2023-06-01 02:32:47 <LACampbell> alpine linux successfully installed on my main PC. so far so good, just gotta sort out flatpak and pulsewire
2023-06-01 02:38:16 <Saijin_Naib> flatpak should be very painless, just follow the instructions on the flatpak site for Alpine
2023-06-01 02:38:30 <Saijin_Naib> As for pipewire, apk add wireplumber, which will pull in the pipewire things you need
2023-06-01 02:38:36 <LACampbell> flatpak is proving a bit painful to me. 
2023-06-01 02:38:41 <Saijin_Naib> Depending upon your environment/DE, you might want to stick your user in the audio group as well
2023-06-01 02:38:47 <Saijin_Naib> LACampbell: That's new. what's going on?
2023-06-01 02:38:49 <LACampbell> followed the instructions on the wiki, but get:
2023-06-01 02:38:51 <LACampbell> flatpak remote-add --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo
2023-06-01 02:38:52 <LACampbell> sorry
2023-06-01 02:39:01 <LACampbell> $ flatpak search signal
2023-06-01 02:39:03 <LACampbell> error: While opening repository /var/lib/flatpak/repo: opening repo: opendir(/var/lib/flatpak/repo): No such file or directoryy
2023-06-01 02:39:11 <Saijin_Naib> https://flatpak.org/setup/Alpine
2023-06-01 02:39:24 <LACampbell> yeap that's the one I followed
2023-06-01 02:39:27 <Saijin_Naib> I'd follow these instructions, make sure your system is up to date, and give a reboot before you proceed further
2023-06-01 02:40:13 <LACampbell> hmmm is it because I installed it as root? I deleted it, re-installed it as a normal user, and I get a different error when remote-adding. that's promising
2023-06-01 02:47:03 <LACampbell> ah added dbus and polit. now I get
2023-06-01 02:47:29 <LACampbell> $ flatpak remote-add --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo
2023-06-01 02:47:32 <LACampbell> error: Flatpak system operation EnsureRepo not allowed for use
2023-06-01 02:47:47 <LACampbell> user is part of the flatpak group
2023-06-01 02:49:30 <LACampbell> ok the alpine wiki had the "--user" flag, while the flatpak page didn't. works when I added that. back to error while searching though. one more restart
2023-06-01 02:51:03 <LACampbell> same error! I'm stumped. sorry for the wall of text
2023-06-01 02:58:10 <psykose> are you even in a dbus session
2023-06-01 02:58:40 <LACampbell> pretty sure I am. how might i tel?
2023-06-01 02:58:53 <LACampbell> it's running in rc-status
2023-06-01 03:05:27 <psykose> that's not the user one
2023-06-01 03:05:35 <psykose> echo $DBUS_SESSION_BUS_ADDRESS
2023-06-01 03:08:05 <LACampbell> unix:path=/tmp/dbus-HPCFHiURjR,guid=915aec3f8c73dc4401b123cc64780afc
2023-06-01 03:08:28 <LACampbell> also, I seem to be able to search and install with alpine if I give every command a --user flag. does everyone else have to do this?
2023-06-01 03:10:23 <LACampbell> can install with the "--user" flags, but can't run anything. damnit
2023-06-01 03:12:20 <psykose> you have a dbus user session then since it's set
2023-06-01 03:12:44 <psykose> you probably broke /var/lib/flatpak somehow
2023-06-01 03:13:06 <psykose> there's various amounts of deleting things in it and `flatpak repair` to run until it works
2023-06-01 03:14:04 <LACampbell> deleting what? I've deleted it and re-installed it from apk pages like 5 times. also tried repair
2023-06-01 03:14:27 <psykose> if i `rm -r /var/lib/flatpak` i get the same error as you
2023-06-01 03:14:33 <psykose> then if i run `doas flatpak repair` after it works again
2023-06-01 03:14:45 <LACampbell> thanks, let me try
2023-06-01 03:15:18 <LACampbell> my god, that worked
2023-06-01 03:15:21 <LACampbell> thank you so much
2023-06-01 03:15:27 <LACampbell> I can run flatpak apps
2023-06-01 03:15:44 <psykose> now iirc it implies --user as long as you never run it as root again
2023-06-01 03:15:51 <LACampbell> yeah don't need the --user flag
2023-06-01 03:15:54 <psykose> so just `flatpak install/update` will work with --user etc
2023-06-01 03:15:59 <psykose> now you can forget about it i guess
2023-06-01 03:18:47 <maria> iirc i had the same issue ootb when i installed flatpak. is this something that can be fixed in the package itself?
2023-06-01 03:20:11 <psykose> it's not a package issue
2023-06-01 03:21:12 <psykose> i never run into it on new installs
2023-06-01 03:21:32 <psykose> every time it happens it's a new person that installs it, runs whatever commands that happen to break it into system mode, then they have to undo that
2023-06-01 03:21:57 <psykose> could be wrong however
2023-06-01 03:22:01 <maria> i'll see if i can reproduce it in a VM
2023-06-01 03:22:17 <LACampbell> yeah like I said before the setup on the flatpak site is subtly different than the setup on the alpine linux wiki
2023-06-01 03:22:23 <LACampbell> look the same at first glance
2023-06-01 03:22:34 <psykose> it would be helpful to have the full shell history of any reproduction
2023-06-01 03:22:36 <maria> it worked without issue on my kde install using discover btw
2023-06-01 03:23:04 <LACampbell> https://wiki.alpinelinux.org/wiki/PipeWire#Running curious why the wiki recommends running pipewire this way, and not as an openrc service?
2023-06-01 03:23:58 <psykose> you can't generally run audio daemons as system services
2023-06-01 03:24:06 <psykose> well, you 'can', every project tells you not to tho
2023-06-01 03:24:35 <LACampbell> huh. that's how I had them set up on my old install, another distro
2023-06-01 03:24:45 <LACampbell> good to know though
2023-06-01 03:24:48 <psykose> first thing i notice on this flatpak article is this flatpak group thing that you don't have to be in
2023-06-01 03:26:47 <LACampbell> can you remove a user from a group in stock alpine? or should I just grab groupdel
2023-06-01 03:27:44 <maria> setup-desktop for plasma doesn't start polkit and elogind which the article for manually setting it up recommends. not having elogind on startup prevents the sddm shutdown and reboot buttons from being used. not sure if this is intended; should i open a bug report?
2023-06-01 03:27:49 <psykose> that wouldn't be any different than editing the text file by hand
2023-06-01 03:28:26 <psykose> maria: they are started automatically afaik
2023-06-01 03:28:48 <psykose> but regardless it's all optional
2023-06-01 03:29:23 <maria> the script only adds dbus and sddm to startup for plasma. for gnome it does add elogind.
2023-06-01 03:32:31 <psykose> what i mean is it's more involved than that and adding elogind doesn't actually do anything iirc, it gets started by itself via pam module from login
2023-06-01 03:32:37 <psykose> could be wrong but that's how it worked the last time i tried
2023-06-01 03:34:01 <maria> should my tests for this stuff in a VM be edge or 3.18?
2023-06-01 03:34:22 <psykose> probably edge
2023-06-01 03:34:31 <psykose> don't think it matters for this
2023-06-01 03:43:53 <LACampbell> how am I supposed to set my $XDG_RUNTIME_DIR? If I am readign this right I need it for pulse audio
2023-06-01 03:43:59 <LACampbell> wiki says elogind will do it for me but no luck
2023-06-01 03:44:04 <LACampbell> maybe I'll just do it manually
2023-06-01 03:46:28 <psykose> elogind does normally do it
2023-06-01 03:46:38 <psykose> ah
2023-06-01 03:46:43 <psykose> apk add util-linux-login
2023-06-01 03:46:44 <psykose> then relog
2023-06-01 03:52:32 <LACampbell> do I need elogind? I ended up deleting it and specifying it manually
2023-06-01 03:54:05 <LACampbell> that having been said I am completely failing to get pipewire working. added "exec /usr/libexec/pipewire-launcher" to my .xinitrc, and "$XDG_RUNTIME_DIR" is set. but "wpctl status" says it can't connect to pipewire
2023-06-01 03:54:57 <psykose> what does `ps aux | grep wire` output
2023-06-01 03:55:41 <LACampbell> I've got /usr/bin/pipewire, /usr/bin/wireplumber, and /usr/bin/pipewire-pulse
2023-06-01 03:56:55 <LACampbell> ok interestingly if I start /usrlib/exec/pipewire-launcher manually ion the command line stuff seems to work
2023-06-01 03:57:08 <psykose> what is your xinitrc
2023-06-01 03:57:32 <psykose> i can already guess what the issue is but i can show it directly in that
2023-06-01 03:58:00 <LACampbell> ... one moment I need a system clipboard
2023-06-01 03:58:58 <LACampbell> https://gist.github.com/LAC-Job-Interview/37a370a8c78dac15ae2c7a90653f3461
2023-06-01 03:59:13 <psykose> well
2023-06-01 03:59:23 <psykose> after you exec dbus-run-session openbox-session
2023-06-01 03:59:25 <psykose> the script ends
2023-06-01 03:59:26 <LACampbell> hey look I posted the coords of where I live. you're all invited to dinner
2023-06-01 03:59:32 <LACampbell> oh
2023-06-01 03:59:43 <psykose> you can't move that up either
2023-06-01 03:59:52 <psykose> all the pipewire stuff needs the dbus user session you get from the exec line
2023-06-01 03:59:59 <psykose> you have to run it after from inside an openbox script or w/e
2023-06-01 04:00:15 <LACampbell> aha, I'll add it to my autostart
2023-06-01 04:00:17 <psykose> it probably has some mechanism for 'run something after launch'
2023-06-01 04:00:17 <psykose> yeah
2023-06-01 04:00:32 <psykose> as for elogind/the xdg thing, you don't explicitly need it no
2023-06-01 04:00:46 <psykose> but you shouldn't manually create an xdg dir because that breaks for people pretty easily
2023-06-01 04:01:14 <psykose> easiest way to get an xdg dir made for you is probably `apk add util-linux-login turnstile; rc-update add turnstiled`, then relog
2023-06-01 04:01:41 <LACampbell> oh I'll need elogind for loginctl. so I will use it
2023-06-01 04:01:44 <psykose> well, also rc-service turnstiled start if not rebooting
2023-06-01 04:01:44 <psykose> sure
2023-06-01 04:02:03 <psykose> the reason you didn't have one is because you didn't have the pam-enabled login(1) so it didn't give you one into the session
2023-06-01 04:02:13 <psykose> default login from bb just logs you in without integrating into anything
2023-06-01 04:04:33 <maria> huh flatpak doesn't have the issue ootb now. i wonder what caused it in the past? it's been a year or so i think and back then i needed to manually create a directory iirc.
2023-06-01 04:08:01 <LACampbell> maria: I thought it might have been an old config of mine, since this home dir was from anotehr linux install. but there was no .config/flatpak
2023-06-01 04:08:33 <LACampbell> psykose: appreciate the help. flatpak apps seem to be picking up mic and stuff now. 
2023-06-01 04:08:52 <psykose> glad it works :)
2023-06-01 04:09:12 <LACampbell> I've run alpine on my server and WSL2 for years. cool to have it on the desktop. all matching
2023-06-01 04:09:14 <psykose> maria: iirc they did fix some stuff in the past few releases over a year or so, though there's still certain ways to break it
2023-06-01 04:09:31 <psykose> it would be nice to get a full shell history when someone does break it (like earlier) but it's impossible to find out otherwise
2023-06-01 04:10:02 <psykose> in my experience it was fixed like a year ago for the easy common case of just adding a repo and using it as a user
2023-06-01 04:40:38 <maria> looks like elogind is started on login it's just the sddm buttons aren't usable.
2023-06-01 04:40:56 <maria> trying to figure out what fixes the firefox CSD buttons now :)
2023-06-01 04:47:43 <maria> installing foliate fixes it so i assume it's one of the dependencies for it
2023-06-01 04:48:55 <maria> it's librsvg
2023-06-01 04:50:21 <psykose> you mean for firefox?
2023-06-01 04:50:24 <maria> yeah
2023-06-01 04:50:54 <psykose> sure, i guess i can add that
2023-06-01 04:51:06 <psykose> next time i upgrade it
2023-06-01 04:54:52 <psykose> hmm
2023-06-01 04:55:03 <psykose> though something makes me think that's not a dep of firefox specifically in this case
2023-06-01 04:55:16 <psykose> which csds are you referring to
2023-06-01 04:57:14 <maria> the close maximise and minimise buttons on the right. by default on plasma at least firefox launches with csd where the tab bar is combined with the window controls instead of ssd. i can take a screenshot with and without the dependency installed to show the difference.
2023-06-01 04:59:45 <maria> https://i.imgur.com/w7iQT0y.png
2023-06-01 04:59:50 <maria> https://i.imgur.com/p7XvdJt.png
2023-06-01 05:02:17 <maria> when launched without the dependency and with MOZ_ENABLE_WAYLAND=1 there's output in the terminal of a gtk warning and being unable to load pixbuf from icon theme. it's weird that that terminal output doesn't appear without MOZ_ENABLE_WAYLAND=1 despite the issue still being present.
2023-06-01 05:17:59 <psykose> yeah, that's not a firefox thing afaict
2023-06-01 05:18:18 <psykose> it's firefox drawing gtk decorations from svg icons, and then gdk-pixbuf requires the svg loader from librsvg
2023-06-01 05:19:00 <psykose> strictly speaking it's a "nothing" dependency, just something you get as part of a complete desktop, but not something you generally specify everywhere all over the place (because this is one of dozens of such things)
2023-06-01 05:20:37 <psykose> i'll still probably add it next upgrade, but everyone would've had librsvg already by installing like anything else as part of a complete desktop
2023-06-01 05:24:07 <maria> okie. sorry for all the noise over so little aha ^^
2023-06-01 05:29:49 <psykose> np :)
2023-06-01 07:11:39 <kjell> Hi. It seems to me that mongodb sever was removed from Alpine Linux. Is there an easier way to install mongodb than compiling it on the system which I guess still would work?
2023-06-01 07:12:51 <kjell> I've seen a guide that adds v3.6 repositories, but is it smart to do that on a 3.18 system?
2023-06-01 07:36:20 <psykose> no
2023-06-01 07:53:11 <pj> Any guide that says "add vX.X repo" and your Alpine version is != vX.X, is not smart
2023-06-01 08:03:01 <Habbie> in general most guides are not smart
2023-06-01 09:26:15 <munchausen> Helloo. I am using an ARM soc that is an A7 with VFPv4 and NEON. Am I right in thinking the armv7 images are compiled with hard float support? Or is it only available in the armhf images? 
2023-06-01 09:27:11 <munchausen> I was looking for a "generic arm" image for armhf, and wondered if that is implied for the armv7 target
2023-06-01 09:30:13 <psykose> armv7 is armv7
2023-06-01 09:30:16 <psykose> armhf is armv6
2023-06-01 09:30:49 <munchausen> Ahh
2023-06-01 09:30:52 <munchausen> Thanks
2023-06-01 09:35:11 <psykose> specifically armv7 is vfpv3-d16 hf and armhf is vfpv2 hf
2023-06-01 09:44:55 <Someone[m]1> Hi, I wanted to use and test the hardened_malloc package system-wide. On the GitHub of the project, they instructed to use /etc/ld.so.preload:
2023-06-01 09:44:55 <Someone[m]1> I tried and do lsof (not busybox lsof) but it seems that the library isn't used. What's the solution?
2023-06-01 09:44:55 <Someone[m]1> > On traditional Linux-based operating systems, hardened_malloc can either be integrated into the libc implementation as a replacement for the standard malloc implementation or loaded as a dynamic library. Rather than rebuilding each executable to be linked against it, it can be added as a preloaded library to /etc/ld.so.preload.
2023-06-01 09:45:37 <psykose> musl doesn't support any global ld.whatever thing
2023-06-01 09:45:57 <psykose> you'd have to LD_PRELOAD it via normal environment using your imagination for how to do that for what you want to run
2023-06-01 09:54:56 <Someone[m]1> also I've found this on dustri.org
2023-06-01 09:54:56 <Someone[m]1> > Note that musl doesn't use /etc/ld.so.conf but /etc/ld-musl-$(ARCH).path instead. You can check that's effectively being used after a reboot by running lsof /usr/lib/libhardened_malloc.so.
2023-06-01 09:54:56 <Someone[m]1> How to use that?
2023-06-01 10:01:54 <psykose> i don't think that does what you want
2023-06-01 10:02:12 <psykose> that's basically an ld_library_path tunable and not something that preloads libraries
2023-06-01 10:18:50 <munchausen> psykose, thanks, I was confused by the presence of a 32 bit HF and a 32 bit *other* but I see now the distinction is only to make clear you dont have a v6 soft float option
2023-06-01 10:19:17 <psykose> i would've personally named them v6hf v7hf or something :p
2023-06-01 10:19:55 <ikke> something something history
2023-06-01 10:25:59 <psykose> armhf was added quite recently for alpine ;p
2023-06-01 10:26:14 <psykose> and i don't think anyone else uses armhf to mean v6hf, do they
2023-06-01 10:26:17 <psykose> debian is v7hf
2023-06-01 12:46:30 <invoked> raspbian did something different/stupid in that regard i can't remember.
2023-06-01 13:05:54 <Habbie> invoked, the pi1 could run debian armel, but that would ignore the hf (hardware floating point); debian armhf meanwhile wasn't compatible with pi1 because other things were missing
2023-06-01 13:06:04 <Habbie> invoked, so they invented their own thing they also called armhf that sat somewhere in between
2023-06-01 13:06:13 <Habbie> invoked, i stand by my vision that they should have called it armpi
2023-06-01 13:07:04 <psykose> iirc debian armhf is v7hf and pi1 is v6hf
2023-06-01 13:07:13 <psykose> so that ofc doesn't work
2023-06-01 13:07:22 <psykose> and alpine armhf is like that 'own thing' in this case since it's also armhf
2023-06-01 13:07:27 <psykose> but i didn't check exactly what raspbian is
2023-06-01 13:09:26 <Habbie> pi1 uname is v5
2023-06-01 13:12:27 <psykose> you mean the kernel they shipped? why? the actual hardware is v6
2023-06-01 13:12:46 <psykose> debian armel is v5 but if they targetted rpi1 what was the point
2023-06-01 13:21:25 <Habbie> i'm 99% sure
2023-06-01 13:21:32 <Habbie> (but not 100%)
2023-06-01 15:02:14 <msiism> minimal: I've found the reason my /etc/default/grub was already fully equipped. It was a left-over from previous installation attempts.
2023-06-01 15:04:50 <minimal> msiism: that's what I suspected
2023-06-01 15:11:03 <msiism> Hm… now grub-install fails with: "error: /usr/lib/grub/modinfo.sh doesn't exist. Please specify --target or --directory."
2023-06-01 15:11:25 <msiism> I did run `grub-install --target=i386-pc /dev/sda`, though.
2023-06-01 15:13:20 <psykose> normally it checks /usr/lib/grub/$target/modinfo.sh
2023-06-01 15:14:11 <msiism>  /usr/lib/grub is empty.
2023-06-01 15:14:14 <minimal> msiism: is the grub-bios package installed?
2023-06-01 15:14:24 <msiism> I just installed grub.
2023-06-01 15:14:45 <minimal> well you're not going to get far without grub-bios package also... ;-)
2023-06-01 15:14:50 <msiism> I mean, it worked yesterday…
2023-06-01 15:15:12 <msiism> But maybe for strange reasons…
2023-06-01 15:15:30 <minimal> so then yesterday grub-bios must have been installed
2023-06-01 15:15:51 <msiism> Well, I didn't install it manually, at least.
2023-06-01 15:16:04 <msiism> But yeah, /usr/lib/grub is now not empty anymore.
2023-06-01 15:16:25 <minimal> but yesterday you ran setup-disk at some time?
2023-06-01 15:16:41 <msiism> I did.
2023-06-01 15:16:51 <minimal> so that likely installed grub-bios
2023-06-01 15:16:53 <msiism> And before installing GRUB.
2023-06-01 15:17:02 <msiism> Hm…
2023-06-01 15:18:05 <minimal> it's going to be hard to document a working set of steps if you don't start from a "blank" machine (i.e. VM) and repeat the same steps in the same sequence...
2023-06-01 15:18:50 <msiism> Agreed.
2023-06-01 15:20:02 <msiism> Okay, GRUB_PRELOAD_MODULES is apparently not needed.
2023-06-01 15:20:51 <msiism> Installation successful.
2023-06-01 15:22:59 <msiism> By the way: I noticed that setup-user doesn't create the user's home directory.
2023-06-01 15:25:15 <psykose> it does? it calls adduser -f .. someuser
2023-06-01 15:25:21 <psykose> that creates a /home/someuse
2023-06-01 15:25:28 <psykose> -g *
2023-06-01 15:27:22 <minimal> setup-user calls "adduser" which should create a home directory unless the "-H" option is specified (which setup-user doesn't)
2023-06-01 15:27:31 <msiism> Okay, I'll try this again as part of the next iteration.
2023-06-01 15:31:01 <msiism> Oh, I think I ran setup-user before the chroot.
2023-06-01 15:31:06 <msiism> That could be the reason.
2023-06-01 15:52:45 <msiism> Okay, I now have a version that works all the time. Just gotta smooth out some rough edges.
2023-06-01 17:18:34 <invoked> Habbie: yeah, it was that namespace collision. thanks, i couldn't remember. it would have made more sense for it to be called 'armpi' or something, at least in 2012.
2023-06-01 17:18:48 <Habbie> yes exactly
2023-06-01 17:18:50 <Habbie> also in 2023
2023-06-01 17:20:14 <invoked> i just remember the confusion it created. people trying to use debian armhf on pi, et al
2023-06-01 17:20:26 <Habbie> worse, people trying to use a few debian armhf packages on pi
2023-06-01 17:20:33 <invoked> right
2023-06-01 17:20:41 <Habbie> i've had to explain this many times
2023-06-01 17:21:02 <invoked> lol, i also remember debian people completely disowning raspi having any similarity to debian
2023-06-01 17:21:26 <Habbie> well, #debian has had plenty of pain trying to support various forks
2023-06-01 17:21:34 <Habbie> and raspbian/raspios is -just- sufficiently different
2023-06-01 17:21:37 <Habbie> like the dhcpcd madness
2023-06-01 17:21:48 <invoked> i don't know how raspbian is now, but it was a mess in, whatever, 2012
2023-06-01 17:22:07 <Habbie> raspbian is dead; raspberry pi os lives
2023-06-01 17:22:17 <Habbie> raspberry pi os 64 bit literally points to debian mirrors for 99% of packages
2023-06-01 17:22:26 <invoked> it just struck me as made by people who never worked on operating systems before
2023-06-01 17:22:32 <Habbie> 32 bit is still mostly that same armhf-should-have-been-armpi thing
2023-06-01 17:22:38 <Habbie> plus modern-ish kernels if you run on a newer pi
2023-06-01 17:25:50 <invoked> i think the raspberry pi org was doing all that stuff for the first time ever
2023-06-01 17:25:57 <invoked> no prior experience
2023-06-01 17:26:10 <invoked> it was like broadcom made this thing for them, now go figure out the rest of it
2023-06-01 17:30:04 <frag> do any laptop have good speakers? mobile phones are quite good right ...
2023-06-01 17:57:50 <Habbie> invoked, yeah
2023-06-01 18:17:18 <scorpion2185[m]1> I cannot compile ikemen-go on the PinePhone
2023-06-01 18:18:07 <Ermine> My condolences
2023-06-01 18:18:22 <scorpion2185[m]1> https://paste.centos.org/view/97ade223
2023-06-01 18:19:34 <scorpion2185[m]1> i guess that the compiled version is 64bit
2023-06-01 18:24:10 <Ermine> Also, how does it relate to Alpine? Isn't it an upstream issue?
2023-06-01 18:26:22 <scorpion2185[m]1> it may be a problem when you are on Alpine
2023-06-01 18:27:29 <scorpion2185[m]1> like xgifwallpaper maybe IDK, it wasn't building on alpine cause some var didn't work 
2023-06-01 18:27:29 <scorpion2185[m]1> I tried on pmso that's why I asked here
2023-06-01 18:28:50 <Ermine> Did you install all dependencies of ikemen listed in their build instructions?
2023-06-01 18:29:12 <scorpion2185[m]1> i think so , it's a list for debian so the names are different
2023-06-01 18:32:53 <Ermine> Did you install mesa-dev? It has definition for GLboolean and other types
2023-06-01 18:33:27 <scorpion2185[m]1> yes i got some missing header or something before
2023-06-01 18:36:01 <scorpion2185[m]1> git libxcursor-dev libxrandr-dev pkgconf alsa-lib-dev xorg-dev mesa-dev gtk+3.0-dev go
2023-06-01 18:37:08 <psykose> they have a 150 line build.sh to run a single go build command
2023-06-01 18:37:16 <psykose> man who writes this shit
2023-06-01 18:40:00 <Ermine> that gl-js package should have included GLES/gl.h in preamble in types_opengles.go . So, upstream issue
2023-06-01 18:41:39 <Ermine> (I'm not well-versed in go though. Please correct me if I'm wrong)
2023-06-01 18:44:02 <psykose> https://img.ayaya.dev/xsu9bB0SfEEI
2023-06-01 18:44:03 <psykose> well that was easy
2023-06-01 18:44:59 <psykose> go build -trimpath -v -tags=gles2 -o Ikemen_GO_Linux ./src/
2023-06-01 18:44:59 <psykose>  mesa-dev libxrandr-dev libxcursor-dev glfw-dev libxi-dev gtk+3.0-dev alsa-lib-dev
2023-06-01 18:45:14 <psykose> then i used the data files from the release tarball since it's missing in the repo for some reason
2023-06-01 18:48:18 <Ermine> Woah
2023-06-01 18:51:33 <txnintn> Hi all tying to do a Sys install, lvm over raid, but I can't seen to keep all the setting during a reboot, what am I missing?
2023-06-01 18:52:19 <ikke> does setup-disk finish succesfully?
2023-06-01 18:52:39 <minimal> txnintn: what settings? Also if you're reboot then you've done the install so how are you "trying to do a Sys install" and failing?
2023-06-01 18:55:03 <txnintn> I have to jump around a little, I setup the raid and lvm, the do the install.
2023-06-01 18:57:23 <txnintn> on a RPI4 using 2 usb drives and booting from a sd card.
2023-06-01 18:57:35 <minimal> can't you specify "USE_RAID=1 USE_LVM=1" to setup-disk for ti to do that instead?
2023-06-01 18:58:04 <minimal> ah, don't think setup-disk is going to handle that for a RPI
2023-06-01 18:58:50 <txnintn> I thing it's disk -r does raid but you have to add the lvm peice.
2023-06-01 19:22:42 <Newbyte> How do I downgrade one particular package? I feel like I asked before, but I don't remember the answer
2023-06-01 19:22:56 <Newbyte> assuming I installed a newer version myself and I want to downgrade to what's in the repo
2023-06-01 19:24:02 <ikke> Newbyte: apk add package=<version>
2023-06-01 19:24:21 <Newbyte> ikke: I tried, but it complains
2023-06-01 19:24:39 <Newbyte> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/OMFhGVPPtSLGAhasMecQkbsi

2023-06-01 19:24:57 <ikke> include the revision
2023-06-01 19:26:20 <Newbyte> thanks, that worked
2023-06-01 19:28:38 <psykose> what is this _p123412523523 thing pma adds
2023-06-01 19:33:30 <scorpion2185[m]1> psykose data files?
