke> not 50mb+
2025-04-04 15:26:58 <DaKnig[m]> I wonder how I could make graphical programs that are compact, really fast and performant on old mobile devices. like, if I run the OS from a slow SD card, and it needs to load a bunch of shared libs that means many random accesses which can take seconds to start up, thats quite slow...
2025-04-04 15:27:36 <DaKnig[m]> in the terminal I could maybe just directly send syscalls to print and read stuff, dont need to link much into the libc...
2025-04-04 15:27:47 <DaKnig[m]> ok thats way offtopic. but if anybody has a pointer please tell me
2025-04-04 18:51:50 <rifux[m]> please help... (full message at <https://matrix.org/oftc/media/v1/media/download/AXoU-7JoeRqQuicb8SkhIqkA0JRyOROAPkwjfRvkX60ORK2QMqHKpVSwBAApie91C1SO-5Wh5JZTZnsrYc4z6Q1CeWSj4p9QAG1hdHJpeC5vcmcvVWVXZVhQbFhiY2RaQkVOR250S3V3S3Nk>)
2025-04-04 18:53:29 <rifux[m]> restarting alsa does nothing
2025-04-04 20:33:26 <WhyNotHugo> how can i react to hotplug events as a non-root user? specifically, run a command when a device is plugged in.
2025-04-04 20:47:41 <rhizoome> WhyNotHugo: generally write udev script that runs as root and pipe the information in a file that the non-root user can read.
2025-04-04 20:48:09 <rhizoome> then poll.
2025-04-04 20:49:28 <rhizoome> WhyNotHugo: actully you can probably do "sudo -u myuser" in the udev event. 
2025-04-04 20:51:55 <rhizoome> WhyNotHugo: ACTION=="add", RUN+="sudo -u myuser myscript"
2025-04-04 20:52:04 <WhyNotHugo> i need to run this as non-root; i want to share the resulting script/binary with others
2025-04-04 20:53:31 <rhizoome> WhyNotHugo: you mean no installation. Then polling /dev is the probably the only option since /dev usually doesn't support iwatch.
2025-04-04 20:54:35 <WhyNotHugo> a one-time installation of dependencies is acceptable.
2025-04-04 20:54:56 <WhyNotHugo> but a udev rule requires root to enable/disable/reconfigure too
2025-04-04 20:56:04 <rhizoome> WhyNotHugo: well you can leave the udev rule there and just do nothing if the user disabled it
2025-04-04 20:57:05 <rhizoome> otherwise probably inotifywait. but the last time I tried my /dev filesystem did not support it. But it was some exotic system. Maybe linux supports inotifywait on /dev.
2025-04-04 21:01:01 <Ermine> or make a daemon that listens on netlink and provides a socket useable by other users
2025-04-04 21:01:24 <rhizoome> WhyNotHugo: nice it work:
2025-04-04 21:01:25 <rhizoome> Setting up watches.
2025-04-04 21:01:25 <rhizoome> /dev/ ATTRIB hidraw4
2025-04-04 21:01:25 <rhizoome> Watches established.
2025-04-04 21:01:25 <rhizoome> » inotifywait /dev
2025-04-04 21:01:44 <rhizoome> (I unplugged hidraw4)
2025-04-04 21:03:10 <WhyNotHugo> nice, inotifywait works. produces a CREATE event
2025-04-04 21:07:21 <rhizoome> WhyNotHugo: even simpler: udevadm monitor works as user.
2025-04-04 21:07:26 <WhyNotHugo> Ermine: looks like I can use libudev to listen on the netlink device directly.
2025-04-04 21:07:46 <WhyNotHugo> haha, perfect timing, yeah, that seems to work
2025-04-04 21:09:07 <rhizoome> WhyNotHugo: if you wanted the to go the netlink route, you would probably do `socket(AF_NETLINK, SOCK_DGRAM, NETLINK_KOBJECT_UEVENT);`
2025-04-05 06:27:42 <rifux[m]> <rifux[m]> "please help..." <- > <@rifux:tchncs.de> please help... (full message at <https://matrix.org/oftc/media/v1/media/download/Ad-s19sGGstFEyaHKwLVkUlfG7EponcS69-YqOP9plkbEorzS0083emNhXtUeSJTyIptcYjDkLeFzw2bJTJPszxCeWTLtAfgAG1hdHJpeC5vcmcvdkhlc3hRaktKY1BuS2x5WENDQWdEaFp5>)
2025-04-05 08:54:56 <jvvv> rifux[m]: that link is very long. for me it spans three lines.
2025-04-05 08:55:49 <longnoserob[m]> restart pulseaudio?
2025-04-05 10:05:24 <sigirnog> "switch_root: can't execute '/sbin/init': Symbolic link loop" where could this error come from? I extracted the initfs and /init is a file, /bin/sh is a symlink to a /bin/busybox, which is a file, whose only dynamically linked library is musl, also a file
2025-04-05 11:52:24 <sertonix[m]> The error happens after the initfs ran. You could try booting with the `single` kernel argument which should out you into a shell before switch_root is executed. Then you can take a look at /newroot/sbin/init
2025-04-05 15:55:11 <rifux[m]> <longnoserob[m]> "restart pulseaudio?" <- pulseaudio? it's always stopped and requires additional configurartion
2025-04-05 15:56:43 <rifux[m]> <jvvv> "リフクス ✿: that link is very long..." <- huh? it's not that big until you don't use huge font size or ultra small display
2025-04-05 15:59:04 <rifux[m]> * additional configurartion to start service
2025-04-05 15:59:38 <rifux[m]> rifux[m]: I have default GNOME installation via `setup-desktop`
2025-04-05 22:00:13 <Kladky> It seems that on my Alpine Edge server, 1-2 weeks ago, my server stopped being able to send ipv6 requests, despite other devices on the network still being able to. Does anyone have any idea why this could be?
2025-04-05 22:01:07 <Kladky> It just uses the networking service, with the following config: https://privatebin.net/?b66076513ffee1d7#Da519kQBHZhVH6qkNZvmJcHCvhtEUc1G37VumYzqUNkG
2025-04-05 22:07:21 <Kladky> Note, that originally, that "pre-up" line was uncommented, but i tried commenting it to see if that would make a difference, which it didn't
2025-04-06 02:00:33 <zayd_webirc> Is there any way to get dictd compiled on Alpine aarch64? I tried and it fails with this https://txt.t0.vc/VUUC while it compiles fine on my Arch install. Doesn't seem like it is but maybe because no glibc?
2025-04-06 05:26:57 <fission> zayd_webirc: I don't think it has anything to do with glibc. maybe missing libtool
2025-04-06 05:29:35 <ikke> It tries to execute an argument as a coman 
2025-04-06 05:29:38 <ikke> Commad
2025-04-06 05:30:14 <fission> there should be libtool before that --mode=...
2025-04-06 05:30:20 <fission> (I guess, I didn't really look carefully)
2025-04-06 05:37:39 <fission> %.o: %.c
2025-04-06 05:37:43 <fission>         $(LIBTOOL) --tag=CC --mode=compile $(CC) -c $(CFLAGS) $< -o $@
2025-04-06 05:37:53 <fission> so yeah. if $(LIBTOOL) doesn't expand to anything, then...
2025-04-06 05:41:35 <fission> zayd_webirc: I had to add:   gcc musl-dev libmaa-dev zlib-dev make flex libtool bison
2025-04-06 05:41:46 <fission> and then run ./configure ; make LEX=flex
2025-04-06 05:41:51 <fission> and it does work
2025-04-06 05:42:14 <fission> or at least builds, I don't even know what this thing does :)
2025-04-06 05:42:16 <fission> alpine-test:~/dictd-1.13.3$ file dictd
2025-04-06 05:42:19 <fission> dictd: ELF 64-bit LSB pie executable, ARM aarch64, version 1 (SYSV), dynamically linked, interpreter /lib/ld-musl-aarch64.so.1, BuildID[sha1]=712e4dc4120b00f6fdd4de25f76fd9e49ffa6e17, with debug_info, not stripped
2025-04-06 05:43:17 <fission> I need to snapshot the base of this install
2025-04-06 06:21:18 <frag> is there a "better" way to see/follow twitter/x.com posts than a browser?
2025-04-06 06:32:44 <fission> I see all the best stuff on twitter with: cat /dev/zero
2025-04-06 08:22:12 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:22:17 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:22:22 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:22:23 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:22:23 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:22:23 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:22:30 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:23:07 <longnoserob[m]> hmm that did not work Kladky 
2025-04-06 08:23:52 <Kladky> What didn't work?
2025-04-06 08:24:54 <longnoserob[m]> 
ACTION uploaded an image: (36KiB) < https://matrix.org/oftc/media/v1/media/download/AemgjUJRM_xeRGB9I4e7i6gHfoHzE_aKOzbMVCcDaKRDkACUumaGt87e-IqoHv_JY8dbA7xsJ8vWsq8o-ugQrP9CeWUkzrFwAG1hdHJpeC5vcmcvc1N3SnZlWkJEeFhHQ3JqdmRkT2ZnZXlT >

2025-04-06 08:25:52 <Kladky> Oh
2025-04-06 08:26:05 <Kladky> Lemme try again
2025-04-06 08:26:24 <Kladky> (Re-sending the following because my bouncer went down)
2025-04-06 08:26:29 <Kladky> It seems that on my Alpine Edge server, 1-2 weeks ago, my server stopped being able to send ipv6 requests, despite other devices on the network still being able to. Does anyone have any idea why this could be?
2025-04-06 08:26:38 <Kladky> It just uses the networking service, with the following config: https://privatebin.net/?b66076513ffee1d7#Da519kQBHZhVH6qkNZvmJcHCvhtEUc1G37VumYzqUNkG
2025-04-06 08:26:50 <Kladky> Note, that originally, that "pre-up" line was uncommented, but i tried commenting it to see if that would make a difference, which it didn't
2025-04-06 08:27:09 <Kladky> -- end resending --
2025-04-06 08:28:46 <longnoserob[m]> anything in the logs?
2025-04-06 08:29:29 <Kladky> Where would the relevant log be?
2025-04-06 08:29:43 <longnoserob[m]> dmesg and logread?
2025-04-06 08:32:49 <Kladky> longnoserob[m]: my dmesg: https://privatebin.net/?b5a7305b80bd1f59#3HvhFvsRmUy7AczopcsDqLPVdXyojyBLiEtNSDhdJzth
2025-04-06 08:33:05 <Kladky> logread errors: "can't find syslogd buffer: No such file or directory"
2025-04-06 08:34:11 <longnoserob[m]> 1698s is a bit short for 1-2 weeks ago stuff..
2025-04-06 08:34:37 <Kladky> Yeah, cause I have rebooted since then
2025-04-06 08:42:23 <Kladky> Anyways, I also changed my /etc/network/interfaces to the following to simplify it, but it didn't make a difference: https://privatebin.net/?9cf3dc9eed488d86#AQAHk59RhxaGiLrwX8reJdvWr2V999faL91txC2w1tnZ
2025-04-06 08:52:42 <longnoserob[m]> did you check if the UFW might be blocking too much?
2025-04-06 08:53:28 <Kladky> I doubt it, since I can't even send requests to ipv6 addresses from the server
2025-04-06 08:53:41 <longnoserob[m]> and btw why do you use a /24 for IPv6?
2025-04-06 08:53:42 <Kladky> but I can check if ufw blocks outbound
2025-04-06 08:54:01 <Kladky> \/24 is for global, right?
2025-04-06 08:54:23 <longnoserob[m]> shouldnt a single host get a /64?
2025-04-06 08:54:48 <Kladky>  /64 is for local
2025-04-06 08:56:44 <Kladky> hmm, checking my laptop `ip a`, both local and global are /64, so i'll try that on my server
2025-04-06 08:57:22 <longnoserob[m]> same here
2025-04-06 09:09:42 <Kladky> Restarted, ipv6 still doesn't work
2025-04-06 09:10:25 <longnoserob[m]> does ip n show you any v6 adresses?
2025-04-06 09:12:37 <longnoserob[m]> do you have access to the router?
2025-04-06 09:13:11 <Kladky> yes and yes
2025-04-06 09:13:47 <Kladky> `ip n` output: https://privatebin.net/?845114e12b4cfb2b#GcRvdC8269GdC5pzCmpXTSijvQ86wcYL5bk1Ve7ZpaLP
2025-04-06 09:15:10 <rhizoome> these are link-local adresses
2025-04-06 09:15:48 <longnoserob[m]> yes, check if your router still has the global ones
2025-04-06 09:18:19 <Kladky> Yep, both the local and global addresses are listed as assigned to the server on the router
2025-04-06 09:19:47 <longnoserob[m]> does pining your sever from the router works?
2025-04-06 09:21:39 <Kladky> I don't have shell access to my router, just a web interface
2025-04-06 09:22:41 <longnoserob[m]> ok than, time to breakout wireshark or tcpdump and look into your local traffic
2025-04-06 09:40:36 <Forza> hi. Is there a variable to check if system is rebooting or shutting down, that can be used in an init script?
2025-04-06 09:41:32 <longnoserob[m]> depends on your init system, I would guess
2025-04-06 09:41:59 <Forza> Default openrc/busybox in Alpine
2025-04-06 09:46:04 <Forza> On Gentoo there is $RC_REBOOT but I think it isn't available.
2025-04-06 09:52:57 <Kladky> Seems like I can ping other v6 devices on the LAN, so it's just the routing to the internet which is the problem
2025-04-06 10:23:47 <Forza> Kladky: I/O error, dev sda, sector 386792248 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
2025-04-06 10:24:22 <Kladky> ?
2025-04-06 10:25:15 <Forza> Kladky: you have io errors on your disk
2025-04-06 10:25:46 <Kladky> That sounds believable
2025-04-06 10:26:22 <Kladky> Shouldn't effect ipv6 though, right?
2025-04-06 10:26:35 <Forza> Kladky: your ip6 config doesn't look right? What's your 'ip -6 route'
2025-04-06 10:27:25 <Kladky> https://privatebin.net/?96a354a28ce06212#LuXyj3chM7fWk7z28xbLHWHeLqmdotPyY1UpyeE4zrt
2025-04-06 10:30:44 <Forza> You have 2a00:23c7:a816:3201::/64 on eth0, but default route doesn't go to an IP within that subnet. 
2025-04-06 10:31:40 <Forza> I suppose it should be possible to use the routers link-local, but maybe the routing isn't set up like that.
2025-04-06 10:32:27 <Kladky> I mean, my laptop's one looks similar, and it can access ipv6 internet just fine
2025-04-06 10:32:42 <Kladky> https://privatebin.net/?d4c35d487c60fa9d#9MgaAcBX6pHUnjHiAo4Wsg6Le67mByvYstjBVURJrAgZ
2025-04-06 10:32:56 <Kladky> But my laptop uses networkmanager instead of the networking service
2025-04-06 10:33:32 <Forza> https://tailscale.com/kb/1121/ipv6
2025-04-06 10:34:36 <Forza> I guess something is not right then with your route and/or firewall
2025-04-06 10:36:12 <Kladky> Well it doesn't seem to be the server's ufw, since connections to it reject, regardless whether the default incoming policy is reject or deny
2025-04-06 10:36:53 <Kladky> And the default outbound policy is allow, so I imagine it's routing
2025-04-06 10:37:33 <Forza> forward?
2025-04-06 10:38:19 <Forza> Can you ping fe80::628d:26ff:feae:7223
2025-04-06 10:39:14 <Kladky> Yep
2025-04-06 10:39:36 <Kladky> And I can ping the server's local ip from my laptop
2025-04-06 10:39:44 <Kladky> Just not the global one
2025-04-06 11:11:33 <Kladky> I think I figured it out
2025-04-06 11:12:09 <Kladky> There are a few chars in the global prefix which are different in the server and router config
2025-04-06 11:12:10 <longnoserob[m]> Kladky: congrats
2025-04-06 11:12:29 <Kladky> So I will update the server's one to match, and see if that fixes it
2025-04-06 11:12:40 <Kladky> longnoserob[m]: I wouldn't jinx it like that lol
2025-04-06 11:26:12 <Kladky> I can now ping v6 addresses!
2025-04-06 11:26:24 <Kladky> Thank you Forza and longnoserob[m] 
2025-04-06 11:26:54 <Forza> Nice!
2025-04-06 11:33:28 <longnoserob[m]> great!
2025-04-06 11:41:23 <prabu> Kladky, i'll appreciate if you verify if the information mentioned in https://wiki.alpinelinux.org/wiki/Configure_Networking is correct from ipv6
2025-04-06 11:44:09 <prabu> there were changes made to that page to make it easier to follow and few things were updated from ifupdown-ng guide..
2025-04-06 12:44:13 <Kladky> Yeah, they are coorect, it's just that my router rotated it's v6 global prefix for some reason
2025-04-06 13:45:06 <herta> Hi all, is it possible to make `networking` service not provide `net` if it didn't start any interfaces (nothing is `auto` in /e/n/i)? I'm thinking that it is more logical, and it will closely match netifrc's behavior in Gentoo, where `net` is satisfied just by `lo` interface which is enabled by default
2025-04-06 13:49:15 <herta> They do it differently for alpine, they have no initscript that configures everything, similar to alpine's /etc/init.d/networking. Instead, they have `net.lo` which provides `net.lo` and `net`, and for other interfaces they symlink it to `net.$iface`, thus every symlink provides only `net.$iface`. Which also allows depending services on particular interface
2025-04-06 13:49:34 <herta> s/for/than/
2025-04-06 13:51:23 <herta> Alpine could be hacked to similar behavior by symlinking `networking` script to `networking.$iface`s but then each of these services provides `net`, which is not desired
2025-04-06 14:10:51 <zayd_webirc> fission: this works after I deleted and re extracted the source tarball, thanks (and sorry for taking so long, was asleep)
2025-04-06 19:36:58 <bleb> is there any documentation of how installation media is produced?
2025-04-06 19:37:20 <Habbie> i bet the software that makes the media is open source
2025-04-06 19:37:22 <bleb> where should I begin if I want to try reproducing an iso from source?
2025-04-06 19:37:42 <Habbie> https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage seems related at least
2025-04-06 19:38:04 <Habbie> or https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image
2025-04-06 19:46:08 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/scripts
2025-04-06 20:17:13 <bitdancer>  /save
2025-04-06 20:17:26 <bitdancer> oops
2025-04-06 20:19:08 <frag> BUSTED
2025-04-06 21:41:34 <Kenneth> Hi, im new to Alpine. Are anyone using Alpine as a daily driver on the Desktop?
2025-04-06 22:21:49 <rhizoome> yes
2025-04-06 22:37:26 <omni> yes
2025-04-06 22:51:04 <fabricionaweb> Yes
2025-04-06 23:02:39 <hazardchem> yes
2025-04-06 23:14:50 <jackson[m]> From Trusted and Vouched Dealers... (full message at <https://matrix.org/oftc/media/v1/media/download/ATt9l5KZ3mztJKWa6_ElQ2XCLjCaDyPQ7U_o6uEtLVQ-SNlMGaAIKuAHkeiG5DemOuxIKjKy-IZ3gSl_xxCKOo1CeWVXutawAG1hdHJpeC5vcmcvYXdJZ1NDZkFoQ3dYQkhDbGJCbkV3Qkhq>)
2025-04-06 23:16:27 <fjdzqoFsDUV2> no thanks, i'm already a customer of Honest Achmeds used Cars and Certificates...
2025-04-07 00:31:10 <the53hornet> Greetings.  I recently upgraded from Alpine 3.19 to 3.21 using the wiki as a guide. Now when I boot my laptop screen goes blank after i915 modesetting kicks in. I think that's what's happening because it's immediately after the udev event processing and if I add `nomodeset` to the boot flags, I get a console. If my laptop is attached to an external DP monitor, I see the whole console as normal there. Just the internal display blanks. Where could
2025-04-07 01:32:58 <fission> zayd_webirc: no worries, thanks for the update
2025-04-07 06:57:36 <omni> the53hornet: your message cut at "Where could"
2025-04-07 06:59:29 <omni> but I too noticed last week that something had changed, without thinking of it I had had `nomodeset` on the cmdline for years but now, with it, I couldn't start sway
2025-04-07 07:00:55 <omni> so I just removed it and that solved it for me, but you seem to have other problems, just thought both could be related to the same change, whatever it is
2025-04-07 07:06:29 <omni> I tried downgrading linux-lts and a few other aports I had in my cache, to figure out what it was that had changed, unfortunately I had not taken any recent rootfs snapshots
2025-04-07 07:12:54 <omni> "a few other aports": sway*1.10.1-r0*apk & wayland*1.23.1-r0*apk, so I ruled out sway, wayland and linux-lts upgrades, but then had to move on with other things
2025-04-07 07:16:08 <rhizoome> I still can't make swaylock not to "crash" every few days. I would like to at least get ctrl-alt-F* back, to switch to the virtual consoles, so I can fix the problem. Does anybody know how that works? It is hyprland.
2025-04-07 07:37:09 <rhizoome> Currently I have to boot another computer login via ssh and issue so
2025-04-07 07:37:29 <rhizoome> some hypctl commands
2025-04-07 07:44:48 <prabu> rhizoome, i use swaylock swaylock version 1.8.2
2025-04-07 07:45:33 <prabu> in 3.21 without any issues in sway
2025-04-07 07:46:36 <prabu> you may want to temporarily disable it in hyprland and see if this is really due to swaylock or something else...
2025-04-07 08:10:21 <rhizoome> prevents starting another swaylock if the screen is already locked. I was able to recude the problem form happending every few hours to a few days.
2025-04-07 08:10:21 <rhizoome> prabu: you mean using another lock-program? hyprland is telling me that swaylock crashed and that I can resolove the problem, by allowing the start of a new swaylock via hyprctl commands. when I examine the situation I always see that two swaylock have been started. Now I am pretty sure I have configured swayidle corretly (I read every bug report on swayidle and swaylock). By adding sleeps to swayidle "commands" and writing a wrapper that 
2025-04-07 08:12:11 <rhizoome> prabu: I am on edge though, but is is also swaylock 1.8.2
2025-04-07 08:16:46 <quinq> Yeah, librewolf broke profiles again
2025-04-07 08:16:52 <prabu> you may want to check alternates like hyprlock to isolate the issue. i never faced issues with swaylock in sway for nearly  a year and half 
2025-04-07 08:17:00 <quinq> Now I understand what some people commented in here before
2025-04-07 08:18:58 <quinq> It's trying to load *firefox* profiles, not its own
2025-04-07 08:25:12 <rhizoome> prabu: yes, I just want access to the virtual consoles frist, otherwise experimenting is really painful. I have to login via ssh everytime it crashes.
2025-04-07 08:29:58 <quinq> sertonix[m], could we rollback librewolf until this is fixed?
2025-04-07 08:30:10 <quinq> (in case you didn't see yet: https://codeberg.org/librewolf/issues/issues/2400)
2025-04-07 08:30:36 <quinq> Either rollback, or apply patches, but not stay in this broken state :D
2025-04-07 08:43:47 <sertonix[m]> quinq: There is already !82382 to fix the issue. Someone just has to hit the merge button.
2025-04-07 08:52:44 <quinq> ohhh, ok sertonix[m] 
2025-04-07 08:52:56 <quinq> Sorry I couldn't see that as I have web browsers problems ;)
2025-04-07 08:52:58 <quinq> Thank you very much!
2025-04-07 11:00:11 <the53hornet> omni: thanks, I guess my client isn't configured right. The last sentence was "Where could I look to start addressing this?" 
2025-04-07 11:01:32 <the53hornet> omni: and thanks for the advice. I've played with a lot of cmdline boot flags over the last couple of weeks. Can't seem to narrow it down.
2025-04-07 11:08:40 <the53hornet> omni: I don't have any root snapshots I can look back on (now I do for next time:)). I did upgrade to edge just to see if something got patched but the problem persists. Couldn't find anything useful in dmesg.
2025-04-07 16:22:25 <Guest12301> elogind installed and running, polkit installed and running... Why does greetd complain that "...[libseat]... No backend was able to open a seat" blablabla "Unable to create seat: Function not implemented"
2025-04-07 17:02:08 <realroot[m]> maybe seatd? idk
2025-04-07 17:44:48 <rhizoome> libseat should detect elogind, system-dbus is running?
2025-04-07 17:55:47 <quinq> sertonix[m], fixed! Thanks again :)
2025-04-07 18:35:17 <rhizoome> Guest12301: actually I just niced that I have set 'export LIBSEAT_BACKEND=logind'
2025-04-07 18:35:23 <rhizoome> noticed
2025-04-07 18:49:01 <WhyNotHugo> I've been thinking that we should have some per-package README files, in the same spirit as BSD ports
2025-04-07 18:49:14 <WhyNotHugo> For example, with details like the one just mentioned about elogind.
2025-04-07 18:49:25 <WhyNotHugo> Lots of Alpine packages have distro-specific instructions.