2023-06-01 19:34:00 <scorpion2185[m]1> i am now building with your command
2023-06-01 19:37:57 <scorpion2185[m]1> what tarball? latest is empty
2023-06-01 19:38:45 <psykose> https://github.com/ikemen-engine/Ikemen-GO/releases/tag/v0.99.0-rc.1
2023-06-01 19:40:07 <Newbyte> <psykose> "what is this _p123412523523..." <- it's a hack to ensure it's newer than your previous build when you build using a local directory
2023-06-01 19:40:26 <scorpion2185[m]1> why is it empty here?
2023-06-01 19:40:34 <psykose> why is what empty
2023-06-01 19:40:37 <Newbyte> if you look closely it's a date
2023-06-01 19:40:44 <scorpion2185[m]1> the tarball
2023-06-01 19:40:51 <psykose> the fuck are you downloading
2023-06-01 19:41:05 <Newbyte> obviously a tarball
2023-06-01 19:41:09 <psykose> i can tell you this is very much not empty https://img.ayaya.dev/6Ul6DONoR7yg
2023-06-01 19:41:15 <Newbyte> there ought to be some empty tarball out there
2023-06-01 19:41:29 <psykose> ah right
2023-06-01 19:41:35 <scorpion2185[m]1> that link on github web page it shows empty too
2023-06-01 19:41:47 <abby> they break the git-archive tarballs: https://github.com/ikemen-engine/Ikemen-GO/blob/v0.99.0-rc.1/.gitattributes#L1
2023-06-01 19:42:30 <psykose> apologies, i meant the 'zip file'
2023-06-01 19:42:35 <psykose> evidently this was impossible to figure out
2023-06-01 19:44:16 <scorpion2185[m]1> data files = data/* ?
2023-06-01 19:44:48 <scorpion2185[m]1> break tarballs? what if you want the source of a release?
2023-06-01 19:45:05 <scorpion2185[m]1> i mean how can you fetch it? out of curiosity
2023-06-01 19:55:02 <scorpion2185[m]1> https://paste.centos.org/view/98677a8f
2023-06-01 19:58:02 <psykose> looks like it's just broken on arm
2023-06-01 19:58:16 <psykose> probably because it doesn't actually include the header
2023-06-01 19:58:25 <psykose> https://github.com/fyne-io/gl-js/blob/master/types_opengles.go#L18
2023-06-01 19:58:30 <psykose> they actually include nothing on linux
2023-06-01 19:59:06 <psykose> it's fixable if you fork the dep and make the changes yourself and change around some imports i guess
2023-06-01 20:03:30 <scorpion2185[m]1> what changes?
2023-06-01 20:05:24 <psykose> #include <GLES2/gl2.h>
2023-06-01 20:08:11 <scorpion2185[m]1> what package should I fork?
2023-06-01 20:08:18 <Ermine> They do not even have issues tracker active
2023-06-01 20:09:33 <psykose> i already linked the file in the repo?
2023-06-01 20:10:31 <scorpion2185[m]1> yes but I am not sure what ikemen-go dep it is
2023-06-01 20:10:45 <psykose> it's literally what i linked
2023-06-01 20:11:16 <scorpion2185[m]1> gl-js?
2023-06-01 20:12:03 <scorpion2185[m]1> IIRC that link is an alpine package also
2023-06-01 20:12:38 <psykose> i am legitimately confused how linking to a random .go file in a repo you have to edit, imported by the thing you are trying to build, can be confused for something else
2023-06-01 20:12:41 <psykose> you'll figure it out
2023-06-01 20:14:26 <psykose> to be exact, it's literally the line i linked
2023-06-01 20:14:39 <psykose> and it's imported from https://github.com/ikemen-engine/Ikemen-GO/blob/b48c80d9ae013c2d4ec15c26ce435ce927505ff9/src/render_gl.go#L11 and https://github.com/ikemen-engine/Ikemen-GO/blob/b48c80d9ae013c2d4ec15c26ce435ce927505ff9/src/system_glfw.go#L9
2023-06-01 20:15:29 <scorpion2185[m]1> ah it's a dep of go , I though a dep like  like glfw-dev
2023-06-01 20:16:47 <scorpion2185[m]1> thanks ok let's see
2023-06-01 20:18:09 <psykose> you can do like ifdef __linux__ iirc, or just #cgo linux CFLAGS: -Dos_linux and ifdef os_linux i guess
2023-06-01 20:18:12 <psykose> a million ways to get that to work
2023-06-01 20:19:36 <scorpion2185[m]1> is just adding that header in`https://github.com/fyne-io/gl-js/blob/master/types_opengles.go#L18` enough?
2023-06-01 20:21:51 <psykose> great opportunity to learn some programming and how all this works and find out :)
2023-06-01 21:14:20 <LACampbell> I've got elogind running, but 'loginctl hibernate' requires root permissions. I think i solved this in other distros by adding my user to the 'power' group, but no such group in alpine. what's the best approach here?
2023-06-01 21:20:05 <psykose> it's some random group i forget
2023-06-01 21:20:32 <LACampbell> really should have saved my groups before I wiped my root partition
2023-06-01 21:26:58 <Saijin_Naib> I think wheel should allow that 
2023-06-01 21:30:02 <LACampbell> definitely not wheel: uid=1000(lewis) gid=1000(lewis) groups=1000(lewis),5(tty),10(wheel),18(audio),23(input),27(video)
2023-06-01 21:30:20 <Saijin_Naib> Plugdev? 
2023-06-01 21:30:48 <Saijin_Naib> I am not sitting at my machines to id my user, but those come to mind
2023-06-01 21:30:58 <abby> plugdev is for pluggable devices generally, i doubt it would affect logind
2023-06-01 21:31:26 <psykose> having polkit running too?
2023-06-01 21:32:11 <LACampbell> polkit is in my rc-status
2023-06-01 21:33:31 <abby> do you have a polkit auth agent running
2023-06-01 21:34:37 <psykose> well then you better go catch it
2023-06-01 22:02:23 <LACampbell> psykose: that was a good dad joke
2023-06-01 22:02:36 <LACampbell> abby: thanks for the recommendation. i will find out what that is and get one up and running in that order
2023-06-01 22:02:42 <psykose> thanks mom\
2023-06-01 22:09:58 <LACampbell> lol
2023-06-01 22:10:25 <LACampbell> ok I think I am going to fix the issue of swap resetting itself on boot. hopefully whatever fixes that will have a positive impact on my logicntl situation
2023-06-01 22:22:55 <LACampbell> ok there was a swap openRC service. that seems to have solved that issue
2023-06-02 07:11:05 <uniqNuuid> apk search does not distinguish installed/not installed?
2023-06-02 07:16:08 <quinq> You have apk-list for that
2023-06-02 07:16:48 <uniqNuuid> i always think that apk list/search has overlap functionality 
2023-06-02 07:29:59 <quinq> Yeah, that's possible
2023-06-02 08:19:02 <jbg> is there any way to configure apk to always ignore a given package (such that it will never be installed even if depended on by another package)?
2023-06-02 08:19:10 <jbg> i.e. the equivalent of always passing --ignore
2023-06-02 08:23:06 <ptrc> jbg: `apk add !package`?
2023-06-02 08:26:34 <jbg> ah cool, seems like it should work. thank you
2023-06-02 08:38:47 <jbg> ptrc: unfortunately that doesn't seem to do the same thing as --ignore
2023-06-02 08:39:00 <jbg> i don't want this package to ever be installed, but i still want to install other packages that depend on it
2023-06-02 08:39:31 <ptrc> not sure what you mean by --ignore, but i suppose in this case you can add an empty virtual package: `apk add -t package`
2023-06-02 08:39:50 <ptrc> this way you do have a package in your system that satisfies the constraints
2023-06-02 10:20:24 <jbg> ptrc: thanks very much, works well
2023-06-02 11:56:17 <munchausen> Hello again.. Is there any way to build an alpine image with a sys install already on it? 
2023-06-02 12:33:41 <WhyNotHugo> psykose: Docker (and also kubernetes and systemd) only use cgroups2.
2023-06-02 12:33:54 <WhyNotHugo> But Alpine's default value for rc_cgroup_mode enables both cgroup1 and cgroup2.
2023-06-02 12:34:15 <WhyNotHugo> I disabled cgroup1 and nothing seems to break -- I don't know what software still relies on cgroup1 (if any).
2023-06-02 14:46:32 <eris> yo
2023-06-02 14:48:49 <lopid> pardon?
2023-06-02 15:14:59 <eris> anyone here know anything about flatpak and steam?
2023-06-02 15:18:11 <ikke> eris: it helps to ask more targeted questions
2023-06-02 15:18:22 <ikke> (I have not used flatpak, ftr)
2023-06-02 15:18:24 <eris> that's true :) 
2023-06-02 15:18:39 <eris> in particular, steam under flatpak doesn't have a functioning UI for me
2023-06-02 15:18:53 <eris> it will load, but the UI only updates whenever the window's size is changed 
2023-06-02 15:19:04 <eris> and otherwise just flickers between two frames 
2023-06-02 16:29:35 <Saijin_Naib> I had to disable Steam UI hardware acceleration to fix that on my i915. Wasn't an issue before the recent UI update 
2023-06-02 16:30:10 <Saijin_Naib> That aside, everything else works fine, including overlay once actually in-game
2023-06-02 16:44:23 <eris> where do i do that? 
2023-06-02 16:47:50 <eris> oh, found it 
2023-06-02 16:47:51 <eris> zero luck
2023-06-02 17:30:09 <witcher> I have a fresh Alpine install and need XDG_RUNTIME_DIR running so I installed elogind. To my understanding XDG_RUNGIME_DIR should just be set when i log in again (given that it's running) but this is not happening for me. What are some thing I could try to fix this?
2023-06-02 17:31:47 <witcher> s/thing/ways
2023-06-02 17:32:11 <ikke> You need ut1l-linux-login
2023-06-02 17:32:34 <ikke> busybox login does not use PAM, so there are no integrations
2023-06-02 17:32:40 <Newbyte> <eris> "it will load, but the UI only..." <- Wayland or X11?
2023-06-02 17:33:19 <witcher> ikke: I see... Thank you for the quick reply!
2023-06-02 17:41:26 <Saijin_Naib> <eris> "zero luck" <- What GPU? And did you mess with any Flatpak settings for it in FlatSeal? 
2023-06-02 17:54:53 <Newbyte> Saijin_Naib: kind of off-topic but your profile picture feels so accurate and I am not even sure why 😄
2023-06-02 18:23:44 <Saijin_Naib> 🤣
2023-06-02 18:23:52 <Saijin_Naib> I'm glad someone appreciates it 
2023-06-02 18:24:40 <Saijin_Naib> It's CC By Whatever, so yoink if you like 
2023-06-02 18:26:42 <Saijin_Naib> Brainworm from last week when someone joined and said that, more or less verbatim 
2023-06-02 19:10:04 <psykose> what's the picture
2023-06-02 19:10:12 <psykose> WhyNotHugo: ah, so it was cgroups1
2023-06-02 19:10:13 <psykose> makes sense
2023-06-02 19:10:20 <psykose> yeah they're both enabled by default for compatibility
2023-06-02 19:10:33 <psykose> you'd have to ask openrc upstream to change the default if you want, but it's pretty easy to make it 2-only
2023-06-02 19:13:34 <eris> ikke: X11
2023-06-02 19:13:51 <eris> Saijin_naib: haven't touched any settings, intel iGPU
2023-06-02 19:14:03 <eris> was also using intel iGPU on my last laptop without issues 
2023-06-02 19:14:59 <Saijin_Naib> Do you have the firmware and xf86 video packages for Intel, and nothing funky on grub boot cmd? 
2023-06-02 19:16:02 <Saijin_Naib> psykose: Steve Buscemi How do you do, fellow kids meme template, but instead it says How Do You Do Fellow Alpine Users 
2023-06-02 19:16:09 <psykose> lmao
2023-06-02 19:16:36 <eris> i have intel-media-driver, libva-intel-driver, linux-firmware-intel, and xf86-video-intel
2023-06-02 19:16:44 <psykose> you probably want to delete the last one
2023-06-02 19:19:23 <eris> xf86-video-intel?
2023-06-02 19:19:24 <eris> why?
2023-06-02 19:21:00 <Saijin_Naib> Not needed, I think, but my desktop has it and works with Steam
2023-06-02 19:21:15 <Saijin_Naib> What generation CPU? I'm using 6th, so quite old
2023-06-02 19:23:12 <eris> 12th gen
2023-06-02 19:23:23 <eris> i7-1280P
2023-06-02 19:23:24 <Saijin_Naib> 
ACTION uploaded an image: (120KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/ygweORyKkewKfUkAIvKbiqBa/image.png >

2023-06-02 19:23:42 <eris> hmmm
2023-06-02 19:23:48 <eris> weird, the UI didn't look like that for me 
2023-06-02 19:36:07 <cockliuser[m]> my kerning got messed up suddenly 😭
2023-06-02 19:36:21 <cockliuser[m]> everything's lumped up together
2023-06-02 20:01:08 <psykose> eris: because it's not what you actually want unless you have like gen4 or lower or whatever
2023-06-02 20:06:45 <eris> oh., fair
2023-06-02 20:08:35 <eris> rebooted, lets see if this works
2023-06-02 20:09:31 <eris> woo, fixed
2023-06-02 20:09:36 <eris> but now the screentearing is AWFUL
2023-06-02 20:09:37 <eris> lmao
2023-06-02 20:10:22 <eris> even with picom running
2023-06-02 20:16:58 <eris> ok, fixed screen tearing
2023-06-02 20:17:02 <eris> onto the games not launching 
2023-06-02 21:04:47 <irreleph4nt> Hi. I have a dockerfile which first installs openssl3 in addition to a small number of build tools, then in a later step installs openssl3-dev before starting a compilation for PHP ...
2023-06-02 21:06:00 <irreleph4nt> The step with openssl3-dev however fails with many errors like this: "ERROR: openssl3-dev<version> trying to overwrite file .... which is also owned by openssl-dev<version>"
2023-06-02 21:06:30 <irreleph4nt> How do I instruct apk to never install openssl, even when solving dependencies?
2023-06-02 21:06:36 <psykose> that's not your issue
2023-06-02 21:07:10 <psykose> your issue is you are trying to use openssl3-dev on like 3.16 or something where everything was still using openssl1.1, so everything else uses openssl-dev instead, and you can't mix the two with the other deps that are still on 1.1
2023-06-02 21:07:53 <irreleph4nt> welp, one of the dependencies I require to build php needs openssl3 ... I am on alpine 3.16
2023-06-02 21:08:02 <psykose> use 3.18 then
2023-06-02 21:08:29 <irreleph4nt> I can't. 3.18 ships a newer version of nginx which does not come with the lua / resty modules I require
2023-06-02 21:08:46 <psykose> then you simply won't be able to build what you're building because it would never work
2023-06-02 21:09:05 <psykose> everything else is built against 1.1 and if you managed to even build something against 3 it would just crash at runtime
2023-06-02 21:09:22 <irreleph4nt> that's unfortunate
2023-06-02 21:09:32 <irreleph4nt> thank you for letting me know, though :)
2023-06-02 21:11:10 <psykose> 3.17 and newer has everything built against 3 instead
2023-06-02 21:15:17 <irreleph4nt> I'll look into the nginx modules and if feasable, upgrade to a newer alpine release ...
2023-06-02 21:16:11 <irreleph4nt> On a smiliar note: As an APKINDEX is constructed from all apk files in a directory, is it possible for an APKINDEX to contain multiple records for the same package when said package exists in multiple versions?
2023-06-02 21:17:47 <irreleph4nt> I am looking into setting up an alpine mirror for the department I work at. The "official" alpine mirror my company runs is broken as APKINDEX files are never refreshed, even if new packages get synced from upstream
2023-06-02 21:22:12 <ptrc> irreleph4nt: yes, if you run something like `apk index *.apk -o APKINDEX.tar.gz`, you can get multiple versions of the same package: https://tpaste.us/1Xa9
2023-06-02 21:25:40 <irreleph4nt> ptrc, thanks for sharing! So I should be able to control the version to install with apk add pkg>=version, correct?
2023-06-02 21:26:10 <ptrc> yes, you can use version constraints to select versions
2023-06-02 22:26:58 <irreleph4nt> Hi. Is there any documentation available to understand musl vs glibc for those new to Alpine? Compiling some PHP extensions appears to require glibc which I am not sure has a correct way for me to provide it.
2023-06-02 22:27:41 <Habbie> musl can mostly fill the role of glibc
2023-06-02 22:27:45 <Habbie> what trouble are you having?
2023-06-02 22:29:15 <psykose> you can't change any of that or 'provide glibc'
2023-06-02 22:29:41 <irreleph4nt> Habbie, I have taken the official php-on-alpine Dockerfile (https://github.com/docker-library/php/blob/master/8.1/alpine3.16/fpm/Dockerfile) and added a number of PHP extensions ot it which I need at work ...
2023-06-02 22:30:11 <psykose> and then
2023-06-02 22:30:37 <irreleph4nt> At the end of compiling (which works) they run a scan for runtime dependencies to then install any missing packages (see here: https://github.com/docker-library/php/blob/master/8.1/alpine3.16/fpm/Dockerfile#L188-L194)
2023-06-02 22:30:53 <psykose> and then
2023-06-02 22:32:07 <irreleph4nt> This breaks for me where lib/libresolv.so.2 can't be satisfied (found a recommended fix to use gcompat). Installing gcompat however fails as I need glibc installed as well, which has conflicting files with gcompat
2023-06-02 22:32:24 <psykose> what does 'added a number of php extensions' mean
2023-06-02 22:32:38 <Habbie> and what does "a scan for runtime depencies" mean
2023-06-02 22:32:40 <psykose> you can't "install glibc" at all, it's not a thing
2023-06-02 22:33:30 <irreleph4nt> psykose, regarding glibc, I am using these packages: https://github.com/sgerrand/alpine-pkg-glibc
2023-06-02 22:33:36 <psykose> yeah you can't use that
2023-06-02 22:33:38 <irreleph4nt> required by the Oracle extension
2023-06-02 22:34:37 <irreleph4nt> Habbie, have a look at the second link I posted, please. It points to a number of rows in the Dockerfile where the compiled files are scanned for linked runtime requirements which have not yet been satisfied
2023-06-02 22:34:44 <psykose> you have a wrong idea of 'required' and how any of this works
2023-06-02 22:34:56 <psykose> but to save you some time, use debian:slim instead
2023-06-02 22:35:05 <dalias> :-p
2023-06-02 22:35:17 <irreleph4nt> I can't. Debian is not available for use at my company
2023-06-02 22:35:25 <irreleph4nt> it's RHEL or Alpine for me
2023-06-02 22:35:32 <psykose> rhel it is then i guess
2023-06-02 22:35:32 <Habbie> rhel has glibc
2023-06-02 22:36:32 <irreleph4nt> psykose, I'd much rather understand what's going on here. Not necessarily with the aim to "fix my build" but so I may explain to my manager what the problem is
2023-06-02 22:36:58 <psykose> you can think of a c library as the platform
2023-06-02 22:37:11 <psykose> you can't mix glibc and musl on the same system (different platforms)
2023-06-02 22:37:23 <psykose> if you want to run things on alpine then more or less you have to build everything against musl
2023-06-02 22:37:45 <psykose> the thing you are installing "php oracle extension" is something you're not actually building, but rather installing some precompiled thing from somewhere that will never run
2023-06-02 22:38:21 <psykose> you can't satisfy that dependency in any way, except to build it yourself somehow or to find one built in a way that works in the first place
2023-06-02 22:38:58 <psykose> so, if you just install it on the rhel image it works. my recommendation is to go do that
2023-06-02 22:40:46 <irreleph4nt> psykose, thank you very much for explaining this. Am I right to assume that any software I can actually compile from source will work fine when compiled against musl, but anything precompiled (go binaries come to mind) will have issues as musl is most likely not used when they were built?
2023-06-02 22:41:04 <psykose> pretty much
2023-06-02 22:41:27 <psykose> statically compiled binaries also work fine (they have no platform dependency), but those are just executables and not modules for stuff
2023-06-02 22:41:45 <psykose> for a musl php there is no way to run anything unless all of them are
2023-06-02 22:42:03 <irreleph4nt> Is there a particular reason why alpine chose musl over glibc?
2023-06-02 22:42:22 <psykose> different platform different tradeoffs
2023-06-02 22:43:07 <irreleph4nt> I'll dig into musl a bit, I think. Now I want to know who else uses it and why :)
2023-06-02 22:43:46 <psykose> the oracle thing specifically doesn't look like something buildable and requires a bunch of random proprietary stuff to run
2023-06-02 22:43:54 <psykose> so i don't think you'll ever get that to run
2023-06-02 22:44:34 <boomanaiden154> Void also has a musl distribution and you can do Gentoo with musl. I think Alpine is probably the biggest distro using it though.
2023-06-02 22:44:43 <Habbie> i'm not entirely sure i should mention this, and i might even be wrong, but if you have a desparate desire to talk to oracle from php on alpine, odbc might help
2023-06-02 22:45:02 <Habbie> boomanaiden154, either alpine or openwrt
2023-06-02 22:45:13 <Habbie> boomanaiden154, depending on how you count :)
2023-06-02 22:45:49 <irreleph4nt> Habbie, I'll look into ODBC then as well to hopefully solve my work problem. Thank you!
2023-06-02 22:46:16 <boomanaiden154> Could be. I never really think about OpenWRT because I pretty much only use Alpine for containers.
2023-06-02 22:46:20 <Habbie> i want to be very clear, it might not be a solution
2023-06-02 22:46:34 <Habbie> odbc is, at best, a painful alternative
2023-06-02 22:47:03 <Habbie> boomanaiden154, yes - that's why i say "depends on how you count"
2023-06-02 22:47:13 <Habbie> boomanaiden154, there are millions, perhaps a billion, openwrt devices out there
2023-06-02 22:47:18 <Habbie> boomanaiden154, but their users don't know that
2023-06-02 22:47:33 <Habbie> boomanaiden154, alpine meanwhile is usually chosen consciously by admins :)
2023-06-02 22:47:40 <irreleph4nt> Our developers created a php lib to talk to Oracle via oci8. I'll ask them to try odbc and see how it goes
2023-06-02 22:48:33 <boomanaiden154> Habbie: There's still probably quite a few people who pull down container images that use Alpine as a base without knowing that. They probably aren't the ones running the majority of them though. Hard to count numbers whatever you look at it, like you mentioned.
2023-06-02 22:48:48 <boomanaiden154> I'd be interested to see if anyone ever does a LLVM libc based distro once it becomes more mature.
2023-06-02 22:48:56 <Habbie> boomanaiden154, i agree, that likely also happens a lot
2023-06-02 22:49:09 <Habbie> llvm libc? why?
2023-06-02 22:50:21 <boomanaiden154> A lot of performance work has gone into it, I like the development methodology, and I do a decent amount of work on LLVM, so I'm biased.
2023-06-02 22:50:28 <Habbie> right!
2023-06-02 22:50:43 <Habbie> i wonder if you could just rebuild glibc with llvm for whatever you're running :)
2023-06-02 22:51:39 <boomanaiden154> LLVM libc is a new libc, not just a clang/LLVM built libc (sorry if this wasn't clear).
2023-06-02 22:52:02 <boomanaiden154> Glibc still isn't buildable with clang and the main person pushing that effort seems to have peeled off recently.
2023-06-02 22:52:09 <boomanaiden154> http://maskray.me/blog/2021-10-10-when-can-glibc-be-built-with-clang
2023-06-02 22:53:13 <Habbie> boomanaiden154, ah! i did assume you mean "glibc built with llvm/clang", my bad
2023-06-02 22:53:31 <Habbie> *meant
2023-06-02 22:58:58 <psykose> google was sad people didn't like the other two libcs they maintain so they made a third one
2023-06-02 22:59:06 <psykose> (just to spoil you there on the background)
2023-06-02 22:59:08 <Habbie> bionic and?
2023-06-02 22:59:20 <psykose> fuschia
2023-06-02 22:59:23 <Habbie> oh yes
2023-06-02 22:59:27 <Habbie> and now the llvm one?
2023-06-02 22:59:34 <Habbie> (fuchsia, i think)
2023-06-02 22:59:41 <psykose> llvm libc seems mostly to be marketed as the overlay mode from what i've seen
2023-06-02 23:00:09 <psykose> where you use it to statically link in some routines on top of whatever actual other libc (glibc) instead of using the other ones
2023-06-02 23:00:22 <Habbie> right
2023-06-02 23:00:23 <psykose> which is mostly a pile of hacks but works for some people
2023-06-02 23:00:48 <psykose> that was already something you could've always done and what everyone did (i.e. as an application developer) so it's not something super novel
2023-06-02 23:00:58 <boomanaiden154> Fuchsia is moving to LLVM libc.
2023-06-02 23:01:39 <boomanaiden154> I'd think Android would so at some point too, but maybe not.
2023-06-02 23:02:53 <psykose> indeed
2023-06-02 23:03:04 <boomanaiden154> But yeah, overlay mode is the big thing currently with Google doing that internally on their server applications for some performance improvements.
2023-06-02 23:03:20 <boomanaiden154> LLVM libc definitely has a lot more focus on performance (and flexibility) compared to Fuchsia/Bionic.
2023-06-03 07:37:07 <qaqland[m]> Does alpine have one man page online like https://man.archlinux.org/ ?
2023-06-03 07:40:22 <ikke> yno
2023-06-03 07:42:23 <scorpion2185[m]1> is that yes or no?
2023-06-03 07:42:58 <ikke> no
2023-06-03 07:43:42 <scorpion2185[m]1> I used apk-files a few times nice.
2023-06-03 07:43:42 <scorpion2185[m]1> An option to download the databse could be added
2023-06-03 07:44:06 <psykose> no, it couldn't
2023-06-03 07:44:25 <scorpion2185[m]1> i mean that i'd like that feature
2023-06-03 07:44:50 <scorpion2185[m]1> and apk-source? to download the APKBUILD and pathes etc.
2023-06-03 07:45:32 <psykose> we already have that it's called looking in the git repo
2023-06-03 07:46:23 <psykose> this isn't some sort of closed-door proprietary thing with compliance to meet
2023-06-03 07:46:30 <psykose> such a feature makes no sense
2023-06-03 07:46:44 <scorpion2185[m]1> it's not the same can you even  clone it?
2023-06-03 07:46:50 <psykose> yes?
2023-06-03 07:46:58 <scorpion2185[m]1> https://git.alpinelinux.org/aports/tree/community/amazfish?h=master
2023-06-03 07:47:07 <psykose> you can indeed clone that repo
2023-06-03 07:47:17 <scorpion2185[m]1> fatal: repository 'https://git.alpinelinux.org/aports/tree/community/amazfish?h=master/' not found
2023-06-03 07:47:46 <psykose> git clone https://gitlab.alpinelinux.org/alpine/aports.git
2023-06-03 07:47:58 <scorpion2185[m]1> immagine being on a tty or something you could use apk-source
2023-06-03 07:48:06 <psykose> the fuck is this scenario
2023-06-03 07:48:18 <Newbyte> scorpion2185[m]1: you could also use git clone
2023-06-03 07:48:43 <psykose> i can think of zero benefits to get-package-source
2023-06-03 07:48:46 <psykose> like actually not a single one
2023-06-03 07:48:51 <qaqland[m]> scorpion2185: take it easy :p
2023-06-03 07:49:05 <scorpion2185[m]1> psykose: you have to clone all?