2025-04-07 19:01:27 <rhizoome> I found more information on why I cannot switch virtual console: https://tpaste.us/N5nm 
2025-04-07 19:03:11 <rhizoome> And it seems elogind should handle that: https://github.com/search?q=repo%3Aelogind%2Felogind%20SwitchTo&type=code
2025-04-07 19:03:55 <rhizoome> I am sure it never reaches polkit
2025-04-07 20:17:58 <fossdd[m]> WhyNotHugo: there are README.alpine files but they are afaik not really in use anymore
2025-04-07 20:18:26 <fossdd[m]> for example: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/80734/diffs#e67b33c8ed4ae700626bef4b063626e967a90db5
2025-04-07 20:19:14 <WhyNotHugo> fossdd[m]: are these installed into a common location? Would be nice if apk could notify of their existance when installing a new package
2025-04-07 20:20:43 <WhyNotHugo> rhizoome: consider if you really need elogind; if you don't have a program that strictly depends on it, life is easier without it
2025-04-07 20:21:04 <ikke> WhyNotHugo: there used to be a package for that
2025-04-07 20:21:21 <WhyNotHugo> what happened that all this went into disuse?
2025-04-07 20:21:34 <rhizoome> WhyNotHugo: I need polkit for all that fency desktop stuff.. or is there a way to run polkit with libseat?
2025-04-07 20:22:04 <rhizoome> s/libseat/seatd/
2025-04-07 20:22:21 <ikke> WhyNotHugo: https://pkgs.alpinelinux.org/package/edge/testing/x86_64/apk-readme
2025-04-07 20:22:25 <WhyNotHugo> rhizoome: afaik, polkit only needs elogind if you want to restrict permission to a users which is currently logged in via elogind. if you configure permissions via group membership you don't need it.
2025-04-07 20:22:31 <WhyNotHugo> that said, i don't use polkit
2025-04-07 20:22:43 <WhyNotHugo> er, i do, but only for modemmanager
2025-04-07 20:24:22 <WhyNotHugo> ikke: nice and simple. can I adopt and revive that?
2025-04-07 20:24:58 <ikke> I don't see why not
2025-04-07 20:25:00 <rhizoome> WhyNotHugo: I'll try it. I defined all my polkit rules as requiring group "wheel". I mix my hyprland setup, with a bit for xfce, thunar, udisksd.. it make same workflows a bit nicer.
2025-04-07 20:25:37 <ikke> WhyNotHugo: https://tpaste.us/xDB1
2025-04-07 20:25:41 <WhyNotHugo> rhizoome: only "subject.active" requires elogind. I removed that from modemmanager rules and it worked fine.
2025-04-07 20:25:57 <WhyNotHugo> ikke: haha, your diff is near-identical to mine XD
2025-04-07 20:26:12 <ikke> heh, no supprise
2025-04-07 20:27:30 <WhyNotHugo> not sure if i want to print the actual README or merely inform of its existance
2025-04-07 20:27:47 <WhyNotHugo> can i make a trigger run only on installation, and not on upgrade?
2025-04-07 20:28:45 <ikke> The trigger runs whenever the trigger file changes
2025-04-07 20:28:53 <ikke> the file it's set to trigger on
2025-04-07 20:29:17 <ikke> So any time something in /usr/share/*/README.alpine changes, it's executed
2025-04-07 20:29:22 <WhyNotHugo> So the readme is unchanged there's no trigger. Acceptable.
2025-04-07 20:29:31 <fossdd[m]> there is https://pkgs.alpinelinux.org/package/edge/testing/x86_64/apk-readme
2025-04-07 20:29:42 <ikke> fossdd[m]: that's what I just pointed to ;-)
2025-04-07 20:29:44 <rhizoome> WhyNotHugo: I am working on some linux audio stuff and some of the programs like to have rtkit. I first started of with seatd and nothing worked. But I wiki reads as if polkit does not work at all with seatd so I hadn't even installed it.
2025-04-07 20:29:58 <fossdd[m]> ikke: ah lol i should read the backlog
2025-04-07 20:32:36 <WhyNotHugo> apk-readme's trigger doesn't work tho. `apk add  autossh-doc` shows nothing
2025-04-07 20:34:36 <ikke> The trigger is `/usr/share/*/README.alpine`, autossh-doc has /usr/share/doc/autossh/README.alpine
2025-04-07 20:34:40 <ikke> There's an extra directory
2025-04-07 20:34:59 <WhyNotHugo> Oh, it's a shell-style glob, not a string-style glob.
2025-04-07 20:35:11 <WhyNotHugo> Does "**" work for this case?
2025-04-07 20:35:25 <ikke> I expect it uses fnmatch
2025-04-07 20:35:41 <ikke> s/expect/suspect/
2025-04-07 20:36:43 <WhyNotHugo> Most README.alpine files seem to match /usr/share/*/*/README.alpine
2025-04-07 20:36:56 <WhyNotHugo> dracut-doc include an README.alpine, a perfect example of such readme files!
2025-04-07 20:37:37 <ikke> WhyNotHugo: https://gitlab.alpinelinux.org/alpine/apk-tools/-/blob/2.14-stable/src/database.c#L1901
2025-04-07 20:38:11 <ikke> So * does not match '/'
2025-04-07 20:38:44 <ikke> And no support for **
2025-04-07 20:41:35 <rhizoome> Executing seatd-0.9.1-r0.post-install
2025-04-07 20:41:35 <rhizoome> Alpine just said this:
2025-04-07 20:41:35 <rhizoome> * If you need seatd-launch SUID binary, install it: apk add cmd:seatd-launch.
2025-04-07 20:41:56 <rhizoome> Mean because you are talking about README.alpine
2025-04-07 20:42:06 <ikke> Yeah, that's an alternative method, using post-install / post-upgrade script
2025-04-07 20:42:15 <WhyNotHugo> /usr/share/*/*/README.alpine still doesn't trigger for /usr/share/doc/autossh/README.alpine
2025-04-07 20:42:34 <WhyNotHugo> rhizoome: that works for one-liners
2025-04-07 20:42:49 <WhyNotHugo> but many packages have much deeper integration and distro-specific details
2025-04-07 20:43:03 <WhyNotHugo> bootloaders are a first-class example of this
2025-04-07 20:43:15 <ikke> WhyNotHugo: did you upgrade apk-readme before installing autossh-doc?
2025-04-07 20:43:34 <WhyNotHugo> abuild -rf && doas apk del apk-readme && doas apk add /home/hugo/packages/testing/x86_64/apk-readme-0.1-r1.apk
2025-04-07 20:43:59 <WhyNotHugo> And then doas apk del autossh-doc && doas apk add autossh-doc
2025-04-07 20:44:58 <ikke> apk info -t apk-readme
2025-04-07 20:46:06 <WhyNotHugo> "apk-readme-0.1-r1 triggers:" (nothing else)
2025-04-07 20:46:29 <ikke> It should list the paths that it has triggers for
2025-04-07 20:46:30 <rhizoome> WhyNotHugo: ok, reboot with seatd, sleep stopped work.. lets see if we can fix that
2025-04-07 20:46:40 <rhizoome> WhyNotHugo: but virtual console switch works
2025-04-07 20:47:05 <WhyNotHugo> rhizoome: elogind also doubles as a power management daemon, so i guess whatever you're using to trigger sleep uses elogind's API
2025-04-07 20:47:56 <WhyNotHugo> ikke: Any ideas what's wrong? https://tpaste.us/yBbp
2025-04-07 20:50:32 <ikke> Not really 
2025-04-07 20:54:25 <rhizoome> WhyNotHugo: polkit generally works, rtkit works, only suspend and reboot stopped working without password. Nice I take that. Thanks for the hint.
2025-04-07 20:54:36 <WhyNotHugo> rhizoome: I use powerctl
2025-04-07 20:54:51 <WhyNotHugo> which is suid-based
2025-04-07 20:56:27 <WhyNotHugo> what were you using for suspend?
2025-04-07 20:56:57 <rhizoome> elogind did that: loginctl suspend
2025-04-07 20:57:22 <rhizoome> is wrote .desktop files for that, I just have to switch commands to powerctl
2025-04-07 20:57:56 <rhizoome> and I do not care about the suspend button on my keyboard, that is just a gimmick
2025-04-07 20:58:29 <WhyNotHugo> on sway, I use "bindsym --locked XF86PowerOff  exec powerctl mem"
2025-04-07 20:59:10 <rhizoome> WhyNotHugo: I searched "powerctl alpine" and second hit was: https://whynothugo.nl/journal/2023/11/19/setting-up-an-alpine-linux-workstation/
2025-04-07 21:18:20 <sertonix[m]> WhyNotHugo: See also https://gitlab.alpinelinux.org/alpine/aports/-/issues/15653 and various linked stuff
2025-04-07 21:23:21 <rhizoome> WhyNotHugo: swayidles before-sleep stopped working. I guess I have to broadcast the sleep on dbus or something.
2025-04-07 21:24:41 <rhizoome> WhyNotHugo: but since I have my down suspend script it does not matter. Now with "bind = ,XF86Sleep, exec, syspend" everything is the same.
2025-04-07 21:29:21 <rhizoome> this system is now really nice. Did you know that with full dynamic preempt I can run Bitwig now with a buffer of 128 bytes, that is only 2.6ms delay. And the sound card in my computer is cheap, you cannot even do that on a mac with an expansive professional soundcard.
2025-04-08 00:58:05 <spikeddrink> does alpine have a package which allows me to make a customized ISO/IMG file(s)?
2025-04-08 01:03:57 <spikeddrink> preferably allowing the functionality of "Live" mode
2025-04-08 01:05:20 <spikeddrink> and maybe also toram option :-)
2025-04-08 10:30:52 <prabu> rhizoome, it's good to know that polkit can be used with seatd and thanks for updating wiki.
2025-04-08 10:31:33 <prabu> i tried using polkit+seatd+mate-polkit with nemo.. unsure how to get it working..
2025-04-08 10:32:10 <prabu> appreciate if you add some notes on wiki about how you're using seatd with polkit. later we can create a new page for polkit
2025-04-08 10:58:29 <rhizoome> prabu: for me it was just installing and starting both. Can you elaborate what difficulty is?
2025-04-08 11:02:27 <prabu> i installed polkit+mate-polkit and started polkit service along with swaymsg exec /usr/libexec/polkit-mate-authentication-agent-1. when i launch nemo and try to access /root folder, the authentication mechanism is not kicking in
2025-04-08 11:02:43 <rhizoome> Ah it might be that mate-polkit deploys polkit rules that use subject.active. In that case we have to make equivalent rules than use a group instead. I‘ll look what mate-polkit does in a few hours.
2025-04-08 11:03:47 <rhizoome> But usually you should get a manual authentication as a fallback, if rule is not met.
2025-04-08 11:04:07 <prabu> i just wanted to test and document how to use a privilege escalation GUI in sway with seatd
2025-04-08 11:05:44 <prabu> i'm not sure if i'm missing something else.. i don't have mate environment installed..just sway is installed https://git.sr.ht/~prabuanand/dotfiles/tree/master/item/etc/apk/world
2025-04-08 11:05:51 <rhizoome> polkit tells you about authentication requests in /var/log/messages. Maybe you start looking there
2025-04-08 11:08:16 <prabu> ok thanks..i'll check again later...
2025-04-08 11:08:37 <prabu> i'm going out now..
2025-04-08 11:08:48 <rhizoome> the env lile LIBSEAT_BACKEND are commented out here https://git.sr.ht/~prabuanand/dotfiles/tree/master/item/home/prabu/.profile. do you set them elsewhere?
2025-04-08 11:09:07 <rhizoome> s/lile/like/
2025-04-08 11:11:18 <prabu> i moved everything here /usr/local/bin/swayWM
2025-04-08 11:16:00 <rhizoome> that looks good, I guess we have to checl what org.freedesktop request you want to handled and why it is not. laters. I am out for a few hours, ping me laters.
2025-04-08 12:40:08 <tomleb> hey folks, after an upgrade recently (well, maybe a few weeks ago) my cursors have different icons and for some reason sometimes the cursor disappears when hovering some elements (eg: images in slides, objects in excalidraw, etc). Any idea? I tried installing another cursor icon package, I tried setting WLR_NO_HARDWARE_CURSORS=1 to see if it makes any changes (it didn't). I'm running sway.
2025-04-08 13:48:31 <sertonix[m]> Does that happen on gtk applications by any chance?
2025-04-08 14:00:12 <tomleb> It happens on firefox, trying to find what firefox is built with.. I don't see it happening with LibreOffice fwiw (which from what I can tell is using their own toolkit)
2025-04-08 14:12:58 <invoked> tomleb: https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/community/firefox
2025-04-08 14:15:12 <sertonix[m]> firefox uses gtk. 
2025-04-08 14:17:15 <sertonix[m]> You could try to set the icon theme that sway uses in gsettings too.
2025-04-08 14:24:42 <tomleb> Tried it and nothing changed. Looking in the sway log, I see "Unable to receive a desktop appearance ....No such interface "org.freedesktop.portal.Settings"" 
2025-04-08 14:24:58 <tomleb> Dunno if related though 
2025-04-08 14:33:15 <rhizoome> tomleb: you need to have gsettings-desktop-schemas installed otherwise gsettings will "know" nothing.
2025-04-08 14:34:07 <rhizoome> probably needs a restart of related processes.
2025-04-08 14:40:43 <tomleb> rhizoome: seems like it's already installed. Does it need a service to run or something else?
2025-04-08 14:43:59 <rhizoome> tomleb: no, then it should have worked. When I wanted to change to dark theme, I hadn't installed it and was confused why it didn't work.
2025-04-08 14:45:22 <rhizoome> tomleb: These people should be talking about your problem: https://www.reddit.com/r/hyprland/comments/1hxt40n/no_custom_cursor_on_gtk_3_applications/
2025-04-08 14:45:46 <rhizoome> tomleb: I have it too and didn't bother to fix it. I am currently trying. I let you know.
2025-04-08 14:48:18 <tomleb> rhizoome: nice thanks, I'll follow that thread. Also same I didn't bother at first but after a few weeks I'd like to fix it lol.
2025-04-08 15:07:53 <rhizoome> tomleb: I haven't solved it,yet. I have another complication: I have hyprland. But this should help you: https://github.com/swaywm/sway/wiki/GTK-3-settings-on-Wayland
2025-04-08 15:08:22 <rhizoome> tomleb: it is basically download and install a theme and set it in both sway and gsettings.
2025-04-08 15:39:09 <rhizoome> tomleb: so I solved it. It is slightly more complex in my case. 1. Download a theme that is availble as xcursors and hyprcursors. 2. set the theme like this:https://tpaste.us/vojo
2025-04-08 15:40:18 <rhizoome> I guess you just have to download one theme and set it in sway, gsettings and the env vars. The env vars are for qt5 applications.
2025-04-08 15:51:18 <rhizoome> tomleb: about your question if gsettings needs a service. The answer is: It needs a backend, but it automatically starts it. I have /usr/libexec/dconf-service running.
2025-04-08 16:53:03 <sertonix[m]> If you don't have dconf installed gsettings still works and doesn't start any service. Maybe it can't live update the config then
2025-04-08 18:07:37 <haoyi> I was wondering what was the best way to automatically update my alpine linux box. I want to know the pros and cons of it and essentially a way to prevent postgresql from updating and messing up the current postgres version I am on since I would have to manually upgrade that.
2025-04-08 18:11:01 <lopid> automatic updates aren't good
2025-04-08 18:23:38 <Saijin_Naib[m]> <haoyi> "I was wondering what was the..." <- maybe the apk cron package or topgrade on cron, with pgsql version pinned in world
2025-04-08 18:24:44 <Saijin_Naib[m]> Keep /etc/apk/repositories pointed to latest-stable instead of a numbered version
2025-04-08 18:25:02 <Saijin_Naib[m]> Works with no issues on my wife's computet
2025-04-08 18:57:48 <seninha> Hi, i installed alpine linux on a Thinkpad X201 Tablet, with an Wacom w8001 tablet, which is connected via serial (not USB) port. And I want to make it to work on X11.
2025-04-08 18:58:55 <seninha> I can run `doas cat /dev/ttyS0` and move the pen over the screen, and it will print binary garbage (which is what i expect, the tablet is working).
2025-04-08 18:59:56 <seninha> I also had run setup-devd and installed xf86-input-wacom and libwacom, rebooted, and the wacom tablet is not listed by “xsetwacom --list”
2025-04-08 19:02:51 <seninha> Gentoo wiki says the kernels needs to have wacom_w8001 driver and attach the tablet with inputattach(1). However, i don't know whether (or how) that applies to Alpine.
2025-04-08 19:11:16 <sertonix[m]> seninga: Could you try installing libwacom-udev and then reboot?
2025-04-08 19:13:09 <sertonix[m]> Oh, lookls like you already did
2025-04-08 19:15:17 <sertonix[m]> The kernel drivers should be included. You can search for "WACOM_W8001" in /boot/config-* to verify
2025-04-08 19:15:23 <seninha> Ok. modprobe/lsmod lists that wacom_w8001 is available; and the inputattach(1) comand is at linuxconsoletools
2025-04-08 19:18:24 <seninha> I tried “inputattach -w8001 /dev/ttyS0” and got this: https://termbin.com/3b7y
2025-04-08 19:18:40 <seninha> Xorg.0.log ^
2025-04-08 19:18:58 <seninha> Xorg segfaults
2025-04-08 20:05:48 <tomleb> sertonix[m]: rhizoome: looks like I have dconf-service running fwiw
2025-04-08 20:15:51 <rhizoome> » gsettings get org.gnome.desktop.interface cursor-theme
2025-04-08 20:15:51 <rhizoome> 'BreezeX-RoséPineDawn'
2025-04-08 20:15:51 <rhizoome> tomleb: you can find out if it worked with
2025-04-08 20:27:53 <the53hornet> I've been doing some monitoring and one thing I noticed watching dmesg for hours is a lot of "HPD interrupt storm detected on eDP-1".
2025-04-08 20:28:19 <the53hornet> eDP-1 is my internal display, which is the one that blanks out at boot time after modesetting kicks in. I'm not sure this is relevant or not, but I'll look for some docs to see what it's about. 
2025-04-08 20:36:32 <tomleb> rhizoome: yeah I had checked that already and it appears correctly. And yet firefox is not using that theme (or any other theme I set).
2025-04-08 20:39:24 <rhizoome> tomleb: for me firefox even uses hyprcursors, ie very good wayland integration. you have GDK_BACKEND=wayland?
2025-04-08 20:51:44 <rhizoome> tomleb: I know this is another rabbit hole, but is use this https://tpaste.us/DrZ0 to switch dark/light themes. All my applications, except that stupid Qt5 application that also ignored the cursor theme react to it. I commented the lines out that I use for xfce applications and the one for the stupid Qt5 app.
2025-04-08 20:52:13 <rhizoome> out
2025-04-08 22:28:06 <seninha> Hi, i installed alpine linux on a Thinkpad X201 Tablet, with an Wacom w8001 tablet, which is connected via serial (not USB) port. And I want to make it to work on X11.   I can run `doas cat /dev/ttyS0` and move the pen over the screen, and it will print binary garbage (which is what i expect, the tablet is working).  I also had run setup-devd and installed xf86-input-wacom and libwacom, rebooted, and the wacom tablet is not listed by “xsetwaco
2025-04-08 22:28:06 <seninha> m --list”.   Gentoo wiki says the kernels needs to have wacom_w8001 driver and attach the tablet with inputattach(1). However, i don't know whether (or how) that applies to Alpine.
2025-04-08 22:28:09 <seninha> (bumping)
2025-04-08 22:28:30 <seninha> modprobe/lsmod lists that wacom_w8001 is available; and the inputattach(1) comand is at linuxconsoletools.  I tried “inputattach -w8001 /dev/ttyS0”, but Xorg segfaults: https://termbin.com/3b7y
2025-04-08 23:44:30 <haoyi> Saijin_Naib[m]: I have it on edge is there anyway to just have postgresql not updated automatically or notify me if postgresql needs an update?
2025-04-08 23:55:07 <Saijin_Naib[m]> <haoyi> "Saijin_Naib: I have it on edge..." <- You can pin a package version with apk add, or editing world
2025-04-08 23:56:32 <Saijin_Naib[m]> https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper
2025-04-08 23:57:06 <Saijin_Naib[m]> I'm not sure how much of a dependency solving nightmare pinning something like pgsql might lead to...
2025-04-09 19:48:14 <realroot[m]> qtcreator seg faults when creating a project how fix it?
2025-04-09 19:48:28 <realroot[m]> missing dep?
2025-04-09 19:49:49 <ikke> A segfault rarely happens just due to a missing dependency
2025-04-09 19:55:24 <realroot[m]> no idea i had one in  tilix cause of that
2025-04-09 19:57:55 <fossdd[m]> it could happen if it tries to dlopen a so
2025-04-09 19:58:00 <fossdd[m]> can you attach a debugger?
2025-04-09 20:01:41 <realroot[m]> https://paste.centos.org/view/2acae9dc
2025-04-09 20:03:14 <realroot[m]> i asked in #Qt too
2025-04-09 20:07:35 <realroot[m]> does qtcreator have debug symbols?
2025-04-09 20:09:32 <ikke> Apparently not
2025-04-09 20:13:14 <realroot[m]> qt6-qtbase-dbg?
2025-04-09 20:13:46 <realroot[m]> it seems so I see new output
2025-04-10 12:01:54 <realroot[m]> qt dev told me this:
2025-04-10 12:05:31 <realroot[m]> 
ACTION uploaded an image: (97KiB) < https://matrix.org/oftc/media/v1/media/download/AZDr6prKfzl3s405I6bMMufSEPa5xHz5xbZIL5Vzb7LDbuG920MZGXf5cR9vkyuX-WBJbseL37LrV1lkzgvtI7lCeWZ7BS8QAG5hZGVrby5uZXQvYXJrY2dhclVpTmltUnR6bWljdUNXd0li >

2025-04-10 14:10:11 <sigirneg> is there an official/easy way to run diskless mode on edge repos?
2025-04-10 14:23:16 <ikke> after installing a release, it's fairly easy to switch it to edge
2025-04-10 14:23:27 <ikke> The problem is the kernel
2025-04-10 14:23:41 <ikke> that needs to be written to the boot media
2025-04-10 16:35:42 <jahway603[m]> Is there a way for a package maintainer to see what fails in the pipeline if they did not package the update? I'm specifically referencing !82521
2025-04-10 17:03:07 <ptrc> jahway603[m]: you can just click the "pipelines" tab of a merge request
2025-04-10 17:03:23 <ptrc> then click through the failed one to find the actual job
2025-04-10 17:03:57 <ptrc> or, i suppose, in this case the entire pipeline failed for some internal reasons
2025-04-10 19:11:16 <rhizoome> WhyNotHugo: no I have found something that does not work with polkit/seatd authentication. You are basically restricted to YES or NO rules, no AUTH_ADMIN rules. https://tpaste.us/QK6Q
2025-04-10 19:11:36 <rhizoome> s/no I/now I/
2025-04-10 19:18:49 <rhizoome> yes all these agents need POLKIT_IS_SUBJECT https://tpaste.us/KxnP
2025-04-10 19:19:44 <simon_> hey
2025-04-10 19:19:59 <simon_> I have a question how to setup ip tables for desktop user?
2025-04-10 19:20:09 <simon_> I want to use alpine as a desktop
2025-04-10 19:20:27 <rhizoome> I wonder what it would take to add that functionality to seatd?
2025-04-10 19:21:01 <simon_> why does ai tell me that only masochists use alpine as desktop?
2025-04-10 19:21:03 <simon_> :D
2025-04-10 19:21:26 <rhizoome> simon_: I don't quite understand your question, most of networking is very independant of the user.
2025-04-10 19:21:42 <simon_> rhizoome:  I want a firewall
2025-04-10 19:21:48 <simon_> against bad peopl
2025-04-10 19:22:04 <rhizoome> simon_: you like to have a firewall GUI?
2025-04-10 19:22:24 <simon_> gui or a configuration with iptables or nft tables
2025-04-10 19:22:33 <ikke> simon_: iptables -I INPUT 1 -j DROP; would block all bad people
2025-04-10 19:22:45 <ikke> (but other usefull traffic as well ;-))
2025-04-10 19:22:48 <simon_> well I also want to play video games :D
2025-04-10 19:23:05 <simon_> use browser and such
2025-04-10 19:23:53 <rhizoome> simon_: well, if no GUI you can set iptable rules with 'doas iptables' or 'sudo iptables' and you can add a open-rc service to save and load your rules at boot and shutown.
2025-04-10 19:23:59 <ikke> add -A INPUT -m conntrack --ctstate ESTABLISHED -j ACCEPT as first rule, and set policy to block, and your mostly done
2025-04-10 19:24:51 <simon_> add -A INPUT -m conntrack --ctstate ESTABLISHED -j ACCEPT as first rule, and set policy to block """"" is this whole thing a command ?? :D
2025-04-10 19:25:25 <rhizoome> simon_: https://wiki.alpinelinux.org/wiki/Iptables
2025-04-10 19:25:31 <simon_> I am very demanding unfortunately cuz network is nothing I know
2025-04-10 19:25:42 <ikke> iptables -I INPUT -m conntrack --ctstate ESTABLISHED -j ACCEPT; iptables -P INPUT DROP;
2025-04-10 19:27:35 <rhizoome> simon_:  do what the wiki says, then add the rule ikke provided and then call "rc-service iptables save"
2025-04-10 19:33:52 <simon_> gnome or plasma
2025-04-10 19:33:54 <simon_> ?
2025-04-10 19:34:51 <ikke> tabs or spaces?
2025-04-10 19:36:45 <simon_> ikke: tabs
2025-04-10 19:50:49 <simon_> what is windowmanager from .xinitrc? is that what gnome is using?
2025-04-10 19:53:16 <ikke> I'm not too familiar with gnome or kde
2025-04-10 19:57:44 <simon_> is setup-apkrepos -cf supposed to run in root?
2025-04-10 19:58:47 <simon_> I guess yes
2025-04-10 19:59:07 <ikke> changing /etc/apk/repositories requires root, yes
2025-04-10 20:00:54 <bent_fingers> any documentation for iconv?
2025-04-10 20:01:01 <bent_fingers> i can't find any
2025-04-10 20:01:38 <simon_> in my /etc/apk/repositories is now 2 repositories one time edge and one time main is that how it is supposed to beß
2025-04-10 20:02:56 <bent_fingers> i'll use gnu's i guess
2025-04-10 20:03:07 <rhizoome> bent_fingers: no musl iconv has no man page, you need browser one on the web
2025-04-10 20:16:20 <rhizoome> simon_: maybe you want follow this guide[1] it explains many details that are special in alpine linux. [1]:https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_repositories
2025-04-10 20:23:19 <WhyNotHugo> rhizoome: yeah, all the agents rely on systemd-logind or elogind
2025-04-10 20:23:53 <WhyNotHugo> seatd has nothing to do with any of this, but you can ask in #kennylevinsen on libera.chat
2025-04-10 20:24:14 <WhyNotHugo> afaik only polkit with group membership rules work without logind, anything else does not.
2025-04-10 20:25:14 <WhyNotHugo> tbf, none of this is documented anywhere, we ought to have a README.alpine in the package, or maybe just a wiki page
2025-04-10 20:25:39 <WhyNotHugo> Oh, it's actually here: https://wiki.alpinelinux.org/wiki/Seatd
2025-04-10 20:25:46 <rhizoome> WhyNotHugo: I'd like to add a reboot similar to sleep via powerctl (without typing a password. Do you have that setup?
2025-04-10 20:26:18 <rhizoome> WhyNotHugo: That was me how added that note on the wiki.
2025-04-10 20:26:27 <rhizoome> s/how/who/
2025-04-10 20:28:27 <simon_> ikke have we blocked downloading from http? 
2025-04-10 20:28:28 <simon_> :D
2025-04-10 20:28:40 <WhyNotHugo> i don't. you can probably configure doas to let you run reboot without passwd
2025-04-10 20:29:01 <WhyNotHugo> ah, nice, thanks for documenting it somewhere ;)
2025-04-10 20:29:07 <WhyNotHugo> * :)
2025-04-10 20:29:32 <rhizoome> WhyNotHugo: Yes, doas rules is what I usually do.
2025-04-10 20:30:15 <WhyNotHugo> logind is such a horrible pile of hacks. I understand why they'd dump all this functionality in one daemon: it's a quick and dirty way to get it done quickly. but I wouldn't call this "designed" at all
2025-04-10 20:30:57 <simon_> iptables -I INPUT -p tcp --dport 80 -m conntrack --ctstate NEW -j ACCEPT
2025-04-10 20:30:57 <simon_>  can I use this?
2025-04-10 20:30:57 <simon_> iptables -I INPUT -p tcp --dport 443 -m conntrack --ctstate NEW -j ACCEPT
2025-04-10 20:31:32 <ikke> simon_: that would allow new connections (from outside to in) on port 80 and 443
2025-04-10 20:31:53 <simon_> and thats http and https?
2025-04-10 20:32:08 <simon_> so a download for example?
2025-04-10 20:32:22 <ikke> No, a download would not be a new connection from outside to in
2025-04-10 20:32:50 <ikke> Your PC would initiate the connection, so it would be ESTABLISHED
2025-04-10 20:33:19 <simon_> but why flatpak steam say Error Download failed gttp error 0 :D
2025-04-10 20:33:25 <simon_> must be the firewall imo
2025-04-10 20:33:38 <simon_> http not gttp
2025-04-10 20:34:47 <rhizoome> WhyNotHugo: btw on the seatd wiki I saw you wanted to remove the environment variables, but there is a second copy of that list of variables. I didn't touch because, I don't know enough about it.
2025-04-10 20:36:36 <WhyNotHugo> yeah, those variables are not needed
2025-04-10 20:36:54 <WhyNotHugo> what i dislike about the wiki is that someone comes around and edits it and there's no review process where one can ask "why did you need this?".
2025-04-10 20:37:11 <WhyNotHugo> it might be that some specific compositor requires this, or some specific scenario, but not we'll never know
2025-04-10 20:37:53 <simon_> ikke so you dont know what it could be?
2025-04-10 20:37:54 <simon_> :D
2025-04-10 20:38:06 <rhizoome> simon_: with iptables -L list your rules. The rules are applied on order topdown. You might have borked something if flatpak does custom rules.
2025-04-10 20:40:03 <simon_> https://paste.debian.net/1368763/ rhizoome 
2025-04-10 20:40:14 <rhizoome> WhyNotHugo: Yes, I still have the LIBSEAT_BACKEND variable for cargo cult purposes, lets me sleep better :-) 
2025-04-10 20:40:53 <simon_> can you open browser or are you fully in a terminal heh
2025-04-10 20:41:04 <WhyNotHugo> i never set any of them, but my setup predates that wiki article's existance
2025-04-10 20:41:05 <simon_> I heard alpine user in terminal only :D
2025-04-10 20:41:24 <rhizoome> WhyNotHugo: the git-annex person, has a wiki that is in git. We could manage it through gitlab :-)
2025-04-10 20:41:38 <WhyNotHugo> simon_: I use the same graphical setup I used on Arch (sway, firefox)
2025-04-10 20:41:54 <WhyNotHugo> simon_: even play steam games and starcraft 2 (although these require flatpak)
2025-04-10 20:42:36 <WhyNotHugo> rhizoome: i don't dislike the idea, but the inertia to move the wiki in such a way doesn't seem minor
2025-04-10 20:42:41 <simon_> WhyNotHugo:  you mean you use the same iptable rules as me?
2025-04-10 20:43:31 <rhizoome> simon_: looks fine. If your webbrowser still works. it is some unrelated problem with flatpak. With iptables -F you can remove the rule and see if flatpak works again.
2025-04-10 20:43:50 <WhyNotHugo> simon_: i was replying to your other message. nvm
2025-04-10 20:45:05 <rhizoome> simon_: wait, actuall you need to change policy too iptables -P INPUT ACCEPT;
2025-04-10 20:46:15 <rhizoome> maybe we have lost simon_ because he locked himself out. I didn't think about the default policy,
2025-04-10 20:46:37 <rhizoome> a reboot will fix it.
2025-04-10 20:47:30 <rhizoome> WhyNotHugo: there is also the handbook, it is in git: https://docs.alpinelinux.org/user-handbook/0.1a/index.html
2025-04-10 20:48:02 <WhyNotHugo> i don't think anyone is currently working on the handbook
2025-04-10 20:48:23 <WhyNotHugo> i found it a bit of a pain to understand, since there's no 1:1 mapping beteween source files and pages, there's lot of inclusions macros and stuff
2025-04-10 20:48:56 <simon_> I am still here
2025-04-10 20:49:14 <simon_> so I need to do iptables -P INPUT ACCEPT;
2025-04-10 20:49:31 <rhizoome> simon_: nice. yes. then everything is accepted again.
2025-04-10 20:49:42 <WhyNotHugo> I like the idea of README.alpine files shipped with packages (in the style of bsd ports), but there's no mechanism to link between pages, which is far from ideal
2025-04-10 20:50:41 <WhyNotHugo> why iptables instead of nftables? The latter seems… easier
2025-04-10 20:50:51 <simon_> jup works thanks
2025-04-10 20:51:49 <simon_> by the way I was previously on chimera linux and had 5 gb ram usage by just using a desktop environment :D
2025-04-10 20:51:58 <rhizoome> WhyNotHugo: we can whip up `aman` alpine-man or something. I just found `apk add docs`, I was looking for this for ages and the search engine just denied me that information.
2025-04-10 20:51:58 <simon_> that was unusual
2025-04-10 20:52:42 <WhyNotHugo> rhizoome: do we have some page with "useful tips for newcomers"? XD
2025-04-10 20:53:17 <simon_> WhyNotHugo: well all I needed was steam and telegram
2025-04-10 20:53:18 <simon_> :D
2025-04-10 20:54:09 <rhizoome> WhyNotHugo: the user manual is the newcomer page for, I find it very well structured and it explains the "alpine" concepts well.
2025-04-10 20:54:19 <rhizoome> for me
2025-04-10 20:54:24 <simon_> I know apk is better than apt pacman eopkg emerge zypper what else exist
2025-04-10 20:55:17 <simon_> but nixos configuration.nix is insane good :D
2025-04-10 20:55:26 <simon_> but I am not that much of a masochist
2025-04-10 20:55:53 <simon_> for me alpine is already good
2025-04-10 20:55:57 <simon_> I hope
2025-04-10 20:55:58 <simon_> ;)
2025-04-10 20:56:24 <simon_> as long as apk doesnt destroy my system
2025-04-10 20:56:29 <simon_> everything is fine
2025-04-10 20:58:13 <simon_> by the way a friend of me want to go away from glibc distros too you think alpine is beginner friendly? he is someone who makes desktops look really good
2025-04-10 20:58:28 <simon_> :D
2025-04-10 20:59:43 <simon_> https://github.com/justincormack/alpine-pkgsrc someone tried this?
2025-04-10 21:00:29 <simon_> I forgot to add dnf to my list of abandon
2025-04-10 21:00:43 <rhizoome> WhyNotHugo: I though the manual is mdbook. There you cannot do anything complicated, a file is page, a folder is chapter and thats it. But it is Antora.
2025-04-10 21:03:24 <simon_> brb
2025-04-10 21:04:54 <simon_> good night
2025-04-10 21:04:56 <simon_> ;)
2025-04-10 21:05:04 <rhizoome> good night
2025-04-11 00:39:42 <cosmic-desktop-woes> Hello all. Seeking guidance on installing the COSMIC desktop environment.
2025-04-11 00:41:34 <cosmic-desktop-woes> I have installed the package "cosmic-session". This installed the related packages for COSMIC. I am currently investigating why it will not start, upon executing "cosmic-comp", "start-cosmic" etc.
2025-04-11 00:43:12 <cosmic-desktop-woes> Have not used a desktop greeter before, so any insight is appreciated.
2025-04-11 00:54:06 <cosmic-desktop-woes> I will check the logs for any responses. Signing out all: thanks in advance.
2025-04-11 03:39:22 <prabu> WhyNotHugo, when i moved from Arch to Alpine, i brought the LIBSEAT_BACKEND variable. I confirm that Alpine linux does not need this variable for seatd+sway
2025-04-11 03:50:52 <prabu> WhyNotHugo, there is a 'patroller' role in wiki who are supposed to review the changes made to wiki. rightfully that role should have few members to review changes to wiki.
2025-04-11 03:52:11 <prabu> if you don't mind, we can ask ikke to add yourself and rhizoome(if he's interested) to that role
2025-04-11 03:54:26 <prabu> a patroller cannot review one's own changes to wiki.. this is a team role in wiki
2025-04-11 07:46:51 <rhizoome> prabu WhyNotHugo it is a little problem we all - definitely I - know far from everything. So everything short of standard MR-review process where the people who know things get pinged, is not going to work well. Having said that I don’t mind a little cargo cult on the wiki. As long as it isn’t completely wrong. Also many thanks to prabu for doing reviews and testing on the wiki. 
2025-04-11 07:49:41 <rhizoome> prabu my note on seatd and polkit isn’t very good, exactly because I don’t know enough about it. I tried to update it, as I learned new things.
2025-04-11 08:35:08 <wyk72> hello !
2025-04-11 08:35:48 <wyk72> I have a question: what is the best method for compiling an old PHP (7.4.33) into Alpine-current (3.21) ?
2025-04-11 08:36:30 <wyk72> should I simply get the sources and compile them (.configure && make) or ahould I go with the APKBUILD route ?
2025-04-11 08:37:06 <wyk72> the (old) APKBUILD for 7.4.33 PHP is usable into alpine-current ?
2025-04-11 08:38:02 <aron> I'd suggest to use APKBUILD, definitely worth a try
2025-04-11 08:39:06 <wyk72> ok, thanks
2025-04-11 08:39:14 <aron> using apk is preferred, since apk will keep track of the installed files, and you can easily clean it up
2025-04-11 08:40:04 <wyk72> I have a lot of issues with openssl: php 7.4.33 supports only 1.1.1 (max)
2025-04-11 08:40:45 <aron> I don't think it will change if you rebuild it
2025-04-11 08:42:00 <wyk72> will try
2025-04-11 08:42:08 <aron> openssl2+ has changed API, thus they are incompatible, and php is unable to use it
2025-04-11 08:42:33 <aron> but sure, give it a try
2025-04-11 08:44:37 <wyk72> many dockerfiles just compile openssl 1.1.1 and install it before php7
2025-04-11 08:45:17 <aron> openssl 1.1.1 is available from pkg
2025-04-11 08:45:55 <aron> why would you compile it?
2025-04-11 08:46:45 <wyk72> I had a lot of issues using old binaries with new kernels
2025-04-11 08:47:13 <wyk72> they usually lack some bits & pieces (fail to load a symbol etc)
2025-04-11 08:47:47 <wyk72> it's much safer to just recompile them 
2025-04-11 08:48:01 <aron> ok, I see, openssl 1.1.1 is removed since alpine 3.19
2025-04-11 08:48:37 <WhyNotHugo> abuild rootbld hangs indefnitely for me for !82599. it hangs after instaling all dependencies
2025-04-11 08:49:07 <WhyNotHugo> apparently stuck running awk
2025-04-11 09:00:57 <ikke> WhyNotHugo: an MR in abuild has been fixed yesterday to address hanga
2025-04-11 09:01:00 <ikke> Hangz
2025-04-11 09:01:22 <ikke> Difference between bb awk and gawk/mawk
2025-04-11 09:27:39 <dngray> ptrc having a look at your comments in there https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112
2025-04-11 09:27:54 <dngray> did you mean like
2025-04-11 09:28:02 <dngray> mv "$builddir" "$srcdir"/"$pkgname"-"$pkgver"/software
2025-04-11 09:28:22 <dngray> im kind of confused about what you meant
2025-04-11 09:50:32 <simon_> does it make sense to take an edge kernel?
2025-04-11 09:50:34 <simon_> on stable
2025-04-11 09:50:57 <simon_> 6.12.21-1-lts have this kernel currently
2025-04-11 09:52:28 <simon_> linux-edge-6.14.0-r0
2025-04-11 09:55:24 <simon_> ( I just install it )
2025-04-11 10:28:19 <rhizoome> simon_ that should work, but generally you have to be careful mixing stable and edge 
2025-04-11 11:05:16 <Bdhfjf> I have 1.95 ram usage on gnome I thought that's a bit suspicious that's why I am here to ask whether it's usual the Internet says it should be 1.45
2025-04-11 11:06:46 <Bdhfjf> But I could also install an arch in a VM to check if it's true
2025-04-11 11:06:50 <Bdhfjf> :D
2025-04-11 11:07:02 <Bdhfjf> Let me know what you think
2025-04-11 11:08:03 <aron> I don't really like when a DE eats my memory :)
2025-04-11 11:08:04 <Bdhfjf> Got to go 
2025-04-11 11:08:24 <aron> 1.5 gigs is utterly wasteful to my taste
2025-04-11 11:08:51 <Bdhfjf> Nah Aron I was yesterday on chimera Linux it had 5.6 GB ram usage in KDE so well that's even more sus
2025-04-11 11:09:04 <aron> lol
2025-04-11 11:09:10 <Bdhfjf> Aron and 1.95??
2025-04-11 11:09:11 <Bdhfjf> Xd
2025-04-11 11:09:27 <aron> for me, a few hundred megs at max
2025-04-11 11:09:46 <Bdhfjf> Aron I am a desktop user ;-) 
2025-04-11 11:09:57 <aron> I have a desktop as well
2025-04-11 11:10:05 <Bdhfjf> But xfce ?
2025-04-11 11:10:10 <aron> I use xfce4
2025-04-11 11:10:21 <aron> simpler, uses much less resources
2025-04-11 11:10:39 <Bdhfjf> But do you think it's virus that gnome has 1.95 GB ram usage or is that usual and expected
2025-04-11 11:11:00 <aron> dunno
2025-04-11 11:11:07 <Bdhfjf> After all gnome gives me best performance ingame 
2025-04-11 11:11:15 <Bdhfjf> 200 fps XD 
2025-04-11 11:11:48 <aron> enjoy then :)
2025-04-11 11:12:10 <Bdhfjf> But still the 1.95 ram usage is very sus for me
2025-04-11 11:12:36 <Bdhfjf> I first handle noticed much ram usage yesterday on chimera Linux 
2025-04-11 11:12:54 <Bdhfjf> Now I see 2 gigs in alpine 
2025-04-11 11:13:01 <Bdhfjf> And think about chimera 
2025-04-11 11:13:01 <Bdhfjf> Xd
2025-04-11 11:15:03 <Bdhfjf> Anyways got togo
2025-04-11 11:15:06 <Bdhfjf> Cya
2025-04-11 11:33:44 <ptrc> dngray: sorry, i meant setting builddir= to that value :p
2025-04-11 11:45:50 <p_6f3Ik7Suw> hmmm. my DE consumes 43MB, a sticky terminal tailing logs, a conky and an i3wm.
2025-04-11 11:46:09 <p_6f3Ik7Suw> and i think even that is quite hungry.
2025-04-11 11:47:09 <p_6f3Ik7Suw> (of which the conky uses 22MB, which is huge for the 13px tall stripe it occupies)
2025-04-11 11:47:22 <aron> lol
2025-04-11 11:48:14 <aron> 100-200 megs for linux + running DE on x86_64 (all services included) is okay for me
2025-04-11 11:53:01 <WhyNotHugo> ikke: oh, thanks
2025-04-11 11:53:06 <WhyNotHugo> i wish i could get rid of gawk entirely
2025-04-11 11:56:40 <WhyNotHugo> the fact that util-linux-dev isn't split is a continuous source of breakage for me, it has me completely fed up
2025-04-11 11:58:17 <xpufx> init script. it's not loading any files..  command_args="$REST_SERVER_OPTS --path '${REST_SERVER_PATH}'"
2025-04-11 11:58:23 <xpufx> where would those variables come from?
2025-04-11 11:59:25 <xpufx> oh conf.d
2025-04-11 15:15:49 <dngray> ptrc: ohh
2025-04-11 16:52:26 <dngray> ptrc when i did that the patch can't be automatically applied
2025-04-11 17:03:54 <ptrc> ah, fair
2025-04-11 17:04:01 <ptrc> i guess you *could* manually edit the paths in the patch
2025-04-11 17:04:05 <ptrc> but you can also just leave it as-is
2025-04-11 17:18:52 <dngray> actually i could get rid of that tools package too
2025-04-11 17:38:02 <lassebq> Hey, so I recently found out about the DPS https://uapi-group.org/specifications/specs/discoverable_partitions_specification/ and I thought to myself this would be a perfect solution for me whenever I'll move my SSD to another system. My current setup involves kernel being booted directly from UEFI, and the boot entry is stored in NVRAM. That means the moment I upgrade my system I would have to manually re-create boot entry from a live e
2025-04-11 17:38:05 <lassebq> nvironment. My expectations were cut short because as it seems the only implementation of this spec is works with systemd initramfs image. My kernel doesn't use one. I've been running default efistub bundled with upstream kernel. So my question is, is there anything else that implements this spec and doesn't depend on systemd?
2025-04-11 17:41:13 <p_6f3Ik7Suw> Unable to decrypt message
2025-04-11 17:41:34 <gentile> what ?
2025-04-11 17:41:57 <ikke> sounds like a matrix error
2025-04-11 17:42:54 <dngray> it does lol
2025-04-11 17:43:18 <gentile> alpine has a matrix server ?
2025-04-11 17:43:38 <nu_> not yet
2025-04-11 17:43:49 <gentile> then why bring matrix up lol
2025-04-11 17:43:57 <nu_> probably the bridge misbehaving
2025-04-11 17:44:25 <nu_> or their client can't decrypt for some reason in the bridged matrix room
2025-04-11 17:46:44 <gentile> but you said alpine doesn't have a matrix channel
2025-04-11 17:46:49 <gentile> why would there be a bridge
2025-04-11 17:47:27 <nu_> having a server and having this irc room bridged into a random matrix server room is not the same^^
2025-04-11 17:48:38 <gentile> oh
2025-04-11 17:48:53 <gentile> yeah i guess
2025-04-11 17:55:43 <invoked> lassebq: that stuff is very early. also, my read of that is the main audience for that stuff is for production deployments, vms and containers primarily, not desktops
2025-04-11 17:55:59 <lassebq> Huh
2025-04-11 17:56:08 <invoked> although maybe it will impact desktops someday
2025-04-11 17:56:23 <lassebq> Why do these hundreds of partition GUIDs exist if not for general use?
2025-04-11 17:56:36 <lassebq> Makes no sense with what you say it's used for
2025-04-11 17:57:18 <invoked> not sure what you mean by hundreds of partition guids
2025-04-11 17:57:25 <lassebq> Or you mean alpine and not the spec?
2025-04-11 17:57:42 <invoked> i'm talking about the uapi project in general.
2025-04-11 17:58:36 <lassebq> It lists UKI specification too.
2025-04-11 17:58:47 <lassebq> UKI is used in a lot of desktop distros
2025-04-11 17:59:01 <invoked> it can be.
2025-04-11 18:08:12 <WhyNotHugo> lassebq: the Discoverable Partitions Specification used to be part of the systemd documentation
2025-04-11 18:08:34 <WhyNotHugo> The UAPI website is basically a rebrand of a lot of systemd specification, but they're generally not standard outside of that ecosystem
2025-04-11 18:08:58 <WhyNotHugo> The DPS in particular is really good one tho, it would be nice to see support for it on alpine.
2025-04-11 18:25:23 <lassebq> WhyNotHugo: Yes. I wish it was supported. Specifically curious about root discovery since that isn't quite possible without initrd image. It would've to be done by some kind of EFI stub
2025-04-11 18:27:27 <WhyNotHugo> doing it in the initrd is the usual approach
2025-04-11 18:31:06 <bent_fingers> ocrmypdf fails to run, erring with https://bpa.st/QLGA
2025-04-11 18:31:29 <bent_fingers> something wrong with the installation?
2025-04-11 18:35:11 <ikke> bent_fingers: tesseract-ocr-data-eng contains that file, you need to manually install it
2025-04-11 18:37:54 <bent_fingers> yup, thanks
2025-04-11 18:39:42 <bent_fingers> there we go
2025-04-11 18:43:54 <f_> >nu_  probably the bridge misbehaving
2025-04-11 18:44:01 <f_> p_6f3Ik7Suw isn't even a matrix user
2025-04-11 18:45:27 <cosmic-desktop-woes> Hello. Has anyone managed to get the COSMIC desktop environment working?
2025-04-11 18:45:57 <cosmic-desktop-woes> I am currently experimenting with getting it working: no results so far.
2025-04-11 18:49:09 <bent_fingers> shouldn't you run popos with that?
2025-04-11 18:50:35 <cosmic-desktop-woes> I see your point: it is integrated with systemd. Noting that it is within the Alpine Linux repositories, under testing, I wondered if anyone has gotten it to work.
2025-04-11 18:51:17 <cosmic-desktop-woes> So far, I have attempted to disable the Rust compile time configurations for systemd, within the project. This involved patching the Cargo.toml.
2025-04-11 18:51:59 <bent_fingers> that's all over my head
2025-04-11 18:52:17 <cosmic-desktop-woes> I then built the project (within a container), using abuild. Installed the package on the host machine, but it still does not launch.
2025-04-11 18:52:30 <bent_fingers> i consider myself fulfilled of accomplishment whenever i manage to tie the laces of my shoes together
2025-04-11 18:52:49 <cosmic-desktop-woes> Rust is new to me, also. I was hoping to get in touch with Clayton Craft, a maintainer for postmarketOS.
2025-04-11 18:53:15 <bent_fingers> instead, you're left with me
2025-04-11 18:53:41 <cosmic-desktop-woes> Aren't we a merry bunch?
2025-04-11 18:54:06 <cosmic-desktop-woes> I hope your day is going well.
2025-04-11 18:54:13 <bent_fingers> not so crestfallen over here
2025-04-11 18:54:22 <bent_fingers> i got a job at Chipotle
2025-04-11 18:54:36 <bent_fingers> my first job in nearly four years
2025-04-11 18:55:07 <cosmic-desktop-woes> What was your most favourite part, on your first day?