2023-06-03 07:49:12 <ikke> --depth=1
2023-06-03 07:49:31 <psykose> or just look at the files you care about or any number of ways to access the content
2023-06-03 07:49:32 <Newbyte> scorpion2185[m]1: yes, it's a monorepo
2023-06-03 07:50:01 <scorpion2185[m]1> Newbyte: you have to browse and search
2023-06-03 07:50:03 <psykose> 'i need to get the exact source+patches only but also i am on a tty' is such a weird usecase i don't even know how you seriously type that and press the enter key
2023-06-03 07:50:18 <Newbyte> scorpion2185[m]1: Why?
2023-06-03 07:50:26 <ikke> even on a tty, it's trivial to clone a git repo
2023-06-03 07:50:41 <scorpion2185[m]1> yes but you have to clone all the aport repo
2023-06-03 07:50:50 <Newbyte> scorpion2185: maybe the real question is, why do you want this?
2023-06-03 07:51:45 <scorpion2185[m]1> i am suggesting a feature that i am sure that it's useful and i'd like but i you don't want it/hate  feel free to don't do it and disagree
2023-06-03 07:51:56 <psykose> you don't have to clone the 'whole' repo actually
2023-06-03 07:52:29 <Newbyte> scorpion2185[m]1: I don't hate the idea, but if you want to a add a feature there should be some justification for it
2023-06-03 07:52:46 <scorpion2185[m]1> e.g. apt-source
2023-06-03 07:53:29 <scorpion2185[m]1> as I said sometimes you want to buidl apcakge for some reason and you can simply apk-source some_package
2023-06-03 08:07:57 <psykose> and you can also do that by checking out the git repo or looking at it and just copying the apkbuild+patches in the folder
2023-06-03 08:08:51 <psykose> there's no unique usecase that this enables, except the one where for some reason one is completely unwilling to use git (that you get to keep around for doing the same thing in the future), is incidentally on a tty as well, ..
2023-06-03 08:09:06 <psykose> we can invent usecases in the sense of a hammer in search of a nail, sure, but that's backwards
2023-06-03 08:09:07 <Newbyte> scorpion2185: if you want this, you could easily implement a script that automatically downloads these files for you
2023-06-03 08:09:36 <psykose> common usecase and why others do it at all: a common requirement of the gpl-* licences is that you redistribute modified source if you patch it
2023-06-03 08:09:51 <psykose> so companies with lawyers create apt-source so you get the patched source since you need it for 'gpl compliance'
2023-06-03 08:09:52 <psykose> etc
2023-06-03 08:10:01 <scorpion2185[m]1> I had to open raw the files and copy paste them
2023-06-03 08:10:18 <scorpion2185[m]1> apt-source it's pretty convenient
2023-06-03 08:10:33 <scorpion2185[m]1> when you rebuild it for some reason to add something etc.
2023-06-03 08:10:39 <psykose> yeah i do that all day
2023-06-03 08:10:46 <psykose> thousands of times a week actually
2023-06-03 08:10:55 <psykose> it's called cd ~/src/aports/some/package; ..
2023-06-03 08:12:05 <scorpion2185[m]1> not all the users have the whole aport in their home
2023-06-03 08:12:15 <psykose> and yet they can acquire it for free
2023-06-03 08:12:25 <scorpion2185[m]1> can you actually git clone the folder?
2023-06-03 08:12:56 <psykose> theoretically sure, https://stackoverflow.com/questions/600079/how-do-i-clone-a-subdirectory-only-of-a-git-repository
2023-06-03 08:13:05 <ikke> You cna do sparse checkouts
2023-06-03 08:13:20 <psykose> that's not exactly only a folder and quite a niche usecase
2023-06-03 08:13:22 <abby> does aports use the ports tree as global state for e.g. package versions to build against?
2023-06-03 08:13:31 <psykose> yes
2023-06-03 08:13:33 <abby> s/package/dependency/
2023-06-03 08:13:37 <psykose> ah
2023-06-03 08:13:38 <psykose> well
2023-06-03 08:13:39 <psykose> kind of
2023-06-03 08:13:57 <psykose> it's the sole repo for anything and everything comes from it more or less
2023-06-03 08:14:20 <psykose> but 'build against' is just what gets installed in the end from the apk repositories on the same system built from the previous
2023-06-03 08:14:24 <psykose> i guess that's just semantics, so yes
2023-06-03 08:15:28 <psykose> the actual apkbuilds are self-contained and don't matter that they're in aports, so no in that sense
2023-06-03 08:15:55 <psykose> from what i see in voidpkgs it looks the same but i didn't look too deeply at how you do it
2023-06-03 08:18:35 <psykose> you can kinda script an apk-source tbh
2023-06-03 08:19:53 <psykose> apk-source $pkg, search all apkindexs for pkg to get repo, fetch https://gitlab.alpinelinux.org/alpine/aports/-/raw/master/$repo/$pkgname/APKBUILD, source apkbuild in shell, echo $source, for urls fetch url, for non-url fetch from / as extra files
2023-06-03 08:20:02 <psykose> idk that sounds pretty easy maybe someone should write it for real
2023-06-03 08:20:35 <psykose> i will do it but only if you pay me $1,000 USD
2023-06-03 08:21:04 <Newbyte> I will do it for 900 Euro
2023-06-03 08:21:23 <psykose> ^ he will fuck it up don't go for this offer
2023-06-03 08:21:50 <psykose> it sounds like something ptrc would write in javascript for fun
2023-06-03 08:22:51 <abby> smh something of this level needs to be written in rust
2023-06-03 08:23:05 <Newbyte> abby: beat me to it
2023-06-03 08:23:37 <ikke> There already exists something in go :)
2023-06-03 08:24:09 <psykose> which thing
2023-06-03 08:24:42 <ikke> parsing apkindexes, parsing APKBUILDs, and do soemthing with that information
2023-06-03 08:24:54 <psykose> smh that is like 2% of the interesting parts here
2023-06-03 08:25:02 <psykose> ah indexes too
2023-06-03 08:25:03 <psykose> right
2023-06-03 08:25:08 <Newbyte> yeah it needs to be rewritten in Rust
2023-06-03 08:25:13 <psykose> that's just a grep ^something:
2023-06-03 08:25:18 <psykose> webscale
2023-06-03 08:25:26 <ikke> 👎
2023-06-03 08:25:34 <psykose> 👍 👍
2023-06-03 08:26:00 <psykose> i remember when i looked at lua-aports and tried to rewrite it
2023-06-03 08:26:07 <psykose> you have no idea how fast i gave up
2023-06-03 08:26:26 <Newbyte> if it's written in Go it can't be reliable 🦀 and blazing fast 🚀
2023-06-03 08:26:41 <Newbyte> psykose: you mean aports-turbo?
2023-06-03 08:26:45 <psykose> yes
2023-06-03 08:26:50 <ikke> that's somethign different
2023-06-03 08:26:51 <psykose> <- bad with names
2023-06-03 08:26:56 <Newbyte> it was rewritten in Python
2023-06-03 08:27:03 <Newbyte> because someone didn't want to debug the Lua version
2023-06-03 08:27:09 <psykose> yeah
2023-06-03 08:28:14 <Newbyte> the Python version has more stars
2023-06-03 08:28:22 <Newbyte> guess it must be superior
2023-06-03 08:28:36 <psykose> i don't think i ever seriously looked at it
2023-06-03 08:29:00 <Newbyte> it's missing the flagging feature
2023-06-03 08:29:18 <psykose> you mean it has the feature, of not having the flagging feature
2023-06-03 08:29:27 <Newbyte> right
2023-06-03 08:29:38 <Newbyte> 🐍🐍🐍🐍
2023-06-03 08:29:39 <psykose> i actually have access to the aports-turbo production instance and you have no idea how tempting it is
2023-06-03 08:29:41 <psykose> to just remove it
2023-06-03 08:29:56 <Newbyte> what you really should do is rewrite it in Rust
2023-06-03 08:30:02 <psykose> yeah i tried see above
2023-06-03 08:30:12 <ikke> Would take another 2 years :P
2023-06-03 08:30:13 <Newbyte> you should just try harder
2023-06-03 08:30:29 <psykose> i would need more stimulants than i possess (zero)
2023-06-03 08:30:55 <Newbyte> if you try really hard and use lube it might work
2023-06-03 08:31:13 <psykose> is that what they do in sweden
2023-06-03 08:31:28 <Newbyte> obviously
2023-06-03 08:32:01 <Newbyte> gotta maintain our rep 💯
2023-06-03 09:14:31 <psykose> ...what rep is that
2023-06-03 09:44:30 <bob> Hola
2023-06-03 09:44:54 <bob> It seems that alpire 7zip doesn't support rar, and I can't find a rar tool with apk search
2023-06-03 09:45:12 <bob> Is it not possible to manipulate rar archives on alpine, or is there a tool that I missed?
2023-06-03 09:46:36 <ikke> bob: can you try bsdtar from libarchive-tools?
2023-06-03 09:47:10 <bob> Bat that's not a tar archive
2023-06-03 09:47:26 <ikke> https://libarchive.org/
2023-06-03 09:47:34 <ikke> "Reads a variety of formats, including tar, pax, cpio, zip, xar, lha, ar, cab, mtree, rar, and ISO images. "
2023-06-03 09:48:02 <ikke> It cannot make rar files, but it can read them
2023-06-03 09:48:59 <bob> I'm trying :)
2023-06-03 09:51:41 <bob> That worked, thanks ikke, I would have never thought to use a tar tool for extracting a rar archive ^^
2023-06-03 09:52:27 <ikke> All the other rar tools have dubious origins (license wise)
2023-06-03 09:52:47 <ikke> bsdtar is like a swiss army knife
2023-06-03 09:53:12 <bob> Yeah, rar format always had this problem in the first place :/
2023-06-03 09:53:49 <psykose> they're not very dubious, rather straight up proprietary :p
2023-06-03 09:54:01 <psykose> i forget what the second alternative was to bsdtar
2023-06-03 09:56:26 <bob> “The unRAR sources cannot be used to re-create the RAR compression algorithm, which is proprietary. Distribution of modified unRAR sources in separate form or as a part of other software is permitted, provided that it is clearly stated in the documentation and source comments that the code may not be used to develop a RAR (WinRAR) compatible archiver.”
2023-06-03 09:56:47 <psykose> yep
2023-06-03 09:57:38 <bob> So can be dumbed down to “can be used for decompression but not compression”
2023-06-03 11:19:34 <msiism> After configuring networking devices, `setup-alpine` updates /etc/hosts using some elaborate sed hackage. Now, I wonder whether I should do that manually when not using `setup-alpine`.
2023-06-03 11:20:22 <msiism> This is what I'm talking about: http://paste.debian.net/plainh/4ff2408c
2023-06-03 11:23:49 <psykose> your /etc/hosts really only has to be
2023-06-03 11:24:03 <psykose> 127.0.0.1 localhost localhost.localdomain $(hostname)
2023-06-03 11:24:11 <psykose> ::1 localhost localhost.localdomain $(hostname)
2023-06-03 11:24:12 <psykose> that's it
2023-06-03 11:24:17 <psykose> not even sure about the .localdomain part
2023-06-03 11:24:47 <psykose> maybe that breaks with some fancy search domain that it also handles, but if you were using that you'd probably already know
2023-06-03 11:26:57 <msiism> Okay.
2023-06-03 11:27:26 <msiism> That's exactly what mine on the installed system says.
2023-06-03 11:28:00 <Hello71> psykose: unarchiver
2023-06-03 11:28:04 <psykose> that's the one
2023-06-03 11:28:55 <Hello71> to download a directory from gitlab you can apparently use git-archive or api (needs authentication?)
2023-06-03 11:35:54 <psykose> iirc git-archive was pretty bad for various resource use? it's been a while
2023-06-03 11:59:10 <quinq> Yeah, I've never seen something using .localdomain, but maybe somebody somewhere thought this was a good idea
2023-06-03 12:01:02 <quinq> Hello71, unarchiver doesn't seem to be packaged though
2023-06-03 12:02:44 <Hello71> psykose: i think the protocol should be reasonably efficient, i doubt it does anything terribly stupid like getting the list of objects first
2023-06-03 12:03:13 <Hello71> quinq: it needs gnustep and some other stuff, not sure it's worth it
2023-06-03 12:03:34 <quinq> Sure, I'm not requesting it at all, one working tool is enough :)
2023-06-03 12:03:37 <psykose> https://img.ayaya.dev/h7ErjxKy7u19
2023-06-03 12:04:00 <psykose> Hello71: ah then it might've been another i was looking at..
2023-06-03 12:09:35 <Hello71> i only know of unrar, 7zip (which is really unrar), libarchive, and unarchiver
2023-06-03 12:11:06 <Hello71> also unrar-free if you count that separately
2023-06-03 12:12:25 <quinq> There's also winrar/wine
2023-06-03 12:13:36 <Hello71> that's arguably just unrar but you also have to pay
2023-06-03 12:14:00 <quinq> :>
2023-06-03 12:20:00 <psykose> new hello71 lore just dropped
2023-06-03 12:20:04 <psykose> he paid the winrar licence
2023-06-03 12:21:29 <quinq> haha
2023-06-03 12:32:47 <quinq> Is there a simple way with ifupdown to trigger dhcp depending on link carrier status instead of always?
2023-06-03 12:53:14 <Fxzxmic> psykose: I would like to inform you that there is a new version of vaultwarden's web vault.
2023-06-03 12:53:46 <psykose> they're tied to eachothers version
2023-06-03 12:55:41 <Fxzxmic> Got it
2023-06-03 12:58:56 <Fxzxmic> Oh, one more question, why put it in the /usr/sbin directory rather than /usr/bin?
2023-06-03 13:00:12 <psykose> how would i know
2023-06-03 13:01:54 <quinq> “web” sounds like it would listen on port 80 by default, which is reserved and needs privileges, so sbin binary
2023-06-03 13:03:20 <psykose> i guess this time it's compatible
2023-06-03 13:04:37 <Fxzx_mic> 😕
2023-06-03 13:04:41 <ikke> quinq: that distinction is arbitrart
2023-06-03 13:04:45 <ikke> arbitrary
2023-06-03 13:08:18 <psykose> that distinction is "sbinmerge and forget about it"
2023-06-03 13:08:21 <psykose> glad we could solve the case
2023-06-03 13:09:43 <quinq> Yes ikke 
2023-06-03 13:09:58 <quinq> Like any rule that doesn't come from physics
2023-06-03 13:11:50 <psykose> socrates has logged on
2023-06-03 13:13:29 <quinq> sarcastic has logged on
2023-06-03 14:25:12 <Piraty> i wonder if reinstalling some of the core packages will break it, like nuke passwd / shadow or the like
2023-06-03 14:25:12 <Piraty> i'm intending to downgrade from edge to latest release , i stopped updating before release was tagged. this will reinstall all the packages if i understand correctly (replace foo (1.33.7-r0 -> 1.33.7-r0) ) 
2023-06-03 14:25:26 <psykose> nope
2023-06-03 14:25:53 <psykose> just change repos and upgrade -a
2023-06-03 14:25:58 <Piraty> ok
2023-06-03 14:26:19 <Piraty> i seem to have some things from edge/testing ...
2023-06-03 14:32:36 <Piraty> (1/1766) here we go
2023-06-03 14:32:51 <ikke> oof, that's quite a lot of packages
2023-06-03 14:32:53 <maria> bigg
2023-06-03 14:33:26 <Piraty> just the usual nodejs all in one development container env /s
2023-06-03 14:34:09 <Piraty> it's my pinephone actually, i'm sick of edge breaking my day
2023-06-03 14:34:36 <pj> "is edge stable for production use?"
2023-06-03 14:35:12 <Piraty> a few years ago when i set it up it was the only way getting a good-enough'ish experience
2023-06-03 14:35:37 <Piraty> also, definition of 'production' may vary
2023-06-03 14:37:32 <psykose> what broke this time
2023-06-03 15:08:51 <mercenary> in 'what broke today', upgrading 3.16 to 3.17 broke surf
2023-06-03 15:10:38 <psykose> surf is in testing/ so you can only use it from edge to begin with
2023-06-03 15:15:59 <mercenary> one of those 'used to work, now it builds fine, but webkit throws x11 errors'.
2023-06-03 15:16:17 <frag> surf works?
2023-06-03 15:16:34 <psykose> do you have GDK_BACKEND set to something other than x11 when opening it
2023-06-03 15:16:36 <frag> usually get a "web crash" in 2sex with webkit
2023-06-03 15:16:38 <mercenary> anyway, that's the universe telling me I should write a proper client instead of using a browser
2023-06-03 15:17:55 <mercenary> env|grep GDK ==> nothing
2023-06-03 15:18:35 <quinq> works here (edge)
2023-06-03 15:18:36 <psykose> set it to x11 and try agane
2023-06-03 15:19:40 <mercenary> same result
2023-06-03 15:19:45 <psykose> interesting
2023-06-03 15:20:31 <quinq> mercenary, what are the errors?
2023-06-03 15:21:27 <sfan5> hi
2023-06-03 15:21:35 <sfan5> I just stumbled upon this https://lists.alpinelinux.org/~alpine/devel/%3C1628515011.zujvcn248v.none%40localhost%3E
2023-06-03 15:21:43 <sfan5> and am wondering: does alpine currently use -no-plt?
2023-06-03 15:21:47 <mercenary> https://dpaste.com/5RUE3MGV2
2023-06-03 15:22:15 <psykose> sfan5: doesn't no
2023-06-03 15:22:22 <sfan5> I see, thanks
2023-06-03 15:25:59 <quinq> Does it happen with another webkit2gtk-based web browser, like epiphany/midori?
2023-06-03 15:29:31 <psykose> sfan5: it's ~mostly quite dubious overall except in the common dynamically-linked-x86-system case (like distros usually are), and e.g. rust funnily does default to that which is a loss everywhere outside of x86 https://github.com/rust-lang/rust/pull/106380
2023-06-03 15:30:20 <psykose> for x86 only it could make sense to set and i'm not opposed to it but i know at least one person would immediately question me on it not being 'simple' and the gain is like 1% binary size
2023-06-03 15:30:21 <psykose> so eh
2023-06-03 17:10:33 <Hello71> it's also probably slightly faster
2023-06-03 17:17:38 <Piraty> (1766/1766) ... now that took some time
2023-06-03 18:20:12 <psykose> Hello71: that probably does a lot of heavy lifting
2023-06-03 18:20:30 <psykose> regardless, enabled
2023-06-03 18:33:29 <Quillith> Hi, would it be possible for Ryujinx to be native for Alpine? or is it not compatible
2023-06-03 18:34:43 <Quillith> https://github.com/Ryujinx/Ryujinx
2023-06-03 18:36:49 <psykose> there's multiple reasons why that would be a disaster that doesn't work
2023-06-03 18:36:56 <psykose> flatpak exists and would let you have it in 3 seconds
2023-06-03 18:38:18 <Quillith> i can't get the flatpak to install firmware on alpine
2023-06-03 18:38:58 <psykose> you should figure out how to do that then
2023-06-03 18:39:22 <Quillith> Well I know how to do it in other OS
2023-06-03 18:40:10 <Quillith> In pop-os I follow the official instructions to select the firmware file via the gui. In Alpine I can a persistent error of 'firmware file does not exist.' Same files, same flatpak install method. So I figured maybe it was an alpine thing and wanted to see if maybe a native install was doable
2023-06-03 18:43:02 <psykose> that seems like a bug in the flatpak or some assumption gone amiss
2023-06-03 18:46:11 <psykose> though i think you just have to put it in  ~/.var/app/org.ryujinx.Ryujinx/config/Ryujinx/system
2023-06-03 18:46:18 <psykose> i don't have any keys to test with
2023-06-03 18:47:20 <Quillith> I uninstalled, removed directories, and reinstalled a few times, no change. I asked in their discord and the only response I got was 'probably an os thing', though I don't know if that's true 
2023-06-03 18:47:27 <Quillith> I'll try moving the firmware files to that directory
2023-06-03 18:48:14 <psykose> yeah
2023-06-03 18:48:15 <psykose> it works
2023-06-03 18:48:21 <psykose> found some cracked keys
2023-06-03 18:48:29 <psykose> also make sure you don't have SDL_VIDEODRIVER=wayland
2023-06-03 18:48:35 <Quillith> Is this for keys or firmware? I 
2023-06-03 18:48:50 <psykose> ah yeah, that was the issue
2023-06-03 18:48:54 <psykose> it prints the first warning thing
2023-06-03 18:48:57 <psykose> then it crashes with =wayland
2023-06-03 18:49:05 <psykose> with =x11 it opens the ui anyway
2023-06-03 18:49:18 <psykose> regardless, you have to put keys in that folder
2023-06-03 18:49:24 <psykose> after you install firmware the same normal way
2023-06-03 18:50:45 <Quillith> I do not have SDL_VIDEODRIVER=wayland . I have WAYLAND_DISPLAY XDG_CURRENT_DESKTOP=SWAY and QT_QPA_PLATFORM="wayland-egl"
2023-06-03 18:51:03 <Quillith> I will try moving my firmware .zip to the directory and see if that helps
2023-06-03 18:51:11 <psykose> make it =x11 then
2023-06-03 18:51:16 <psykose> that won't
2023-06-03 18:51:19 <psykose> the issue is none of that
2023-06-03 18:51:26 <psykose> it just won't open without the sdl_videodriver i guess
2023-06-03 18:52:55 <psykose> https://img.ayaya.dev/pBu0pnaNbsiB
2023-06-03 18:55:58 <Quillith> And did you install firmware (not keys) via gui or through directly moving the firmware into the directory? I put "exec export SDL_VIDEODRIVER=xll" into my sway config file and tried to use the gui to install firmware but I get the same error
2023-06-03 18:56:16 <Quillith> And I refreshed my sway config to make sure it loaded
2023-06-03 18:56:58 <Quillith> The terminal output isn't super helpful, it just says "|E| Application CreateErrorDialog: Firmware file does not exist."
2023-06-03 18:58:18 <psykose> x11 not xll
2023-06-03 18:58:26 <Quillith> ?
2023-06-03 18:58:32 <Habbie> x one one
2023-06-03 18:58:34 <Habbie> not x el el
2023-06-03 18:58:38 <psykose> fix your fonts
2023-06-03 18:58:38 <ikke> Habbie: beat me to it
2023-06-03 18:58:43 <Quillith> right, x 1 1
2023-06-03 18:58:57 <psykose> that said also paste the log of it starting
2023-06-03 18:59:37 <psykose> and then try  `export QT_QPA_PLATFORM=` (nothing)
2023-06-03 19:00:06 <psykose> you can just export this in a shell you open and `flatpak run ..` it from there, nothing else needed
2023-06-03 19:00:14 <psykose> no restarts or anything
2023-06-03 19:00:39 <Quillith> https://bpa.st/RSKSC
2023-06-03 19:01:31 <psykose> env | grep -i way
2023-06-03 19:02:20 <Quillith> https://bpa.st/YRPGY
2023-06-03 19:02:52 <psykose> now `SDL_VIDEODRIVER=x11 flatpak run org.ryujinx.Ryujinx`
2023-06-03 19:04:09 <Quillith> Aborted, SDL2 initialization failed with error 'x11 not available'
2023-06-03 19:04:18 <psykose> apk add xterm; xterm
2023-06-03 19:04:20 <psykose> does that start
2023-06-03 19:04:49 <Quillith> I ran that command and now there is a white terminal open 
2023-06-03 19:04:52 <psykose> hrm
2023-06-03 19:05:04 <psykose> that's weird because the exact same thing works for me
2023-06-03 19:05:05 <Quillith> but it did install libxpm libxaw and xterm
2023-06-03 19:05:11 <psykose> ..which ryujinx do you have installed
2023-06-03 19:05:23 <psykose> flatpak list | grep ryu  
2023-06-03 19:05:52 <Quillith> 1.1.859
2023-06-03 19:06:00 <psykose> same as me
2023-06-03 19:06:03 <psykose> that makes no sense then
2023-06-03 19:06:34 <psykose> i guess for some reason the xwayland socket or whatever isn't passed into flatpak or similar
2023-06-03 19:07:16 <Quillith> If the issue is wayland do you think trying to switch to i3wm would help?
2023-06-03 19:07:49 <Quillith> I don't know what that would involve but it's better than dual-booting pop-os just to use ryujinx
2023-06-03 19:09:04 <psykose> it's not, everything else works fine
2023-06-03 19:09:11 <psykose> does the steam flatpak work
2023-06-03 19:09:14 <psykose> it mostly requires the same stuff
2023-06-03 19:09:26 <psykose> i guess you could paste the full `env`, and `ps aux | grep portal`
2023-06-03 19:09:28 <Quillith> I'll try the empty QT_QPA_PLATFORM you mentioned and see if that works
2023-06-03 19:09:34 <psykose> and what wm you're running
2023-06-03 19:09:35 <Quillith> oh ok I'll grab that output
2023-06-03 19:09:51 <psykose> try that too sure, but you didn't have it set in the above env
2023-06-03 19:11:39 <Quillith> I didn't? weird, i have "exec export QT_QPA_PLATFORM="wayland-egl"" set in my sway config file
2023-06-03 19:12:11 <Quillith> https://bpa.st/27NI4
2023-06-03 19:13:24 <Quillith> https://bpa.st/R6LEE
2023-06-03 19:15:11 <Quillith> Yes steam flatpak works fine, I can play games without problems
2023-06-03 19:16:44 <psykose> https://bpa.st/YRPGY didn't have it
2023-06-03 19:16:55 <psykose> idk where you ran that but i hope it was the same session as everything else
2023-06-03 19:17:24 <psykose> that looks ok
2023-06-03 19:17:27 <psykose> can't think of anything else
2023-06-03 19:17:39 <Quillith> I haven't logged off if that's what you mean
2023-06-03 19:18:09 <Quillith> Well, I appreciate the help troubleshooting. I'll keep my dual-boot, maybe it'll fix itself eventually
2023-06-03 19:24:18 <psykose> well, something quick
2023-06-03 19:24:19 <psykose> uuhh
2023-06-03 19:25:38 <psykose> env -i HOME="$HOME" USER="$USER" PATH="$PATH" DISPLAY="$DISPLAY" WAYLAND_DISPLAY="$WAYLAND_DISPLAY" XDG_RUNTIME_DIR="$XDG_RUNTIME_DIR" sh
2023-06-03 19:25:43 <psykose> in that shell, run command
2023-06-03 19:27:15 <Quillith> I ran that, then immediately after I ran the Ryujinx command in the same terminal. Unfortunately the error persisted
2023-06-03 19:27:58 <Quillith> I don't know if it's relevant but my cursor gets blurry when it goes into the Ryujinx window. It could just be the warning that "couldn't determine monitor DPI: Wayland not yet supported" though
2023-06-03 19:31:07 <Quillith> what the shit it's working now?? 