2025-04-11 18:55:34 <user0> hello
2025-04-11 18:55:40 <bent_fingers> haven't started yet. my fav part will be when i get paid
2025-04-11 18:56:31 <cosmic-desktop-woes> Ola! And yes, I remember feeling like a king, buying pizza with my first paycheck.
2025-04-11 19:01:58 <voidastro[m]> conflicts on alpine edge: libprotobuf-24.4-r4
2025-04-11 19:02:49 <voidastro[m]> libprotobuf-25.6-r0  and libprotobuf-24.4-r4   :  libprotoc-25.6-r0  and android-tools-35.0.2-r7
2025-04-11 19:19:15 <simon_> hey 
2025-04-11 19:20:32 <simon_> one question, so my alpine runs but I need to activate cpu power to get great performance. is there other linux maybe better for gaming? or does it matter to play games with cpu power?
2025-04-11 19:21:12 <simon_> well I think you understand zero what I am gluglu about
2025-04-11 19:21:15 <simon_> :D
2025-04-11 19:22:14 <realroot[m]> do you have the microcode?
2025-04-11 19:22:45 <simon_> realroot[m]: which microcode?
2025-04-11 19:22:56 <simon_> but I think it does automatically install or not
2025-04-11 19:23:18 <realroot[m]> intel-ucode or amd-ucode
2025-04-11 19:23:30 <simon_> does these packages really exist?
2025-04-11 19:23:45 <realroot[m]> yes
2025-04-11 19:23:51 <simon_> I seee
2025-04-11 19:24:00 <mio> voidastro[m]: android-tools has already been rebuilt with newer protobuf, but may not be uploaded yet for x86 and riscv64, maybe check back later
2025-04-11 19:24:18 <simon_> but amd-ucode is rather for gpu or not?
2025-04-11 19:24:25 <simon_> sorry cpu
2025-04-11 19:25:24 <realroot[m]> yes it's for cpu
2025-04-11 19:26:02 <simon_> should I just update my stable versionn to bleeding edge or first try intel-ucode
2025-04-11 19:26:40 <realroot[m]> here it made it a lot better
2025-04-11 19:27:13 <mio> cosmic-desktop-woes: try opening an issue and cc the maintainer (@craftyguy) https://gitlab.alpinelinux.org/alpine/aports/-/issues
2025-04-11 19:28:21 <cosmic-desktop-woes> Thanks mio: I will explore this now.
2025-04-11 19:29:20 <simon_> I try
2025-04-11 19:29:54 <realroot[m]> not sure if it needs reboot
2025-04-11 19:30:44 <simon_> I reboot allways by hardware stuff is rebooting bad on linux? :d
2025-04-11 19:31:47 <simon_> someone wanna play dota 2 ?? :F
2025-04-11 19:31:48 <simon_> :D
2025-04-11 19:44:19 <simon_> is it safe to use alpine as a normal person as a desktop or would you say its hard to answer? :D the only thing I was able to do as security measure was setting up a firewall through your help I dont know whether everything what you can do else would just be paranoia
2025-04-11 19:45:56 <simon_> I mean I heard there is se linux or apparmor what ever
2025-04-11 19:46:23 <ikke> Those can be really painful though, not really suitable for desktops
2025-04-11 19:47:40 <simon_> you mean the apps I mentioned are not really good for desktops?
2025-04-11 19:48:02 <simon_> OHHH wow
2025-04-11 19:48:07 <ikke> selinux and apparmor are not apps, they are security frameworks
2025-04-11 19:48:13 <simon_> 200+ fps with intel-ucode
2025-04-11 19:48:36 <ikke> These frameworks require you (someone) to specify what applications are allowed to do, and anything else is blocked (in enforcing mode)
2025-04-11 19:48:52 <ikke> That means a lot of things can break if they are not described properly
2025-04-11 19:50:27 <Saijin_Naib[m]> simon_ I have used Alpine for a desktop OS for almost five years now. I can't be happier, really
2025-04-11 19:50:37 <Saijin_Naib[m]> Firewall, I use UFW and GUFW for ease of use currently
2025-04-11 19:50:49 <Saijin_Naib[m]> Very, very simple to add new rules, view them, edit them, toggle them, etc
2025-04-11 19:51:10 <Saijin_Naib[m]> I want to learn awall/acf, but that is to manage my family's computers remotely
2025-04-11 19:51:30 <Saijin_Naib[m]> You don
2025-04-11 19:52:00 <Saijin_Naib[m]> t need something like clamav or other malware scanner unless you expect to work in a mixed environment where you might pass along infected files to Windows PCs unwittingly
2025-04-11 19:52:59 <Saijin_Naib[m]> along ikke's point, some apps will make use of things like bubblejail or landlock to ensure a level of security for the app itself. Warpinator, for instance, does
2025-04-11 19:53:21 <Saijin_Naib[m]> Without requiring everything that apparmor/selinux do to work properly
2025-04-11 19:53:42 <Saijin_Naib[m]> Think of a mangaged device like a phone, and that is more the intended usage of those frameworks. Trusted device, trusted code, etc
2025-04-11 20:22:49 <simon_> can I set cpupower to autostart with sudo?
2025-04-11 20:36:43 <simon_> it works good I think
2025-04-11 20:36:59 <simon_> sometimes only 100 fps
2025-04-11 20:37:04 <simon_> but that is fine
2025-04-11 20:37:18 <simon_> dota 2 has become a huge game
2025-04-11 20:38:41 <simon_> did I already ask whether someone of you tried the packagemanager of netbsd pkgsrc on alpine? :D
2025-04-11 20:39:01 <simon_> oh thats maybe offtopic
2025-04-11 20:39:08 <simon_> well all I do is offtopic damn
2025-04-11 20:42:57 <Saijin_Naib[m]> cpupower comes with an openrc system service, so you just start that
2025-04-11 20:43:37 <Saijin_Naib[m]> like doas rc-update add cpupowerd
2025-04-11 20:43:52 <Saijin_Naib[m]> then doas rc-service cpupowerd start
2025-04-11 20:43:53 <Saijin_Naib[m]> or reboot
2025-04-11 20:44:22 <simon_> I had bad expericen with such stuff, on void linux I used their sv stuff and it spammed on my console every  second :D
2025-04-11 20:44:29 <simon_> the command of cpu power
2025-04-11 20:44:30 <simon_> xd
2025-04-11 20:44:40 <Saijin_Naib[m]> I have never seen that
2025-04-11 20:44:44 <Saijin_Naib[m]> YMMV
2025-04-11 20:45:08 <simon_> is rc-update like doing one time start on bootup or would it do start start start start
2025-04-11 20:46:36 <simon_> cpupowerd doesnt exist its cpupower I guess
2025-04-11 20:47:57 <Saijin_Naib[m]> one time start at whatever runlevel
2025-04-11 20:48:06 <Saijin_Naib[m]> default is what is not specified, worked fine there for me
2025-04-11 20:48:20 <Saijin_Naib[m]> ah, you can find it with rc-service -l | grep cpup
2025-04-11 20:48:29 <Saijin_Naib[m]> I might have mis-typed what the service name was, sorry
2025-04-11 20:48:52 <Saijin_Naib[m]> dont have it on this machine, doesn't do much for my N3450
2025-04-11 20:48:54 <simon__>  * Running cpupower -c all frequency-set --governor ondemand ...                                                                                                                                                                        [ !! ]
2025-04-11 20:48:54 <simon__>  * ERROR: cpupower failed to start
2025-04-11 20:49:16 <Saijin_Naib[m]> So, sounds like your CPU is using the intel governors
2025-04-11 20:49:29 <Saijin_Naib[m]> So, CPUPower isn't going to be terribly helpful, I don't think
2025-04-11 20:49:39 <Saijin_Naib[m]> You can force stuff using grub, I guess
2025-04-11 20:49:59 <simon__> I do all the time this doas cpupower frequency-set -g performance
2025-04-11 20:50:24 <simon__> because my cpu allways want to make my cpu slower for some reason :D
2025-04-11 20:51:30 <simon__> after exectuting ```cpupower frequency-set -g performance``` the cpu runs on higher mHz ingame
2025-04-11 20:51:49 <simon__> without throtteling
2025-04-11 20:52:51 <simon__> Saijin_Naib[m],  could I just edit the service or openrc? XD
2025-04-11 20:53:31 <Saijin_Naib[m]> You can edit I think a config file
2025-04-11 20:53:47 <Saijin_Naib[m]> I personally don't want a spaceheater, nor to pay for my CPU running full-power all day haha
2025-04-11 20:54:04 <Saijin_Naib[m]> Trying to recall where the config is....
2025-04-11 20:55:32 <simon__> Saijin_Naib[m],  but is there another thing I could do to not make the cpu being throtteled to 800 mHz?
2025-04-11 20:55:59 <simon__> like all the time I have such a weird intel i7 9700t prozessor :D
2025-04-11 20:56:43 <Saijin_Naib[m]> It should dynamically scale based upon demand and thermal overhead
2025-04-11 20:56:53 <Saijin_Naib[m]> If it cant scale up on schedutil, your thermals are messed up
2025-04-11 20:57:13 <Saijin_Naib[m]> my N3450 idles at just under 800Mhz, but will clock up to 2.4GHz when it is cold enough
2025-04-11 21:06:00 <simon__> do you think flatpak is a problem for gaming? XD
2025-04-11 21:06:14 <Saijin_Naib[m]> https://pkgs.alpinelinux.org/contents?name=cpupower&repo=community&branch=edge&arch=x86_64
2025-04-11 21:06:19 <simon__> the powersave mode is good
2025-04-11 21:06:31 <Saijin_Naib[m]> Check the top config, see if that lets you change what the policy is for the service
2025-04-11 21:06:47 <Saijin_Naib[m]> I don't know, most of my gaming is via flatpak and seems to work fine?
2025-04-11 21:06:56 <Saijin_Naib[m]> Lutris and Steam are all flatpak for me here
2025-04-11 21:07:07 <Saijin_Naib[m]> Same goes for Steam Link
2025-04-11 21:07:11 <Saijin_Naib[m]> :shrug:
2025-04-11 21:07:16 <Saijin_Naib[m]> What isn't right for you?
2025-04-11 21:07:42 <simon__> sometimes the fps drop when not cpupower activated from 120 to 50 :D
2025-04-11 21:08:11 <Saijin_Naib[m]> That sounds like your system chosing a lower power profile for some reason
2025-04-11 21:08:32 <Saijin_Naib[m]> If you aren't on AC, it might, or if there isn't a lot of thermal headroom, it might
2025-04-11 21:08:39 <Saijin_Naib[m]> Laptop gonna laptop, I guess
2025-04-11 21:09:07 <Saijin_Naib[m]> https://pkgs.alpinelinux.org/package/edge/testing/x86_64/gamemode
2025-04-11 21:09:15 <Saijin_Naib[m]> gamemode is in testing, but no idea how to make it work
2025-04-11 21:09:32 <Saijin_Naib[m]> That should, in theory, make a lot of this not your job to manage manually
2025-04-11 21:10:01 <simon__> gamemoderun %command% in your launch settings of your game
2025-04-11 21:11:06 <Saijin_Naib[m]> Ahhh, okay
2025-04-11 21:12:26 <simon__> I thought maybe this flatpak has more operations
2025-04-11 21:12:35 <simon__> which could stress the processor
2025-04-11 21:20:10 <Saijin_Naib[m]> My understanding is that any overhead is basically noise-floor/statistically insignificant at this point
2025-04-11 21:24:15 <simon__> how to get these testing repo beside my main repo?
2025-04-11 21:24:35 <simon__> well technically I think gamemode is not worth it 
2025-04-11 21:24:39 <simon__> cpupower
2025-04-11 21:24:40 <simon__> :D
2025-04-11 21:25:30 <simon__> what would be another idea to change the desktop environment from gnome to kde for example
2025-04-11 21:25:33 <simon__> and see if its the same
2025-04-11 21:26:45 <Saijin_Naib[m]> So, what I do is I copy /etc/apk/world to a backup, date and name of DE
2025-04-11 21:27:13 <Saijin_Naib[m]> then I move a copy of /etc/apk/world from setup-alpine with just minimal/console packages back
2025-04-11 21:28:06 <Saijin_Naib[m]> then apk fix, reboot, then setup-desktop, reboot
2025-04-11 21:29:32 <simon__> cp /etc/apk/world /etc/apk/world.bak like this
2025-04-11 21:33:16 <Saijin_Naib[m]> Yep!
2025-04-11 21:33:39 <Saijin_Naib[m]> But I have a world copy before adding any DE
2025-04-11 21:34:13 <Saijin_Naib[m]> I copy that back to world to run fix against to drop back to console-only
2025-04-11 21:34:30 <Saijin_Naib[m]> You might need to trim yours manually...
2025-04-11 21:35:51 <simon__> Saijin_Naib[m],  sound complicated
2025-04-11 21:35:52 <simon__> xd
2025-04-11 21:36:33 <simon__> couldnt I just delete first gnome and then install kde?
2025-04-11 21:36:50 <simon__> and also "wipe" userprofiel
2025-04-11 21:36:56 <Saijin_Naib[m]> Sure
2025-04-11 21:37:20 <Saijin_Naib[m]> But you may find that hard since installed programs may have depends on gnome
2025-04-11 21:38:05 <Saijin_Naib[m]> Any way you go, backup world 🤣
2025-04-11 21:40:36 <simon__> Saijin_Naib[m], world is backed up
2025-04-11 21:40:53 <simon__> now setup-alpine? :D
2025-04-11 21:41:01 <Saijin_Naib[m]> Aight, go nuts 🤣
2025-04-11 21:41:10 <Saijin_Naib[m]> Nah, setup-desktop
2025-04-11 21:41:52 <simon__> xfce bad or? :D
2025-04-11 21:42:15 <simon__> wait but will setup-desktop wipe gnome?
2025-04-11 21:42:16 <simon__> :D
2025-04-11 21:42:36 <Saijin_Naib[m]> I use XFCE for me, GNOME for family
2025-04-11 21:42:53 <simon__> dont you have a smashed screen on xfce
2025-04-11 21:42:58 <Saijin_Naib[m]> I do not believe setup-desktop will remove, just add
2025-04-11 21:43:17 <Saijin_Naib[m]> Eh? Not sure I understand
2025-04-11 21:43:32 <simon__> like a screen sliced into 2 pieces when you play video games
2025-04-11 21:43:37 <simon__> cuz its x11 :D
2025-04-11 21:43:58 <Saijin_Naib[m]> Oh! Tearing. Dont care about that, haha
2025-04-11 21:44:24 <Saijin_Naib[m]> Gave up on caring when I had to give up my 120Hz CRT
2025-04-11 21:44:55 <simon__> but dont you find the sliece in 2 pieces annoying
2025-04-11 21:44:55 <simon__> xd
2025-04-11 21:45:07 <simon__> is xfce4 now wayland`??
2025-04-11 21:45:13 <Saijin_Naib[m]> 🤷 been 20 years now
2025-04-11 21:45:26 <Saijin_Naib[m]> Xfce has limited/experimental wayland
2025-04-11 21:45:34 <simon__> did you tried it
2025-04-11 21:45:52 <Saijin_Naib[m]> No, I want my panels
2025-04-11 21:46:06 <Saijin_Naib[m]> And xfwm4 and its controls
2025-04-11 21:47:24 <Saijin_Naib[m]> My wife's laptop (same as mine) on GNOME is very smooth to use, so I guess their work on wayland and dynamic triple buffer et al is solid
2025-04-11 21:47:53 <simon__> does alpine stable have allready tripple buffering
2025-04-11 21:47:55 <simon__> xd
2025-04-11 21:48:31 <Saijin_Naib[m]> Mmm... not yet? I think that landed in GNOME 48
2025-04-11 21:48:53 <simon__> is gnome 48 in testing? 
2025-04-11 21:50:32 <simon__> ohh yes
2025-04-11 21:50:33 <simon__> I see
2025-04-11 21:50:34 <Saijin_Naib[m]> Well, community, but edge instead of latest-stable
2025-04-11 21:51:21 <simon__> shall I just upgrade the system to edge?
2025-04-11 21:52:10 <Saijin_Naib[m]> Uh, your call. I use it, but not for family
2025-04-11 21:52:55 <Saijin_Naib[m]> I find it to be quite stable generally, but every so often I need to do some extra work
2025-04-11 21:55:28 <simon__> extra work of what ?
2025-04-11 21:55:29 <simon__> :D
2025-04-11 21:56:19 <simon__> https://paste.debian.net/1368984/ this correct
2025-04-11 21:56:20 <simon__> ?
2025-04-11 21:58:16 <simon__> was right
2025-04-11 21:59:32 <simon__> https://paste.debian.net/1368985/ shit
2025-04-11 22:00:09 <simon__> gn8 
2025-04-11 22:00:18 <simon__> will try tomorrow B)
2025-04-11 22:05:20 <Simon_> Wish me luck that edge doesn't destroy my system :-)
2025-04-11 22:05:58 <Simon_> You said you used it for 10 years ?
2025-04-11 22:06:26 <Simon_> But how exactly do you know that there is a bug
2025-04-11 22:06:44 <Simon_> Or did never a bug happen
2025-04-11 22:08:51 <Simon_> Anyways maybe I will wake up tomorrow and everything will be fine 
2025-04-11 22:08:53 <Simon_> I hope
2025-04-11 22:08:57 <Simon_> :D
2025-04-11 22:09:01 <Simon_> Gn8
2025-04-11 22:12:17 <Saijin_Naib[m]> Five years, hahq
2025-04-11 22:14:47 <Saijin_Naib[m]> Issues I had were breakage from updating packages during a major upgrade on edge of stuff like icu, etc
2025-04-11 22:14:59 <Saijin_Naib[m]> Usually resolved by waiting
2025-04-11 22:34:24 <teapot9> hi, will apk overwrite changes I make to /etc/mkinitfs/features.d/* during upgrades?
2025-04-11 22:36:00 <rhizoome> teapot9: apk creates a file called .apk-new if it wants to save to a modified file. 
2025-04-11 22:36:27 <rhizoome> $filename.apk_new
2025-04-11 22:38:16 <rhizoome> only in /etc of course. Everywhere else it will replace.
2025-04-11 23:41:25 <cosmic-desktop-woes> To anyone investigating why the COSMIC desktop environment isn't working on Alpine Linux, I point towards this file I found in a Gentoo overlay. https://github.com/fsvm88/cosmic-overlay/blob/main/cosmic-de/cosmic-session/files/no_journald-systemctl.patch
2025-04-11 23:41:53 <cosmic-desktop-woes> I have yet to apply this patch. Will report back.
2025-04-12 00:15:36 <cosmic-desktop-woes> Hello. To anyone who has been able to get COSMIC working on Alpine Linux: I would like to know. I will read the logs tomorrow, so please send a message here if you have been able to get it working.
2025-04-12 00:30:36 <teapot9> rhizoome: great, thanks!
2025-04-12 08:40:30 <simon_> tripple buffering technology of gnome is great
2025-04-12 08:40:31 <simon_> :D
2025-04-12 08:41:22 <simon_> I dont even have to turn on performance mode
2025-04-12 09:12:31 <Infiniti> Hello everyone. I am seeking an avenue to use snapd on Alpine Linux and my search has led me here, I would greatly appreciate any advice, thanks in advance.
2025-04-12 09:15:31 <simon_> A friend of me has a question. I dont understaand the question well but is alpine linux having a musl Kernel and whether everything in Alpine is in userspace musl compiled?
2025-04-12 09:15:50 <simon_> Infiniti, you could try pkgsrc
2025-04-12 09:15:51 <simon_> :D
2025-04-12 09:21:48 <rhizoome> Infiniti: I don't think that is possible, but I haven't tried. I'd say, use flatpak and if something is missing make your own flatkap manifest. I build my own flatpaks, for stuff that need different runtimes. After all Alpine is DIY.
2025-04-12 09:28:02 <sigirneg> ikke: >The problem is the kernelthat needs to be written to the boot media
2025-04-12 09:28:12 <sigirneg> yes, kernel, initramfs and modloop
2025-04-12 09:28:40 <sigirneg> i can fetch the kernel from apk but the initramfs and modloop I have to build them myself, right?
2025-04-12 09:29:15 <rhizoome> Infiniti: I assume the fastest route to get snap working is: Install nix on alpine (there are packages) and then installing snapd on nix. But the multiple levels of indirection, might cause some headache.
2025-04-12 09:31:38 <sigirneg> also the package repo at <media>/apks. I think the apps there aren't installed from cache on boot. Maybe I can set that dir as apkcache? Or would that break things?
2025-04-12 09:51:05 <rhizoome> WhyNotHugo: I am back on elogind - sorry! :-) - It turns out, every single problem I had was because polkit would not read: /usr/share/polkit-1/actions/org.freedesktop.policykit.policy I don't know why it didn't, but now I was able to remove a ton of config and custom hacks.
2025-04-12 10:00:52 <rhizoome> simon_: If you want to play with performance stuff even more you can try to add preempt=full to GRUB_CMDLINE_LINUX_DEFAULT (if you use grub). It is full "realtime" (realtime is very vague term). It is only useful if the application you use actually tries to use "realtime". I use a DAW on linux and it uses it. Usually for linux-audio you are happy if a buffer size of 1024 bytes (20ms delay) works without glitches. With preempt full I can go down 
2025-04-12 10:00:52 <rhizoome> to 128 bytes (2.6ms delay). Not that I really want and need that, but it is nice how the kernel improved.
2025-04-12 10:01:44 <rhizoome> I think some games try to use realtime. You see it htop if the nice level is lower.
2025-04-12 10:04:52 <rhizoome> https://wiki.alpinelinux.org/wiki/PipeWire#Realtime_scheduling
2025-04-12 10:25:29 <ikke> sigirneg: there is the update-kernel script that can help
2025-04-12 10:58:10 <simon_> rhizoome, isnt rt a kernel?
2025-04-12 10:58:19 <simon_> rt-linux
2025-04-12 10:58:47 <simon_> how to add something to grub_cmdline_linux_default
2025-04-12 10:58:48 <rhizoome> simon_: not anymore. dynamic realtime is not in main kernel and can be enabled or disabled at boot.
2025-04-12 11:00:14 <rhizoome> simon_: in /etc/defaults/grub
2025-04-12 11:00:39 <rhizoome> simon_: and then call update-grub
2025-04-12 11:02:38 <rhizoome> simon_: https://tpaste.us/BMnq
2025-04-12 11:02:57 <rhizoome> s/not/now/
2025-04-12 11:05:16 <sigirneg> thank you, I'll look into it
2025-04-12 11:06:25 <rhizoome> simon_: https://tpaste.us/bQ5e
2025-04-12 11:11:11 <simon_> https://paste.debian.net/1369087/ rhizoome 
2025-04-12 11:11:24 <simon_> this is how my looks like
2025-04-12 11:11:39 <simon_> this folder /etc/default/grub
2025-04-12 11:12:33 <rhizoome> yes there: for example GRUB_CMDLINE_LINUX_DEFAULT="modules=sd-mod,usb-storage,ext4,nvme quiet rootfstype=ext4 preempt=full"
2025-04-12 11:14:10 <simon_> rhizoome, I assume there is a comma between ext4 and preemt=full?
2025-04-12 11:14:20 <simon_> GRUB_CMDLINE_LINUX_DEFAULT="modules=sd-mod,usb-storage,ext4,nvme quiet rootfstype=ext4,preempt=full"
2025-04-12 11:14:29 <rhizoome> simon_: no, spaces
2025-04-12 11:15:00 <rhizoome> simon_: otherwise you would break your rootfstype
2025-04-12 11:15:02 <simon_> thanks will try
2025-04-12 11:15:26 <rhizoome> simon_: its x=bla y=fasel
2025-04-12 11:52:58 <simon_> rhizoome, its not really better :D
2025-04-12 13:38:11 <|cos|> 
ACTION has been running alpine in a wifibox-bhive on an old macbook for about a year. after my upgrade to freebsd 14.2-p3 broke wifi, i've now reinstalled to run alpine directly on the metal.

2025-04-12 13:39:02 <|cos|> with that said. here i am with no wifi, and no deep understanding of how to fix that issue.
2025-04-12 13:41:08 <|cos|> i've tried b43 with fwcutter and built wl which required a possibly faulty one line path (adm/unaligned.h does not exist here)
2025-04-12 13:42:17 <|cos|> my device is a 14e4:43a0 and it obviously used to work with alpine. would anyone know how to point me in the direction of true wisdom?
2025-04-12 13:43:08 <|cos|> (apart from maybe not accepting to be gifted apple hardware in the future)
2025-04-12 13:43:45 <rhizoome> simon_: it depends on the applications. for it to correctly work the application has to move realtime tasks (audio and graphics) into a separete thread and set a higher priority on that thread. browsers and audio applications usually do that correctly. the internet said some games also do it, but games is really not my thing. you can also make things worse with realtime, if you for example set a thread that does big computations to realtime, 
2025-04-12 13:43:45 <rhizoome> everything else suffers.
2025-04-12 13:51:12 <|cos|> wifibox-alpine does document 'wl
2025-04-12 13:52:05 <|cos|> ' as the driver i am after. i guess that means i would wish to obtain an older kernel than 6.12.21-r1 to build it against.