2023-06-03 19:32:27 <Quillith> I did the env -i stuff again just in case, but even though it didn't work last time now it installed the firmware. Unless I copied and pasted wrong the only thing different is that I'm tiling my windows differently. I don't get it, but thank you
2023-06-03 19:33:49 <Quillith> Okay yes the first time it ran the env -i stuff early before the sh, when I tiled differently it pasted the length of the code then ran after the sh. Now it works! I'll put it in the discord to try the env -i solution in case anyone else runs into the same error
2023-06-03 19:34:38 <Quillith> Oh does this mean that there's something wrong with my configs that I should change? like does this indicate a deeper problem I should address
2023-06-03 19:57:19 <psykose> just means one of the things in the env from when it doesn't work interferes
2023-06-03 19:57:32 <psykose> could just bisect it, but it's one of the =wayland / qt / sdl related vars
2023-06-03 20:54:35 <Quillith> Okay, thank you for your help!
2023-06-04 11:03:29 <solfisher> hi, i'm having some issues with booting alpine on a raspberry pi 4. udhcpc loops on "read error: network is down, reopening socket" and doesn't connect to the internet; trying to replace dhcp with a static ip in /etc/network/interfaces just gets reverted on the next boot
2023-06-04 11:14:52 <thought> I need some help writing an init script. The more I read between alpine and gentoo, the less I'm confident about it.
2023-06-04 11:15:12 <thought> To get ffmpeg to be supervised, on a normal installation, do I simply set the supervisor=supervise-daemon variable?
2023-06-04 11:15:32 <thought> With that in mind, I don't need to write start() stop(), etc, I can just set the command and args in variables, as wlel?
2023-06-04 11:15:34 <thought> well*
2023-06-04 11:16:39 <thought> Is calling another shell script a bad idea for supervision? I've currently got this running in a cronjob, so it's calling something in /usr/local/bin
2023-06-04 11:49:59 <thought> Uh oh. I need to mkdir -p "$SOMESTUFF" before my command is issued, for sorting purposes. I think I can offload that to cron instead, for daily cleanup of the main dir
2023-06-04 11:50:34 <dewdude_> I mean there may be a more optimal way of doing things...but if it works. it works
2023-06-04 11:51:02 <dewdude_> that's the magic of linux. there are 300 ways of doing something and usually they'll all work as well as the others it's just optimization and initial effort
2023-06-04 14:03:43 <quinq> Doing apk upgrade -l, I get the following error message:
2023-06-04 14:03:44 <quinq> ERROR: unable to select packages:
2023-06-04 14:03:57 <quinq> And for example:
2023-06-04 14:03:57 <quinq>   pcre (virtual):
2023-06-04 14:03:57 <quinq>     provided by: pcre
2023-06-04 14:03:57 <quinq>     note: please select one of the 'provided by' packages explicitly
2023-06-04 14:03:57 <quinq>     required by: pcre-dev-8.45-r3[pcre=8.45-r3]
2023-06-04 14:04:18 <quinq> So I try to do what apk tells me, to select one of the provided by packages, that is pcre here
2023-06-04 14:04:41 <quinq> $ doas apk upgrade -l pcre
2023-06-04 14:04:48 <quinq> But I get the exact same error message
2023-06-04 14:05:09 <quinq> How is that problem supposed to be solved?
2023-06-04 14:18:58 <qman> pcre is a virtual package, so it's likely a version conflict of the actual underlying packages
2023-06-04 14:19:42 <qman> Would need to see the full output to make a determination, also, use a pastebin for anything more than 2 or 3 lines
2023-06-04 14:23:24 <qman> Such conflicts usually happen when you pin some packages to a specific version or from a specific source 
2023-06-04 14:23:54 <qman> Like installing one package from edge on a 3.x system, or keeping an older package from a previous version 
2023-06-04 14:40:48 <quinq> I see, so the error message is inaccurate
2023-06-04 14:41:00 <quinq> Thanks, I'll continue investigating in that deriction :)
2023-06-04 16:09:39 <kitzman> are there differences between glibc and musl open flags?
2023-06-04 16:11:47 <kitzman> 9pfuse from plan9port on alpine (aarch64), can't properly mount filesystems. 'unexpected open flags requested=0440000 unhandled=0400000'
2023-06-04 16:15:59 <kitzman> i vaguely remember patching this around 1 year ago
2023-06-04 16:37:27 <noanon> what are my options if a package I'd like to use does not exist for the arch I'm using? I'm looking at pkgs.alpinelinux.org and I can't even figure the contact email of the guy who produces this package, to ask him
2023-06-04 16:37:54 <ptrc> noanon: you can open an issue on https://gitlab.alpinelinux.org/alpine/aports/-/issues
2023-06-04 16:38:02 <ptrc> which package are you referring to?
2023-06-04 16:38:53 <noanon> ptrc: I want dosbox for x86, but I see it can only exist for x86_64
2023-06-04 16:39:55 <ptrc> noanon: so, if you click the link 'Git repository' on pkgs.a.o and open the APKBUILD, you can see both the maintainer and a comment explaining why it's disabled on x86
2023-06-04 16:40:10 <ptrc> if you open an issue on gitlab, someone might take a look at it and try to fix it
2023-06-04 16:40:35 <kwadroke[m]> isn't there a libretro-dosbox for x86?
2023-06-04 16:41:30 <kwadroke[m]> https://pkgs.alpinelinux.org/package/v3.18/community/x86/libretro-dosbox
2023-06-04 16:41:35 <noanon> kwadroke[m]: I found something like that while searching, but it did show me only a .so file in the package listing
2023-06-04 16:41:52 <ptrc> that would require a libretro frontend, such as retroarch
2023-06-04 16:41:54 <kwadroke[m]> yeah,
2023-06-04 16:42:09 <ptrc> and you wouldn't really be able to use it standalone
2023-06-04 16:43:23 <noanon> so I'm out of luck with dosbox?
2023-06-04 16:47:57 <ptrc> well, if you don't want to use libretro, then you can still open an issue
2023-06-04 16:48:52 <noanon> I've never used libretro so I guess I can at least try figuring it out
2023-06-04 16:49:24 <noanon> but as you pointed, since there's a reason for that arch being excluded, I don't think stating that fact in a separate issue would change the status quo
2023-06-04 16:49:50 <noanon> I was just too dumb to figure out how to get to the APKBUILD myself and see that
2023-06-04 16:50:38 <noanon> just trying to make use of this old hardware that's like 20+ years, only alpine managed to boot and install on it
2023-06-04 16:58:32 <noanon> also, both firefox and midori fail at startup with "Illegal instruction" message
2023-06-04 16:59:15 <noanon> I wonder if this is related to glx Xorg module being disabled, otherwise X won't start, and what web browser that's not too RAM heavy can I use
2023-06-04 16:59:23 <noanon> hopefully not dillo :D
2023-06-04 17:00:30 <quinq> Web browsers are monsters, so don't get your hopes up
2023-06-04 17:00:44 <quinq> Not using javascript helps a bit
2023-06-04 17:00:51 <noanon> I can use links fine, but not all sites are friendly
2023-06-04 17:00:57 <quinq> yep
2023-06-04 17:01:27 <noanon> this computer has like 380MB RAM only
2023-06-04 17:01:51 <quinq> :)
2023-06-04 17:01:55 <kwadroke[m]> Illegal instruction usually means that your CPU is missing an instruction that the application is built with
2023-06-04 17:02:29 <kwadroke[m]> I have the same issue on my 1st Gen Intel Atom and Lagrange. I had to download the APKBUILD file and compile it locally
2023-06-04 17:02:47 <kwadroke[m]> Maybe try Surf browser?
2023-06-04 17:03:58 <noanon> second alpine linux day, there's much to learn ahead
2023-06-04 17:08:08 <noanon> same problem mentioned for dosbox-staging, no package for x86 :(
2023-06-04 17:08:18 <noanon> I want my master of orion 2 on this computer 
2023-06-04 17:09:33 <kwadroke[m]> isn't there a newer native engine for Master of Orion1/2?
2023-06-04 17:11:23 <kwadroke[m]> FreeOrion
2023-06-04 17:12:12 <kwadroke[m]> it might a bit more processing power than that machine can give though
2023-06-04 17:12:45 <kwadroke[m]> openmoo2 might be an option - written in Python
2023-06-04 17:13:34 <kwadroke[m]> oh, osgameclones lists openmoo2 as unplayable :(
2023-06-04 17:14:36 <noanon> never heard of those, I run the dos one with dosbox
2023-06-04 17:15:01 <noanon> freeorion might be that foss project that's only inspired by moo but has no tangencies
2023-06-04 17:15:18 <noanon> i shall look at openmoo2 now 
2023-06-04 17:16:13 <noanon> no commits in 6 years, prolly dead
2023-06-04 17:16:21 <kwadroke[m]> yeah saw that
2023-06-04 17:16:28 <kwadroke[m]> at least it's python3 though
2023-06-04 17:17:58 <noanon> I don't know what else can I do on such an ancient computer
2023-06-04 17:18:08 <noanon> maybe read ebooks
2023-06-04 17:28:31 <kwadroke[m]> I use mine for email, chat, calendar. Granted mine has 1GB of RAM
2023-06-04 17:28:47 <kwadroke[m]> my eeepc only has 512M though
2023-06-04 17:30:25 <kwadroke[m]> firefox is a pain on the eeepc though
2023-06-04 17:30:39 <kwadroke[m]> surf is somewhat managable
2023-06-04 17:31:30 <noanon> maybe I shall try to get qemu and run msdos inside it, for master of orion
2023-06-04 17:59:26 <Hello71> kitzman: you can try strace
2023-06-04 18:03:49 <thought> My ffmpeg-as-a-service is running comrades
2023-06-04 18:03:51 <thought> we did it
2023-06-04 18:04:06 <thought> sadly, it means i don't get nice dates automade for me
2023-06-04 18:04:31 <thought> because ffmpeg won't premake the directory that i tell it to make
2023-06-04 18:46:15 <psykose> kwadroke[m]: you should've reported that you know :p
2023-06-04 18:46:18 <psykose> should be fixed
2023-06-04 18:55:13 <kwadroke[m]> figured it was more of an issue with my computer than the way it was compiled
2023-06-04 18:56:51 <kwadroke[m]> I didn't have to change anything in the APKBUILD file besides the pkgrel to get it to work
2023-06-04 18:58:04 <kwadroke[m]> btw, lagrange 1.16.3 is out now
2023-06-04 18:58:09 <psykose> the build system checks if -msse4.1 works on the current machine
2023-06-04 18:58:14 <psykose> then just uses it
2023-06-04 18:58:18 <psykose> yes yes, one of those
2023-06-04 18:58:20 <kwadroke[m]> ah
2023-06-04 18:59:54 <Newbyte> > <@kwadroke:matrix.org> btw, lagrange 1.16.3 is out now
2023-06-04 18:59:54 <Newbyte> speaking of Gopher, do you know any cool capsules?
2023-06-04 19:01:30 <kwadroke[m]> I mostly visit Chilly Weather - gemini://gemi.dev/cgi-bin/chilly.cgi & Station gemini://station.martinrue.com/
2023-06-04 19:01:57 <kwadroke[m]> don't do much with Gopher, mostly Gemini
2023-06-04 19:15:54 <msiism> FYI: Here's a less WIP version of my write-up on Alpine with FDE and a custom partition layout: https://www.msiism.org/files/doc/alpine-linux-fde-custom.html
2023-06-04 19:16:06 <msiism> I'll try to keep this URL intact.
2023-06-04 19:16:29 <psykose> does that ever work out
2023-06-04 19:16:43 <psykose> i feel like everyone's urls die eventually
2023-06-04 19:16:51 <psykose> unless they're on wordpress
2023-06-04 19:17:36 <ikke> cool urls never change
2023-06-04 19:18:23 <msiism> I just said that because I removed the old write-up and the new one is under a more adequate URL.
2023-06-04 19:18:55 <msiism> I don't see why this shouldn't work out as long as I don't run out of money.
2023-06-04 19:19:00 <msiism> ;)
2023-06-04 19:23:31 <psykose> to avoid running out of money, simply acquire more money than you lose money
2023-06-04 19:23:32 <psykose> hot tip
2023-06-04 19:24:00 <msiism> 
ACTION makes a note.

2023-06-04 19:24:17 <ikke> subscribe
2023-06-04 19:25:29 <psykose> thank you for subscribing to cat facts
2023-06-04 19:25:34 <psykose> did you know cats are believed to be the only mammals who don't taste sweetness
2023-06-04 19:25:50 <ikke> poor cats
2023-06-04 19:25:51 <ikke> more
2023-06-04 19:26:02 <msiism> What if you run them with -v?
2023-06-04 19:26:17 <psykose> sadly i don't actually know many cat facts so that would just be google
2023-06-04 19:26:22 <ikke> :D
2023-06-04 19:26:42 <ikke> Did you know cats have a higher core temperature and that's why they seek warm areas
2023-06-04 19:28:00 <psychedelicsShopNembutal[m]> https://skicannabinoids.store/product/%e5%9c%a8%e7%b7%9a%e8%b3%bc%e8%b2%b7%e6%88%8a%e5%b7%b4%e6%af%94%e5%a6%a5%ef%bc%88%e5%87%ba%e5%94%ae%e6%88%8a%e5%b7%b4%e6%af%94%e5%a6%a5%ef%bc%89/
2023-06-04 19:28:43 <psykose> psychedelics shop nembutal
2023-06-04 19:28:44 <psykose> lmao
2023-06-04 19:29:08 <psykose> who the fuck is buying sodium pentobarbital
2023-06-04 19:33:56 <msiism> I still wonder why I get "service `cron' does not exist" when I try to add "cron" to runlevel default during the setup.
2023-06-04 19:34:18 <msiism> I mean, I've got crond, so I'm wondering what "cron" is anyway.
2023-06-04 19:34:19 <ikke> it's crond
2023-06-04 19:34:29 <ikke> Oh, you get that as a message
2023-06-04 19:34:47 <ikke> But you say: "when I try to add cron"
2023-06-04 19:35:22 <msiism> `setup-alpine` has a list of services it will try to add if they exist. And that list has "cron" and "crond" in it.
2023-06-04 19:36:32 <ikke> It does check if the service exists, though
2023-06-04 19:36:42 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-alpine.in#L197
2023-06-04 19:37:02 <msiism> Right.
2023-06-04 19:37:19 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/commit/bca5e2171c0c430438d09feea508c9d264afdb4c
2023-06-04 19:37:25 <ikke> My suspicion
2023-06-04 19:37:31 <ikke> cron was renamed to crond
2023-06-04 19:37:54 <msiism> Oh, that would make sense.
2023-06-04 19:38:32 <ikke> (I love being able to read that from the commit message)
2023-06-04 19:38:54 <msiism> 
ACTION sees it now.

2023-06-04 20:48:07 <vhns> Hi all. I'm currently trying to instal Alpine through libvirtd, but am running into an issue. I am using the "alpine-virt" ISO, but, upon booting, I get the error message "Mounting boot media failed.". I couldn't really seem to find other people having this issue online. I'd be glad to provide more info, if required.
2023-06-04 21:17:25 <txnintn> I have done a clean install of alpine linux on my 4gb RPI4, Dual 3.1 usb thumb drives, LVM over Raid1.  Will not boot from usb, but I copied md1 to a mmcblk boots from it.  Any suggestions?
2023-06-04 21:20:14 <txnintn> Mdadm details shows that metadata is .90.
2023-06-04 21:27:41 <eris> hey, how do you get the microphone to work on alpine...? is there a group i need to add my user to?
2023-06-04 21:34:37 <ptrc> eris: are you using pipewire/pulseaudio or something, or just raw alsa?
2023-06-04 21:34:45 <eris> pulseaudio
2023-06-04 21:34:54 <eris> the devices show up to applications, but produce no audio 
2023-06-04 21:34:58 <eris> webcam works 
2023-06-04 21:35:21 <herkulessi[m]> txnintn: Have you enabled USB-Booting in the Firmware of your raspberry pi?
2023-06-04 21:38:13 <bowenp> Yes, I keep it on a 8mb sd...just in case.  
2023-06-04 21:39:07 <bowenp> eris, what is the platform,?
2023-06-04 21:39:40 <eris> x86_64 laptop
2023-06-04 21:39:42 <eris> framework 
2023-06-04 21:43:16 <bowenp> are do you have anyother sound apps installed?
2023-06-04 21:43:56 <eris> hm?
2023-06-04 21:44:06 <eris> I can use sound just fine on firefox and all that 
2023-06-04 21:44:12 <eris> It's just the microphone that doesn't work 
2023-06-04 21:50:54 <bowenp> no alsamixer?
2023-06-04 22:07:58 <eris> ii have alsamixer and pamixer
2023-06-04 22:08:15 <eris> nvm, no pamixer
2023-06-04 22:10:31 <eris> adjusting the mic boost, captire and internal mic boost do nothing 
2023-06-04 22:19:08 <bowenp> is pamixer a command line mixer?
2023-06-04 22:32:00 <ovf> in a few not particularly wayland specific applications, the equivalent of https://0x0.st/Hb6X.c fails with 'libEGL debug: bad context attribute 0x31b3'. i don't know enough about egl to judge whether that's the correct behaviour, but eglinfo seems to show https://registry.khronos.org/EGL/extensions/KHR/EGL_KHR_create_context_no_error.txt as present for me.
2023-06-04 22:35:58 <psykose> which glfw do you have
2023-06-04 22:46:36 <ovf> glfw-wayland
2023-06-04 22:47:24 <psykose> use the non-wayland one
2023-06-04 22:47:25 <psykose> enjoy
2023-06-04 22:48:05 <psykose> pretty sure i've had the same issue for 2 years
2023-06-04 22:49:39 <psykose> the fact they have separate compile-time selected builds for which window to make is hilarious
2023-06-04 22:51:38 <psykose> they fixed it on master so it will become one build eventually
2023-06-04 22:51:41 <psykose> hopefully also this issue
2023-06-04 22:53:12 <psykose> hmm
2023-06-04 22:54:06 <ovf> ...enjoy 'X11: The DISPLAY environment variable is missing'? :-)
2023-06-04 22:55:16 <ovf> also i'm somewhat certain this problem is below glfw since it just checks that the extension is present and then sets the relevant attrib. i can probably reproduce it without glfw, just not looking forward to initialising wayland egl context by hand
2023-06-04 22:55:49 <psykose> it's not present without glfw-wayland
2023-06-04 22:56:13 <psykose> insofar as things being broken to start with it
2023-06-04 22:56:20 <psykose> obviously by sidestepping a lot of stuff
2023-06-04 23:01:02 <ovf> glfw (not -wayland) wants to only deal with x11 which i don't have, so indeed i'm not encountering the egl problem because glfw fails to initialise
2023-06-04 23:03:49 <psykose> strangely the only relevant thing i can find is https://gitlab.freedesktop.org/mesa/mesa/-/issues/4586
2023-06-04 23:04:05 <psykose> i guess you can start with a mesa issue and a similar egl-only context init
2023-06-04 23:06:45 <ovf> thanks! i couldn't find that one. then of course i've found like half a dozen of wayland egl boilerplate examples none of which work in sway (i think mostly because of wl_shell dying in the meantime)
2023-06-04 23:07:36 <ovf> this seems to be the relevant bit on the mesa side: https://gitlab.freedesktop.org/mesa/mesa/-/blob/main/src/egl/main/eglcontext.c#L327 but i don't see anything offending
2023-06-04 23:09:49 <psykose> yeah it looks like it should work
2023-06-04 23:10:10 <psykose> i also tested latest glfw just in case it was not just something below it, but no, indeed it is
2023-06-04 23:10:41 <psykose> wonder why nothing else gets this issue
2023-06-04 23:12:42 <ovf> mesa-dbg also seems to be a 'thin' one, without args/locals. this is going to be fun :-)
2023-06-04 23:23:21 <maria> why is glfw enabling EGL_CONTEXT_OPENGL_NO_ERROR_KHR? i've not seen it before
2023-06-04 23:32:38 <ovf> it's not glfw itself, it's application authors, through glfwWindowHint. but i've got at least two applications here that do this (one without source code and perhaps not using glfw)
2023-06-04 23:38:53 <ovf> ok. i think this and the subsequent ifs should be swapped around: https://github.com/glfw/glfw/blob/master/src/egl_context.c#L643
2023-06-04 23:40:50 <ovf> by default major version is 1 (https://gitlab.freedesktop.org/mesa/mesa/-/blob/main/src/egl/main/eglcontext.c#L613) and needs to be set to at least 2 by the relevant thing that glfw does after no_error, thus the latter request fails.
2023-06-04 23:41:23 <ovf> i'll report the bug to glfw tomorrow
2023-06-04 23:46:00 <psykose> glad you figured it out :)
2023-06-05 04:31:38 <somercet1> oof. upgraded ncurses, nss and musl, and now firefox crashes after the ProfileManager, after the profile browser window opens, but before it can show anything
2023-06-05 04:32:19 <somercet1> gonna restart services, see if it's dbus or some such. mate-terminal and hexchat still work
2023-06-05 04:32:43 <psykose> sounds nss related but it works for me
2023-06-05 04:36:23 <somercet1> hmmm, nope, FF is still non compos mentis
2023-06-05 04:43:52 <psykose> easy way to test that is
2023-06-05 04:43:53 <psykose> doas apk add -i 'nss<3.90' -X http://dl-cdn.alpinelinux.org/alpine/v3.18/community
2023-06-05 04:43:56 <psykose> (undo it after obviously)
2023-06-05 04:44:15 <psykose> doesn't sound related to any of the things you say upgraded though
2023-06-05 04:44:45 <psykose> except that
2023-06-05 04:46:30 <somercet1> hang on...
2023-06-05 04:49:15 <somercet1> working
2023-06-05 04:49:51 <somercet1> ty
2023-06-05 04:49:55 <psykose> hm :-)
2023-06-05 04:50:04 <psykose> now i am confused why it doesn't break it for me
2023-06-05 04:51:22 <somercet1> nspr didn't upgrade
2023-06-05 04:51:34 <psykose> why would it
2023-06-05 04:51:38 <somercet1> also, I am on sway/wayland, if that makes a diff.
2023-06-05 04:51:54 <psykose> or do you mean you have some.. older than 4.35 nspr
2023-06-05 04:52:25 <somercet1> no, it's 4.35.
2023-06-05 04:52:41 <psykose> what do you mean then
2023-06-05 04:53:00 <somercet1> They are usually installed as a pair, nss and nspr
2023-06-05 04:53:28 <somercet1> I thought maybe there was an nspr upgrade I didn't get
2023-06-05 04:53:38 <psykose> not the case, no
2023-06-05 04:54:44 <psykose> you can run `{ cat /etc/apk/repositories; apk list -O; apk version } > log` and send it to me i guess
2023-06-05 04:55:46 <somercet1> preferred pastebin?
2023-06-05 04:56:41 <somercet1> https://pastebin.com/htVDzarw
2023-06-05 05:01:27 <psykose> pipewire-media-session was removed so you have to delete that
2023-06-05 05:01:30 <psykose> wouldn't cause this issue tho
2023-06-05 05:01:37 <psykose> same for udevil
2023-06-05 05:07:03 <somercet1> hmmm
2023-06-05 05:08:32 <psykose> also not sure why lame is downgraded but also wouldn't cause the issue
2023-06-05 05:08:43 <psykose> (same for libproxy)
2023-06-05 05:10:31 <somercet1> probably the lame rdeps: firefox deadbeef ffmpeg yt-dlp mpv
2023-06-05 05:12:58 <somercet1> brb coffee
2023-06-05 05:14:15 <psykose> they don't hold it back or anything
2023-06-05 05:14:52 <psykose> the libs were split so they depend on lame-libs instead, nothing depends on lame, but having lame installed would also be fine
2023-06-05 05:15:10 <psykose> unless you did lame<lame-3.100-r3 or something it doesn't make sense
2023-06-05 05:15:24 <psykose> what's the output of `apk add 'lame>lame-3.100-r3'`
2023-06-05 05:22:30 <somercet1> added lame-libs, upgraded lame
2023-06-05 05:23:57 <psykose> that makes no sense unless you had added it as lower before :p
2023-06-05 05:24:41 <psykose> what's in your /etc/apk/world
2023-06-05 05:26:44 <somercet1> https://pastebin.com/WTATKFP0
2023-06-05 05:27:24 <psykose> nothing weird in there
2023-06-05 05:28:18 <psykose> now if you remove the <3.90 from nss does firefox crash again
2023-06-05 05:30:07 <somercet1> nice and hard, just as before
2023-06-05 05:31:53 <psykose> no idea why tbh but nobody else i asked can reproduce it either
2023-06-05 05:31:54 <psykose> just works
2023-06-05 05:32:06 <psykose> perhaps by magic moving to wireplumber and rebooting works
2023-06-05 05:32:58 <somercet1> magic?
2023-06-05 05:35:30 <somercet1> nope
2023-06-05 05:35:36 <somercet1> still crashes
2023-06-05 06:02:48 <thought> I might write a cron script to clean up the last day's directory structure from my currently firehose output, so if I'm in here again, prepare for stupid questions.
2023-06-05 06:04:08 <somercet1> 
ACTION puts on seatbelt

2023-06-05 06:04:13 <somercet1> ready
2023-06-05 08:05:58 <LACampbell> am I supposed to be using eudev or mdev or udev. 
2023-06-05 08:06:02 <LACampbell> for x11
2023-06-05 08:11:04 <psykose> 1 and 3 are the same
2023-06-05 08:11:10 <psykose> that one
2023-06-05 08:13:32 <ncopa> x11 needs libudev for hotplugging input devices, so you need eudev
2023-06-05 08:51:35 <LAC-Laptop> slightly regretting alpine linux on the desktop. borked my system, udev was uninstalled and now I can't connect to the internet.
2023-06-05 08:53:04 <ncopa> how got udev uninstalled?
2023-06-05 08:53:55 <LAC-Laptop> I did it. was messing around with my systems because it had issues with some peripherals that worked fine on other machines
2023-06-05 08:54:26 <LAC-Laptop> trying to get my system to recognise wlan0 or eth0, but apparently it was udev that found that
2023-06-05 08:59:35 <LAC-Laptop> ok I modprobed and got eth0 up. feel like a hacker
2023-06-05 09:01:03 <lopid> did you do it at a terminal?
2023-06-05 09:01:54 <psykose> setup-devd udev after you get internet back
2023-06-05 09:02:51 <LAC-Laptop> lopid: yeah, well my peripherals don't work when I startx lol
2023-06-05 09:03:37 <lopid> then you *are* a hacker
2023-06-05 09:08:53 <LAC-Laptop> god damn right I am. my ethernet cable was broken or something, so I had to modprobe my chipset, just to get wlan0 up, and I couldn't use setup-interfaces, so I had to get it an address with dhcpd... dhcpdsdcd... something. and then I could get udev and probe it. and my peripherals appear to be working!
2023-06-05 09:09:24 <lopid> calm down, neo
2023-06-05 09:11:45 <LAC-Laptop> my trackball still works after hibernating, and it remembered it was in lefthand mode.
2023-06-05 09:11:48 <LAC-Laptop> I am the chosen one
2023-06-05 09:23:46 <dekedro> I did hit firefox issue as well, downgrading nss helped
2023-06-05 09:24:18 <somercet1> dekedro, are you on wayland, X11, or XWayland?
2023-06-05 09:24:53 <psykose> i have a strong hunch that a rebuild downgrade won't fix it
2023-06-05 09:25:06 <psykose> or/and that it'll be broken again on next firefox upgrade (that will need 3.90 anyway)
2023-06-05 09:26:20 <psykose> lets see tho
2023-06-05 09:26:21 <dekedro> sway, issue present on both Wayland and XWayland
2023-06-05 09:29:29 <somercet1> hmm, ty
2023-06-05 09:30:26 <dekedro> also last message in stderr before ff closes is "Illegal instruction", haven't seen that before
2023-06-05 09:30:40 <psykose> can you get the x/5i $pc of that
2023-06-05 09:30:57 <psykose> from gdb
2023-06-05 09:32:17 <psykose> and the backtrace too i guess with musl-dbg firefox-dbg just to see where it specifically is by chance
2023-06-05 09:38:26 <psykose> ulimit -c unlimited to make a core to make it easier
2023-06-05 09:43:39 <somercet1> psykose, how do you load /usr/lib/debug files?
2023-06-05 09:43:54 <psykose> you don't, they get loaded by gdb automatically
2023-06-05 09:45:17 <dekedro> yeah, for some reason gdb doesn't like firefox-dbg files: https://pastebin.com/qhjexmUQ
2023-06-05 09:46:29 <psykose> the entire backtrace is in nss so that's not even relevant
2023-06-05 09:46:37 <psykose> thanks
2023-06-05 09:48:49 <dekedro> np
2023-06-05 09:48:56 <psykose> hmm
2023-06-05 09:49:05 <psykose> what cpu do you have
2023-06-05 09:49:35 <dekedro> i7-4770HQ
2023-06-05 09:49:36 <psykose> and /proc/cpuinfo
2023-06-05 09:49:39 <psykose> sigh
2023-06-05 09:49:40 <psykose> well
2023-06-05 09:49:47 <psykose> the issue is that is adx (broadwell/ryzen)
2023-06-05 09:49:59 <psykose> something in nss either misdetects a feature or just uses that unconditionally
2023-06-05 09:50:03 <psykose> that makes it easier to track down
2023-06-05 09:50:32 <psykose> proc/cpuinfo puts those in 'adx' flag afaik
2023-06-05 09:53:26 <dekedro> hmm, I see no 'adx', https://pastebin.com/bw0zQeSq
2023-06-05 09:53:31 <psykose> nss/lib/freebl/verified/curve25519-inline.h
2023-06-05 09:53:36 <psykose> has.. that all over the place
2023-06-05 09:53:38 <psykose> no checks anywhere
2023-06-05 09:57:49 <psykose> can you test https://img.ayaya.dev/ZShN357Tf8ug
2023-06-05 10:04:51 <dekedro> yup, with this patch firefox works as before
2023-06-05 10:05:05 <psykose> ty
2023-06-05 10:05:06 <dekedro> so my cpu is too old for firefox now? :)
2023-06-05 10:05:14 <psykose> it's not firefox, but nss
2023-06-05 10:05:23 <psykose> they'll probably fix it
2023-06-05 10:05:32 <psykose> broadwell is way too new as an auto baselie
2023-06-05 10:05:45 <psykose> then again, their build system is quite weird, and this is only in the gyp one i think
2023-06-05 10:07:30 <dekedro> cool, thanks for figuring this out
2023-06-05 10:08:43 <psykose> i guess i should add some weird qemu-x86_64 -cpu low-baseline to a list of manual tests to run myself :p
2023-06-05 10:12:13 <somercet1> yah, no adx here either
2023-06-05 10:21:45 <justagoose> guuys the speed-fapping carnival is on in the harbour tonght,
2023-06-05 10:21:55 <justagoose> you going to participate?
2023-06-05 10:23:36 <psykose> the what
2023-06-05 10:24:16 <pj> the speed-fapping
2023-06-05 11:12:31 <somercet1> ... maybe this topic is for #OT?
2023-06-05 11:14:15 <psykose> it's a spam bot pretty sure
2023-06-05 11:36:16 <duckl1ng[m]> ```diff
2023-06-05 11:36:41 <duckl1ng[m]> 
ACTION sent a diff code block: https://matrix.org/_matrix/media/v3/download/matrix.org/xfptPEhJyiIiybqbwTeXHoyh

2023-06-05 11:36:59 <duckl1ng[m]>  * ```git... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/FXcUyIRwSIhSfQCKZnWbnOyV>)
2023-06-05 11:37:21 <psykose> ?
2023-06-05 11:37:25 <lopid> saved
2023-06-05 11:46:49 <duckl1ng[m]> oopsie, wrong room
2023-06-05 12:01:41 <raspbeguy> hello, how can I set mtu on alpine? I tried this https://termbin.com/hrvv but when I do `ifup eth2` I get "ip: ioctl 0x8922 failed: Invalid argument". The interface is up but has the wrong mtu (1500).
2023-06-05 12:09:26 <raspbeguy> oh I think I figured out that my hardware doesn't support such a MTU... Sad, but that's the answer
2023-06-05 12:12:22 <psykose> sometimes it just is that easy
2023-06-05 12:12:45 <psykose> whadya want to set it for? you cant really have >1500 across the internet anyway, it just auto fragments
2023-06-05 12:12:52 <psykose> jumbo is just for local networking stuff
2023-06-05 12:13:05 <psykose> or something like that
2023-06-05 12:14:14 <ikke> Or if you have dedicated lines  support it
2023-06-05 12:14:15 <raspbeguy> psykose, I am trying to expose iscsi targets
2023-06-05 12:17:35 <raspbeguy> this is a NUC with a single interface, and I had 2 unused USB to RJ45 adapters that I pluged in it, so I could do cheap multipath on 2 GB links
2023-06-05 12:18:17 <psykose> ah
2023-06-05 12:18:25 <psykose> never done anything like that
2023-06-05 12:18:57 <raspbeguy> never did either, but it's interresting to try
2023-06-05 12:24:24 <nangel> interesting ... apparently there are some usb-rj45 adapters that support jumbo frames.  I did not expect that.
2023-06-05 12:27:44 <em6> hey, i apparently need to install "libmvec" library for renoise. is this possible? I have installed gcompat and reaper works all fine, but not renoise
2023-06-05 12:29:23 <em6> here is the errorhttps://mclo.gs/t0SM5iW
2023-06-05 12:29:31 <em6> oh sorry
2023-06-05 12:29:31 <em6> https://mclo.gs/t0SM5iW
2023-06-05 12:30:03 <em6> sorry, am a bit noob
2023-06-05 12:32:28 <Hello71> do jumbo frames even help that much if gro/gso are enabled
2023-06-05 12:35:22 <Hello71> it's 4.4% more bandwidth for TCP over IPv4
2023-06-05 12:53:57 <raspbeguy> nangel, mind to share?
2023-06-05 12:55:12 <nangel> amazon search shows that plugable and some other companies make them - search "usb rj45 jumbo frames" they all look to be usb-c though. :(
2023-06-05 13:38:08 <thought> Who's still using alpine for a desktop?
2023-06-05 13:38:20 <ikke> Plenty of people are
2023-06-05 13:38:26 <aron> me
2023-06-05 13:38:46 <thought> ikke, I'm asking in the tone that aron responded to
2023-06-05 13:38:54 <thought> aron: what's your apk list like
2023-06-05 13:39:29 <aron> I'm at alpine-3.15 right now
2023-06-05 13:40:22 <quinq> Why “still” though? Has a random internet person officially deprecated Alpine for a desktop usage?
2023-06-05 13:41:26 <ovf> my impression was that the edge package base has been expanding quite rapidly recently, which could be both a symptom of and a prerequisite for growing desktop use
2023-06-05 13:41:51 <thought> quinq, nope: just saying I've asked before.
2023-06-05 13:42:13 <aron> thought: why is apk list relevant?
2023-06-05 13:42:38 <thought> aron: I'm just curious what you're running for your desktop experience, and wasn't going to ask "what for this? what2 for that?"
2023-06-05 13:42:55 <ovf> that'd be /etc/apk/world, not apk list
2023-06-05 13:43:05 <aron> yep :)
2023-06-05 13:43:40 <aron> I prefer xfce4
2023-06-05 13:44:22 <thought> Wait a MOMENT
2023-06-05 13:44:41 <thought> What the hell am I thinking
2023-06-05 13:45:18 <thought> Why's it called world?
2023-06-05 13:45:31 <quinq> ok thought :)
2023-06-05 13:45:56 <thought> lmao, you can stop waiting
2023-06-05 13:45:58 <ovf> it's a set of package constraints the describe the entirety of your system, i.e. your world
2023-06-05 13:46:40 <thought> Very interesting.
2023-06-05 13:47:08 <quinq> https://yewtu.be/64Aoy91d_SM
2023-06-05 13:48:24 <aron> quinq: did I answer your question?
2023-06-05 13:48:47 <quinq> No :(
2023-06-05 13:48:48 <thought> I think phrasing the question I was about to ask solved my problem: I was thinking about this nvr setup, and about what happens if ffmpeg errors out because the net's down or the camera gets unplugged, and I think I can just increase the restart delay and max restarts to be satisfactory
2023-06-05 13:49:18 <aron> quinq: oh, my mistake, it was thought :P
2023-06-05 13:49:28 <quinq> :>
2023-06-05 13:49:28 <thought> aron: almost. what's the audio stack like
2023-06-05 13:49:51 <aron> alsa
2023-06-05 13:50:05 <quinq> thought, my answer would be I do, and it works well for me, though I'm not sure if that really helps you
2023-06-05 13:50:18 <quinq> I use alsa and sndio and it works in that regard
2023-06-05 14:24:13 <somercet1> Just installed wireplumber, and apps no longer restore the last volume state. How do I enable that?
2023-06-05 14:24:53 <psykose> restore it in what sense
2023-06-05 14:26:13 <psykose> it does a bunch of per-stream whatever state management by default, but maybe you're thinking of something specific
2023-06-05 14:26:46 <somercet1> on invocation, mpv would restore the sound level to the last time it was executed.
2023-06-05 14:28:02 <psykose> i.e. before, each time you opened mpv, the 'mpv app volume' would be where you left it
2023-06-05 14:28:05 <psykose> now it defaults to something
2023-06-05 14:28:11 <somercet1> top volume
2023-06-05 14:28:16 <somercet1> yes
2023-06-05 14:29:26 <psykose> not entirely sure, i always expected it to reset and got exactly that (because i have volume=30 set in mpv conf)
2023-06-05 14:29:35 <psykose> you can see if https://pipewire.pages.freedesktop.org/wireplumber/configuration/main.html has something but i am not sure
2023-06-05 14:30:01 <psykose> all other applications remember their volume state it seems, so maybe this is actually intended for mpv volume
2023-06-05 14:30:42 <psykose> (everything else i just tested keeps volume across close-reopen)
2023-06-05 14:49:17 <somercet1> hmm
2023-06-05 16:45:36 <Quillith> hey, what's a good way to extract rar files on Alpine? Internet says to use unrar but it doesn't seem to be in the repo
2023-06-05 16:45:55 <ikke> Quillith: bsdtar from libarchive-tools
2023-06-05 16:46:01 <Quillith> Thank you!
2023-06-05 17:01:04 <thought> ikke, you don't like 7z?
2023-06-05 17:01:11 <ikke> thought: try it
2023-06-05 17:01:23 <thought> try bsdtar or 7z?
2023-06-05 17:01:26 <ikke> 7z
2023-06-05 17:01:31 <thought> i swear i remember using p7z
2023-06-05 17:01:39 <thought> back when alpine was the main driver
2023-06-05 17:03:24 <ikke> 7e060e5e91b48712c5da470ccf4a84f6af8611e0
2023-06-05 17:03:41 <ikke> 2b3abc74c5f33e4ab6b2b23a1716c813af73edc1
2023-06-05 17:03:45 <psykose> (does the same thing as 7zip)
2023-06-05 17:03:57 <psykose> both of them use the same unrar code for rar so it was only mistakenly enabled tho
2023-06-05 17:04:26 <psykose> unrar code as in from unrar
2023-06-05 17:10:00 <thought> oh, what da hell
2023-06-05 17:10:30 <thought> does bsdtar support lzma?
2023-06-05 17:10:50 <aron> yes, it does
2023-06-05 17:10:57 <ikke>  https://libarchive.org/
2023-06-05 17:25:09 <jvoisin> p7z has horrible code, and is a thing from the past that shouldn't be used, now that 7z works on linux.
2023-06-05 17:25:31 <ikke> jvoisin: that's why it was removed
2023-06-05 17:32:22 <psykose> 7z also has horrible code
2023-06-05 17:32:31 <psykose> it is just like 5% less scuffed
2023-06-05 17:32:42 <psykose> and seemingly works without random bug reports from people lol
2023-06-05 17:33:50 <jvoisin> glory be
2023-06-05 17:33:57 <maria> psykose: i figured out the flatpak problem i used to have. installing flatpak, adding flathub as user, installing discord as user, and launching discord results in an error about being unable to open /var/lib/flatpak/repo. adding flathub to system and then removing it fixes it. i have no idea why the discord flatpak does this.
2023-06-05 17:34:59 <psykose> bizarre!
2023-06-05 17:35:11 <psykose> that sounds easy enough to try reproduce elsewhere and report upstream
2023-06-05 17:38:18 <maria> oh no it's even worse. running doas flatpak info fixes it too.
2023-06-05 17:50:19 <thought> Why is p7z so bad?
2023-06-05 17:53:07 <jvoisin> the code is bad
2023-06-05 17:53:33 <thought> That's what I mean: Why is its code bad?
2023-06-05 17:53:42 <thought> What is your favorite code project?
2023-06-05 18:04:57 <psykose> yours
2023-06-05 19:29:57 <maria> hi i'm trying to launch minecraft 1.12.2 through prismlauncher and i'm getting errors about openal. without configuration it segfaults and with native libraries enabled it says it's unable to locate the openal library. i have openal-soft installed and am unsure of what to do next to get this to work. does someone here know or could help me?
2023-06-05 19:31:18 <psykose> the post-install tells you why that wouldn't work
2023-06-05 19:31:20 <psykose> not really fixable
2023-06-05 19:31:37 <maria> that specified it was for version of minecraft 1.19 and newer, though.
2023-06-05 19:31:55 <maria> i assume older versions were fine
2023-06-05 19:31:56 <psykose> no, for "all" and that 1.19 and newer have no workaround
2023-06-05 19:32:04 <psykose> well, in that case the workaround would
2023-06-05 19:32:27 <psykose> have to lower lwjgl for older ones too iirc
2023-06-05 19:33:27 <maria> 1.12.2 is already using lwjgl 2. i'll try some different versions of it.
2023-06-05 19:36:00 <NepNepdmsalwaysopen[m]> I mean, strictly speaking there is always a workaround, it's technically possible to recompile and replace all native libraries with musl-compatible versions 
2023-06-05 19:36:00 <NepNepdmsalwaysopen[m]> Though of course that would be highly impractical 
2023-06-05 19:36:41 <maria> there's no difference between them. enabling native libraries for openal allows the game to launch there's just no audio.
2023-06-05 19:38:02 <psykose> the audio thing seems separate
2023-06-05 19:38:20 <psykose> or well, that's a random guess
2023-06-05 19:38:21 <psykose> no clue
2023-06-05 19:44:29 <maria> i figured it out. it's a bit cursed. i made a symlink from libopenal.so.1 to libopenal.so
2023-06-05 19:44:41 <maria> works fine now lol
2023-06-05 19:44:52 <psykose> you can add openal-soft-dev to get that anyway
2023-06-05 19:45:10 <psykose> but yes, some stuff opens the bare name with no version
2023-06-05 19:45:14 <psykose> bad practice and yes
2023-06-05 19:45:16 <psykose> yet&
2023-06-05 19:46:09 <maria> i imagine the answer is no but is there a way to install -dev packages automatically like you can with -doc? setting up my dev environment was a bit annoying due to needing to specify them all manually.
2023-06-05 19:46:21 <psykose> nope
2023-06-05 19:46:54 <maria> well thanks anyway for letting me know openal-soft-dev provides the file needed :)
2023-06-05 19:46:56 <psykose> well, finding like 15 -dev packages is probably quite easy compared to anything else hehe
2023-06-05 19:47:12 <psykose> you can also use https://pkgs.alpinelinux.org/contents?file=libopenal.so&path=&name=&branch=edge
2023-06-05 19:47:33 <maria> thanks!
2023-06-05 20:29:42 <WhyNotHugo> What's the recommended approach for cross-compiling a package for aarch64 on x86_64?
2023-06-05 20:29:54 <WhyNotHugo> I mean, are there handy scripts to do the lifting for me? :D
2023-06-05 20:29:58 <ikke> qemu-user
2023-06-05 20:30:00 <ikke> :P
2023-06-05 20:30:06 <psykose> `CBUILD=aarch64 abuild rootbld`
2023-06-05 20:30:23 <psykose> not really guaranteed to work obviously
2023-06-05 20:30:28 <psykose> and that is qemu-user
2023-06-05 20:30:50 <psykose> (apk add qemu-aarch64 qemu-openrc; service qemu-binfmt start)
2023-06-05 20:31:08 <psykose> technically that is not cross compiling
2023-06-05 20:31:22 <psykose> actual cross compiling doesn't have any support
2023-06-05 20:31:38 <psykose> you'd have to manage your own --sysroot with apk --root and probably use clang+lld (which does cross)
2023-06-05 20:31:41 <psykose> but all manual
2023-06-05 20:33:24 <WhyNotHugo> Oh, I expected something that used qemu under the hood, thanks.
2023-06-05 21:01:40 <WhyNotHugo> Didn't expect to see hyprland in aports
2023-06-05 21:04:57 <psykose> got tired of seeing requests constantly
2023-06-05 21:09:32 <WhyNotHugo> Why does it blink so much like a light tube from an abandoned warehouse in a horror movie?
2023-06-05 21:10:45 <psykose> flashy effects
2023-06-05 21:11:17 <abby> that's the snappiness
2023-06-06 01:45:58 <batap> Hi, I’m new to alpine and I’m not experienced with Linux in general. I have an IO ERROR with apk when I do "apk add" or "update" etc... I don't know what I can do
2023-06-06 02:24:42 <Saijin_Naib> Can you paste the exact error and the commands you're executing? 
2023-06-06 02:50:29 <Saijin_Naib> Anyone on Edge x86_64 using picard and getting a crash of the entire DE (maybe DM?) when launching picard?
2023-06-06 03:50:15 <Someone[m]1> alpine linux edge always have problem with some mesa-dri-gallium and mesa-va-gallium lol
2023-06-06 03:50:29 <Someone[m]1> I had to roll back to v3.18 for a day
2023-06-06 04:01:48 <Saijin_Naib> Ah, maybe. I'm amazed that I can't even catch anything from the console before I log in, even. I guess I'd have to kill lightdm and try again...
2023-06-06 05:31:45 <quinq> Saijin_Naib, I have that exact problem, but on another distro on another laptop
2023-06-06 05:31:52 <quinq> (picard problem)
2023-06-06 05:32:15 <quinq> And since a few months
2023-06-06 05:32:30 <quinq> It works on my main laptop though, so didn't take the time to debug…
2023-06-06 06:44:17 <somercet1> hyprland looks like the Enlightenment of Wayland.
2023-06-06 10:10:03 <oleovit> hi. since chromium v.114, the context-menu is flickering, even if you disable the #chrome-refresh-2023
2023-06-06 10:10:36 <oleovit> does anyone experience this too? wayland, sway
2023-06-06 10:34:23 <somercet1> oleovit, I get it sometimes when the menu is low enough that it disappears below the bottom screen edge.
2023-06-06 10:34:37 <somercet1> that's in FF.
2023-06-06 10:40:30 <oleovit> yep, experienced a similiar behavior in firefox too, long ago. i guess, in chromium, this has to to with their unnecessary "refresh". but it can be anything. wayland, drivers, my ~ flags, .. et cetera. browsers are complex. very. i dont know.
2023-06-06 10:49:35 <oleovit> anyway, a f.. norovirus is forcing me to logout. sick. let's see. i guess, this is not that alpine related, so i dont want to waste dev-time. have a nice day/night and thx again for all the work to everyone.
2023-06-06 11:40:09 <Saijin_Naib> quinq: any idea of what the cause and work around might be? 
2023-06-06 12:35:00 <irreleph4nt> Hi. I am trying to add the msgpack extension to the official php 8.1 alpine-based Docker Image. Compiling the extension works fine, but when I try to enable it or run "ldd msgpack.so" the symbols of other php modules (like zend or session) can't be found so msgpack can't be enabled
2023-06-06 12:36:06 <irreleph4nt> This is the official Dockerfile I am using: https://github.com/docker-library/php/blob/master/8.1/alpine3.16/fpm/Dockerfile
2023-06-06 12:56:12 <ovf> irreleph4nt: any particular reason you're doing that instead of using alpine's php81-pecl-msgpack?
2023-06-06 13:00:20 <ovf> seeing that it's not using alpine's php distribution whatsoever, i guess the answer is that you have to manually put all of msgpack's dependencies where php expects to find them. none of this bears any relation to alpine.
2023-06-06 13:08:00 <irreleph4nt> ovf, reason is that I need to install a number of extensions that are not delivered with standard php. Namely pdo_oci / oci8 and pdo_sqlsrv / sqlsrv.
2023-06-06 13:08:02 <irreleph4nt> I considered using the php81 packages from alpine, but given how these packages will update in version rather silently, I don't think I will be able to set up repeatable builds of php + extensions in a Dockerfile.
2023-06-06 13:08:48 <irreleph4nt> I am doing this for work BTW, so I can't have the build of an ext fail just because alpine upgraded the php81 packages
2023-06-06 13:10:08 <ovf> are you managing security updates yourself then?
2023-06-06 13:10:24 <irreleph4nt> ovf, I was going to do that. yes
2023-06-06 13:10:53 <irreleph4nt> Given how I am behind a company proxy, my docker builds also can't use pecl or pear. So the standard ways for getting these extensions installed don't wortk for me
2023-06-06 13:14:39 <mercenary> doesn't pecl/pear respect http_proxy?
2023-06-06 13:16:40 <irreleph4nt> It may. But the pecl / pearl URLs are blocked in our Docker build environment. I hence won't be able to connect.
2023-06-06 13:18:23 <mercenary> heh. Alsp it doesn't apparently, but has yet another knob https://github.com/mlocati/docker-php-extension-installer/issues/332
2023-06-06 13:23:19 <irreleph4nt> Does PECL support installing from a local tgz package? I seem to recall that ability to be documented for PEAR
2023-06-06 13:39:09 <dalias> i'm trying to use the arm-none-eabi cross compiler package and getting errors like this:
2023-06-06 13:39:27 <dalias> /usr/lib/gcc/arm-none-eabi/12.2.0/../../../../arm-none-eabi/bin/ld: warning: /usr/lib/gcc/arm-none-eabi/12.2.0/../../../../arm-none-eabi/lib/thumb/v7e-m/nofp/libg_nano.a(libc_a-memcpy-stub.o) uses 2-byte wchar_t yet the output is to use 4-byte wchar_t; use of wchar_t values across objects may fail
2023-06-06 13:39:46 <dalias> looks like newlib was miscompiled with a bad nondefault abi .. ?
2023-06-06 13:47:13 <dalias> yes the APKBUILD is explicitly building it wrong .... grrrr
2023-06-06 13:47:21 <dalias> export CFLAGS_FOR_TARGET="-Os -ffunction-sections -fdata-sections -fshort-wchar"
2023-06-06 13:48:10 <dalias> WHY
2023-06-06 13:50:19 <aron> what are you want to achive?
2023-06-06 13:50:47 <aron> some context would help
2023-06-06 13:55:32 <ovf> dalias: why are you using the _nano version?
2023-06-06 13:56:09 <ovf> please read the comment at https://git.alpinelinux.org/aports/tree/community/newlib/APKBUILD#n54
2023-06-06 14:01:20 <dalias> i'm just trying to build firmware that uses it
2023-06-06 14:04:29 <ovf> specifically newlib_nano?
2023-06-06 14:06:41 <dalias> yes it specifically uses --specs=nano.specs --specs=nosys.specs
2023-06-06 14:06:44 <dalias> but here's the problem
2023-06-06 14:06:49 <dalias> only for the final link *grrrr*
2023-06-06 14:06:56 <dalias> wonder how this works on other systems
2023-06-06 14:13:11 <dalias> ok the specs file is broken
2023-06-06 14:13:28 <dalias> it has -isystem =/include/newlib-nano %(nano_cpp_unique_options) but there's no such directory in the package
2023-06-06 14:13:52 <dalias> no, there is under include..
2023-06-06 14:13:56 <dalias> but it's not working 
2023-06-06 14:22:33 <ovf> i guess nano.specs should pass on -fshort-wchar?
2023-06-06 14:23:51 <dalias> yes
2023-06-06 14:23:59 <dalias> and it's broken in some other ways too i think
2023-06-06 14:24:02 <dalias> that's only a warnng
2023-06-06 14:24:06 <dalias> i'm also getting..
2023-06-06 14:24:23 <dalias> (.text.get_eit_entry+0x84): undefined reference to `__exidx_start'
2023-06-06 14:24:35 <dalias> and other errors
2023-06-06 14:25:57 <dalias> hacking in a --sysroot to use picolibc seems to make it build fine but i'm not sure if it actually works :-p
2023-06-06 16:57:34 <psykose> you should open an issue about it then
2023-06-06 17:04:31 <quinq> No, Saijin_Naib, sorry :/
2023-06-06 17:04:40 <quinq> I'm just using it on my main music machine
2023-06-06 17:07:54 <psykose> it sounds like it uses non-nano specs for the whole build and then nano for linking which is obviously not going to work and the actual issue
2023-06-06 17:08:02 <lopid> your main music machine? down with the scene, like a sex machine? (get down, get on up)?
2023-06-06 17:09:22 <dalias> psykose, yes but i changed the CFLAGS to use the nano specs for the whole build and it's still broken
2023-06-06 17:09:40 <psykose> i see no open issue about it so i guess it doesn't exist
2023-06-06 17:09:41 <psykose> :-)
2023-06-06 17:09:49 <dalias> i'm just going to copy the target includes/libs from binaries elsewhere and use --sysroot :-p
2023-06-06 17:10:05 <dalias> because from the perspective of the host/distro they're data not code
2023-06-06 17:10:17 <dalias> and there's no harm in just using ones from a foreign source as such
2023-06-06 17:12:56 <quinq> lopid, not quite :)
2023-06-06 17:13:28 <lopid> oh
2023-06-06 17:34:13 <Saijin_Naib> quinq: all good. I'm going to try the Flatpak and see if that also breaks things 
2023-06-06 17:34:59 <Saijin_Naib> I just switched off the Flatpak to try and have all software from APK if possible, so if this issue has been around for a bit, that's probably why I missed it 
2023-06-06 17:40:07 <psykose> picard always worked for me fine
2023-06-06 17:40:21 <psykose> in wayland anywa
2023-06-06 17:43:25 <quinq> Yeah, it works fine always until it doesn't :D
2023-06-06 18:38:25 <irreleph4nt> Hi. Yet another PHP question: Given how I was unable to get msgpack to work, I switched to the official alpine php81* packages. Out of curiosity I ran ldd on the .so files of these after installation and they all report a number of missing symbols. Is that to be expected?