2025-04-12 13:53:21 <|cos|> no older kernel sources are installable using `apk`? that's ow i understand the output of `apk search`.
2025-04-12 13:59:55 <simon_> rhizoome, not sure if I asked already but did someone tried pkgsrc ? :D
2025-04-12 14:02:04 <simon_> n
2025-04-12 14:02:08 <simon_> hexchat is for kde?
2025-04-12 14:02:18 <simon_> hexchat looks like crap on gnome
2025-04-12 14:02:18 <simon_> :D
2025-04-12 14:02:26 <simon_> crap ware
2025-04-12 14:04:01 <simon_> I cant eat honey milk flesh and eggs again wtf how can it be that I have problems with these all stuff to eat
2025-04-12 14:04:04 <simon_> :(
2025-04-12 14:04:10 <simon_> I am now vegan
2025-04-12 14:04:11 <simon_> lol
2025-04-12 14:04:54 <simon_> is there security tools which can increase my game performance? XD
2025-04-12 14:05:13 <simon_> like apparmor
2025-04-12 14:05:16 <simon_> or selinux
2025-04-12 14:11:40 <rhizoome> simon_: about pkgsrc, probably people aren't too interested because there is nixpkgs and it has alpine package: https://search.nixos.org/packages
2025-04-12 14:12:19 <rhizoome> https://pkgs.alpinelinux.org/packages?name=nix
2025-04-12 14:30:13 <simon_> rhizoome,  isnt nixpkg an installation that doesnt last long?
2025-04-12 14:30:24 <simon_> its nixshell only or not
2025-04-12 14:31:40 <rhizoome>  simon_: it can do that, it can basically do everything. It can be an OS, it can be package manager, it can be declarative shell enivronment and build system. 
2025-04-12 14:32:46 <simon_> how would you for example get with nix rstudio?
2025-04-12 14:34:59 <rhizoome> simon_: doas apk add nix; user logout / login; nix-env --install --attr nixpkgs.rstudio
2025-04-12 14:35:16 <simon_> brb
2025-04-12 14:36:57 <simon_> oh there is not rstudio in nix :D
2025-04-12 14:37:58 <simon_> nix-env --install --attr rstudio
2025-04-12 14:38:02 <simon_> this should be right
2025-04-12 14:38:21 <rhizoome> simon_: there is https://search.nixos.org/packages?channel=24.11&from=0&size=50&sort=relevance&type=packages&query=rstudio
2025-04-12 14:38:34 <rhizoome> simon_: probably not the rstudio you meant.
2025-04-12 14:39:02 <ava_mateo21[m]> Hi
2025-04-12 14:39:17 <simon_> btw has someone answered in the meantime my one question ? :D I could not roll up in chat"A friend of me has a question. I dont understaand the question well but is alpine linux having a musl Kernel and whether everything in Alpine is in userspace musl compiled?"
2025-04-12 14:39:44 <simon_> rhizoome, the problem was rather the nixpkgs.rstudio * ;)
2025-04-12 14:39:50 <simon_> it is just rstudio
2025-04-12 14:40:08 <simon_> nevermind also wrong
2025-04-12 14:40:19 <simon_> https://paste.debian.net/1369115/
2025-04-12 14:41:44 <rhizoome> simon_: I forgot you have to add channel.
2025-04-12 14:41:45 <rhizoome> nix-channel --add https://nixos.org/channels/nixpkgs-unstable nixpkgs
2025-04-12 14:42:25 <rhizoome> nix-channel --update
2025-04-12 14:42:43 <simon_> error: cannot connect to socket at '/nix/var/nix/daemon-socket/socket': No such file or directory
2025-04-12 14:43:00 <rhizoome> simon_: there is even more, yes
2025-04-12 14:43:20 <rhizoome> rc-update add nix-daemon
2025-04-12 14:43:40 <rhizoome> rc-service nix-daemon start
2025-04-12 14:44:00 <simon_> error: cannot open connection to remote store 'daemon': error: reading from file: Connection reset by peer
2025-04-12 14:44:01 <rhizoome> now add the channal again, then profit.
2025-04-12 14:44:36 <rhizoome> to add the daomon and start it you need to be root
2025-04-12 14:44:54 <rhizoome> doas rc-service nix-daemon start
2025-04-12 14:44:54 <rhizoome> doas rc-update add nix-daemon
2025-04-12 14:45:08 <simon_> yes did do that
2025-04-12 14:45:39 <simon_> maybe log out login now?
2025-04-12 14:45:46 <simon_> or reboot
2025-04-12 14:48:32 <rhizoome> simon_: yes, relogin 
2025-04-12 14:50:06 <simon_> did not work rhizoome 
2025-04-12 14:50:07 <simon_> :D
2025-04-12 14:50:41 <simon_> error: cannot open connection to remote store 'daemon': error: reading from file: Connection reset by peer
2025-04-12 14:50:44 <simon_> maybe reboot
2025-04-12 14:50:58 <rhizoome>  4303 root      0:00 /usr/sbin/nix-daemon
2025-04-12 14:50:58 <rhizoome> » ps ax | grep nix-daemon
2025-04-12 14:51:13 <simon_>  9933 root      0:00 /usr/sbin/nix-daemon
2025-04-12 14:51:13 <simon_> 12344 simon     0:00 grep nix-daemon
2025-04-12 14:52:36 <rhizoome> simon_: I can't rember that I did more than that, but the environment gets updated somehow. I might as well need reboot. 
2025-04-12 14:53:49 <rhizoome> simon_: yes /etc/profile.d/nix-daemon.sh updates your environment. maybe that only becomes active after a reboot.
2025-04-12 14:56:18 <simon_> maybe with doas?
2025-04-12 14:56:21 <simon_> sudoer?
2025-04-12 14:56:22 <simon_> :D
2025-04-12 14:56:51 <simon_> ~ $  nix-channel --list
2025-04-12 14:56:51 <simon_> nixpkgs https://nixos.org/channels/nixpkgs-unstable
2025-04-12 14:56:51 <simon_> unstable https://nixos.org/channels/nixpkgs-unstable
2025-04-12 14:57:40 <simon_> ~ $ nix-channel --update nixpkg
2025-04-12 14:57:40 <simon_> error: filesystem error: read_symlink: No such file or directory [/home/simon/.local/state/nix/profiles/channels/nixpkgs]
2025-04-12 14:59:19 <rhizoome> nix-channel --add https://nixos.org/channels/nixpkgs-unstable nixpkgs
2025-04-12 14:59:19 <rhizoome> nix-channel --update
2025-04-12 14:59:19 <rhizoome> these should be done as your user
2025-04-12 14:59:30 <simon_> error: cannot open connection to remote store 'daemon': error: reading from file: Connection reset by peer
2025-04-12 15:01:34 <rhizoome> simon_: I got the missing piece: "doas adduser simon nix"
2025-04-12 15:02:17 <rhizoome> then you need to logout once more, sorry. It was too long ago when I did that (last week :-))
2025-04-12 15:03:43 <simon_> ~ $ nix-channel --update
2025-04-12 15:03:45 <simon_> unpacking 2 channels...
2025-04-12 15:03:47 <simon_>  oups
2025-04-12 15:03:49 <simon_> my bad
2025-04-12 15:03:51 <simon_> xd
2025-04-12 15:04:16 <simon_> ~ $ nix-env -iA nixpkgs.rstudio
2025-04-12 15:04:16 <simon_> installing 'RStudio-2024.12.1+563'
2025-04-12 15:04:27 <rhizoome> nix-channel --remove unstable
2025-04-12 15:04:41 <simon_> it instally
2025-04-12 15:04:44 <simon_> installing
2025-04-12 15:04:44 <simon_> lol
2025-04-12 15:05:18 <simon_> are these things compiled against musl?
2025-04-12 15:05:25 <simon_> pkgs
2025-04-12 15:05:34 <rhizoome> simon_: no that is all glibc
2025-04-12 15:06:03 <simon_> ohhh thats bad isnt it?
2025-04-12 15:06:08 <rhizoome> simon_: it gets installed into /nix, so it does not clash
2025-04-12 15:06:42 <simon_> how to open rstudio
2025-04-12 15:06:59 <rhizoome> hash -r; rstudio
2025-04-12 15:07:30 <rhizoome> if rstudio has desktop file it should appear in desktop environment
2025-04-12 15:08:02 <rhizoome> the next time you can start it with just "rstudio" of course
2025-04-12 15:08:28 <rhizoome> hash -r just updates the binary file search list in your shell
2025-04-12 15:08:37 <simon_> https://paste.debian.net/1369118/
2025-04-12 15:08:40 <simon_> oh errors
2025-04-12 15:09:16 <simon_> there is no desktop file
2025-04-12 15:09:55 <rhizoome> simon_: no problem you have duplicate some dri drives and mesa. a moment I made note somewhere
2025-04-12 15:11:42 <rhizoome> nix-env --install --attr nixpkgs.mesa
2025-04-12 15:11:42 <rhizoome> nix-env --install --attr nixpkgs.adwaita-qt
2025-04-12 15:12:14 <rhizoome> nix-env --install --attr nixpkgs.gnome-themes-extra
2025-04-12 15:13:07 <rhizoome> mesa is for GPU acceleration and the other two so that the GUI does not look crappy (in most cases ie gnome and kde (qt))
2025-04-12 15:14:11 <simon_> still errors appear :D
2025-04-12 15:14:23 <simon_> https://paste.debian.net/1369120/
2025-04-12 15:14:53 <simon_> does it matter if errors appear but it works<?
2025-04-12 15:17:08 <simon_> but in the app lol > 1+1
2025-04-12 15:17:08 <simon_> Character set is not UTF-8; please change your locale 
2025-04-12 15:17:08 <simon_> Warning message:
2025-04-12 15:17:08 <simon_> [1] 2
2025-04-12 15:19:28 <rhizoome> It has desktop file. I just tested it https://gist.github.com/rhizoome/f7410d8b429d78a6386f809fcda1b4b7
2025-04-12 15:21:54 <xse> i had to cut power to an rpi 'cause ssh wasn't working: https://tmp.krkrkr.org/alpine_ssh.log
2025-04-12 15:22:44 <xse> the power cycle fixed it but if someone has any clue about what causes that i'm all ears
2025-04-12 15:24:00 <simon_> rhizoome, maybe because I did nix-env --iA rstudio instaead of nix-env --install --attr nixpkgs.rstudio?
2025-04-12 15:24:27 <simon_> how to remove rstudio?
2025-04-12 15:25:10 <rhizoome> nix-env --uninstall rstudio
2025-04-12 15:25:35 <rhizoome> nix-collect-garbage --delete-old
2025-04-12 15:25:42 <rhizoome> nix-store --gc
2025-04-12 15:25:54 <simon_> warning: selector 'rstudio' matched no installed derivations
2025-04-12 15:26:57 <rhizoome> nix-env --uninstall RStudio
2025-04-12 15:27:17 <rhizoome> nix-env --query --installed --out-path
2025-04-12 15:27:54 <rhizoome> simon_: I made a bunch of scripts that make things easier. I send them.
2025-04-12 15:30:57 <simon_> rhizoome, I cant find rstudio in gnome search xd
2025-04-12 15:31:29 <simon_> ima rebot
2025-04-12 15:31:32 <simon_> boot
2025-04-12 15:33:01 <simon_> cant find rstudio
2025-04-12 15:33:03 <simon_> :(
2025-04-12 15:33:21 <rhizoome> simon_: see direct messages
2025-04-12 15:36:32 <rhizoome> simon_: It is nix-env --uninstall RStudio (appearantly install is not case sensitive and uninstall is casesensitive)
2025-04-12 15:37:11 <rhizoome> simon_: did you see the direct messages?
2025-04-12 15:37:50 <simon_> ye
2025-04-12 15:39:10 <simon_> noo /bin/sh: cd: can't cd to /home/simon/bin: No such file or directory
2025-04-12 15:39:11 <simon_> :D
2025-04-12 15:40:05 <rhizoome> mkdir -p ~/bin
2025-04-12 15:58:13 <simon_> ~ $ nix-list
2025-04-12 15:58:15 <simon_> ./bin/sh: nix-list: Permission denied
2025-04-12 15:58:17 <simon_> works
2025-04-12 15:58:25 <simon_> but why it ask me for doas
2025-04-12 15:58:26 <simon_> xd
2025-04-12 16:59:17 <cosmic-desktop-woes> Hello. Does anyone know how to enable the framebuffer console within Alpine Linux?
2025-04-12 17:05:27 <p_6f3Ik7Suw> it's on by default? what do you mean exactly?
2025-04-12 17:06:10 <p_6f3Ik7Suw> i have a console when i boot, and i get a login prompt that is in the console when doing so. i did not do anything special to achieve this, it was the default.
2025-04-12 17:09:27 <bent_fingers> enable?
2025-04-12 17:09:45 <bent_fingers> the fb console is what you see when you press your power button
2025-04-12 17:10:14 <bent_fingers> p_6f3Ik7Suw: how do you manage to remember your name?
2025-04-12 17:11:05 <p_6f3Ik7Suw> i don't need to remember my name
2025-04-12 17:11:28 <p_6f3Ik7Suw> and other people can just hilite me with stf if they wanna ping me
2025-04-12 17:11:51 <bent_fingers> i use tab
2025-04-12 17:12:45 <cosmic-desktop-woes> I think there is a special framebuffer console, implemented within a kernel module. fbcon.
2025-04-12 17:12:45 <p_6f3Ik7Suw> well, you don't know me so i wouldn't expect you to address me directly
2025-04-12 17:19:45 <bent_fingers> cosmic-desktop-woes: it's the same one
2025-04-12 17:20:20 <bent_fingers> fbcon is what you see when you boot
2025-04-12 17:22:56 <cosmic-desktop-woes> Try executing "modprobe fbcon". I'm not certain here, but I believe there are two terminals: one is the Linux console, operating keyboard and display; one is the framebuffer terminal. Can anyone confirm?
2025-04-12 17:29:57 <bent_fingers> same thing man
2025-04-12 17:32:25 <simon_> there are 2 graphic drivers for amdgpus if I am not incorrect
2025-04-12 17:32:29 <simon_> I need both of them
2025-04-12 17:32:35 <simon_> but where is amdvlk driver?
2025-04-12 17:32:43 <simon_> and radeon vulkan ? :D
2025-04-12 17:38:31 <simon_> rhizoome, you active? :D
2025-04-12 18:38:53 <realroot[m]> rebuilding qt-creator package is taking eternity 😮‍
2025-04-12 18:46:20 <realroot[m]> can somebody fix the problem in alpine package maybe?
2025-04-12 18:46:31 <realroot[m]> ncopa
2025-04-12 18:46:43 <realroot[m]> should be the maintainer
2025-04-12 18:50:18 <bent_fingers> it's saturday
2025-04-12 18:51:39 <fossdd[m]> <realroot[m]> "can somebody fix the problem..." <- please open a issue
2025-04-12 18:58:15 <realroot[m]> <fossdd[m]> "please open a issue" <- https://gitlab.alpinelinux.org/alpine/aports/-/issues/17075
2025-04-12 18:58:57 <realroot[m]> @fossdd:matrix.org I was told in devel that I just to rebuild it so maybe someone can start the rebuilding
2025-04-12 19:44:50 <|cos|> i summarized my understanding on https://www.netizen.se/docs/alpine-a1502emc2678
2025-04-12 19:45:52 <|cos|> would absolutely love if someone could give me any pointers on how to get broadcom wifi working (again)
2025-04-12 19:48:32 <Saijin_Naib[m]> If anyone else is looking, I'm trying to update our pinta package to v3.0, but it fails to build because it can't find the csproj for the add-ons/plug-ins functionality during build after configure
2025-04-12 19:48:37 <Saijin_Naib[m]> Not sure what leads to that
2025-04-12 19:50:02 <Saijin_Naib[m]> /usr/lib/dotnet/sdk/9.0.104/NuGet.targets(461,5): error MSB3202: The project file "/home/USERNAME/Sources/aports/community/pinta/src/pinta-3.0/installer/addins/Pinta.AddinUtils.csproj" was not found.
2025-04-12 19:50:16 <Saijin_Naib[m]> And indeed, that file does not exist
2025-04-12 20:08:46 <Saijin_Naib[m]> whoop, apparently them stating dotnet8 as minimum does not imply dotnet9 is supported, haha
2025-04-12 20:10:37 <Saijin_Naib[m]> bah, still failson the same file, just takes longer
2025-04-12 20:23:58 <fossdd[m]> i would start with looking where this file is supposed to be coming from
2025-04-12 20:26:10 <Saijin_Naib[m]> No idea. It seems to be part of the make directive somehow
2025-04-12 20:27:09 <Saijin_Naib[m]> https://github.com/search?q=repo%3APintaProject%2FPinta%20Pinta.AddinUtils.csproj&type=code
2025-04-12 20:27:54 <WhyNotHugo> rhizoome: i appreciate ou sharing your notes. it's useful research to know what works and what doesn't
2025-04-12 20:33:45 <Saijin_Naib[m]> Ah, using the wrong source archive
2025-04-12 20:33:55 <Saijin_Naib[m]> this should do it! exciting! Can not wait to try the new release
2025-04-12 20:53:55 <Saijin_Naib[m]> Bah, still fails
2025-04-12 20:54:11 <Saijin_Naib[m]> There must be something not happening during ./configure for it to keep dropping the AddIns
2025-04-12 22:00:54 <haoyi> \quit
2025-04-12 22:56:44 <haoyi> Hello I was wondering if anyone has a patch for java-service-wrapper for musl
2025-04-12 23:00:12 <ptrc> what the hell even is java-service-wrapper
2025-04-12 23:00:21 <ptrc> the only thing that i can find is some corporate bullshit with a free trial
2025-04-12 23:00:28 <haoyi> https://sourceforge.net/projects/wrapper/
2025-04-12 23:00:47 <haoyi> when I try compiling it it says something like gnu/libc-version.h not found
2025-04-12 23:02:51 <ptrc> you can just remove the include + replace the call to `gnu_get_libc_version()` with something hardcoded
2025-04-12 23:04:27 <ptrc> or well
2025-04-12 23:04:41 <ptrc> you can probably get rid of the entire "does ftell cause memory leak" function
2025-04-12 23:05:03 <ptrc> remove the entire `ifdef LINUX` block
2025-04-12 23:05:04 <ptrc> should work
2025-04-12 23:06:29 <haoyi> what would I do in this section lol
2025-04-12 23:06:30 <haoyi>     mbstowcs(versionW, gnu_get_libc_version(), 10);
2025-04-12 23:06:30 <haoyi>     _stscanf(versionW, TEXT("%d.%d.%d"), &vmaj, &vmin, &vrev);
2025-04-12 23:06:30 <haoyi>     return ((vmaj == maj && vmin == min &&  vrev >= rev) || (vmaj == maj && vmin > min) || vmaj > maj);
2025-04-12 23:08:25 <haoyi> you are right I could probably take away the entire block
2025-04-12 23:23:10 <haoyi> looks like I got it to compile
2025-04-12 23:23:13 <haoyi> thanks ptrc 
2025-04-12 23:24:19 <ptrc> you're welcome!
2025-04-12 23:25:38 <haoyi> for some reason the patches I made by just removing the blocks didn't apply idk if its because its svn but if i renamed ifdef -> ifndef it did the trick
2025-04-13 02:36:28 <ava_mateo21[m]> This company is worth more than an investment platform (Https://gaubertpower.ltd).... (full message at <https://matrix.org/oftc/media/v1/media/download/AW2jPn762TjvC8MpemlZca8ZkbhByBhqbkEULPhTEpprLiVoPNLP4gRXI0ZfALeLMtrTzAeAdbcQcxGj447qlmpCeWdRpsNwAG1hdHJpeC5vcmcvSVlKRUlIcFhvR092cUJsd1BCcGpQQ1BK>)
2025-04-13 04:17:56 <shadow255> Hi folks. I'm having an issue trying to use networkmanager-l2tp wherein I can set up a new connection successfully, but it's not possible to activate it. In network-manager-applet the network is greyed out and unclickable. Any pointers to fixing this would be appreciated.
2025-04-13 08:14:56 <simon_> rhizoome, 
2025-04-13 08:24:48 <simon_> I am back
2025-04-13 08:24:50 <simon_> :D
2025-04-13 08:43:50 <simon_> https://wiki.alpinelinux.org/wiki/Radeon_Video does anyone understand what is written here?
2025-04-13 09:03:44 <rcf> simon_: what is unclear? 
2025-04-13 09:04:00 <simon_> idk how to make radeon vulkan work :D
2025-04-13 09:05:38 <rhizoome> simon_: it only one driver
2025-04-13 09:05:48 <rhizoome> simon_: but you need to install some firmware
2025-04-13 09:06:03 <rhizoome> simon_: the easiest, you install all
2025-04-13 09:06:16 <simon_> which one is all ? :D
2025-04-13 09:06:45 <simon_> 249 KiB
2025-04-13 09:06:45 <simon_> seems to be installed linux-firmware-amd-20250311-r0 installed size:
2025-04-13 09:07:38 <simon_> I have rx 7600 amd gpu
2025-04-13 09:07:39 <rhizoome> simon_: just install all packages on that wiki page and reboot.
2025-04-13 09:08:50 <simon_> ohhh I see 
2025-04-13 09:08:50 <rhizoome> simon_: I does not matter what you have. If you install  linux-firmware-amdgpu,  linux-firmware-radeon, mesa-dri-gallium, mesa-va-gallium: 
2025-04-13 09:09:04 <rhizoome> you have covered everything.
2025-04-13 09:09:15 <simon_>  xf86-video-ati this one should be radeon vulkan I guess
2025-04-13 09:09:16 <simon_> xd
2025-04-13 09:09:24 <simon_> ok I reboot
2025-04-13 09:09:31 <rhizoome> simon_: there is nothing extra for vulkan
2025-04-13 09:10:58 <simon_> but how can proprietary and opensource driver work at the same time
2025-04-13 09:10:58 <simon_> xd
2025-04-13 09:11:17 <rhizoome> simon_: there are no proprietary drivers, this is alpine linux.
2025-04-13 09:12:02 <simon_> do I need kms to be radeon or amdgpu?
2025-04-13 09:12:08 <simon_> Install either the Free Software driver xf86-video-ati or the proprietary amdgpu firmware linux-firmware-amdgpu.
2025-04-13 09:12:09 <simon_> modprobe fbcon might be needed to avoid black screen when leaving Xorg. 
2025-04-13 09:12:26 <simon_>  I installed both
2025-04-13 09:12:27 <rhizoome> simon_: you can ignore KMS
2025-04-13 09:12:27 <simon_> lol
2025-04-13 09:12:40 <simon_> well I did kernel mode setting to amdgpu
2025-04-13 09:12:43 <simon_> is that false xd
2025-04-13 09:13:15 <rhizoome> simon_: the only question is are you running X or Wayland?
2025-04-13 09:13:20 <simon_> wayland
2025-04-13 09:13:53 <rhizoome> simon_: then you can delete xf86-video-ati again
2025-04-13 09:15:49 <simon_> hmm it is still not loading vulkan
2025-04-13 09:15:51 <simon_> ingame
2025-04-13 09:16:00 <simon_> or it did already compile vulkan
2025-04-13 09:16:01 <simon_> dunno
2025-04-13 09:16:20 <rhizoome> simon_: there is nothing special for vulkan
2025-04-13 09:16:42 <rhizoome> simon_: why do you think it is not working?
2025-04-13 09:17:37 <rhizoome> simon_: your game does not select vulkan? Installed via flatpak or steam, right?
2025-04-13 09:17:55 <simon_> rhizoome, is there another way to install it ? XD
2025-04-13 09:19:51 <rhizoome> simon_: You have installed vulkan-loader, I am not sure if that is needed, but that is the only thing, but it is not amd specific.
2025-04-13 09:20:45 <rhizoome> simon_: you don't install in another way, you debug the problem. Lets see if we can find a test-program. I bet the problem is your game. My DAW used vulkan right away.
2025-04-13 09:21:21 <simon_> nahh not the game :D
2025-04-13 09:22:34 <simon_> dota 2 did allways use vulkan compiling on the nixos for example
2025-04-13 09:22:40 <rhizoome> simon_: no I mean you don't install drivers in another way. We find out what is missing and fix it.
2025-04-13 09:23:38 <simon_> maybe kms is the problem for worse performance
2025-04-13 09:23:39 <rhizoome> simon_: lets start by "apk add vulkan-tools" and then start vulkaninfo
2025-04-13 09:24:34 <simon_> https://paste.debian.net/1369240/
2025-04-13 09:24:55 <simon_> lol
2025-04-13 09:25:04 <simon_> it reffers to nix
2025-04-13 09:25:31 <simon_> maybe cuz we installed mesa in nix?
2025-04-13 09:25:31 <simon_> xd
2025-04-13 09:25:38 <rhizoome> simon_: see this shows you that you are trying to run vulkaninfo of nix. In that case you can install whatevery you want in alpine and it is not going to work
2025-04-13 09:26:36 <rhizoome> simon_: I currently have the same problem. So if you give me a minute, I will sort it out on my machine and tell you what is needed to fix that.