2023-06-06 18:44:03 <psykose> yes, the symbols come from php
2023-06-06 18:44:14 <psykose> the extensions are dlopened by php
2023-06-06 18:44:22 <psykose> ldd'ing any python .so module does the same
2023-06-06 18:44:24 <psykose> etc
2023-06-06 18:45:34 <bl4ckb0ne> the s in php stands for security
2023-06-06 18:47:41 <abby> php: hypertext preprocesecuritysecurityor
2023-06-06 19:49:40 <irreleph4nt> Thanks! I am not entirely sure I understand how any of that translates into missing symbols, but I'll take it :D
2023-06-06 19:54:28 <psykose> they come from php
2023-06-07 00:00:11 <dalias> psykose, i don't see any -fshort-wchar in debian's nano.specs file either. it looks the same. so i wonder if someone put a bunch of CFLAGS in alpine's newlib_nano build that weren't supposed to be there...
2023-06-07 00:00:27 <psykose> i don't see an open issue about it either
2023-06-07 00:00:36 <dalias> ok :)
2023-06-07 00:01:09 <dalias> i'm trying to figure out what the actual problem is tho before i report a problem different from the underlying issue
2023-06-07 00:02:22 <psykose> it probably was just accidentally added
2023-06-07 00:02:24 <psykose> (2 years ago)
2023-06-07 00:02:33 <dalias> ?
2023-06-07 00:02:41 <psykose> the flag
2023-06-07 00:02:52 <psykose> though that would surprise me as i'd have expected someone else to have had it break something
2023-06-07 00:02:57 <dalias> i didn't see it changing in the history but i might have overlooked it
2023-06-07 00:03:00 <psykose> so not really sure if indeed as you say it's the underlying issue
2023-06-07 00:03:04 <dalias> i doubt anything uses wchar in this kind of embedded
2023-06-07 00:03:10 <dalias> but it puts an ABI tag on the elf files
2023-06-07 00:03:15 <dalias> that makes warnings at link time
2023-06-07 00:03:27 <psykose> ah
2023-06-07 00:03:32 <psykose> it was indeed 4 years on initial creation
2023-06-07 00:03:46 <dalias> i suspect old binutils didnt complain
2023-06-07 00:03:57 <psykose> though initially in cflags and not _for_target
2023-06-07 00:04:01 <dalias> other stuff is broken too tho -- it might be a mismatched libgcc or something
2023-06-07 00:04:10 <dalias> uhg, in CFLAGS is even worse..
2023-06-07 00:04:15 <dalias> that breaks host ABI
2023-06-07 00:04:19 <psykose> yeah, that one was just wrong
2023-06-07 00:04:38 <dalias> it might not be a bad idea to trap that somewhere
2023-06-07 00:04:43 <dalias> #error in headers if it's been set
2023-06-07 00:05:12 <psykose> yes
2023-06-07 00:05:28 <psykose> you have to be quadruple sure it's not some niche valid use though :p
2023-06-07 00:07:16 <dalias> it's not. you can't use libc if you're passed CFLAGS that change the ABI to one that mismatches :-p
2023-06-07 00:20:03 <Saijin_Naib> psykose: I don't have a working Wayland install to test... Something to try 
2023-06-07 00:20:17 <psykose> yeah don't really a suggestiong
2023-06-07 00:20:26 <psykose> i mean, opening an application and X crashing is firmly an X bug
2023-06-07 00:20:30 <psykose> doesn't really matter what it does
2023-06-07 00:20:50 <psykose> so the questions are 'why does it crash (is it the python qt bindings or qt codepath itself, and what part of x fails)
2023-06-07 00:21:15 <psykose> could also be the xorg wm crashing i guess
2023-06-07 00:26:08 <psykose> that said i havent used xorg in like 6 years and firmly Don't Care if it's broken somewhere or not personally
2023-06-07 00:26:37 <psykose> there's probably a soup of random bugs i ran into 10 years ago that are still there
2023-06-07 00:27:01 <psykose> unlike wayland where every single bug i ever encountered got fixed at some point
2023-06-07 00:27:21 <psykose> (doesn't stop new ones from happening :-) )
2023-06-07 00:35:59 <Saijin_Naib> Flatpak kills XORG, too
2023-06-07 00:36:03 <Saijin_Naib> No easy fix
2023-06-07 00:36:14 <psykose> yeah something sounds giga broken
2023-06-07 00:36:18 <psykose> which wm again
2023-06-07 00:36:23 <Saijin_Naib> If XFCE was Wayland, I'd be in Wayland
2023-06-07 00:36:44 <psykose> in roughly 2 years it probably would be
2023-06-07 00:36:45 <Saijin_Naib> XFCE4WM, lightdm (if that matters) 
2023-06-07 00:36:59 <psykose> xfwm4 has a wayland port somewhere in some progress
2023-06-07 00:37:34 <Saijin_Naib> Yeah, I saw! And many tools are getting compatible. Causes some rage in that community 🙄
2023-06-07 00:37:45 <vyivel> https://github.com/adlocode/xfwm4/tree/wayland this?
2023-06-07 00:37:55 <psykose> ye
2023-06-07 00:37:59 <Saijin_Naib> Hmm 🤔, I guess I can muck around and go GDM+GNOME to finish testing 
2023-06-07 00:38:25 <psykose> i'm happy it is wlroots
2023-06-07 00:38:46 <Saijin_Naib> All I know is each one is super different. Why is that a good one? 
2023-06-07 00:39:07 <Saijin_Naib> I'm not questioning your judgement, I trust that, I'm just curious 
2023-06-07 00:40:41 <psykose> well there's no other one that is usable (unless i'm forgetting)
2023-06-07 00:40:52 <Saijin_Naib> 🤣 
2023-06-07 00:40:57 <psykose> gnome's mutter and kde's kwin implement their own wayland compositor stack from scratch
2023-06-07 00:41:12 <psykose> so i'm happy they picked wlroots instead of making their own like the other de's :-)
2023-06-07 00:41:36 <vyivel> libweston \o/
2023-06-07 00:41:40 <vyivel> there's also smithay
2023-06-07 00:41:47 <psykose> yes, those two as well
2023-06-07 00:42:47 <psykose> does anything use those? the latter is also an entire wayland toolkit, so clients use it, but i haven't seen a wm aside from cosmic
2023-06-07 00:42:58 <psykose> former is.. uhh idk who uses any part of weston except industrial
2023-06-07 00:43:18 <psykose> 'industrial' meaning any business stuff, appliance devices
2023-06-07 00:43:33 <Saijin_Naib> Is weston Maynard? 
2023-06-07 00:43:47 <psykose> yes
2023-06-07 00:43:48 <psykose> https://github.com/raspberrypi/maynard
2023-06-07 00:43:50 <psykose> apparently
2023-06-07 00:44:08 <psykose> ah, there's Mir too
2023-06-07 00:44:26 <psykose> as a standalone thing
2023-06-07 00:44:40 <Saijin_Naib> Mir was Ubuntu, right? Didn't that stall? 
2023-06-07 00:45:25 <psykose> no, still maintained
2023-06-07 00:47:19 <Saijin_Naib> Ahh... I have a half decade of this to catch up on, apparently 
2023-06-07 01:11:47 <Guest2367> Hi, I am trying to use android-tools on alpine linux, but I the the error: "Error relocating /usr/bin/fastboot:<unreadable hash>. symbol not found"
2023-06-07 01:12:17 <Guest2367> After some searching, it seems like it may be a musl versus glibc issue. Any ideas for a workaround?
2023-06-07 01:12:34 <psykose> paste the actual error
2023-06-07 01:13:17 <kwadroke[m]> installed android-tools earlier on my pinephone via apk add
2023-06-07 01:13:17 <Guest2367> Error relocating /usr/bin/fastboot: _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE15_M_replace_coldEPcmPKcmm: symbol not found
2023-06-07 01:13:29 <psykose> where did you get android-tools from
2023-06-07 01:13:43 <Guest2367> apk add android-tools
2023-06-07 01:13:51 <psykose> paste `cat /etc/apk/repositories`
2023-06-07 01:14:47 <Guest2367> http://dl-cdn.alpinelinux.org/alpine/edge/main
2023-06-07 01:14:50 <Guest2367> http://dl-cdn.alpinelinux.org/alpine/edge/community
2023-06-07 01:14:52 <Guest2367> http://dl-cdn.alpinelinux.org/alpine/edge/community
2023-06-07 01:14:54 <Guest2367> http://dl-cdn.alpinelinux.org/alpine/edge/community
2023-06-07 01:14:56 <Guest2367> http://dl-cdn.alpinelinux.org/alpine/edge/testing
2023-06-07 01:15:19 <psykose> did you forget to upgrade to edge with `-a`
2023-06-07 01:15:24 <psykose> just run `apk upgrade -a`
2023-06-07 01:16:09 <Guest2367> Okay, what does that option do?
2023-06-07 01:16:35 <psykose> you need it to upgrade to the new packages in a new repository, otherwise it keeps the ones in the old one
2023-06-07 01:17:37 <Guest2367> Okay, cool. I see it in the help menu now. I was running "apk --help" instead of "apk upgrade --help"
2023-06-07 01:17:47 <Guest2367> Thanks, I'll try it
2023-06-07 01:56:12 <martin1> I think I have a different nickname now, but I just asked about android-tools. I did the "apk upgrade -a", which solved the error that I was having, although I am not sure that android-tools is using the usb properly. Now my fastboot command is just waiting at "waiting for any device"
2023-06-07 01:57:55 <psykose> you probably want android-udev-rules for the devices to get assigned to plugdev group, and to add yourself to plugdev group
2023-06-07 01:58:02 <psykose> otherwise only as root can you access the devices afaik
2023-06-07 02:02:51 <martin1> Yes, using root worked!
2023-06-07 02:03:05 <martin1> Awesome, thanks again
2023-06-07 05:37:01 <ricardo> hello!
2023-06-07 05:39:23 <richirm> hello!
2023-06-07 05:40:15 <richirm> I have a question, in case some kind soul wants to solve it for me.
2023-06-07 05:40:54 <richirm> Where does Alpine Linux store time zone information?
2023-06-07 05:41:05 <richirm> Thank you so much!
2023-06-07 05:42:22 <ikke> richirm: /etc/localtime is a symlink to a timezone
2023-06-07 05:42:55 <psykose> nowhere, unless you mean /usr/share/zoneinfo from tzdata, or /etc/localtime that you can symlink to there, or setup-timezone that can either do that for you or install only a single one by cp instead of ln and delete tzdata after
2023-06-07 05:47:49 <richirm> I use Alpine from Docker ( "docker run -di --name alpine alpine; docker exec -ti alpine sh;" ). It would be nice if the time zone could be set somehow.
2023-06-07 05:50:50 <psykose> apk add alpine-conf; /sbin/setup-timezone
2023-06-07 05:51:25 <ikke> or apk add tzdata && ln -s /etc/zoneinfo/<timezone>
2023-06-07 05:52:54 <psykose> missing an argument there pretty sure
2023-06-07 05:56:30 <richirm> Thank you so much!
2023-06-07 05:56:34 <richirm> What I'm looking for is a command that returns the time zone in a string. In Fedora 38, which is the Linux distribution that I use to work, it is done like this:
2023-06-07 05:57:07 <richirm> 03timedatectl | awk '/Time zone:/ {print $3}'
2023-06-07 05:57:41 <psykose> returns what timezone in a string?
2023-06-07 05:57:46 <psykose> the current one?
2023-06-07 05:57:54 <richirm> For example, Europe/Madrid
2023-06-07 05:58:11 <psykose> readlink /etc/localtime | cut -d/ -f5-
2023-06-07 05:58:47 <psykose> won't work if it doesn't exist or uses /etc/zoneinfo type instead
2023-06-07 06:00:09 <psykose> this isn't distro specific, use your imagination :)
2023-06-07 06:00:12 <psykose> fedora has the same symlink
2023-06-07 06:00:51 <richirm> Sorry, I'm a newbie to Alpine... :)
2023-06-07 06:01:13 <psykose> the only part specific to alpine is timedatectl doesn't exist because there is no systemd
2023-06-07 06:01:31 <psykose> everything else is mostly generic linux
2023-06-07 06:03:27 <richirm> Forgive my ignorance, but when you're installing Alpine, why doesn't the installer ask for the time zone?
2023-06-07 06:04:43 <richirm> I am very glad to be here with expert Linux users.
2023-06-07 06:05:18 <psykose> it does
2023-06-07 06:08:08 <richirm> I use dockerized Alpine: 03"docker run -di --name alpine alpine; docker exec -ti alpine sh;"01, and I don't see any timezone when I get into the container with sh.
2023-06-07 06:08:51 <psykose> to put it bluntly- you are starting a shell in a container
2023-06-07 06:08:53 <psykose> not running an installer
2023-06-07 06:09:09 <psykose> i think you don't understand what docker is or what you are doing if you expect that to do anything except give you a shell
2023-06-07 06:09:52 <psykose> the "installer" is setup-alpine (a shell script, provided by the alpine-conf package, not installed in the container image by default because there is no use in even running it there)
2023-06-07 06:10:11 <psykose> for container uses you can just `apk add tzdata` and set TZ= env var
2023-06-07 06:10:38 <psykose> e.g. TZ=Europe/Amsterdam
2023-06-07 06:11:53 <psykose> if you "docker run -di --name fedora fedora; docker exec -ti fedora sh;" that's not going to run you an installer either
2023-06-07 06:12:46 <psykose> docker exec runs a command in container context, sh in that case, and sh is just a shell
2023-06-07 06:14:15 <richirm> But couldn't there be a way to set the Timezone when you do "docker pull alpine"? I have knowledge of Docker for a medium level user. I'm happy to talk to you, @psykose
2023-06-07 06:14:32 <psykose> no
2023-06-07 06:14:48 <psykose> docker pull alpine tells the docker daemon to pull a container image and save it into the store
2023-06-07 06:14:59 <psykose> it doesn't then also create a container and run some arbitrary installer and ask you stuff
2023-06-07 06:15:02 <psykose> that makes no sense
2023-06-07 06:17:10 <ikke> That kind of configuration is typically done in an application specific entrypoint that gets run when you start the xotbai4
2023-06-07 06:17:24 <ikke> container*
2023-06-07 06:28:58 <richirm> So in order to have a Timezone configured on my Alpine, I have to create an image based on Alpine's Docker image, and add to the Dockerfile the necessary commands to initialize the Timezone. Would this be correct?
2023-06-07 06:30:09 <psykose> sure
2023-06-07 06:30:55 <ikke> That's a pretty common workflow for docker, not specific to alpine 
2023-06-07 06:33:04 <psykose> you've probably done it a few thousand times by now as a medium level docker user with experience, as that's how you generally work with dockerfiles and making containers in the docker ecosystem
2023-06-07 06:33:17 <psykose> there's slightly different approaches outside of docker but fundamentally it's mostly the same
2023-06-07 06:45:36 <richirm> Thank you for your answers, [psykose]
2023-06-07 06:46:10 <richirm> Thanks [ikke]
2023-06-07 06:54:04 <richirm> Thanks and bye!! 👋
2023-06-07 07:01:06 <abby> could also just mount localtime from the host
2023-06-07 07:01:57 <psykose> does docker pull voidlinux do that?
2023-06-07 07:02:18 <abby> only if you run the installer in there
2023-06-07 07:02:31 <psykose> that sounds like bloat
2023-06-07 07:02:46 <abby> time is bloat
2023-06-07 07:07:23 <psykose> if you just don't look at the time it won't go forward
2023-06-07 07:15:08 <kmmndr> Hi all :-)
2023-06-07 07:16:58 <psykose> what's up
2023-06-07 07:17:28 <kmmndr> I'm in love with Alpine Linux :-D
2023-06-07 07:18:13 <psykose> don't think you can marry that one sadly
2023-06-07 07:18:44 <kmmndr> :-D
2023-06-07 07:23:34 <kmmndr> I'm encountering a strange thing: I'm unable to play any youtube video, on all installed browsers (firefox alpine's package or chromium/brave as flatpak). The page is loaded properly, but video seems constantly loading
2023-06-07 07:24:30 <psykose> sounds like some youtube thing where they broke something or geo / blocking-stuff related
2023-06-07 07:27:26 <kmmndr> same thing with dailymotion or vimeo
2023-06-07 07:29:02 <kmmndr> It smells as a browser issue as mpv is able to play without any problem
2023-06-07 07:30:55 <psykose> could be
2023-06-07 07:30:58 <mercenary> graphics acceleration things?
2023-06-07 07:31:14 <psykose> if flatpak doesn't work though it sounds like gpu stuff
2023-06-07 07:31:33 <psykose> though well, by default nothing uses hardware accelerated decode or whatevver
2023-06-07 07:31:47 <psykose> this sounds like something that would print a log when failing
2023-06-07 07:32:02 <psykose> what's your system
2023-06-07 07:43:40 <kmmndr> It's a framework laptop with 12th gen intel cpu
2023-06-07 07:47:29 <psykose> what wm are you running
2023-06-07 07:47:51 <kmmndr> gnome on wayland
2023-06-07 07:48:50 <psykose> does a different one work
2023-06-07 07:54:02 <kmmndr> I haven't tryed another wm. I'm not a big video consumer but it was working properly last time I watched a video (maybe weeks ago)
2023-06-07 12:41:57 <eletrotupi> hello friends
2023-06-07 12:42:30 <eletrotupi> I'm trying to debug why screensharing on firefox & sway doesn't work
2023-06-07 12:42:41 <eletrotupi> but I can't seem to find any dbus log whatsoever, any tips?
2023-06-07 12:47:51 <psykose> what version of firefox and stuff
2023-06-07 12:50:09 <psykose> check https://github.com/emersion/xdg-desktop-portal-wlr/wiki/%22It-doesn't-work%22-Troubleshooting-Checklist and make sure you have that +wireplumber etc
2023-06-07 12:50:18 <psykose> check https://mozilla.github.io/webrtc-landing/gum_test.html for the easy way to check capture
2023-06-07 12:50:33 <psykose> and specifically it was broken until a few weeks ago due to a bug in libwebrtc but it's been fixed for a while
2023-06-07 12:56:01 <eletrotupi> figure it out, by manually running both the desktop-portal and desktop-portal-wlr with a -r
2023-06-07 12:56:17 <eletrotupi> yeah, it apparently worked fine now
2023-06-07 12:56:39 <eletrotupi> maybe the means to activate those two services are broken somehow
2023-06-07 12:56:53 <eletrotupi> probably locally, since it worked on another alpine machine I had
2023-06-07 12:58:18 <psykose> they're meant to auto launch but sometimes it doesn't happen
2023-06-07 12:58:43 <psykose> glad that works now
2023-06-07 13:25:31 <eletrotupi> yeah, weird
2023-06-07 13:25:42 <eletrotupi> it's dbus who spawn those, right?
2023-06-07 13:25:51 <eletrotupi> well, it should be, I guess
2023-06-07 13:26:09 <eletrotupi> dbus is always broken here for some weird reason
2023-06-07 13:26:28 <eletrotupi> or I didn't really setup in a correct way to make it work 100% of the time
2023-06-07 13:43:14 <eddsalkield> yesterday I had to force reboot and now fsck is failing, but whenever it runs with error "Error relocating /sbin/fsck: mnt_cache_set_sbprobe: symbol not found"
2023-06-07 13:43:48 <eddsalkield> Whenever I try to start the networking service it says "ERROR: fsck failed to start" with the same reason
2023-06-07 13:44:09 <eddsalkield> I've mounted each block device on a recovery USB and run fsck, which reports no problems
2023-06-07 13:44:22 <eddsalkield> anyone have an idea as to why this symbol is now not being found?
2023-06-07 13:45:48 <psykose> does it come with more output than that like a missing solib
2023-06-07 13:46:19 <eddsalkield> no the missing symbol is the only output when I run `fsck`
2023-06-07 13:46:47 <psykose> that sounds like your /sbin/fsck is from util-linux 2.39 but /lib/libmount.so.1 is from 2.38 then
2023-06-07 13:47:34 <eddsalkield> yes that's exactly right
2023-06-07 13:47:37 <psykose> what's in your /etc/apk/repositories
2023-06-07 13:48:02 <eddsalkield> only edge/main,community,testing
2023-06-07 13:48:06 <eddsalkield> I'm not mixing major versions
2023-06-07 13:48:13 <psykose> that makes little sense then
2023-06-07 13:48:14 <psykose> no idea
2023-06-07 13:48:34 <psykose> what does `apk version` say, and does libmount upgrade on upgrade
2023-06-07 13:48:43 <psykose> if you ever remade the initfs both would get moved into it at once
2023-06-07 13:48:44 <psykose> hrm
2023-06-07 13:49:24 <eddsalkield> apk version says `libmount-2.38.1-r4` is installed, `2.39-r6` is available
2023-06-07 13:49:41 <psykose> and if you upgrade
2023-06-07 13:49:47 <eddsalkield> don't have a network connection
2023-06-07 13:49:52 <eddsalkield> but I could get the package on a usb and bring it across
2023-06-07 13:50:06 <psykose> hrm, no, more interested in how it would happen normally
2023-06-07 13:50:14 <psykose> no clue why it would be held back before
2023-06-07 13:50:47 <psykose> if just `apk add 'libmount>=2.39'` fails etc
2023-06-07 13:52:02 <eddsalkield> I think I can fix it from here, is there anything you want to know before I continue?
2023-06-07 13:54:13 <psykose> don't think so, but do run that command i guess
2023-06-07 13:54:41 <psykose> well, i guess it won't matter without a whole index
2023-06-07 13:57:33 <eddsalkield> under that command libmount upgraded fine
2023-06-07 13:58:53 <psykose> i actually forgot to ask for world too
2023-06-07 13:58:54 <psykose> oh well
2023-06-07 13:59:00 <psykose> but yeah it makes no sense
2023-06-07 13:59:13 <psykose> either there was a bug somewhere or you had like <2.39 of something
2023-06-07 13:59:15 <psykose> or
2023-06-07 13:59:15 <psykose> ah
2023-06-07 13:59:25 <psykose> did you upgrade from 3.1X->edge without -a?
2023-06-07 13:59:59 <eddsalkield> actually yes i think so
2023-06-07 14:00:02 <psykose> that would keep '3.18 libmount', and it would just never get changed
2023-06-07 14:00:03 <eddsalkield> what does -a do?
2023-06-07 14:00:12 <psykose> changes to latest from available repos
2023-06-07 14:00:26 <psykose> i.e. libmount was in 3.18, you changed repos to edge, there is a 'newer libmount' in edge
2023-06-07 14:00:32 <psykose> but, things only pull it in via so: dep
2023-06-07 14:00:35 <psykose> and that one is the same
2023-06-07 14:00:41 <psykose> no version constraints anywhere on anything
2023-06-07 14:00:49 <psykose> so it just never gets upgraded to 2.39 of edge
2023-06-07 14:00:54 <psykose> because it's not tracking edge
2023-06-07 14:00:59 <psykose> -a makes it track the new repo too
2023-06-07 14:01:07 <psykose> also allows downgrades etc
2023-06-07 14:01:15 <psykose> in laymans terms anyway
2023-06-07 14:01:20 <psykose> but yes, it's required when swapping repos
2023-06-07 14:01:52 <psykose> same between releases and so on
2023-06-07 14:04:01 <psykose> it swaps repos anyway to satisfy explicit version/soname constraits to be clear, but it won't otherwise, and nothing requests >=2.39 and the soname didn't change, etc
2023-06-07 14:08:17 <eddsalkield> thanks!
2023-06-07 14:08:32 <eddsalkield> also unrelated but why isn't testing built for major versions?
2023-06-07 14:13:29 <psykose> just isn't
2023-06-07 15:11:52 <wener> Hi, dose apk auidt the binary ?
2023-06-07 15:13:19 <psykose> in what sense
2023-06-07 15:13:25 <psykose> what binary
2023-06-07 15:13:29 <psykose> what audit
2023-06-07 15:13:47 <wener> e.g. some malware changed /bin/bash, I want to apk audit report the change
2023-06-07 15:14:09 <psykose> no
2023-06-07 15:14:37 <ikke> not apk audit?
2023-06-07 15:14:50 <wener> recently, due to mem failed, my disk report some error, I hope I can ensure the binary is ok
2023-06-07 15:16:00 <psykose> apk audit is just for the protected dirs (i.e. /etc)
2023-06-07 15:16:04 <wener> apk audit seems a great tool for this, it's audit 
2023-06-07 15:16:11 <ikke> apk audit --full
2023-06-07 15:16:21 <psykose> til
2023-06-07 15:16:25 <psykose> yeah that works
2023-06-07 15:16:33 <wener> thanks  
2023-06-07 15:16:37 <ikke> also --system
2023-06-07 15:16:39 <psykose> yep
2023-06-07 15:16:54 <psykose> damn that's useful
2023-06-07 15:17:02 <psykose> i keep forgetting the little things :-)
2023-06-07 15:17:13 <ikke> to be fair, I just checked apk audit --help :P
2023-06-07 15:17:18 <psykose> also just reinstalling every bin is easy (apk fix '*')
2023-06-07 15:17:26 <psykose> though ofc you will then have 500 .apk-new files for /etc
2023-06-07 15:17:59 <wener> wow, thanks for apk fix '*'
2023-06-07 15:18:22 <wener> seems --full=--system
2023-06-07 15:18:30 <wener> but not
2023-06-07 15:20:02 <ikke> https://man.archlinux.org/man/apk-audit.8.en
2023-06-07 15:20:24 <ikke> (doesn't list all options)
2023-06-07 15:20:38 <ikke> man apk-audit does if you have apk-doc installed
2023-06-07 15:21:28 <wener> never installed man in alpine, my bad
2023-06-07 15:22:36 <wener> report `A /bin/ash`, but `apk info -W /bin/ash` symlink to busybox, seems apk audit do not handle symlink
2023-06-07 15:23:22 <ikke> if you use --full, it reports all adding files, and those symlinks are not tracked by apk, so it shows them as added
2023-06-07 15:23:30 <ikke> all added files*
2023-06-07 15:47:29 <foo845829> anyone know the status of https://dso.docker.com/cve/CVE-2023-2650 fix for Alpine?
2023-06-07 15:48:21 <ikke> openssl-3.1.1 with the fix is available
2023-06-07 15:48:34 <foo845829> its not available in edge or 3.18 though, right?
2023-06-07 15:48:49 <ikke> It is
2023-06-07 15:49:27 <foo845829> http://dl-cdn.alpinelinux.org/alpine/edge/main/aarch64/ only has openssl-3.1.0 not 3.1.1
2023-06-07 15:49:37 <foo845829> openssl-3.1.0-r5.apk                               15-May-2023 08:38    317K
2023-06-07 15:50:01 <psykose> that's from arm builders being down
2023-06-07 15:50:08 <psykose> back eventually
2023-06-07 15:51:05 <foo845829> :)
2023-06-07 16:29:09 <NepNepdmsalwaysopen[m]> i don't mean to put any pressure, especially considering the package in question (synapse) has only been outdated for a day, this is just a question: edge packages are generally updated as soon as the maintainer gets to it, without a set schedule, right?