2025-04-13 09:27:28 <simon_> I guess uninstalling mesa from nix would work
2025-04-13 09:27:28 <simon_> xd
2025-04-13 09:29:10 <simon_> rhizoome, I wonder why wouldnt we just straight install nix without systemd?
2025-04-13 09:29:11 <simon_> :D
2025-04-13 09:29:11 <rhizoome> simon_: yes nix-del mesa might be a shortterm workaround, but if you want run GPU stuff in nix, you are going to need it
2025-04-13 09:29:39 <simon_> https://paste.debian.net/1369241/
2025-04-13 09:30:06 <rhizoome> simon_: you mean the nix-daemon? with daemon it work better.
2025-04-13 09:30:58 <simon_> you see no loader found :D
2025-04-13 09:31:06 <rhizoome> simon_: I saw your paste. Currently doing much the same on my machine. give me a minute.
2025-04-13 09:32:47 <simon_> vulkan-utility-libraries-1.4.304.1-r0
2025-04-13 09:32:47 <simon_> vulkan-validation-layers-1.4.304.1-r0
2025-04-13 09:32:51 <simon_> maybe these two
2025-04-13 09:32:59 <simon_> or
2025-04-13 09:33:00 <rhizoome> simon_: yes try them.
2025-04-13 09:33:03 <simon_> mesa-vulkan-layers-24.3.4-r0
2025-04-13 09:33:04 <simon_> this
2025-04-13 09:34:09 <simon_> will reboot
2025-04-13 09:35:36 <simon_>  345 doas apk add mesa-vulkan-layers-24.3.4-r0 vulkan-utility-libraries-1.4.304.1-r0 vulkan-validation-layers-1.4.304.1-r0
2025-04-13 09:35:39 <simon_>  346 doas apk add mesa-vulkan-layers vulkan-utility-libraries vulkan-validation-layers
2025-04-13 09:35:41 <simon_> one of them fixed it
2025-04-13 09:35:43 <simon_> xd
2025-04-13 09:35:53 <rhizoome> simon_: why not waiting. I was just going tell you "apk add mesa-vulkan-ati"
2025-04-13 09:36:15 <rhizoome> simon_: no reboot need just install mesa-vulkan-ati
2025-04-13 09:36:25 <simon_> the other ones I  do not need?
2025-04-13 09:36:31 <simon_> or great to have anyways
2025-04-13 09:37:16 <rhizoome> simon_: yes only mesa-vulkan-ati 
2025-04-13 09:37:39 <rhizoome> simon_: but you can keep the vulkan utilities for testing
2025-04-13 09:37:59 <simon_> I think the xf86-video-ati is still good to have for apps like steam
2025-04-13 09:38:07 <simon_> because steam reacts worse without it
2025-04-13 09:39:22 <rhizoome> simon_: really? then leave it in. maybe XWayland needs it. I don't undertand XWayland well.
2025-04-13 09:40:27 <rhizoome> simon_: I am going to fix the nix / alpine clash when mesa is installed in nixpkgs, but I need to get lunch first. I'll keep you posted.
2025-04-13 09:43:43 <simon_> rhizoome, well no not really
2025-04-13 09:43:44 <simon_> xd
2025-04-13 09:45:44 <simon_> https://imgur.com/a/xjSlvcc rhizoome look I click but it doesnt respond
2025-04-13 09:45:44 <simon_> xd
2025-04-13 09:46:57 <rhizoome> simon_: I read about it. XWayland only needs drm/mesa, if you really need xf86-video-ati something is strage or text is wrong.
2025-04-13 09:47:35 <simon_> did you see the video? rhizoome 
2025-04-13 09:47:36 <simon_> :D
2025-04-13 09:50:04 <rhizoome> simon_: yes and I don't understand what exactly is happening. I need to go for lunch. If this is a flatpak, all the drivers have do be in the flatpak anyways, since musl and glibc won't work togehter. So need to go to lunch.
2025-04-13 09:51:23 <simon_> rhizoome, dont you use steam?
2025-04-13 09:51:34 <rhizoome> simon_: that is why I told you you don't need anything else for a flatpak. If there kernel drivers are there, you can't do more. 
2025-04-13 09:51:59 <simon_> uff so flatpak steam is borked
2025-04-13 09:52:51 <rhizoome> user space drivers aren't going to change anything. I use a flatpak for my DAW and vulken worked directly, without me installing mesa-vulkan-ati, which I need for vulkan on alpine.
2025-04-13 09:56:35 <simon_> do you think kms might be a problem?
2025-04-13 10:41:26 <simon_> rhizoome, do you think nixos has more performance?
2025-04-13 10:41:26 <simon_> :d
2025-04-13 10:41:28 <simon_> :D
2025-04-13 10:44:11 <simon_> will try without ingame overlay maybe  that helps
2025-04-13 11:12:35 <lopid> it appears that debug logging has been left enabled in the most recent /etc/nextcloud/config.php
2025-04-13 11:23:19 <lopid> though i don't see it in the repo source, so maybe it was just my update that added it 
2025-04-13 12:08:53 <simon_> can I install kde5 on alpine
2025-04-13 12:08:57 <simon_> or rather not
2025-04-13 12:11:56 <rhizoome> simon_: https://wiki.alpinelinux.org/wiki/KDE
2025-04-13 12:12:36 <simon_> rhizoome, but thats kde 6
2025-04-13 12:12:38 <simon_> ew
2025-04-13 12:12:39 <simon_> :D
2025-04-13 12:13:27 <simon_> or has alpine stable still kde5>?
2025-04-13 12:14:06 <rhizoome> simon_: there is no kde5
2025-04-13 12:14:14 <simon_> rhizoome, rip
2025-04-13 12:14:17 <simon_> :(
2025-04-13 12:15:11 <simon_> I have a major problem with these bleeding edge ware
2025-04-13 12:15:56 <simon_> everytime I upgrade I feel like fps is getting worse
2025-04-13 12:16:22 <elagost> simon_: kde6 is >1year old, if you want kde5 you want alpine 3.19, which is unsupported.
2025-04-13 12:17:01 <simon_> can you port kde 5 to edge?
2025-04-13 12:17:55 <simon_> or am I better of on another distro because you wont put kde5 to edge ? :DD
2025-04-13 12:18:13 <elagost> not likely since it would involve rolling back all the libraries, all the applications, etc. that have moved on; you're welcome to try and host your own personal repo (I host my own, it's pretty easy)
2025-04-13 12:21:16 <elagost> I think it would be as simple as checking out the pre-kde6 commit of aports, rename all the packages and dependencies to avoid conflicts (add a -kde5 suffix), then abuild them all in the right order. Simple enough but would be time consuming.
2025-04-13 12:22:17 <simon_> elagost, that is actually a bit too hard for me heh :D
2025-04-13 12:22:27 <simon_> I think I will just go back to slackware
2025-04-13 12:22:50 <simon_> with the knowledge that rhizoome gave me that will be really cool on slack
2025-04-13 12:22:55 <simon_> (for me at least)
2025-04-13 12:23:06 <simon_> not for people with new hardware
2025-04-13 12:24:01 <simon_> or I will try nixos they actually use  too kde5
2025-04-13 12:24:04 <simon_> but nixos is strange
2025-04-13 12:25:12 <simon_> see ya on the other end guys thanks for the try
2025-04-13 12:25:29 <rhizoome> simon_: a) you don't need to use nixpkgs b) is fine, just debug the steam flatpak
2025-04-13 12:25:46 <rhizoome> s/is fine/everything is fine/
2025-04-13 12:29:47 <rhizoome> sigh
2025-04-13 12:38:47 <rhizoome> that was some hours of my life. I hope he at least learned something.
2025-04-13 13:47:20 <simon_> rhizoome: is nixos an american spy distribution? :D
2025-04-13 13:47:48 <simon_> anyways nixos has kde
2025-04-13 13:47:49 <simon_> 5
2025-04-13 16:01:58 <realroot[m]> waht builder does testing? edge?
2025-04-13 16:02:07 <realroot[m]> https://build.alpinelinux.org/
2025-04-13 16:08:54 <ikke> realroot[m]: yes, testing is only built for edge
2025-04-13 17:38:52 <dngray> ptrc ugh this is stumping me https://gitlab.alpinelinux.org/dngray/aports/-/jobs/1808549#L126
2025-04-13 17:39:17 <dngray> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112/diffs?commit_id=24d43cdfc0e96a7d65abedad95dc15c5c28bc0cf i tried this
2025-04-13 17:50:21 <dngray> oh yeah i tried that and it didn't seem to work
2025-04-13 17:50:54 <dngray> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112/diffs?commit_id=a5b15c27f1d783490b9245d23bd16a90b53fb889
2025-04-13 17:56:26 <dngray> https://gitlab.alpinelinux.org/dngray/aports/-/jobs/1808574#L131 then tells me no make rule target
2025-04-13 17:56:34 <dngray> so im kinda confused
2025-04-13 17:57:55 <ikke> dngray: algitbot is a bot, it just told you the commit message :P
2025-04-13 17:58:05 <dngray> yeah i know that
2025-04-13 17:58:12 <dngray> but i did try various things
2025-04-13 17:58:29 <ikke> It sounded like you responded to algitbot suggesting you something :P
2025-04-13 17:59:29 <dngray> oh nah i was messing around with those lines saying what i tried
2025-04-13 17:59:52 <dngray> the error apkbuild is giving is really unhelpful
2025-04-13 18:00:14 <dngray> https://github.com/leetronics/infnoise/blob/master/software/Makefile.linux#L50-L56
2025-04-13 18:00:36 <ikke> dngray: make sure you're in the correct dir
2025-04-13 18:00:54 <ikke> only the default functions (prepare, build, check, package) cd to $builddir
2025-04-13 18:02:00 <dngray> oh i wonder if that is the reason
2025-04-13 18:02:14 <ikke> That would explain the error from make
2025-04-13 18:03:10 <dngray> cd "$builddir"/software
2025-04-13 18:03:15 <dngray> should be right it works for the above step
2025-04-13 18:03:33 <dngray> in build
2025-04-13 18:22:36 <ikke> dngray: in build, you do -f Makefile.linux
2025-04-13 18:23:03 <ikke> Perhaps do a cp software/Makefile.linux software/Makefile in prepare
2025-04-13 18:38:30 <realroot[m]> is there a way to see the list of incoming packages in https://build.alpinelinux.org/?
2025-04-13 18:38:47 <ikke> realroot[m]: no
2025-04-13 19:01:55 <dngray> oh yeah that might be a better idea
2025-04-13 19:27:40 <ikke> dngray: it seems like PREFIX should include DESTDIR
2025-04-13 19:28:00 <dngray> ohhh yeah that makes sense
2025-04-13 19:28:35 <dngray> so that should be like
2025-04-13 19:28:40 <dngray> make PREFIX=/usr install infnoise
2025-04-13 19:28:47 <dngray> if im not being dumb
2025-04-13 19:28:58 <ikke> make PREFIX="$pkgdir"/usr
2025-04-13 19:29:24 <ikke> https://tpaste.us/gbjr works except that there is no init.d file
2025-04-13 19:29:50 <dngray> oh well i can do the init.d file manually
2025-04-13 19:30:07 <dngray> or patch the makefile
2025-04-13 19:30:10 <dngray> maybe that would be better
2025-04-13 19:30:18 <dngray> seems it installs the systemd one
2025-04-13 19:30:21 <dngray> so we dont want that
2025-04-13 19:32:13 <ikke> the -doc package is also empty
2025-04-13 19:33:21 <dngray> oh yeah i think i got rid of that mmm
2025-04-13 19:33:53 <dngray> i should patch that makefile to have a doc function
2025-04-13 19:33:59 <dngray> that installs the man file
2025-04-14 03:25:52 <swee> Hi, I'd like to report that the kmod-bash-completion package in the edge branch is a little broken
2025-04-14 03:26:12 <swee> https://lounge.swee.codes/uploads/cfe14fc31f59f60b/image.png 
2025-04-14 04:14:43 <ikke> A file conflict with bash completion
2025-04-14 08:12:01 <|cos|> I've reworked https://www.netizen.se/docs/alpine-a1502emc2678 since posting it here during the weekend.
2025-04-14 08:13:14 <|cos|> Any advise on how to get wifi running under alpine would be more than welcome!
2025-04-14 08:49:34 <|cos|> It seems I found a work-around. Am still curious of what would be the "correct" solution to run alpine on a macbook.
2025-04-14 11:21:30 <Kladky> I noticed that after upgrading my server to openssh 10, i can no longer connect to it from my laptop, which uses openssh 9. Is this intended behaviour from openssh?
2025-04-14 11:21:45 <Kladky> The error is "kex_exchange_identification: read: Connection reset by peer"
2025-04-14 11:22:58 <aron> it happens from time to time, they removed the KEX that were used by your laptop
2025-04-14 11:23:10 <aron> this is unrelated to alpine, tho
2025-04-14 11:23:45 <aron> upgrade you laptop as well, and generate new keys
2025-04-14 11:24:32 <aron> s/you/your/
2025-04-14 11:25:38 <Kladky> might be a while, since my laptop is on stable, not edge
2025-04-14 11:28:17 <aron> your server is on edge?
2025-04-14 11:28:54 <Kladky> yep
2025-04-14 11:29:18 <aron> you know the saying: if you are on edge, you may cut yourself!
2025-04-14 11:30:37 <aron> you can try to downgrade your server's openssh from latest stable, but I wouldn't recommend it
2025-04-14 11:30:55 <aron> and there is nothing else I can think of
2025-04-14 11:31:37 <Kladky> i haven't had any issues with edge so far on my server
2025-04-14 11:31:55 <Kladky> been using it for 1.5 years now
2025-04-14 11:32:19 <Kladky> my laptop however...
2025-04-14 11:33:28 <Kladky> from looking at the changelog, it seems that the kex methods have just changed
2025-04-14 11:33:51 <Kladky> the default ones at least
2025-04-14 11:34:08 <Kladky> so should i just be able to change the default methods on the client to match?
2025-04-14 11:35:27 <aron> best case scenario: your client's kex method is not enabled on the server
2025-04-14 11:36:13 <aron> ofc you can try to adjust your client as well
2025-04-14 11:37:06 <aron> just make sure they have a common method
2025-04-14 11:42:21 <ikke> Kladky: I think it should be sufficient to restart sshd
2025-04-14 11:59:17 <Kladky> Seems you were right ikke, thanks!
2025-04-14 11:59:29 <Kladky> I can't believe I didn't try that lol
2025-04-14 11:59:47 <aron> oh, I thought you did :P
2025-04-14 12:00:06 <aron> glad it's fixed
2025-04-14 13:25:07 <dngray> hmm ikke i'm still getting an error
2025-04-14 13:25:27 <dngray> install -m 0644 init_scripts/infnoise.gentoo.openrc /etc/init.d/infnoise
2025-04-14 13:25:36 <dngray> install: can't create '/etc/init.d/infnoise': File exists
2025-04-14 13:25:40 <dngray> note i patched the makefile
2025-04-14 13:26:20 <dngray> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112/diffs?commit_id=136f35be7908478396781ac72c0ba0bef0d406ce because we obviously don't need the systemd file
2025-04-14 13:26:20 <ikke> How did you patch it?
2025-04-14 13:26:29 <dngray> but im unsure about the prefix it doesn't make any sense becuase that is /usr
2025-04-14 13:27:17 <dngray> oh i forgot to get rid of that systemd unit hehe
2025-04-14 13:27:57 <dngray> i think the install command is being installed not to the right place though
2025-04-14 13:29:59 <ikke> dngray: when you set PREFIX, DESTDIR is ignored 
2025-04-14 13:30:55 <dngray> oh so i don't need the dest dir?
2025-04-14 13:31:07 <dngray> how does it know where to install the openrc script tho
2025-04-14 13:33:44 <dngray> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112/diffs?commit_id=5a9518b09e1964be4e7294288a26f98d718e130e#098fe97bf1a9e9ad50786e5c2dbf7cf3621e30d3_0_20 that's the bit im not sure about
2025-04-14 13:34:38 <dngray> $(PREFIX)/etc/init.d/infnoise doesn't make sense
2025-04-14 13:42:43 <dngray> maybe i can't do it becuase like can't use $pkgdir" in the bash script
2025-04-14 13:42:51 <dngray> should that install command be in the make file or be in the APKBUILD
2025-04-14 13:43:26 <dngray> like for example
2025-04-14 13:43:28 <dngray> install -Dm755 software/init_scripts/infnoise.gentoo.openrc "$pkgdir"/etc/init.d/infnoise
2025-04-14 13:44:18 <dngray> oh actually it should be
2025-04-14 13:44:29 <dngray> because otherwise the openrc subpkg won't have anything in it
2025-04-14 13:46:54 <ikke> dngray: you install it in the package() function, then the default openrc split function will do the rest
2025-04-14 13:47:16 <dngray> yup which is why i should not try to install the openrc script in the makefile because yeah
2025-04-14 14:06:21 <dngray> right so anyway seem to be having some trouble with my lib subpkg now
2025-04-14 14:06:29 <dngray> /usr/bin/abuild: cd: line 2680: can't cd to /builds/dngray/aports/testing/infnoise/pkg/libinfnoise: No such file or directory
2025-04-14 14:06:53 <dngray> which is weird it worked for the package() function
2025-04-14 14:07:04 <dngray> i think its copying the name from libinfnoise() {
2025-04-14 14:11:10 <ikke> You _can_ do it from the makefile, you just need to make sure it's in the right directory
2025-04-14 14:20:31 <realroot[m]> is it possible that qt-creator will be built for stable?
2025-04-14 14:22:02 <rifux[m]> How to setup pppoe on Alpine Linux? Reproducing what Alpine Linux Wiki says doesn't work
2025-04-14 14:23:55 <dngray> are you having some error?
2025-04-14 14:24:08 <dngray> i have used pppoe on alpinelinux before
2025-04-14 14:47:04 <rifux[m]> 
ACTION uploaded an image: (12KiB) < https://matrix.org/oftc/media/v1/media/download/AYnBZzqUi0n3OtvfqsXWxFh7eZwrDDF1NJrlWOOk4vrQScfOhKcAQAtyI2jtdMO9u8NUNeDAZ_zadUx2IfrisGlCeWfN2qNwAHRjaG5jcy5kZS8xOWJmOGIxMjAyMDc3NzQxMjY3ODYzM2NiYjI3YzY1ZTM5MmQzMTkwMTkxMTc5MzEzODYyNDYyNjY4OA >

2025-04-14 14:47:04 <rifux[m]> <dngray> "are you having some error?" <- rifu@nihon ~> cat /etc/ppp/peers/provider... (full message at <https://matrix.org/oftc/media/v1/media/download/AQD1aYu8j5FF_05MLcDI1kGlB2fCtPR3g5ypgaT5ZquqPSdZ0Enm3-lePOhDOmJXDIzTovg3I5tVF4Eg3VeNnKtCeWfN2rCgAG1hdHJpeC5vcmcvdHRuRHdRSUdsckVGam1TVXhBUWxueWRo>)
2025-04-14 14:47:05 <rifux[m]> hope you can see this image
2025-04-14 14:47:51 <dngray> hmm maybe wrong group or something you'd probably need to check dmesg
2025-04-14 14:47:58 <dngray> for something a bit more useful
2025-04-14 14:51:33 <rifux[m]> dngray: [ 1644.618046] r8169 0000:02:00.0 eth0: Link is Down
2025-04-14 14:51:33 <rifux[m]> [ 1648.085253] r8169 0000:02:00.0 eth0: Link is Up - 100Mbps/Full - flow control off
2025-04-14 14:52:15 <dngray> does rp-pppoe have any logs
2025-04-14 14:52:18 <rifux[m]> dngray: what do you mean by group?
2025-04-14 14:52:29 <dngray> you might need to enable that in the config
2025-04-14 14:52:36 <rifux[m]> dngray: how to check?
2025-04-14 14:53:20 <dngray> i think its like a config option like
2025-04-14 14:53:21 <dngray> logfile /var/log/pppoe-server-log
2025-04-14 14:53:32 <dngray> and the "debug" keyword
2025-04-14 14:53:54 <rifux[m]> dngray: but how?
2025-04-14 14:53:59 <dngray> https://www.mankier.com/8/pppoe
2025-04-14 14:56:28 <dngray> in the peers file https://wiki.alpinelinux.org/wiki/PPP
2025-04-14 14:56:44 <dngray> https://salsa.debian.org/debian/ppp/-/raw/master/debian/extra/peers-pppoe like that one
2025-04-14 14:57:13 <longnoserob[m]> cant networkmanager handle pppoe?
2025-04-14 14:57:35 <buinb> networkmanager can't handle much in general
2025-04-14 14:58:15 <dngray> you can do ppp with network manager
2025-04-14 14:58:57 <dngray> though i have not done that
2025-04-14 15:02:41 <PureTryOut> リフクス ✿: since this is an IRC-first channel please don't use non-IRC features like replies and multi-line messages
2025-04-14 15:03:26 <abby> ironically, your message used the matrix display name as the nick you're replying to
2025-04-14 15:03:58 <dngray> lol because im on matrix lol probably
2025-04-14 15:05:28 <rifux[m]> <dngray> "and the "debug" keyword" <- https://privatebin.net/?b0441397d1ba04c1#6Rz8w9o55GfmtWpBA4QzHLZj7osKU3KvMyGQBzRtBSDF
2025-04-14 15:11:04 <PureTryOut> dngray: according to your Matrix ID on the Matrix side, you're not? You're not joined in #_oftc_#alpine-linux:matrix.org 
2025-04-14 15:12:16 <abby> i assume the message was for rifux <PureTryOut> リフクス ✿: since this
2025-04-14 15:24:54 <rifux[m]> <rifux[m]> "https://privatebin.net/?..." <- pls help
2025-04-14 16:17:17 <dngray> i run my own heisenbridge
2025-04-14 16:17:35 <dngray> so probs just looks like a normal irc client
2025-04-14 16:24:39 <XFaCiEr> What is a heisenbridge?
2025-04-14 16:32:45 <rhizoome> probably a matrixbridge
2025-04-14 16:43:02 <hsp> I have had the following error with apk upgrade for a long time. apk fix does not help. ==> 1 error; 472 MiB in 203 packages
2025-04-14 16:43:12 <hsp> $ apk fix
2025-04-14 16:43:14 <hsp> (1/1) Reinstalling kmod-bash-completion (34.2-r0)
2025-04-14 16:43:16 <hsp> ERROR: kmod-bash-completion-34.2-r0: trying to overwrite usr/share/bash-completion/completions/insmod owned by bash-completion-2.14.0-r1.
2025-04-14 16:43:18 <hsp> ERROR: kmod-bash-completion-34.2-r0: trying to overwrite usr/share/bash-completion/completions/rmmod owned by bash-completion-2.14.0-r1.
2025-04-14 16:43:20 <hsp> 1 error; 472 MiB in 203 packages
2025-04-14 17:05:30 <mio> hsp: #17074
2025-04-14 17:11:18 <hsp> algitbot, What can I do?
2025-04-14 17:12:28 <hsp> mio,  What can I do?
2025-04-14 17:13:14 <mio> wait, should be fixed when the related MR is merged
2025-04-14 17:13:27 <hsp> ok
2025-04-14 17:43:01 <sertonix[m]> |cos|: uppercase letters should probably be allowed in the hostname. I opened a MR to fix that. Not sure if I can fix the other stuff
2025-04-14 17:48:20 <|cos|> sertonix[m]: thanks for looking and acting on what you can do. appreciated!
2025-04-14 18:54:03 <camelia> hii!
2025-04-14 18:54:23 <camelia> did anyone manage to install the alpine cloud image on an OVH dedicated server by any chance?
2025-04-14 18:54:59 <camelia> to be fair i don't have much experience with cloud images and I'm not sure what to enter in different fields :/
2025-04-14 19:06:25 <xyh> alpine plan to drop x11?
2025-04-14 19:10:22 <sewn> why?
2025-04-14 19:12:01 <bent_fingers> i can install tor and use regular firefox to access the onion network?
2025-04-14 19:12:30 <p_6f3Ik7Suw> yes you can
2025-04-14 19:12:51 <bent_fingers> inversely, i can use just the tor-browser (from flatpak) and use that without installing the `tor` package at all?
2025-04-14 19:12:58 <p_6f3Ik7Suw> probably yes
2025-04-14 19:13:03 <bent_fingers> thanks man
2025-04-14 19:13:42 <p_6f3Ik7Suw> really dpends on the flatpack itself.
2025-04-14 19:13:46 <p_6f3Ik7Suw> if it bundles tor or not
2025-04-14 19:19:24 <bent_fingers> yay!!!
2025-04-14 19:20:20 <bent_fingers> neat implementation
2025-04-14 19:20:37 <bent_fingers> only need to install tor, cp the sample config, and start the openrc service
2025-04-14 19:24:27 <bent_fingers> time to worsen my criminal record
2025-04-14 19:25:12 <dwfreed> O.o
2025-04-14 19:26:10 <f_> o.O
2025-04-14 19:26:57 <p_6f3Ik7Suw> worsen the criminal record means you get caught. why use tor then if you wanna get caught?
2025-04-14 19:27:04 <bent_fingers> it's fine, felonies don't mean anything anymore, can even become a US president, nobody cares
2025-04-14 19:27:33 <p_6f3Ik7Suw> there's a world outside the us...
2025-04-14 19:28:23 <bent_fingers> they can't stop felons from running the world, though
2025-04-14 20:15:57 <ssm_> bent_fingers: are you saying the only thing that can beat a bad felon is a good felon
2025-04-14 20:23:25 <bent_fingers> i'm too pessimistic to say that
2025-04-14 20:24:10 <bent_fingers> most felons here just get their records expunged anyhow, it's all perfunctory
2025-04-14 20:53:24 <Apachez> d
2025-04-14 21:48:10 <bent_fingers> p
2025-04-14 22:15:38 <omni> sounds offtopic
2025-04-14 22:23:37 <swee> are there any extra things I have to do for Qt5 apps to match KDE6 settings on Alpine?
2025-04-14 22:23:47 <swee> I use Kvantum, also has kvantum-qt5 installed, but qt5 apps always go to Breeze and use the wrong cursor size (I have it set to 72 but it keeps going to default)
2025-04-15 04:01:35 <bent_fingers> what do `fwupd-grub` and `fwupd-openrc` do?
2025-04-15 04:04:00 <bent_fingers> i see
2025-04-15 04:04:17 <bent_fingers> well, i don't know what fwupd-grub does
2025-04-15 08:06:53 <Alphare[m]> Hello all. I am having trouble building a new docker base image for testing a Rust-augmented Python project. I wanted to start small and just create an image for the cargo test bit, but some tests include embedding Python to test our memory sharing stuff we do. I get the following failure: https://paste.debian.net/plain/1369549
2025-04-15 08:08:24 <Alphare[m]> I'm using the python:alpine3.21 base image with python3, python3-dev, libgcc, build-base and make installed, as well as rustup/cargo. We currently use Debian images but wanted to get smaller images, hence the move to Alpine, but this is a new problem for me.
2025-04-15 08:15:06 <ikke> Alphare[m]: alpine itself doesn't ship python 3.13 yet
2025-04-15 08:15:51 <ikke> So python3-dev only has libpython3.12.so
2025-04-15 08:16:02 <Alphare[m]> I've only just realized that now, but I'm trying to point PyO3 now to 3.12 (which is installed according to apk info -L python3-dev)
2025-04-15 08:17:47 <ikke> The python:alpine3.21 image comes with python 3.13
2025-04-15 08:18:31 <Alphare[m]> python --version is Python 3.13.2
2025-04-15 08:18:54 <Alphare[m]> but then I don't have the .so, do I?
2025-04-15 08:19:02 <ikke> Correct 
2025-04-15 08:19:15 <Alphare[m]> I tried with LD_LIBRARY_PATH="/usr/lib" PYO3_PYTHON=python3.12, still no luck
2025-04-15 08:19:28 <Alphare[m]> I do have the right .so in /usr/lib
2025-04-15 08:19:45 <Alphare[m]> and python3.12 is a valid interpreter for 3.12.9
2025-04-15 08:21:02 <ikke> You could use the python:3.12-alpine3.21 image
2025-04-15 08:22:19 <Alphare[m]> in this particular case I don't need a specific Python, so it's probably fine
2025-04-15 08:22:47 <Alphare[m]> I don't understand why pointing PyO3 to the "correct places" doesn't work, but this is starting to feel more like a PyO3 problem
2025-04-15 08:22:59 <Alphare[m]> here I go rebuilding another image ;)
2025-04-15 08:46:54 <Alphare[m]> I used alpine:3.21 and installed python3 and python3-dev, but I get the same error though it's now looking for -lpython3.12
2025-04-15 08:47:04 <Alphare[m]> by default I mean
2025-04-15 09:04:26 <AndroUser> Hello
2025-04-15 09:05:22 <AndroUser> How to install alpine-linux on Tablet?
2025-04-15 09:06:00 <AndroUser> I can't install, I have a Android tablet
2025-04-15 09:06:16 <Ermine> no way
2025-04-15 09:07:09 <Ermine> but you can try to port postmarketos, which is a derivative of alpine focused on such devices
2025-04-15 09:08:31 <Ermine> But android devices are a dark land when it comes to their hardware support, so here be dragons (and a huge amount of effort to make them work)
2025-04-15 09:09:19 <ikke> Alphare[m]: It's tricky because python in the image is compiled separately from libpython provided by alpine itself 
2025-04-15 09:09:28 <ikke> Not sure if that matters 
2025-04-15 09:12:39 <AndroUser> Okay, maybe I should seek help elsewhere and come back when I have time
2025-04-15 09:13:27 <AndroUser> Because I come from China and my English is average, I will seek help from others
2025-04-15 09:13:40 <AndroUser> Thank you
2025-04-15 09:15:08 <Alphare[m]> @ikke to me it looks like it's not actually looking in the right folders for libs
2025-04-15 09:15:40 <Alphare[m]> because if you have a different version of Python you get "undefined symbol" errors, and even then, only in different versions of Python itself
2025-04-15 09:15:54 <ikke> Alphare[m]: check the environment variables in the image 
2025-04-15 09:16:38 <Alphare[m]> there are almost none
2025-04-15 09:19:05 <Alphare[m]> ldconfig -v doesn't print anything
2025-04-15 09:22:49 <Alphare[m]> maybe it's gcc that doesn't play well with alpine
2025-04-15 09:32:33 <ikke> Where do you get gcc from?
2025-04-15 09:32:35 <Alphare[m]> mold doesn't like it either :(
2025-04-15 09:32:58 <buinb> Alphare[m], ldconfig is just a simple script that links library names to actual filename
2025-04-15 09:38:43 <Alphare[m]> no matter the linker used, I get the same error
2025-04-15 09:38:47 <Alphare[m]> which is good, in a sense
2025-04-15 09:39:06 <Alphare[m]> but this is really something I've never had to deal with, so I don't know the tools to debug this effectively
2025-04-15 09:47:22 <Alphare[m]> Looking at the musl manual, in the section about LD_LIBRARY_PATH: "This variable is completely ignored in programs invoked setuid, setgid, or with other elevated capabilities"
2025-04-15 09:58:44 <Alphare[m]> no such luck there either
2025-04-15 10:00:57 <Alphare[m]> (I mean by using a non-privileged user)
2025-04-15 10:01:25 <Alphare[m]> I've also tried writing to /etc/ld-musl-x86_64.path
2025-04-15 10:01:43 <Alphare[m]> but nothing I do seems to have any effect whatsoever
2025-04-15 10:02:12 <Alphare[m]> I might come back after lunch break with some strace digging, but this is starting to feel like a strong L for alpine
2025-04-15 11:39:07 <Alphare[m]> wow.
2025-04-15 11:39:22 <Alphare[m]> there is a separate env variable for Rust doc tests
2025-04-15 11:41:36 <Alphare[m]> sorry about the fuss, this seems to work fine now. This was all very confusing
2025-04-15 11:44:48 <ikke> Alphare[m]: glad you found it
2025-04-15 11:48:00 <Alphare[m]> thank you for your help
2025-04-15 11:50:47 <dngray> so i kinda copied the alpinelinux package with the libinfnoise https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=infnoise#n28 that includes a header file and library
2025-04-15 11:50:51 <dngray> maybe that should be a -dev subpkg
2025-04-15 11:50:57 <dngray> https://github.com/leetronics/infnoise/blob/master/software/Makefile.linux#L43-L48
2025-04-15 11:51:12 <dngray> is that the norm in alpinelinux?
2025-04-15 11:52:07 <dngray> the other question i had was regarding the patching of the makefile for the openrc script
2025-04-15 11:52:43 <dngray> because i can't put it under the other functions because the prefix will be different
2025-04-15 11:53:00 <dngray> the prefix for the others is set to PREFIX="$pkgdir"/usr
2025-04-15 11:53:18 <dngray> so if i want to set that to PREFIX="$pkgdir"/etc i'm going to need to have that in it's own makefunction
2025-04-15 11:53:45 <dngray> is there a defacto standard for what that function should be named like
2025-04-15 11:53:46 <dngray> make -f Makefile.linux PREFIX="$pkgdir"/etc install init
2025-04-15 11:53:48 <dngray> "init"
2025-04-15 12:02:41 <ikke> dngray: if the makefile does not include installing an openrc file, I'd just copy it manually to the correct location (instal ..)
2025-04-15 12:03:04 <dngray> yeah i guess that's less lines tbh
2025-04-15 12:03:17 <dngray> patching the makefile with its own function is going to be more lines of stuff
2025-04-15 12:04:45 <buinb> that would also be a one-line change
2025-04-15 12:05:06 <buinb> But here it's a package addition, so it makes more sense to manage it in the package
2025-04-15 12:07:34 <dngray> well technically no they include the openrc file
2025-04-15 12:07:39 <dngray> but i do have a patch which changes a path in it
2025-04-15 12:07:56 <dngray> and another patch which removes the systemd units from the general make install command
2025-04-15 12:28:30 <buinb> Ah
2025-04-15 12:28:45 <buinb> Then I'd patch the Makefile (but that's just a question of taste)
2025-04-15 12:30:28 <dngray> https://gitlab.alpinelinux.org/dngray/aports/-/jobs/1810862
2025-04-15 12:30:33 <dngray> i think i did everytrhing i was told
2025-04-15 12:31:53 <buinb> :)
2025-04-15 17:13:28 <bent_fingers> anyone know what fwupd-grub does?
2025-04-15 17:48:00 <XFaCiEr> Sounds like a Grub add-on that runs the fwupd service
2025-04-15 17:49:34 <|cos|> while on the topic of grub, i've spent days attempting to get my alpine system to boot.
2025-04-15 17:51:42 <|cos|> according to the LVM_on_LUKS wiki page, one should enable "modules=sd-mod,usb-storage,ext4,cryptsetup,keymap,cryptkey,kms,lvm". would that be correct informatin? and in that case, how can one install those kernel modules which are not in linux-lts?
2025-04-15 17:53:25 <XFaCiEr> |cos|: are you trying to do a completely encrypted system?
2025-04-15 17:55:09 <|cos|> XFaCiEr: i am. i need separate partitions though. running setup-alpine works, but setup-disk leaves me one not booting. :(
2025-04-15 17:55:39 <XFaCiEr> "need separate partitions"?
2025-04-15 17:55:46 <XFaCiEr> As in... EFI?
2025-04-15 17:57:33 <|cos|> separate / and /home are important to me.
2025-04-15 17:58:41 <XFaCiEr> You could do that with one partition and BTRFS
2025-04-15 17:58:42 <bent_fingers> XFaCiEr: but how?
2025-04-15 17:58:55 <XFaCiEr> bent_fingers: pardon me?
2025-04-15 17:59:07 <bent_fingers> it queries the internet and downloads firmware during boot?
2025-04-15 17:59:44 <bent_fingers> i don't understand how it would work
2025-04-15 17:59:47 <|cos|> XFaCiEr: not different datasets, or whatever the linux terminology is. different partitions.
2025-04-15 17:59:55 <XFaCiEr> No idea, there are similar packages for UEFI
2025-04-15 18:00:02 <XFaCiEr> as EFI applications
2025-04-15 18:00:39 <XFaCiEr> |cos|: You're coming from BSD?
2025-04-15 18:01:28 <|cos|> XFaCiEr: FreeBSD are on all my primary systems, yes.
2025-04-15 18:02:18 <XFaCiEr> And these two LUKS containers use different passphrases/keyfiles?
2025-04-15 18:03:43 <|cos|> as someone pointed out in a dm. i do mean separate logical volumes, not separate actual partitions.
2025-04-15 18:03:59 <XFaCiEr> Oh... "Good luck with all that"
2025-04-15 18:04:40 <|cos|> well. "thanks."
2025-04-15 18:05:04 <XFaCiEr> I don't touch LVM when ZFS exists
2025-04-15 18:05:16 <|cos|> getting my system up and running is my thing, for sure. i though my question here was about correcting the alpine wiki.
2025-04-15 18:05:43 <|cos|> zfs doesn't really exist in the linux world, one could argue. it's incompatible with gpl, right?
2025-04-15 18:06:23 <XFaCiEr> If people are willing to slap proprietary Nvidia blobs into their Linux kernels, FLOSS ZFS modules are more acceptable.
2025-04-15 18:06:51 <XFaCiEr> |cos|: Furthermore, Alpine has ZFS in the repos, and uses Linux-LTS by default
2025-04-15 18:08:13 <XFaCiEr> |cos|: Reminder that realtime support was not upstreamed until recent months
2025-04-15 18:11:03 <|cos|> XFaCiEr: so your suggestion is that i boot from zfs, with an encrypted dataset as its root partition? i doubt that will involve quite a bit more of grub work than getting luks to function.
2025-04-15 18:11:39 <XFaCiEr> I am not suggesting anything, just that personally I don't use LVM, and try to use ZFS as much as possible
2025-04-15 18:12:10 <|cos|> i'm quite sure the tricky part which is standing between me and alpine is luks
2025-04-15 18:12:34 <XFaCiEr> And also that OpenZFS on Linux despite being out-of-tree is a perfectly valid usecase, and "really exists" in the Linux world. Even FreeBSD's current OpenZFS implementation is derived from the zfs-on-linux project
2025-04-15 18:13:14 <XFaCiEr> |cos|: As for the LUKS stuff, really depends if you are handling it within Grub or the initramfs
2025-04-15 18:13:38 <XFaCiEr> But you already said you are doing a full encryption so that non-encrypted boot partition is out of the question
2025-04-15 18:14:19 <|cos|> i'm happy to go with whatever brings me a booting laptop. my current attempt has an unencrypted /boot, which i try to share with the EFI-partition.
2025-04-15 18:15:14 <|cos|> my usecase is allowing me to forget the laptop on a train. no advanced threats.
2025-04-15 18:15:39 <XFaCiEr> In that case, my approach would be to boot just the kernel from the boot{loader,manager} and handle the LUKS decryption from the initramfs
2025-04-15 18:15:44 <bent_fingers> advanced threats?
2025-04-15 18:15:54 <XFaCiEr> But that should be a feature of mkinitrd
2025-04-15 18:16:11 <bent_fingers> just use a tiling window manager with custom keybindings. nobody will know how to use it
2025-04-15 18:16:13 <XFaCiEr> Sorry, mkinitfs... went Slackware-brained for a second
2025-04-15 19:08:35 <|cos|> darn! i found the error, and oddly enough it annoyed me even though i can boot now.
2025-04-15 19:09:13 <|cos|> these old eyes, and this apple retina screen with its tiny, tiny console font...
2025-04-15 19:09:59 <|cos|> seems i slipped in the grub-style cryptoroot rather than linux-spelled cryptroot as kernel argument
2025-04-15 19:49:47 <realroot[m]> how install libsomething from community/main after I installed it from masked egde?
2025-04-15 19:50:26 <Saijin_Naib[m]> The short answer is don't mix stable/edge
2025-04-15 19:50:40 <Saijin_Naib[m]> what does that line look like in world?
2025-04-15 19:50:55 <Saijin_Naib[m]> Can you pin the other repo and apk fix it?
2025-04-15 19:52:40 <realroot[m]> pin the repo?
2025-04-15 19:53:57 <realroot[m]> package is in `  http://alpinelinux.../v3.21/main`
2025-04-15 19:55:49 <Saijin_Naib[m]> https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper
2025-04-15 19:55:59 <Saijin_Naib[m]> The section under World talks about that
2025-04-15 19:56:09 <Saijin_Naib[m]> What is /etc/apk/repositories pointing towards?
2025-04-15 19:58:11 <realroot[m]> to main and community with edge masked
2025-04-15 19:58:14 <realroot[m]> before i ran apk add libsom@edge
2025-04-15 19:58:17 <realroot[m]> apk fetch ; apk add and it is done
2025-04-15 19:58:20 <realroot[m]> but i wonder if apk can do it automatically
2025-04-15 20:00:25 <realroot[m]> found it apk add libsome -u
2025-04-16 03:12:35 <ang1e> I'm fixing my alpinelinux machine with system rescue rn. I mounted boot to /mnt/alpine/boot and I decrypted the rest of my filesystem to /dev/mapper/alps and mounted to /mnt/alpine.
2025-04-16 03:13:51 <ang1e> I got the rest of my packages fixed, but I'm having an issue with apk fix and grub. it still detects it's a mounted mapper partition, so I get "error: failed to get canonical path of `/dev/mapper/alps`."
2025-04-16 04:45:32 <systemdlete> I dunn did it.  I actually built a package!   (whooptee)  I've managed to modify the APKBUILD and patch files for the virtualbox-guest-additions and successfully built the apk package files.  I just completed 7.1.6, when I noticed that virtualbox has releaseed 7.1.8 (or are about to).
2025-04-16 04:45:45 <systemdlete> I am building and testing the 7.1.8 package now.
2025-04-16 04:46:34 <systemdlete> I am new to doing this, esp. on alpine linux, so I am not sure what the protocol is.  I know there are at least a couple people who have done the port (most recent in repo is 7.0.22), and I really don't mean to duplicate or jump their work.
2025-04-16 04:47:57 <systemdlete> I'd imagine that now I need to create a branch and make a PR.  I've been using fossil because I find it less confusing than git, and there is a git bridge for it, so I want to know if there are any issues with using fossil locally here and bridging changes to the git repo.  
2025-04-16 04:48:26 <systemdlete> But I can try to do this with git I suppose.  
2025-04-16 04:48:54 <systemdlete> (I meant to post this in alpine-devel, sorry)
2025-04-16 05:36:09 <frag> any news on rpi5 crashing problem?
2025-04-16 05:40:39 <|cos|> good morning! thanks for your help yesterday everyone. it took all of my mental energy, but i got booting from a /boot within a luks container working in the end.
2025-04-16 05:44:19 <|cos|> i also learned that i can't do that with the freebsd-wifibox-alpine kernel, but need to figure out how that one/the wl driver are made to actually work.
2025-04-16 05:45:49 <|cos|> would anyone know of any documentation of the alpine kernel build process, other than the "Kernels", "How_to_build_the_Alpine_Linux_kernel" and "Custom_Kernel" wiki pages?
2025-04-16 05:45:54 <|cos|> specifically, i can
2025-04-16 05:46:34 <|cos|> not figure out how to set _flavour and have it result in a package with my own suffix
2025-04-16 06:59:25 <|cos|> no custom kernel building strictly required, it turned out. i've got wifi AND luks, an a silly macbook. yey!
2025-04-16 07:00:58 <|cos|> so now i'm just lost in alpine. a bit frustrated over having to have added passwords on non-login accounts.
2025-04-16 07:02:21 <|cos|> i take it shoehorning in sudo is the wrong approach in this ecosystem. anyone got any hints on how to get something like pam_ssh_agent_auth working with become?
2025-04-16 07:02:34 <|cos|> with become, i mean `doas`
2025-04-16 10:42:39 <aron> what is this situation about CVE's funding? will CVE be down?
2025-04-16 10:43:13 <lopid> strange place to ask
2025-04-16 10:43:54 <aron> why? alpine's one strength is security, CVE is highly related to security
2025-04-16 10:44:21 <lopid> same argument about relevance for anything that has had a CVE
2025-04-16 10:45:09 <ikke> https://krebsonsecurity.com/2025/04/funding-expires-for-key-cyber-vulnerability-database/
2025-04-16 10:46:33 <simon_> rhizoome:  you online bro
2025-04-16 10:46:43 <simon_> I am now on void linux
2025-04-16 10:46:52 <simon_> not sure if my distro is smart
2025-04-16 10:47:00 <rhizoome> simon_: ok
2025-04-16 10:47:16 <simon_> rhizoome:  but there is one good thing about it I have stable fps ingame
2025-04-16 10:47:51 <simon_> I somehow think that wayland has been the problem of my performance on all new distros and  several other factors
2025-04-16 10:48:04 <simon_> like kernel or musl
2025-04-16 10:48:12 <aron> ikke: thanks, so no resoultion yet
2025-04-16 10:48:44 <simon_> also it turned out that systemd could also be a factor of the performance deficit
2025-04-16 10:49:46 <simon_> after all I am not willing to dig deeper on why I have bad performance on alpine debian chimera redhat opensuse or arch.
2025-04-16 10:50:44 <simon_> I knowwhat also could be a factor that wayland simply is faster than my screen such that it looks laggy even though it is totally fluent?
2025-04-16 10:51:14 <lopid> there have been reports of wayland introducing some screen lag
2025-04-16 10:51:14 <ikke> aron: I came accross https://www.thecvefoundation.org/ as well
2025-04-16 10:52:05 <quinq> https://gcve.eu/ too, etc.
2025-04-16 10:52:33 <quinq> simon_, what's “bad performance” here?
2025-04-16 10:52:38 <aron> ah, cool
2025-04-16 10:52:58 <aron> afterall, this was the right place to ask
2025-04-16 10:54:16 <lopid> you can look at it that way
2025-04-16 10:54:58 <quinq> You can always go to your bakery and ask for a good stew recipie, you might have *some* answer
2025-04-16 11:08:17 <WhyNotHugo> ocasionally, `apk upgrade -a` will pull ALL of the linux-firmware packages
2025-04-16 11:08:27 <WhyNotHugo> I can't reproduce this consistently, butit happens once every few weeks
2025-04-16 11:09:47 <WhyNotHugo> apk shows that it is a dependency of linux-edge
2025-04-16 11:14:34 <aron> install linux-firmware-none, it provides linux-firmware-any
2025-04-16 12:14:28 <taterspie[m]> Hi, is it possible to run the cryptsys option under the setup-disk script when you have multiple partitions on disk already, without wiping all those partitions?
2025-04-16 12:14:28 <taterspie[m]> I tried lvm on luks guide on the wiki multiple times but it always fails to create the boot partition properly
2025-04-16 13:00:38 <|cos|> taterspie[m]: i've installed alpine during the past few days and concluded that manual setup is required. ask me while my knowledge is still fresh.
2025-04-16 13:07:35 <taterspie[m]> <|cos|> "taterspie: i've installed alpine..." <- By manual, are you referring to the luks on lvm page on the wiki?
2025-04-16 13:09:04 <|cos|> taterspie[m]: i refer to setup need to be done manually. gimme a couple of minutes and i'll post my notes.
2025-04-16 13:11:44 <taterspie[m]> Okay, thanks
2025-04-16 13:18:38 <|cos|> taterspie[m]: You can follow along the creation at https://www.netizen.se/docs/arriving_to_alpine but waiting until I've at least copy'n'pasted my successful commands is advised.
2025-04-16 13:19:39 <simon_> quinq: well basically bad performance is that I have 180-200 fps and I think that my screen can not understand this many fps and therefore there is a problem between my screen and new software to make it overperform than my screen can understand
2025-04-16 13:20:12 <simon_> but on the new kde xorg it is not better I only have good performance on old kde and old cinnamon
2025-04-16 13:21:23 <simon_> btw can docker somehow do telegram gui?
2025-04-16 13:21:34 <simon_> or rstudio gui?
2025-04-16 13:21:41 <simon_> that would be insane cool
2025-04-16 13:21:41 <simon_> :D
2025-04-16 13:21:54 <simon_> but I am not sure if that is overkill if there is already nix
2025-04-16 13:21:59 <quinq> Ah, yeah, your driver sould use the data provided by your monitor for adapting the output framerate
2025-04-16 13:23:17 <simon_> quinq:  afaik the wayland protocol doesnt care anylonger about low performing screens. and afaik the xorg can only do display 60 fps
2025-04-16 13:23:38 <simon_> I am not willing to buy a new screen for wayland
2025-04-16 13:24:46 <simon_> and  i will not use vsync xd vsync in games is horrible
2025-04-16 13:27:46 <simon_> quinq:  what would you do to install telegram on your desktop os? or rstudio is nix already enough to isolate it to not conflict it or to not get a virus?
2025-04-16 13:28:26 <simon_> I have heard docker is somehow a solution
2025-04-16 13:33:22 <quinq> I wouldn't, I try to avoid proprietary software
2025-04-16 13:35:36 <taterspie[m]> <|cos|> "taterspie: You can follow..." <- Thanks for posting that, I'll have another try today
2025-04-16 13:35:36 <taterspie[m]>  Also kinda neat that I can see you editing the document in real time
2025-04-16 13:42:05 <|cos|> taterspie[m]: Good look! I've posted the commands now. Am not fully happy with the document yet, but will need to do a few other things before getting back to it. At least it now contains the steps I took. Please let me know if anything is unclear!
2025-04-16 13:44:40 <taterspie[m]> Awesome, thanks :) 
2025-04-16 13:44:40 <taterspie[m]> I'll try and give some feedback but there's also a chance that I just ran off into the woods
2025-04-16 13:59:56 <WhyNotHugo> aron: thanks, that works…
2025-04-16 14:00:01 <WhyNotHugo> but I have the following installed already: linux-firmware-amdgpu linux-firmware-intel linux-firmware-mediatek  linux-firmware-other  linux-firmware-rtl_bt
2025-04-16 14:01:25 <WhyNotHugo> and it has the same provides directive
2025-04-16 14:12:25 <aron> dunno, I was just guessing, it could be a bug
2025-04-16 14:23:57 <simon_> quinq: but if you needed an proprietary app which way would you install it like nix or docker?