2023-06-07 16:29:33 <ikke> correct
2023-06-07 16:30:41 <psykose> sure i'll update it
2023-06-07 16:30:42 <psykose> :-)
2023-06-07 16:51:02 <NepNepdmsalwaysopen[m]> <ikke> "correct" <- alright cool
2023-06-07 16:52:55 <NepNepdmsalwaysopen[m]> <psykose> "sure i'll update it" <- oh thank you :>
2023-06-07 16:53:34 <psykose> should be there
2023-06-07 16:55:49 <NepNepdmsalwaysopen[m]> yep! it is
2023-06-07 19:35:53 <nortti> hi. I'm trying to boot 3.18.0 x86_64 .iso in qemu using `qemu-system-x86_64 -accel kvm -cdrom alpine-standard-3.18.0-x86_64.iso` and it fails with a lot of error messages of the form "error: /sbin/udhcpc: No such file or directory", leading eventually to it to drop me to a recovery shell
2023-06-07 19:36:15 <nortti> https://ahti.space/~nortti/u/Kuvakaappaus_2023-06-07_22-36-00.png
2023-06-07 19:36:36 <ikke> nortti: you need to give the VM more memory
2023-06-07 19:36:51 <ikke> The default amount is too little
2023-06-07 19:36:55 <nortti> ah, I see. thank you
2023-06-07 19:50:36 <blt[m]> wondering, as root why doing a doas -u user is asking me the root password ? is it by design or is there a way to avoid that
2023-06-07 19:51:20 <psykose> it prompts for the current user password
2023-06-07 19:51:35 <psykose> not the root one
2023-06-07 19:51:41 <ikke> so, by design
2023-06-07 19:51:53 <ikke> blt[m]: su would ask for the target users password
2023-06-07 19:52:00 <ikke> (except for root)
2023-06-07 19:52:14 <ikke> as root, it would let you just switch user
2023-06-07 19:52:33 <blt[m]> it is not what I experiment here
2023-06-07 19:53:13 <blt[m]> am trying to pass some commmand as ceph user (no shell) and I have to put the root user when I try a doas -u ceph
2023-06-07 19:54:13 <psykose> what user are you running it as
2023-06-07 19:54:21 <psykose> i.e. whoami
2023-06-07 20:00:26 <blt[m]> as root
2023-06-07 20:00:31 <blt[m]> doas -u ceph echo "muvra" is asking root password for me
2023-06-07 20:00:37 <psykose> yes, that's what we said
2023-06-07 20:00:39 <psykose> you are root
2023-06-07 20:00:44 <psykose> so it asks for the root password
2023-06-07 20:00:46 <psykose> i.e. current user
2023-06-07 20:00:55 <psykose> it does not prompt for the target password
2023-06-07 20:01:02 <psykose> that wouldn't make sense; that's what su does already
2023-06-07 20:01:09 <psykose> you can exclude root if you want
2023-06-07 20:01:12 <psykose> permit nopass root
2023-06-07 20:01:22 <psykose> then it prompts nothing for doas -u whatever as root
2023-06-07 20:01:26 <psykose> since you're root anyway
2023-06-07 20:01:30 <blt[m]> ha ok I understant
2023-06-07 20:01:52 <psykose> i generally add exactly that to doas.conf
2023-06-07 20:01:53 <psykose> :)
2023-06-07 20:01:56 <blt[m]> yes I thought it would not ask any password as root (I was indeed not expected to be ask for the targeted user password as I am root)
2023-06-07 20:04:40 <blt[m]> very nice, I just tested 🙂
2023-06-07 20:49:53 <eggbean> Is possible to add a package without its dependencies? I am trying to install something which is just a script but it's adding a load of stuff I don't need as that stuff is being exposed by the container docker host
2023-06-07 20:51:43 <ptrc> eggbean: if it's just a handful of dependencies, you could add matching virtual packages for them with `apk add -t pkgname`
2023-06-07 20:52:25 <ptrc> otherwise, you could also try and modify the package if you build it yourself 
2023-06-07 20:53:06 <ptrc> in general, ( to the best of my knowledge ), apk ensures every package in /etc/apk/world has its dependencies met on every transaction
2023-06-07 20:53:16 <ikke> yes
2023-06-07 20:53:21 <ikke> that's how apk works
2023-06-07 20:53:28 <ikke> fundamentally
2023-06-07 20:53:38 <ptrc> so even if you could add it and somehow not add the dependencies, they would get installed anyway on any other add/upgrade/del
2023-06-07 20:55:42 <eggbean> Thanks. Just trying to find out what a virtual package it
2023-06-07 20:55:47 <eggbean> s/it/is
2023-06-07 20:55:56 <ikke> It's a package that does not have any content
2023-06-07 20:56:04 <ikke> so not really what you want
2023-06-07 20:56:20 <ikke> it's more like the opposite, you can add dependencies to a virtual package
2023-06-07 20:56:43 <eggbean> It's just a script, so I could just wget it from github on build
2023-06-07 20:57:36 <eggbean> It's fzf-tmux. apk goes ahead and installs tmux and libevent etc
2023-06-07 21:07:14 <psykose> as is required to actually use it, with tmux
2023-06-07 21:07:21 <psykose> it's quite useless by itself without.. tmux
2023-06-07 21:07:46 <psykose> but no, in general you can't do that
2023-06-07 21:08:02 <psykose> the -t trick works but then eventually you'll break something else with it, and then that's your fault :-)
2023-06-07 21:08:42 <ikke> how would it work? You could only do the opposite, only install dependencies while masking the actual package
2023-06-07 21:09:43 <psykose> it can't
2023-06-07 21:12:56 <ptrc> ikke: if there's only one dependency, you just add that dependency as a virtual, then apk thinks it's already installed
2023-06-07 21:14:09 <ikke> oh yeah, that would work'
2023-06-07 21:14:13 <ikke> masking each dependency
2023-06-07 22:51:48 <frag> do you think a usb to ethernet will work if it says it has drivers on flash memort, and compatible with linux? (doesnt it depends on which distro??)
2023-06-07 22:53:22 <psykose> no
2023-06-07 22:53:26 <psykose> it should work
2023-06-07 22:54:01 <frag> psykose: cool cool :)
2023-06-08 01:55:47 <LACampbell> how can I stop pipewire launcher wiping my alsa volume levels? I have alsa as an openrc service, and it restores the levels correctly. then I start x, which starts pipewire-launcher, and it resets them
2023-06-08 01:56:07 <LACampbell> having real difficulty with sound in general. not 100% sure where pipewire launcher should be, but only seems to work in my WM autostart
2023-06-08 01:57:15 <Someone[m]1> pipewire is really bad
2023-06-08 01:57:51 <LACampbell> well what else works with modern normie apps? they expect some like pulse audio like, not alsa
2023-06-08 02:00:30 <abby> i don't think you can use alsa and pipewire together like you are
2023-06-08 02:01:06 <LACampbell> fair enough. my end goal is basically I want to turn my machine on and not have to manually reset the volume. not sure what the solution is
2023-06-08 02:01:20 <abby> you shouldn't need alsa to do that
2023-06-08 02:01:27 <abby> pipewire should do that itself already
2023-06-08 02:01:57 <LACampbell> ok, so I should disable the alsa startup service? I'll try that and restart
2023-06-08 02:05:08 <Someone[m]1> <LACampbell> "well what else works with modern..." <- I thought browser "just play" audio and I'm using alsa now
2023-06-08 02:05:34 <LACampbell> electron apps. stuff like signal, discord, steam etc
2023-06-08 02:06:25 <Someone[m]1> LACampbell: I'm using them in a browser
2023-06-08 02:06:41 <LACampbell> abby: how are you starting pipewire? it's not remember my audio levels at all. (I assume alsamisxer still shows audio levels)
2023-06-08 02:06:45 <Someone[m]1> but I think it is wise to use in app because I'm using the testing repo
2023-06-08 02:07:00 <abby> i'm not on alpine but i just start pipewire with my wm
2023-06-08 02:07:12 <LACampbell> Someone[m]1: yeah I'd love to but I'm not there yet. I had all this stuff working in my old install of antoher distro, wish I'd written it down
2023-06-08 02:07:25 <LACampbell> ok I am starting pipewire with my wm too, so that's not what I'm doing wrong
2023-06-08 02:07:27 <abby> and then pipewire launches wireplumber and pipewire-pulse for me
2023-06-08 02:07:49 <Someone[m]1> I'm using sway lol
2023-06-08 02:07:54 <Someone[m]1> it is really hard to set up
2023-06-08 02:08:16 <LACampbell> ok I've got wire plumber running too
2023-06-08 02:09:12 <LACampbell> ok I've got wire plumber running too. but audio levels are not being remembered. hmm I'll try one more restart then that's enough sysfiddling for today
2023-06-08 02:09:30 <Someone[m]1> also element-desktop is working and it is electron app
2023-06-08 02:10:59 <LACampbell> huh. discord (electron) won't pick up anything without pipewire IIRC. I'll try and kill it off I guess
2023-06-08 02:11:00 <abby> do you have pipewire-alsa installed? i think for alsa-mixer to show things correctly, you'd need that
2023-06-08 02:11:13 <abby> is pipewire-pulse running?
2023-06-08 02:11:20 <abby> things like discord will only know how to speak pulse
2023-06-08 02:11:52 <LACampbell> abby: yes that's installed, pulse is running, wpctl status is showing all devices
2023-06-08 02:12:01 <Someone[m]1> LACampbell: discord is really violating user's privacy, I suggest running it in the browser
2023-06-08 02:12:05 <LACampbell> *pipewire-pulse is I mean
2023-06-08 02:12:13 <LACampbell> Someone[m]1: won't screenshare properly then
2023-06-08 02:12:21 <Someone[m]1> yes
2023-06-08 02:13:07 <LACampbell> ok I'm not RMS, if I want to interact with the realworld I need to run this crapware unfortunately
2023-06-08 02:14:47 <quinq> “realworld” is out the window ;)
2023-06-08 02:14:59 <LACampbell> semirealworld
2023-06-08 02:15:18 <LACampbell> business world? you know what I mean
2023-06-08 02:16:27 <quinq> I suppose you mean constraints you have to deal with for some reason, yeah
2023-06-08 02:17:01 <LACampbell> huh, there's a pipewire group. maybe it would help if my user was apart of that
2023-06-08 02:17:42 <quinq> Did you have a look at the wiki (just in case)?
2023-06-08 02:18:48 <LACampbell> yeah multiple times. guess I'll give it another look
2023-06-08 02:19:31 <quinq> Well, if you already followed the instructions and it didn't work, there must be a problem
2023-06-08 02:20:24 <quinq> That was to make sure you were not starting from scratch without it
2023-06-08 02:21:22 <LACampbell> yeah but IME It's always worth while reading stuff carefully. 
2023-06-08 02:24:30 <LACampbell> I give up for today. will just have to remember this when I restart for now. thanks for the suggestions everyone
2023-06-08 02:27:39 <quinq> good luck :)
2023-06-08 02:48:27 <Someone[m]1> Anyone know how to type another language (Vietnamese) on wayland (sway)?
2023-06-08 02:48:34 <Someone[m]1> using fcitx5
2023-06-08 02:48:52 <Someone[m]1> I don't understand what their guides are saying
2023-06-08 07:21:23 <kmmndr> Hi all :-)
2023-06-08 09:48:55 <ktz> hey good people, a quick question. apk hasn't given me any updates for days now, at this point I'm almost certain something is wrong with it. can't pinpoint or fix the problem. I've had my cache cleaned with the purge flag, tried other repo mirrors but still no luck. Has anyone gone through this maybe? any ideas?
2023-06-08 09:54:07 <lopid> are you trying to use an old version that just has no updates?
2023-06-08 09:54:39 <aron> what's in /etc/apk/repositories?
2023-06-08 09:55:48 <ktz> aron:
2023-06-08 09:55:53 <ktz> aron: http://dl-cdn.alpinelinux.org/alpine/edge/main http://dl-cdn.alpinelinux.org/alpine/edge/community http://dl-cdn.alpinelinux.org/alpine/edge/testing  # https://www.arvanta.net/mps/aports-local # /home/ktz/.local/dist/out/packages/root # /home/ktz/.local/share/containers/storage/overlay/56c00e60e36ae29f14103f7019d8c10cf753a7e25793b95bc0bf5c74c9425ae2/diff/root/packages/testing
2023-06-08 09:56:01 <ktz> not sure how to format this properly, excuse me
2023-06-08 09:56:21 <ktz> edge/main, edge/community and edge/testing essentially
2023-06-08 09:56:44 <aron> so, briefly: you are using edge
2023-06-08 09:56:52 <ikke> ktz: what arch?
2023-06-08 09:56:57 <ktz> aarch64
2023-06-08 09:57:22 <ikke> Right, builders are currently unreachable, so packages have not been able to build 
2023-06-08 09:57:35 <ktz> ohhh I see now
2023-06-08 09:57:52 <ktz> okay that solves it, at least as far as I'm concerned
2023-06-08 09:58:24 <ktz> I assume the build system will most likely be back online soon
2023-06-08 09:59:33 <ktz> thanks
2023-06-08 10:20:02 <ikke> ktz: we do our best, but it's out of our control
2023-06-08 12:36:10 <txnintn> Still can not get an setup-alpine lvn over raid1 to boot, if I dd /dev/md0 to a mmcblk0p1, it will boot off the mmcblk.  WTF?
2023-06-08 12:38:31 <minimal> txnintn: is this a RPI or a PC with a SDcard?
2023-06-08 12:40:00 <txnintn> RPI4, two 64GB USB 3.1 drives, and a mmcblk /w alpine setup image.
2023-06-08 12:40:33 <minimal> and mmcblk0p1 is the RPI's fat boot partition?
2023-06-08 12:41:09 <minimal> boot partition/fat boot fs
2023-06-08 12:41:13 <txnintn> Yes, fat32 300mb partition, the rest is ext4
2023-06-08 12:41:35 <minimal> so you're expecting the RPI's own bootlaoder to understand RAID?
2023-06-08 12:42:27 <txnintn> No I expect alpine bootloader to understand raid.
2023-06-08 12:42:53 <minimal> txnintn: what Alpine bootloader? it is a Raspberry PI - it has it's own bootloader
2023-06-08 12:43:59 <minimal> a RPI has a *builtin* bootloader that uses a partition with a FAT fs to provide it with additional files to boot
2023-06-08 12:45:13 <txnintn> It trying this other ways, manually, to load raid support you have to add apk packages of mdadm, dosfstools, and rpibootloader.
2023-06-08 12:46:34 <minimal> you can have the *initramfs* once loaded, handle RAID for the rootfs etc, but that is *after* the RPI's own bootloader has loaded some files including the kernel and initramfs from the boot partition
2023-06-08 12:46:53 <minimal> so don't "RAID" the boot partition
2023-06-08 12:48:49 <txnintn> So how do I get it to work, Setup-Alpine is Raiding md0 and md1.  It should be a mirrored lvm Logical Volume.
2023-06-08 12:49:19 <minimal> txnintn: I assume setup-alpine using RAID was designed for PCs, not for RPIs
2023-06-08 12:49:51 <minimal> what should be a mirrored LVM LV?
2023-06-08 12:51:17 <txnintn> In lou of the /dev/md0
2023-06-08 12:51:29 <minimal> what?
2023-06-08 12:52:35 <txnintn> The boot partion should be a mirrored LVM LV, my understading then it can boot from either drive.
2023-06-08 12:53:31 <minimal> as I already explained the Raspberry Pi's *builtin* bootloader only supports a "simple" FAT-formatted partition for booting from
2023-06-08 12:53:41 <minimal> no RAID, no LVM, etc
2023-06-08 12:54:03 <txnintn> crap
2023-06-08 12:54:24 <minimal> I don't believe they ever advertised/claimed it could do RAID and LVM...
2023-06-08 12:55:22 <txnintn> no they dont, do only choices is one usb with plain jane fat and a mmcblk to back it up
2023-06-08 12:55:24 <minimal> now you could use RAID+LVM for everything except the boot partition
2023-06-08 12:55:57 <txnintn> yeah, it is going to be my home automation server
2023-06-08 12:55:58 <minimal> as that would be handled by Alpine's initramfs and Alpine itself once the initial booting occurs
2023-06-08 12:56:46 <txnintn> Sounds like someone need to working on setup-alpine script for RPI raid.
2023-06-08 12:57:06 <minimal> txnintn: so when do you start on it?
2023-06-08 12:58:02 <txnintn> don't understand the question
2023-06-08 12:58:09 <minimal> RAID support wouldn't be useful for many/most people using RPI so that's probably why setup-alpine doesn't support it
2023-06-08 12:58:44 <txnintn> World is changing...hehe
2023-06-08 12:58:48 <minimal> txnintn: you said "Sounds like someone need to working on setup-alpine script for RPI raid", I am suggesting that you are that "someone" as you have a need for it
2023-06-08 13:01:57 <txnintn> I was in the computer gig for well 20+ years. Information from a mfg standpoint need to be clear.  Otherwise they will move to a different product line.
2023-06-08 13:02:52 <minimal> txnintn: what was'nt clear about RPI? They never indicated they supported RAID booting
2023-06-08 13:03:27 <txnintn> so why wasn't the script adjusted for it.
2023-06-08 13:05:11 <minimal> txnintn: I guess because no-one in Alpine ever thought that someone would try and RAID a RPI. Alpine is *not* a commercial product, it is a opensource project run by volunteers
2023-06-08 13:05:56 <minimal> so why don't you "volunteer" to fix it?
2023-06-08 13:09:16 <txnintn> I'm <- building a house, I'm the carpenter, electician, plumber, etc...  I don't have time to "volunteer".   
2023-06-08 13:09:45 <txnintn> I'm volunteered up to my eyeballs...
2023-06-08 13:09:54 <txnintn> hahah
2023-06-08 13:10:06 <minimal> txnintn: but you have time to complain about other volunteers not have fixed something that they probably were not aware of?
2023-06-08 13:10:41 <minimal> I volunteered to answer your question and to try and help you, maybe I shouldn't have bothered?
2023-06-08 13:11:48 <txnintn> I'm not complain, I'm just pointed out a issue.  well scroll up, for bowenp. thats me.
2023-06-08 13:12:41 <txnintn> So I haven't don IRC in 30+ year, just a little rusty.
2023-06-08 13:14:10 <minimal> if you open an Issue for this on Gitlab then it can be tracked and it might, at some point in the future, be fixed
2023-06-08 13:16:41 <txnintn> Ok, in the mean time, it need to get this fixed, I will be ready to start running low voltage wiring in about two weeks, and I want to live test before insulation is blown in.
2023-06-08 13:22:27 <txnintn> ur ARM infrastructure is unreachable at the moment, so CI jobs will time-out and packages will not be updated until the servers are back.
2023-06-08 13:52:21 <msiism> Is it still true that setup-alpine is unable to set up encrypted swap partitions?
2023-06-08 14:09:08 <minimal> msiism: in which context? swap-on-LUKS? swap-on-LVM-on-LUKS?
2023-06-08 14:14:56 <msiism> Good question. I mean, if you tell setup-disk to encrypt the disk. I don't exactly remember how that worked.
2023-06-08 14:26:01 <jhjacobs[m]> hmz
2023-06-08 14:47:15 <minimal> msiism: you can tell setup-disk to use LVM, or you can tell it to use CRYPT, or you can tell it to use both (so LVM-on-LUKS)
2023-06-08 14:49:52 <minimal> msiism: however you haven't been specific about exactly how you want to use encryption - with LVM or without LVM?
2023-06-08 14:57:54 <msiism> Well, I'd prefer LVM on LUKS.
2023-06-08 14:59:49 <minimal> when setting up LVM setup-disk appears to create a swap LV, not a partition, so I assume if you put LVM on LUKS that the same would apply and therefore the swap would be encrypted
2023-06-08 15:01:20 <msiism> Okay, that makes sense. What about not using LVM, but only LUKS?
2023-06-08 15:03:39 <minimal> you just said you preferring LVM on LUKS...
2023-06-08 15:08:26 <msiism> Yeah, the reason I'm asking is that I don't want to make false statements about what Alpine does by default in my guide for FDE on BIOS/MBR.
2023-06-08 15:08:44 <msiism> I'm currently going through setup-disk.
2023-06-08 15:18:44 <msiism> Hm… seems to me, if you only have USE_CRYPT set (no LVM), only the root partition will be encrypted, which kind of makes sense too.
2023-06-08 15:21:06 <msiism> But okay, the way I'm saying it actually covers both cases well.
2023-06-08 15:56:56 <eggbean> ikke: About the fzf-tmux package - confused about what I see in the linked repository from the packages page... Does fzf-tmux need bash, as there seems to have been an attempt to have it run in ash?
2023-06-08 15:59:17 <ikke> eggbean: the script uses bash arrays for one
2023-06-08 15:59:33 <eggbean> Okay, so it requires bash. Cheers
2023-06-08 16:07:40 <oleovit> hi. abort: https://git.alpinelinux.org/aports/commit/?id=190c146c766bf5bc7e241d236fb833c342db044c . i had to open an issue on mpv's git, and recogniced, that mpv --version and the mpv log is showing the wrong version (apk info/policy not). see beginning of the logfile: https://0x0.st/HcKt.txt
2023-06-08 16:08:47 <ikke> oleovit: it has already been built
2023-06-08 16:10:03 <ikke> oleovit: are you referring to " mpv 0.35.0-UNKNOWN"?
2023-06-08 16:10:34 <psykose> that's what it always says
2023-06-08 16:11:15 <psykose> i.e. the VERSION has that string in it
2023-06-08 16:11:46 <psykose> guess i'll just overwrite it
2023-06-08 16:13:05 <psykose> lightly amended
2023-06-08 16:18:51 <oleovit> yes, thx ikke & psykose. the mpv dev was a little confused by this. i understand.
2023-06-08 17:06:16 <userdocs> now it will say mpv 0.35.0-PSYKOSE?
2023-06-08 18:09:58 <quinq> *lightly* amended, it'll say 0.35.0-KNOWN
2023-06-08 18:11:49 <saba> any kind of ETA on when arm builds will be working?  The only thing I've seen is the message in bottom corner of the gitlab
2023-06-08 18:13:46 <pj> no
2023-06-08 18:15:02 <ikke> saba: It's an issue with the provider hosting them, so not really somethign we can control or give an ETA about
2023-06-08 18:16:04 <psykose> eta wen
2023-06-08 18:16:06 <saba> ok thanks. hope you get credit for free month or however long it ends up being down
2023-06-08 18:16:21 <psykose> it's already free
2023-06-08 18:22:25 <aldcor> hello! How can I fully get rit of pipewire files. I removed it with apk but I get this  pw.loop [loop.c:67 pw_loop_new()] 0x7fa7a329d5c0: can't make support.system handle: No such file or directory
2023-06-08 18:23:08 <aldcor> i got sound working, however I got mic problems. The error above is shown when I try to play file which is microphone test
2023-06-08 18:23:29 <aldcor> it looks like it's using pw.loop which is pipewire plugin or smth like that
2023-06-08 18:24:56 <psykose> how are you playing it
2023-06-08 18:33:19 <aldcor> psykose: aplay test.wav or mpv test.wav gives same error
2023-06-08 18:33:35 <psykose> apk list -I | grep wire
2023-06-08 18:33:38 <psykose> ps aux | grep wire
2023-06-08 18:34:40 <aldcor> oh yes, i got libs regdb an wireplumber
2023-06-08 18:34:47 <psykose> sounds fine
2023-06-08 18:34:48 <aldcor> thanks, useful commands!
2023-06-08 18:34:55 <psykose> can you actually pastebin that
2023-06-08 18:35:02 <aldcor> but, i don't need pipewire-libs
2023-06-08 18:35:05 <aldcor> gona remove
2023-06-08 18:35:19 <psykose> you can't remove that and they don't do anything
2023-06-08 18:35:24 <psykose> now can you pastebin that
2023-06-08 18:37:33 <aldcor> psykose: https://0x0.st/Hcb9.txt
2023-06-08 18:38:00 <psykose> and the other thing
2023-06-08 18:38:11 <psykose> and `ls -la /usr/share/alsa/alsa.conf.d/ /usr/lib/alsa-lib/ /etc/alsa/conf.d/`
2023-06-08 18:38:30 <aldcor> 4963 <username>     0:00 grep wire
2023-06-08 18:39:46 <aldcor> psykose: https://0x0.st/Hcbp.txt
2023-06-08 18:40:00 <psykose> looks fine
2023-06-08 18:40:04 <psykose> what does `aplay -l` say
2023-06-08 18:40:34 <quinq> Where does all that come from? Mine are empty
2023-06-08 18:42:37 <aldcor> psykose: https://0x0.st/Hcbf.txt
2023-06-08 18:43:13 <psykose> yeah everything looks fine
2023-06-08 18:43:29 <aldcor> do I have pipewire plugin somewhere?
2023-06-08 18:43:32 <psykose> aplay only sees the hardware directly so aplay would output to it normally
2023-06-08 18:43:32 <psykose> nope
2023-06-08 18:43:41 <aldcor> ls
2023-06-08 18:43:50 <psykose> no idea why it would get routed to pipewire, except if you haven't rebooted and there's weird state left
2023-06-08 18:44:09 <aldcor> oh, I rebooted often today
2023-06-08 18:44:35 <psykose> did you reboot after removing all of that
2023-06-08 18:45:07 <aldcor> `all that` being pipewire ?
2023-06-08 18:45:35 <aldcor> i removed pipewire many days ago (via apk del) 
2023-06-08 18:47:10 <psykose> if you just aplay a file now can you post a paste of it
2023-06-08 18:47:26 <aldcor> also when I do `alsamixer` I get this ALSA lib pulse.c:242:(pulse_connect) PulseAudio: Unable to connect: Connection refused. Workaround is this - alsamixer --card=1
2023-06-08 18:47:47 <aldcor> i dont have puleaudio.. 
2023-06-08 18:48:10 <aldcor> i installed it but then I removed it, seems like somewhere in system are leftovers
2023-06-08 18:48:48 <psykose> you do have alsa-plugins-pulse which makes it output to pulseaudio by default
2023-06-08 18:49:39 <aldcor> alright! removed that
2023-06-08 18:50:06 <aldcor> aplay *.wav doesn't give an error
2023-06-08 18:50:22 <aldcor> however, mpv *.wav gives same error: pw.loop [loop.c:67 pw_loop_new()] 0x7fc9bcf8ba80: can't make support.system handle: No such file or directory
2023-06-08 18:51:15 <psykose> that probably needs --ao=alsa
2023-06-08 18:51:21 <psykose> what's in your default config
2023-06-08 18:51:58 <aldcor> i don't have default mpv config
2023-06-08 18:52:04 <aldcor> or do I .. i never made file
2023-06-08 18:52:10 <psykose> mpv --log-file=log somefile.wav
2023-06-08 18:52:12 <psykose> pastebin that
2023-06-08 18:53:00 <aldcor> omg thats insane file
2023-06-08 18:53:30 <aldcor> https://0x0.st/HcbJ.txt
2023-06-08 18:55:11 <psykose> that outputs to alsa fine
2023-06-08 18:55:17 <psykose> and the error you say you see isn't there
2023-06-08 18:55:32 <aldcor> that error shows up when I use mpv
2023-06-08 18:55:40 <aldcor> not when I use aplay
2023-06-08 18:56:00 <psykose> the mpv log doesn't have that error
2023-06-08 18:56:06 <aldcor> i mean, i am not lying.. i can make video 
2023-06-08 18:56:44 <psykose> [ao/alsa] opening device 'default'
2023-06-08 18:56:45 <psykose> [ao/alsa] Hardware PCM card 1 'HDA Intel PCH' device 0 subdevice 0
2023-06-08 18:56:47 <psykose> looks fine to me
2023-06-08 18:56:52 <psykose> no errors in it anywhere
2023-06-08 18:57:31 <psykose> which would mean the 'error: pw.loop' you see is not from mpv, because if it was it would be in the log
2023-06-08 18:58:18 <psykose> idk tho, i haven't used bare alsa in 14 years
2023-06-08 18:58:21 <psykose> maybe someone else knows
2023-06-08 18:58:26 <psykose> miserable broken audio stack :D
2023-06-08 18:58:38 <aldcor> yes, I have no idea where that error originate from, i'm just saying, it's there in mpv :)
2023-06-08 19:10:54 <aldcor> it may be broken.. as long as I can run spotify_player, I'm happy man :)
2023-06-08 19:11:35 <psykose> you could take a video but mpv says it's fine
2023-06-08 19:12:42 <aldcor> ah.. then I have to figure out how to make video with wayland :D 
2023-06-08 19:13:29 <aldcor> btw, when I play any video with mpv, same error shows up
2023-06-08 19:16:06 <psykose> yeah, it's doing the same thing
2023-06-08 19:17:16 <psykose> wf-recorder is pretty easy to record but you can't record anything on wayland without pipewire
2023-06-08 19:17:29 <aldcor> what?! /O\
2023-06-08 19:18:11 <aldcor> even if I don't need sound when recording?