2025-04-16 14:24:10 <simon_> I think nix might be good if it isolate the app
2025-04-16 14:30:39 <WhyNotHugo> simon_: nix does not provide any isolation
2025-04-16 14:30:49 <WhyNotHugo> it merely install into a different directory
2025-04-16 14:31:19 <WhyNotHugo> for some proprietary app you'll want to use docker, podman or flatpak
2025-04-16 14:31:26 <WhyNotHugo> or build yuor own sandbox using bwrap+helpers
2025-04-16 14:34:45 <simon_> WhyNotHugo:  sounds  horrible
2025-04-16 14:36:01 <simon_> sad that there is not an easy way to isolate like on android
2025-04-16 14:36:22 <WhyNotHugo> i just listed three ways
2025-04-16 14:36:29 <WhyNotHugo> android runs each app as a differnt user, you can do that too
2025-04-16 14:38:37 <simon_> do you know subuser the app from docker
2025-04-16 14:39:58 <simon_> I will try subuser maybe this is good
2025-04-16 14:39:58 <simon_> :D
2025-04-16 14:46:41 <simon_> subuser much too complicated
2025-04-16 14:46:47 <simon_> damn
2025-04-16 14:49:32 <quinq> WhyNotHugo, I would use a chroot
2025-04-16 14:49:34 <quinq> simple
2025-04-16 14:50:16 <simon_> chroot for what exactly?
2025-04-16 14:51:03 <simon_> brb I got to relogin
2025-04-16 14:51:06 <quinq> For running your proprietary application, which I suppose is built dynamically against libraries it doesn't actually provide
2025-04-16 14:52:46 <rhizoome> quinq WhyNotHugo: unshare wasn't listed yet.
2025-04-16 14:52:50 <simon_> that did not go as planned xd
2025-04-16 14:53:40 <WhyNotHugo> a chroot might not suffice. you're still in the same namespace for other shenanigans
2025-04-16 14:53:53 <WhyNotHugo> like, a chrooted program can send SIGKILL to your compositor
2025-04-16 14:54:42 <quinq> I think that the goal was to run it without polluting the rest of the system
2025-04-16 14:55:01 <simon_> kinda yes
2025-04-16 14:56:18 <simon_> I will just install nix
2025-04-16 14:56:18 <simon_> xd
2025-04-16 14:56:56 <rhizoome> quinq WhyNotHugo: sandboxes can also do harm, if for example a browser doesn't get access to BPF, you might be worse off.
2025-04-16 14:57:12 <invoked> not using telegram seems the path of least resistance
2025-04-16 14:57:44 <rhizoome> not using computers :-)
2025-04-16 14:57:50 <invoked> even better
2025-04-16 14:58:07 <taterspie[m]> |cos|: what are the `_efipart` and `_lvmpart` variables supposed to represent? I'm a little confused by the 1 and 2 that's currently there
2025-04-16 15:07:26 <WhyNotHugo> invoked: telegram-desktop is FLOSS (GPL-3.0-or-later WITH OpenSSL)
2025-04-16 15:11:08 <quinq> And where do you connect to with it?
2025-04-16 15:11:43 <WhyNotHugo> I connect it via my ISP's proprietary infrastructure to some proprietary server. From my proprietary hardware.
2025-04-16 15:12:44 <quinq> Sad.
2025-04-16 15:14:17 <taterspie[m]> |cos|:  never mind, figured it out, I just hand to change them to p1 and p2 since I'm doing this on an nvme drive
2025-04-16 15:25:06 <simon_> couldnt I just take qubes os to make what I desire what I talked about?
2025-04-16 15:46:18 <taterspie[m]> |cos|:  I've managed to install alpine with your guide, after entering my encrypted volume password it boots up but it appears /root /boot and /home don't mount so it won't even start a tty session to log in
2025-04-16 16:01:11 <|cos|> taterspie[m]: glad you figured out the partitions. too bad it didn't get you all the way. :(
2025-04-16 16:02:25 <|cos|> taterspie[m]: i might have manually added filesystems to /etc/fstab and forgotten about documenting it.
2025-04-16 16:03:28 <|cos|> taterspie[m]: do you literally mean "/root" or "/" by "/root"? it's the superusers home directory? the actual root filesystem gets mounted?
2025-04-16 16:07:11 <|cos|> 
ACTION would assume logging in on a tty should work without home directories and what not, but you get to a login prompt taterspie[m] and not past it?

2025-04-16 16:10:08 <taterspie[m]> 
ACTION uploaded an image: (49KiB) < https://matrix.org/oftc/media/v1/media/download/Ac7ZYzmHnT0LEaMshjF4Nf2pD5dF927A8nZVT88V1kLmbRpNGdcTiRI8auryvcFUaFEyVPW0v1SWcpA4Yt6iT5VCeWh3ZtWAAG1hdHJpeC5vcmcvRnFscUNVbk9nTkxLaWxXeldLeEZrTUdN >

2025-04-16 16:10:23 <taterspie[m]> Sorry for the blurry picture, I'm on my phone my right
2025-04-16 16:11:55 <|cos|> ah. the actual root filesystem does not get mounted.
2025-04-16 16:12:29 <taterspie[m]> Maybe I do have to update fstab
2025-04-16 16:12:51 <|cos|> taterspie[m]: are you able to mount it manually from that initial ramdisk environment? `cryptsetup luksOpen /dev/your-puks-part some-name`
2025-04-16 16:13:01 <|cos|> Nah. This is before fstab gets relevant.
2025-04-16 16:13:14 <|cos|> *luks
2025-04-16 16:15:16 <|cos|> if there's no cryptsetup, the error lies in creation of the mkinitfs
2025-04-16 16:15:56 <taterspie[m]> Hmm nope, /dev/mapper doesn't even exist either
2025-04-16 16:16:08 <taterspie[m]> Cryptsetup is still there though
2025-04-16 16:17:23 <taterspie[m]> I don't even see any  other nvme0n1## partitions either in /dev/
2025-04-16 16:17:46 <|cos|> ah!! you might have followed my guide too closely :
2025-04-16 16:18:02 <|cos|> i don't have any nvme devices on my machine, but you do.
2025-04-16 16:18:42 <taterspie[m]> Aaah okay
2025-04-16 16:18:53 <|cos|> the missing features ought to be in the mkinitfs.bak file
2025-04-16 16:21:44 <taterspie[m]> I'll have to go through the install again, but I'll double check there and see what I find
2025-04-16 16:31:50 <taterspie[m]> |cos|:  the .bak file is just empty before and after running the `mkinitfs -c ...` command
2025-04-16 16:33:48 <taterspie[m]> I'll try adding nvme and see how that goes
2025-04-16 16:42:31 <taterspie[m]> Hmmm, one step closer, for some reason the install forgot my root password and account+password
2025-04-16 17:15:11 <|cos|> taterspie[m]: i understand that as you actually getting booted all the way into a login prompt?
2025-04-16 17:16:25 <taterspie[m]> Aaah yeah, adding nvme to the features got me to the login screen
2025-04-16 17:16:50 <taterspie[m]> It seems a lot of things didn't make it to the final install though
2025-04-16 17:18:00 <taterspie[m]> Accounts, passwords, doas is missing, can't seem to get apk to work, wifi settings stuck but I can't seem to do anything with the internet in general though
2025-04-16 17:21:07 <taterspie[m]> When I tried setting up the keyboard settings with setup-keymap, no options show up, just an error
2025-04-16 17:21:42 <|cos|> taterspie[m]: odd. in the installer or post-reboot?
2025-04-16 17:23:53 <taterspie[m]> Post reboot
2025-04-16 17:25:36 <taterspie[m]> I went though all those steps during the install, not sure what happened after that
2025-04-16 17:46:39 <|cos|> taterspie[m]: you didn
2025-04-16 17:46:51 <|cos|> 't perhaps forgot to do them in the chroot?
2025-04-16 18:58:54 <taterspie[m]> |cos|: aaaah I think I got it, I ran setup-alpine first and hit ctrl+c before setup-disk, then proceeded with the rest of the guide
2025-04-16 18:58:55 <taterspie[m]> I'll reinstall it a couple times to make sure it's working properly and I'll post the script I used for you to look over
2025-04-16 19:32:51 <realroot[m]> can i install libc6-compat 32 bit on x86_64?
2025-04-16 19:33:49 <realroot[m]> which is part of gcombat i think
2025-04-16 19:35:08 <realroot[m]> libdl.so.2: path not found I think that it's libc6 32 bit
2025-04-16 19:44:10 <realroot[m]> how can I get rid of that file somehow?
2025-04-16 19:44:31 <realroot[m]> in $pkgdir/.control.<packager>
2025-04-16 19:44:42 <realroot[m]> *<package>
2025-04-16 19:55:38 <taterspie[m]> |cos|: Here's the script I used: https://files.sustainabilityexchange.ca/data/Open%20Source/Script-Alpine-LVMLUKS
2025-04-16 19:55:39 <taterspie[m]> I tried with luks2 just for fun but apparently it needs a cryptodisk module
2025-04-16 19:56:14 <realroot[m]> i see that other libs are link to gcompat lib so i link it
2025-04-16 19:59:11 <frag> is there a way to copy to primary from busybox vi? or would the best way be sed -n 10p file|xsel?
2025-04-16 20:01:41 <realroot[m]> primary clipboard?
2025-04-16 20:01:55 <taterspie[m]> |cos|: also thanks so much, last time I tried doing this I spent nearly 5 days and got nowhere at all, this was huge
2025-04-16 20:02:24 <frag> realroot yes
2025-04-16 20:02:41 <frag> or selection, not sure what the absolute definition is
2025-04-16 20:05:53 <|cos|> taterspie[m]: that script looks a bit familiar. :)  glad to have been to some help!
2025-04-16 20:07:43 <taterspie[m]> Hahaha yeah, I did all my partitioning from gparted from a linux mint iso so I cut those parts out
2025-04-16 20:08:27 <taterspie[m]> Also added efibootmgr earlier because I was getting errors
2025-04-16 20:09:35 <taterspie[m]> As well as nvme to the mkinitfs modules since I was working with an nvme drive
2025-04-17 06:14:21 <faithy3> Hey there! I'm trying to boot Alpine (standard, x86, burned to a CD-R) off of an old Compaq Armada laptop with a 700mhz Pentium III (for fun!). Booting fails however, I just see it go "Illegal instruction" "Mounting boot media failed." before booting me to a rescue shell.
2025-04-17 06:15:45 <faithy3> As far as I can tell Alpine does support instruction sets this old, so I'm not really sure how to go about debugging or resolving this problem. If anyone has ideas I'd really appreciate it!
2025-04-17 06:19:06 <ikke> faithy3: since recent, at least SSE2 is required 
2025-04-17 06:19:37 <faithy3> ah, i see.
2025-04-17 06:20:11 <dwfreed> 3.17+
2025-04-17 06:20:21 <dwfreed> according to some random reddit thread
2025-04-17 06:21:35 <faithy3> i guess i could grab 3.16 to at least have *a* linux running (My primary motivation right now is to get the drive imaged, and clonezilla failed to boot lmao)
2025-04-17 06:22:10 <ssm_> faithy3: openbsd or netbsd may work
2025-04-17 06:22:44 <faithy3> I feel like i always should have been dreading the day that the solution to my problem is a BSD
2025-04-17 06:23:12 <ssm_> gentoo as well, maybe; though I'd at the very least want a build server for hardware that old (maybe the binary packages work anyways...?)
2025-04-17 06:23:44 <faithy3> yeah, I was considering that! gentoo + distcc'ing from my desktop to boost compile time
2025-04-17 06:24:10 <faithy3> probably my best option if i want something modern to run on this old thing
2025-04-17 06:24:50 <ssm_> having used alpine and openbsd extensively you'll find very little different for basic usecases moving between them. netbsd is a bit more unusual
2025-04-17 06:25:19 <faithy3> though, I spent like a few weeks trying to get distcc working on a cluster of office computers running Arch once and could *never* get it to actually work but maybe it'll work for me this time lol
2025-04-17 06:25:34 <ssm_> even the installers are quite similar
2025-04-17 06:26:00 <faithy3> i see. I might try it, at the very least so I don't have to spend a whole day waiting for compiles
2025-04-17 06:26:31 <faithy3> thanks for the help y'all!
2025-04-17 06:34:00 <|cos|> faithy3: i'd say that computer is ideal for minix3, if doing things for fun. unfortunately not much is happening with that operating system these days.
2025-04-17 06:44:19 <hazardaj_nombroj> weird, normally firefox gets updated in the stable repos as well as the edge one, but right now v3.21 is still on firefox (and librewolf) 136 whereas edge is on 137
2025-04-17 10:58:49 <realroot[m]> how make link in package?
2025-04-17 10:59:18 <ptrc> you mean a symbolic link? you can just do `ln` in the package() function
2025-04-17 10:59:35 <realroot[m]> `package() { ln -s "$pkgdir/usr/share/some/file" "$pkgdir/usr/bin/file" }`
2025-04-17 11:01:48 <realroot[m]> i mean that ^. if i do that it will link to /home/user/path/packge...
2025-04-17 11:19:51 <ptrc> yeah, so you don't do that
2025-04-17 11:20:23 <ptrc> you can just do `ln -s /usr/share/some/file "$pkgdir"/usr/bin/file`
2025-04-17 12:48:54 <fabricionaweb> I like this app, but it seems not a good name - its a single letter - to be on aports https://github.com/natesales/q 
2025-04-17 14:13:43 <xpufx> "q-dns-git in the AUR"
2025-04-17 14:13:57 <xpufx> q is still unfortunate for a network app.  too much like queue
2025-04-17 14:28:16 <abby> q for query
2025-04-17 15:39:06 <dngray> ptrc i'm a little confused https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112/#note_498192
2025-04-17 15:39:50 <dngray> are you suggesting to move
2025-04-17 15:39:51 <dngray> make -f Makefile.linux PREFIX="$subpkgdir"/usr install install-lib
2025-04-17 15:40:13 <dngray> into package() and then "move" the files that were installed in the subpackage with this "amove" command?
2025-04-17 15:40:21 <dngray> seems like double work
2025-04-17 15:40:36 <dngray> guess i can it's only one .h file and a .so file
2025-04-17 15:40:37 <dngray> https://github.com/leetronics/infnoise/blob/master/software/Makefile.linux#L43-L48
2025-04-17 15:43:20 <dngray> a
2025-04-17 15:43:31 <dngray> also abuild failed when i didn't cd with the $builddir https://gitlab.alpinelinux.org/dngray/aports/-/jobs/1813796#L142
2025-04-17 15:51:01 <ikke> dngray: the typical use case is where `make install` installs everything already in the `package()` function, and then you use amove to move things from the pkgdir to the subpkgdir
2025-04-17 15:51:42 <ikke> Some split functions like openrc and dev already know what to move, so in that case, you just install them in the package() function as well and you don't need to implement those functions manually
2025-04-17 16:19:51 <dngray> ah that makes more sense
2025-04-17 16:23:46 <dngray> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/82112/diffs?commit_id=f2c97b1d15575a67814f24dbfb5072a86cef51a0 guess we will see if that works :P
2025-04-17 16:29:42 <dngray> seems to work, and the package contains what it should
2025-04-17 16:32:31 <dngray> oh i could problably shorten those two lines
2025-04-17 16:32:38 <dngray> to one make command
2025-04-17 16:33:28 <ikke> yes
2025-04-17 16:33:49 <dngray> is there any way to simplify "cd "$builddir"/software"
2025-04-17 16:33:55 <dngray> "cd software" didn't work
2025-04-17 16:36:11 <dngray> hmm and maybe i should move the install openrc line to the makefile in a patch rather than have that in the apkbuild as it's not like i am providing the openrc script. upstream provides that
2025-04-17 16:36:27 <dngray> they just choose to only install the systemd unit, which i patched out anyway because we don't have systemd
2025-04-17 16:36:58 <ikke> It happens more often that you have to manually install files provided by the project, so I would not worry too much about that
2025-04-17 16:39:06 <dngray> ah okay.
2025-04-17 16:39:36 <dngray> ye ah because someone in here said they usually patch the makefile if the project includes the openrc script
2025-04-17 16:39:42 <dngray> which would kind of be my preference
2025-04-17 16:40:01 <ikke> dngray: it would make sense if you can upstream that patch
2025-04-17 16:40:03 <dngray> ie they supply a bunch https://github.com/leetronics/infnoise/tree/master/software/init_scripts and the gentoo one works fine
2025-04-17 16:40:28 <dngray> that's fair i guess
2025-04-17 16:40:44 <dngray> but they installed the systemd unit under their main install function
2025-04-17 16:41:55 <dngray> maybe i should move their systemd install commands to an install: systemd
2025-04-17 16:42:02 <dngray> then i can try to upstream it for people who want that
2025-04-17 18:59:54 <realroot[m]> what do you install to hgave clamav? there are some paclages
2025-04-17 19:00:38 <realroot[m]> clamav istalled some i gues that's what i need
2025-04-17 20:21:27 <b00jum> any one is poking at openrc user services? if you are using pam-rundir.so for runtime dir creation / removal, can i please see you pam setup?
2025-04-17 20:23:33 <rhizoome> b00jum: I am using it. also with standard pam-rundir. I am starting openrc --user way later in the exec of hyprland.
2025-04-17 20:24:18 <rhizoome> exec-once = openrc --user gui
2025-04-17 20:26:09 <rhizoome> all I did was "apk add pam-rundir"
2025-04-17 20:27:57 <b00jum> can you please share you pam config? alpine is using base_session which includes pam_rundir as optional, i set it up in /etc/pam.d/openrc-user and /etc/pam.d/sshd as required - and the user dir under /run/user is not being created. the xdg_runtime_dir is set later (via /etc/profile), and this part work just fine. also, if i create /run/user/666 by hand, all is well. but i can
2025-04-17 20:27:59 <b00jum> get this dir to be automagically created on boot / removed on shutdown.
2025-04-17 20:33:21 <rhizoome> b00jum: sorry, for asking more question, but I really don't know what I should show you. It all as installed by apk. you only get and pam-rundir actually sets xdg_runtime_dir, do you override the runtime dir pam-rundir set, you are looking in the wrong place?
2025-04-17 20:34:31 <rhizoome> b00jum: it only works on login shells though, ssh shell won't get (just because you mentioned ssh)
2025-04-17 20:34:39 <rhizoome> it
2025-04-17 20:36:19 <b00jum> ssh is getting it as well, you can multiplex the standard oprnrc `user` service as `user.your_user`...
2025-04-17 20:37:18 <b00jum> i'm really sorry but i'm about to take my doggie for a walk, i'll be back in a little bit. failing that, if you don't mind i would love to pick your brain about your setup tomorrow.
2025-04-17 20:37:37 <b00jum> in case i miss you tonight - have a awesome one!
2025-04-17 20:40:49 <rhizoome> b00jum: when ssh session get xdg_runtime_dir then you mesed around with it. or you doing ssh -t user@host 'sh -l'
2025-04-17 21:58:18 <b00jum> rhizoome: you still about? and to answer your question, i didn't had to mess with anything at all. conviniently, openrc comes with a user service. you can multiplex it as user.your-user and it *should* start your services at boot - no login required.
2025-04-17 22:00:07 <rhizoome> b00jum: sorry, I did not setup something so complicated. I just start openrc --user gui in exec of the compositor.
2025-04-17 22:02:00 <b00jum> got you! that's fair. my setup works if i create /run/user/666 manually - services are started on the server, no login required. the culprit seem to be pam_rundir.so. it haven't been updated in years, got some issues and pr open.
2025-04-17 22:02:35 <b00jum> someone involved in atrix liunx forked it and - apparently - mad it work but i haven't try it yet.
2025-04-17 22:03:20 <b00jum> so just to confirm: you installed openrc-user, and run openrc --user from whatever is starting your desktop env?
2025-04-17 22:05:17 <rhizoome> b00jum: exactly. also in the pam config of alpine pam-rundir gets executed on a login. So I assume if you start your services earlier, they cannot have / know the rundir.
2025-04-17 22:06:59 <rhizoome> b00jum: I think, if you really need to start the services before you login, you have to disable pam-rundir. It probably will remove the rundir, the services are already using when you login.
2025-04-17 22:07:55 <rhizoome> b00jum: I guess you need to switch to mkrundir and run it before the services are started
2025-04-17 22:08:39 <b00jum> i very much think you are on the ball here! removing pam_rundir i mean. i'll test with a plain openrc service creating / removing the required dirs.
2025-04-17 22:09:09 <b00jum> ooo! mkrundir! that i haven't came across. thanks for the tip, i shall investigate!
2025-04-18 03:16:25 <PhillipBeliever[m]>  Hello @everyone want to introduce y'all to our RIMS & TIRES supplier.... (full message at <https://matrix.org/oftc/media/v1/media/download/AVnPjpFm5xAvwaWdrcdbKn2bhciK_lvPrEpSSpTfKVM5_z_JlX8kFE1wYmXGJNtEL6tYP_iz7IUH94JvUbgqY-xCeWjv7G5gAG1hdHJpeC5vcmcvVGZ2TndiTXFMUmxJWVRMdUJmUmR6cmlX>)
2025-04-18 09:19:24 <f_> can anyone ban this spambot? ^
2025-04-18 15:15:53 <b00jum> rhizoome: for what it's worth - i got the user services working at boot just fine. pam is needed (pam_rundir.so), the issue was busybox. now my rpi cluster has rootless superpowers ;-)
2025-04-18 16:23:39 <rhizoome> b00jum: great! 
2025-04-18 16:24:26 <rhizoome> b00jum: with rootless you mean: you can add stuff without needing to doas?
2025-04-18 16:39:48 <b00jum> rhizoome: exactly. my particular use-case here was rootles podman run containers managed as openrc user services. as in, the containers and services doesn't require root and still starts at boot regardless if the user is logged in or not, headless servers.
2025-04-18 16:41:09 <rhizoome> b00jum: very nice. is that starting at boot part of alpine oder did you add your own openrc script? if it part of alpine is there a wiki entry?
2025-04-18 16:41:58 <rhizoome> s/oder/or/ ... mixing in german :-)
2025-04-18 16:45:53 <b00jum> rhizoome: in a nutshell: make sure you're using a full shell, busybox applets won't cut it; enable your user service at boot (https://wiki.gentoo.org/wiki/OpenRC#At_boot_auto-start); install alpine's openrc-user (no setup required); and ensure that you are using something different that busybox login (https://wiki.alpinelinux.org/wiki/PAM)
2025-04-18 16:47:46 <b00jum> rhizoome: if you want to run rootless containers like this (at least via podman with pasta) you need to also rewrite net-online system service as a user service (user service can't depend on system services)
2025-04-18 16:48:43 <rhizoome> b00jum: oh, am sorry. I would have known that one has to install util-linux-login. I did not make the connection.
2025-04-18 16:50:18 <b00jum> rhizoome: yeah, a few moving parts but in the end it wasn't really difficult to set up.
2025-04-18 16:51:03 <b00jum> rhizoome: if this is something you want to have a look at i could potentially write it up and stick it in a gist somewhere?
2025-04-18 16:51:54 <b00jum> rhizoome: although not today as there some things i need to cater to.
2025-04-18 16:54:51 <rhizoome> b00jum: I currently don't need it, all my podman stuff is hacked with a custom supervisor. I will use it the next time a need to build a server. but the people who are pushing openrc user services in edge (I don't who it) would probably like a extension here: https://wiki.alpinelinux.org/wiki/OpenRC#User_services
2025-04-18 16:57:11 <b00jum> rhizoome: good point! i'll see to it over the next few days. also, custom supervisor? now, that sound sexy! would you mind to elaborate?
2025-04-18 17:12:42 <rhizoome> b00jum: well a custom supervisor was an overstatment, just "sudo -u user podman compose up -d" calls from openrc and let compose do the supervision. so if I change something I will do podman compose up and if I reboot openrc will do it. I am very lazy...
2025-04-18 21:28:09 <nilfsuser5678> I can't boot any alpine-uboot tarball newer than 3.18 on an odroid c4 because of config_efi_zboot. I've tried looking at the u-boot apkbuild to see how it works for the supported boards and haven't found anything there. 
2025-04-18 21:29:47 <nilfsuser5678> Has anyone here dealt with the config_efi_zboot and 'bad linux arm64 image magic' error from extlinux?
2025-04-18 22:02:33 <b00jum> it's been a while i played with odroid but i seem to recall `bad linux arm64 image magic`. if memory serves me right, alpine kernel is configured with CONFIG_EFI_ZBOOT=y, which older versions o u-boot does not support.
2025-04-18 22:09:33 <nilfsuser5678> b00jum: I'm using the latest release of u-boot from the amlogic custodian tree (2025.01). I'll try using the mainline 2025.04 release.
2025-04-18 22:16:04 <b00jum> nilfsuser5678: what's your boot command? i think booti only supports flat / Image, to go with zboot you need to use bootefi.
2025-04-18 22:16:47 <b00jum> nilfsuser5678: all this is from memory so take it with a grain of salt ;-)