2023-06-08 19:18:18 <psykose> (which then requires wireplumber too, which by default would grab every audio device, so you have to configure it to not do stuff to keep using bare alsa)
2023-06-08 19:18:21 <psykose> pipewire started for video
2023-06-08 19:18:35 <psykose> they only added audio to it later
2023-06-08 19:19:57 <aldcor> 
ACTION going back to ubuntu

2023-06-08 19:21:24 <aldcor> btw, psykose are you one of alpine devs or maintainers?
2023-06-08 19:21:24 <psykose> it's the same there, this isn't distro specific
2023-06-08 19:21:41 <aldcor> it was pathetic joke attempt :)
2023-06-08 20:01:36 <kmmndr> aldcor: nice to hear that you like spotify_player too, I'm currently upgrading our package to latest release :-)
2023-06-08 20:03:59 <aldcor> kmmndr: it was such a relief to install and `launch` it compared to voidlinux..  I had to do some wizardry deep in files and without help of staff I wouldn't be able to run it. 
2023-06-08 20:04:13 <aldcor> here I launched it.. it asked for username and password and done :)
2023-06-08 20:04:37 <aldcor> installation was probably ~5 secs
2023-06-08 20:04:59 <aldcor> it felt a bit magical and super hacky :) 
2023-06-08 20:05:49 <aldcor> since I'm on windows managers for years, spotify_player is very nice addition
2023-06-08 20:06:12 <kmmndr> Nice to hear too, as we are probably all loving Alpine Linux here :-)
2023-06-08 20:08:48 <kmmndr> psykose: we discussed yesterday about youtube video not playing on any browser, it was related to pulse (remove pulseaudio, install pipewire-pulse and it's working now). Thanks again for your help :-)
2023-06-08 20:09:47 <psykose> glad it works
2023-06-08 20:15:12 <aldcor> well, I was convinced not to use pipewire stuff.. but maybe pipewire would solve my `mic not working` issue too
2023-06-08 20:17:19 <kmmndr> aldcor: which wm are you using ?
2023-06-08 20:17:53 <kmmndr> It's working great on gnome with pipewire
2023-06-08 20:19:32 <aldcor> I used i3 for many years but when switched to alpine, I couldn't solve screen tearing issue which happened when watching videos (i got nvidia) and so solution is wayland. I use riverwm (wayland based) and got no issues. Also, font renders better. Text looks crisp, easier on eyes. Some kind of magic
2023-06-08 20:20:36 <psykose> i thought the meme was you were supposed to pretend wayland is garbage and everything is broken? TIL
2023-06-08 20:20:42 <aldcor> i tried bspwm which I also like
2023-06-08 20:21:41 <aldcor> oh, i don't know that meme :)
2023-06-08 20:25:45 <kmmndr> Then you may try pipewire, as you may be at a one package install step from your perfect setup ;-)
2023-06-08 20:26:44 <aldcor> if I install pipewire now, will I still need ~/.asoundrc file ?
2023-06-08 20:27:00 <aldcor> for now, I got only alsa
2023-06-08 20:27:10 <aldcor> can pipewire work without pulseaudio?
2023-06-08 20:27:32 <psykose> if anything it conflicts with it
2023-06-08 20:27:56 <psykose> you don't need ~/.asoundrc either 
2023-06-08 20:28:09 <kmmndr> pipewire is compatible with pulse, it is a compatible replacement (pipewire-pulse)
2023-06-08 20:28:31 <psykose> `apk add pipewire pipewire-alsa pipewire-jack pipewire-pulse wireplumber`
2023-06-08 20:28:50 <psykose> then you have to start 3 daemons with your choice of starting daemons i guess
2023-06-08 20:29:02 <aldcor> omg, demons in house again
2023-06-08 20:29:07 <psykose> /usr/libexec/pipewire-launcher starts all 3 with no supervision
2023-06-08 20:29:33 <psykose> pipewire-spa-bluez for bluetooth too more or less
2023-06-08 20:29:45 <psykose> and with that alsa+pulse+pipewire+jack clients should all work
2023-06-08 20:29:49 <psykose> has for me for years anyway
2023-06-08 20:30:18 <aldcor> so.. do I need alsa in setup you mention? Or I replace alsa with pulse?
2023-06-08 20:30:31 <aldcor>  /o\
2023-06-08 20:30:36 <aldcor> 
ACTION panicking

2023-06-08 20:30:44 <psykose> no
2023-06-08 20:30:56 <psykose> as in it doesn't need any alsa config or extra stuff installed except for that
2023-06-08 20:31:33 <psykose> but idk what you already have and if there's special migration involved from what you already did
2023-06-08 20:32:11 <aldcor> psykose: i have only alsa (for sound)
2023-06-08 20:32:23 <aldcor> so, I assume I need pulseaudio too or only pipewire-pulse ?
2023-06-08 20:32:37 <psykose> i already told you what to install
2023-06-08 20:33:04 <aldcor> 
ACTION palms sweaty

2023-06-08 20:40:35 <maria> mom's spaghetti 
2023-06-08 20:53:26 <Saijin_Naib> Yes someone did it 🤣
2023-06-08 20:55:12 <aldcor> :D
2023-06-08 21:04:21 <aldcor> ok.. it is recording mic but not my external mic (which is part of headphones). It's recording that terrible laptop mic. Also, when I plug in headphones I expect laptop's speakers to be muted but they don't become muted
2023-06-08 21:04:45 <Saijin_Naib> Wireplumber installed? 
2023-06-08 21:05:22 <aldcor> yap, i see : wireplumber-0.4.14-r2 x86_64 {wireplumber} (LGPL-2.1-or-later) [installed]
2023-06-08 21:08:44 <psykose> you can see what inputs/outputs you have in the usual set of configuration apps, like pavucontrol
2023-06-08 21:11:38 <aldcor> ok, i fixed sound output via alsamixer but.. how can I change which mic is being used?
2023-06-08 21:12:37 <psykose> you open something like pavucontrol and change the default input
2023-06-08 21:12:50 <aldcor> i don't have thing like that
2023-06-08 21:13:04 <psykose> if only you could apk add things like that
2023-06-08 21:13:24 <psykose> (i use pulsemixer myself but anything works)
2023-06-08 21:14:02 <aldcor> im just confused since i don't have pulse but.. i still can use pulse thingies
2023-06-08 21:14:22 <psykose> the same pulse configuration applications work with pipewire because it implements the same api
2023-06-08 21:15:46 <psykose> alsamixer doesn't work though afaik and would only be able to control volume, so it works you have something else set up wrong and you're not actually using pipewire
2023-06-08 21:15:51 <psykose> probably related to some other setup you did
2023-06-08 21:16:24 <aldcor> yes I can control volume with alsamixer, nothing else
2023-06-08 21:17:52 <aldcor> ok, in input devices tab I see Microphone (unplugged), i got it plugged tho
2023-06-08 21:19:04 <psykose> what do you see in the configuration tab
2023-06-08 21:19:41 <psykose> (screenshot)
2023-06-08 21:20:15 <aldcor> before I do that, i will go through all options myself and see if something works
2023-06-08 21:20:22 <aldcor> now spotify_player doesn't work
2023-06-08 21:21:31 <psykose> ps aux | grep wire
2023-06-08 21:22:46 <aldcor>  0:08 /usr/bin/pipewire, 0:03 /usr/bin/wireplumber, 0:08 /usr/bin/pipewire-pulse, 0:00 grep wire. Ok, I noticed I haven't removed ~/.asoundrc file. I will do that
2023-06-08 21:23:39 <aldcor> ok, now spotify_player works
2023-06-08 21:24:45 <psykose> what headset do you have actually
2023-06-08 21:25:04 <psykose> and how does it plug in
2023-06-08 21:34:39 <aldcor> psykose: hyperx cloud alpha. It's very basic setup. One wire for headphones, one wire for mic. I plug them in laptop.. that's it
2023-06-08 21:35:20 <aldcor> i also can confirm that it works perfectly fine. I use that headset on other computer daily
2023-06-08 21:35:35 <psykose> not entirely sure then
2023-06-08 21:39:36 <aldcor> ahh.. :/
2023-06-08 21:46:19 <Guest2533> Hi, I have two bluetooth keyboards, and I am able to pair, trust, and connect to each of them using bluetoothctl. My old simple keyboard works great, but my newer logitech keyboard cannot successfully send any keystrokes to my alpine linux laptop. Any thoughts on trouble-shooting? I saw online someone recommended "modprobe uhid", but that resulted in an error.
2023-06-08 21:47:06 <psykose> you probably upgraded your kernel but didn't reboot yet
2023-06-08 21:48:07 <Guest2533> I don't think I have upgraded anything. How would I do that?
2023-06-08 21:48:42 <psykose> by rebooting
2023-06-08 21:49:21 <Guest2533> Okay, I'll try it. Brb
2023-06-08 21:53:42 <martin1> Hi, I'm back about the bluetooth keyboards. I rebooted, but the same result.
2023-06-08 21:53:51 <psykose> what is the 'modprobe error'
2023-06-08 21:55:45 <martin1> Oh interesting, I don't get the modprobe error on alpine.First, I tried on postmarketOS, and I got "modprobe: module uhid not found in modules.dep"
2023-06-08 21:56:39 <psykose> what kernel was that on postmarketos
2023-06-08 21:56:43 <martin1> Wow, now the keyboard works!
2023-06-08 21:57:32 <martin1> Do you know a command to reveal the kernel?
2023-06-08 21:57:35 <psykose> uname -a
2023-06-08 21:57:58 <martin1> 6.3.3
2023-06-08 21:58:07 <psykose> that is not the output of uname -a, sorry
2023-06-08 22:00:17 <martin1> Linux fairphone-fp4 6.3.3 #1-postmarketos-qcom-sm6350 SMP Wed May 24 14:47:34 UTC 2023 aarch64 GNU/Linux
2023-06-08 22:00:48 <aldcor> huh.. fairphone?
2023-06-08 22:01:46 <martin1> Yes, I am trying to use postmarketOS with Fairphone. I couldn't get my keyboard to work, so I tried it on my alpine laptop, andthe keyboard wouldn't work there either
2023-06-08 22:01:58 <ptrc> well, your kernel package is linux-postmarketos-qcom-sm6350 from pmaports
2023-06-08 22:02:17 <psykose> that has CONFIG_UHID disabled
2023-06-08 22:02:24 <psykose> open an issue in pmaports if you want
2023-06-08 22:02:38 <psykose> specifically here https://gitlab.com/postmarketOS/pmaports/-/issues about https://gitlab.com/postmarketOS/pmaports/-/blob/master/device/community/linux-postmarketos-qcom-sm6350/config-postmarketos-qcom-sm6350.aarch64
2023-06-08 22:03:35 <psykose> you can see # CONFIG_UHID is not set in it, so uhid doesn't exist
2023-06-08 22:04:03 <martin1> I will do that. Thanks very much!
2023-06-09 01:08:59 <hippocrampus> hi, anyone here know if alpine has functionality like gentoo-style USE flags?
2023-06-09 01:09:43 <hippocrampus> for building packages from source locally using ie aports
2023-06-09 01:10:12 <hippocrampus> i read thru the wiki and documentation i could find, and although alpine was (at one point) based on gentoo, i saw nothing confirming if this was possible or not
2023-06-09 01:37:59 <psykose> no
2023-06-09 01:38:57 <hippocrampus> alright, thank you
2023-06-09 02:15:04 <Someone[m]1> <hippocrampus> "i read thru the wiki and..." <- alpine isn't based on gentoo, the package manager use binary, not source
2023-06-09 02:15:40 <abby> it *was* based on gentoo https://en.wikipedia.org/wiki/Alpine_Linux#History
2023-06-09 02:15:49 <Someone[m]1> you can download the source code, refer to the APKBUILD file
2023-06-09 02:16:09 <Someone[m]1> abby: *was*
2023-06-09 02:16:41 <abby> neither me nor hippo claimed it is
2023-06-09 02:17:15 <Someone[m]1> gentoo USE flag is basically instruction to pass --use-feature when you "make"
2023-06-09 02:17:21 <Someone[m]1> s/instruction/instructions/
2023-06-09 02:17:21 <psykose> we know
2023-06-09 02:17:41 <abby> also hippocrampus left a while ago
2023-06-09 02:20:29 <Someone[m]1> why don't Alpine ship linux-hardened?
2023-06-09 02:21:03 <Someone[m]1> it is a "libre alternative" to grsecurity
2023-06-09 02:21:17 <psykose> 
ACTION squints

2023-06-09 02:21:19 <psykose> is it 2017
2023-06-09 02:21:25 <Someone[m]1> or at least toggle on some security features
2023-06-09 02:21:34 <abby> what about a use flag
2023-06-09 02:21:39 <Someone[m]1> psykose: https://github.com/anthraxx/linux-hardened
2023-06-09 02:21:51 <psykose> 
ACTION adds to veto list

2023-06-09 02:21:54 <psykose> ty
2023-06-09 03:50:35 <Saijin_Naib> psykose: Do I have enough credits to bug you over a post-build issue 😛
2023-06-09 03:50:40 <Saijin_Naib> packaging CoBang
2023-06-09 03:50:54 <Saijin_Naib> https://github.com/hongquan/CoBang
2023-06-09 03:51:23 <Saijin_Naib> All swimmingly, just at runtime I'm missing single_version, which looks like some unpackaged py3 module for Poetry-based projects... am I getting into a mega can of worms here?
2023-06-09 03:52:15 <psykose> that and kiss_headers are unpackaged yeah
2023-06-09 03:55:48 <Saijin_Naib> Damn...
2023-06-09 03:55:54 <Saijin_Naib> Alright, this might have to be a project for another day
2023-06-09 03:56:19 <Saijin_Naib> I tried it via flatpak, and CoBang is insanely good at capturing QR/barcodes, even from my crappy webcam in a dark room
2023-06-09 03:56:28 <psykose> i can do it if you want
2023-06-09 03:57:23 <Saijin_Naib> Nah, that's alright. I think I can handle those two to package
2023-06-09 03:57:39 <Saijin_Naib> Its just a bigger project than I can do tonight
2023-06-09 03:57:49 <psykose> sure thing
2023-06-09 03:57:51 <Saijin_Naib> I'll definitely ask for help if I get stuck, as you well know 😛
2023-06-09 03:58:14 <Saijin_Naib> Thanks, as ever
2023-06-09 08:00:01 <LACampbell> how many of you are using the default /bin/ash? I'm trying to be hardcore but there's a few things from the ol bashrc I am struggling to port over
2023-06-09 08:01:16 <dwfreed> imo ash is not a user shell
2023-06-09 08:02:45 <LACampbell> yeah I had no issues with it in my servers or wsl2 dev environments. but as a desktop thing.. some things get annoying
2023-06-09 08:03:39 <LACampbell> I'll switch to bash
2023-06-09 08:03:55 <psykose> i don't use ash but i don't miss anything when i use ssh to something else
2023-06-09 08:04:03 <psykose> so i could totally use it i guess
2023-06-09 08:04:18 <psykose> a working tab and C-r is all i really use
2023-06-09 08:04:23 <LACampbell> I miss being able to name terminal windows after the last command entered. makes it really easy to find them with rofi
2023-06-09 08:04:36 <LACampbell> I also miss the tab working inside quote marks. I have a lot of files with spaces in them
2023-06-09 08:04:53 <dwfreed> tab completion should escape spaces
2023-06-09 08:05:07 <dwfreed> if it doesn't, that's another reason to use bash instead :P
2023-06-09 08:05:09 <LACampbell> it does, but then you have to escape spaces yourself sometimes
2023-06-09 08:05:29 <psykose> https://img.ayaya.dev/qsHquigHeElB
2023-06-09 08:05:30 <psykose> works ok
2023-06-09 08:05:52 <psykose> if you open a quote it works too (though doesn't close it)
2023-06-09 08:06:37 <LACampbell> if you open a quote it adds the nasty escaped space syntax
2023-06-09 08:06:46 <LACampbell> in bash, with quotes it just puts spaces, don't have to escape them
2023-06-09 08:08:25 <psykose> ah you mean that
2023-06-09 08:08:25 <psykose> sure
2023-06-09 08:10:20 <LACampbell> ahh this is much nicer. haven't run into any "I need glibc" issues yet
2023-06-09 09:24:54 <ncopa> im using /bin/ash as my default shell
2023-06-09 09:31:01 <donoban> im playing with xonsh since 1 month ago :)
2023-06-09 10:03:11 <ovf> https://xkcd.com/686/
2023-06-09 10:31:00 <Someone[m]1> <LACampbell> "how many of you are using the..." <- me
2023-06-09 11:07:18 <aldcor> good day! Thank you, Alpine-linux. Really enjoying it so far and here's how my setup looks. Enjoying spotify_player too! :) https://0x0.st/Hcmt.png
2023-06-09 11:08:06 <psykose> a lot of brad artists eh
2023-06-09 11:08:06 <psykose> :D
2023-06-09 11:09:34 <aldcor> yeah, search result function is working :D
2023-06-09 11:37:36 <psykose> incidentally the whole sound thing reminds me of https://listman.redhat.com/archives/fedora-devel-list/2008-January/msg00861.html
2023-06-09 11:45:45 <aldcor> "Software is hard.  The way to fix it is to fix it, not sweep it under the rag." That's the only option I'm left with, it seems (regarding mic).
2023-06-09 11:49:57 <aldcor> i just pretend I didn't buy 80 euro headset with mic ;)
2023-06-09 12:25:11 <eris> im back, still trying to debug steam 
2023-06-09 12:25:15 <eris> found more logs, somehow 
2023-06-09 17:21:58 <frag> i think someone said alpine doesnt work on pinebook pro a year-ish ago, is that still the case?
2023-06-09 17:22:17 <psykose> might actually have changed but not sure
2023-06-09 17:22:26 <psykose> i somewhat doubt it idk
2023-06-09 17:22:30 <psykose> i forget what the situation was like
2023-06-09 17:22:43 <psykose> iirc it was still some stuff downstream
2023-06-09 17:22:50 <frag> how different is postmarketOS? i remember they said they used that
2023-06-09 17:22:52 <psykose> for u-boot at least
2023-06-09 17:23:07 <psykose> i wouldn't say it's very different depending on how you want to use it
2023-06-09 17:23:11 <psykose> you could literally just use it to install
2023-06-09 17:23:22 <psykose> then just delete all their repos and keep just the kernel package from their repo etc
2023-06-09 17:23:27 <ikke> it's mostly alpine with some things on top for device compattibility
2023-06-09 17:23:33 <psykose> +- also the plumbing for initfs etc
2023-06-09 17:23:37 <ikke> righgt
2023-06-09 17:23:47 <psykose> it basically is just alpine with device packages, then some other stuff for uis and the like
2023-06-09 17:24:06 <frag> hmm
2023-06-09 17:24:11 <psykose> the forked packages are mostly about like phone hardware support and not something forced to install
2023-06-09 17:24:29 <frag> bought TWO new laptops .. pinebook pro .. well one USED thinkpad l480
2023-06-09 17:25:10 <frag> i havent tried any much yet, but im not so sure pinebook feels great, or maybe it was manjaro/kde ...
2023-06-09 17:25:39 <psykose> it's just a standard rk3399 so
2023-06-09 17:25:57 <psykose> that's basically "rpi with a screen"
2023-06-09 17:26:13 <frag> i use rpi for my main machine :D
2023-06-09 17:26:22 <psykose> not terrible or anything, just yknow
2023-06-09 17:26:23 <frag> but big tv screen .. 
2023-06-09 17:26:23 <psykose> ah :)
2023-06-09 17:26:25 <psykose> haha
2023-06-09 17:26:49 <frag> trying to cut screen time, maybe can do it if i have to use minimalist laptop i duno
2023-06-09 17:27:06 <frag> rpi did not work at cutting screen time :p
2023-06-09 18:05:19 <frag> btw, they sell a travel-charger for thinkpads ... why not use that all the time? its better/faster to charge with the regular one or something?
2023-06-09 18:05:32 <frag> always found them a bit big
2023-06-09 18:05:41 <ikke> what's special about a travel charger?
2023-06-09 18:05:48 <ikke> less power?
2023-06-09 18:05:55 <frag> its smaller, not sure about power
2023-06-09 18:06:07 <ikke> usually smaller means less power
2023-06-09 18:06:58 <frag> also the "big box" is on the plug so only one wire, much cleaner imo..
2023-06-09 18:07:23 <psykose> you can use it if you want afaik
2023-06-09 18:12:05 <frag> :)
2023-06-09 19:19:39 <frag> hmm cant find the nvme (possibly with some converter) on my thinkpad l480 when trying the installer ..
2023-06-10 07:44:30 <jhjacobs[m]> Is Alpine “oficially” on lemmy somewhere?
2023-06-10 07:44:52 <ikke> Is lemmy a social network?
2023-06-10 07:44:53 <ikke> no
2023-06-10 12:20:54 <msiism> How far back does x86 compatibility go for Alpine's "x86" variant?
2023-06-10 12:22:00 <ikke> 686 nowadays
2023-06-10 12:22:07 <ikke> used to be 586 for a long time
2023-06-10 12:22:08 <infinite[m]> msiism: i586 up to pentium4
2023-06-10 12:22:21 <infinite[m]> such as go rust is pentium4
2023-06-10 12:22:37 <infinite[m]> i can prove it by using i586 machine
2023-06-10 12:22:58 <ikke> i586 is best effort
2023-06-10 12:23:02 <ikke> https://gitlab.alpinelinux.org/alpine/tsc/-/issues/35
2023-06-10 12:23:24 <ikke> sse2 must be available
2023-06-10 12:24:20 <msiism> Okay, so 686 is Pentium 4 and such?
2023-06-10 12:24:26 <ikke> "continue with i586 as a best effort, specify that a platform must actually implement i586 correctly to be supported (e.g. no Vortex86), and some packages may require i686 if upstreams only support i686 (e.g. those with JITs)."
2023-06-10 12:24:38 <infinite[m]> pentium4 include sse2
2023-06-10 12:25:12 <infinite[m]> ikke: yes vortex86dx is i586
2023-06-10 12:25:19 <ikke> infinite[m]: but not really
2023-06-10 12:25:32 <ikke> "
2023-06-10 12:25:34 <ikke> For what it’s worth, your Vortex86 CPU has incomplete i586 support"
2023-06-10 12:39:01 <infinite[m]> <ikke> "For what it’s worth, your Vortex..." <- did you mean mmx?
2023-06-10 12:39:10 <ikke> It's a quote
2023-06-10 12:39:24 <ikke> https://gitlab.alpinelinux.org/alpine/tsc/-/issues/20#note_180544
2023-06-10 12:40:10 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues/13016
2023-06-10 12:41:34 <ikke> "This core seems to be from an i686 device which lacks CMOV, part of the i686 base spec.  Your other device lacks support for lock cmpxchg8b, which is part of the i586 base spec"
2023-06-10 12:47:02 <infinite[m]> interesting
2023-06-10 13:01:00 <LonePlanet> ping
2023-06-10 13:09:37 <frag> pong
2023-06-10 13:51:08 <quinq> peng
2023-06-10 13:56:29 <Newbyte> pog
2023-06-10 14:16:47 <msiism> Does the "extended" image still fit on a CD?
2023-06-10 14:42:20 <_7> hello all, this is powers. i did a little rebranding to unify my main-alias
2023-06-10 15:37:14 <uuidNuniq> _7: who cares the branding 
2023-06-10 16:13:44 <_7> uuidNuniq: for spoof mitigation, so that there isn't much confusion to confirm its me sending code/patch/suggestion 
2023-06-10 16:15:11 <_7> in reality, it goes farther than just a name (i.e. gpg key), but for those just wanting an at-a-glance confirmation it makes it easier
2023-06-10 16:16:43 <_7> i plan on sharing code extensively, so im setting everything up
2023-06-10 20:05:06 <port19> Hi, I'm having a problem installing to the pinebook pro. The pinebook pro is an aarch64 device. On the websites under downloads there exists both a standard image and a "generic arm" image. I should note that I have tow-boot, a u-boot fork that works quite well, flashed to SPI. Booting the standard image gets me up to the point where it loads the lts kernel, then black screen. The "generic arm" image does the normal short display
2023-06-10 20:05:06 <port19> flickering that is normal when booting other operating systems, but then nothing more. How should I proceed?
2023-06-10 20:09:47 <psykose> find the postmarketos image that can boot on it, probably
2023-06-10 20:11:59 <port19> Can I then just switch repos to the alpine upstream or should I stick to pmos?
2023-06-10 20:13:19 <psykose> it does use the alpine repos
2023-06-10 20:13:46 <psykose> + their repos that have a few packages on them
2023-06-10 20:13:57 <psykose> it's not a completely forked distro
2023-06-10 20:14:20 <port19> Oh, then that's probably the safest bet. I'll keep you updated on my progress
2023-06-10 20:29:33 <port19> Aight pmos boots fine. Now to connect to wifi could be a challenge, I don't see iwctl. Maybe their wiki has infos, I'll look 
2023-06-10 20:34:04 <port19> oh, they use network manager. I should be fine from here, cya 
2023-06-10 21:42:10 <PureTryOut> eh they left already. "pmOS" doesn't use networkmanager, some UI's do. The user can just replace it if they feel it's necessary though
2023-06-10 21:49:03 <frag> is there a easy terminal way to connect wifi?
2023-06-10 21:49:30 <psykose> i just use iwctl
2023-06-10 21:50:32 <frag> *takes notes*
2023-06-11 18:59:39 <novorol[m]> anyone got some dash vs ash benchmarks?
2023-06-11 20:45:48 <zcrayfish> busybox ash (which is based on dash) has a lot more configuration options than dash . . . . so hopefully any benchmarks specify the configuration they are using.
2023-06-11 22:57:20 <idkrn[m]> <zcrayfish> "busybox ash (which is based on..." <- Aren't they both based on an older ash?
2023-06-11 22:58:13 <zcrayfish> Probably
2023-06-11 23:00:40 <zcrayfish> https://github.com/brgl/busybox/blob/master/shell/ash.c 
2023-06-11 23:01:01 <zcrayfish> lines 25-27 have enough of the history
2023-06-12 06:39:47 <LACampbell> alpine desktop is going well. played minecraft other day and its now picking up usb drives automatically
2023-06-12 06:41:49 <LACampbell> just gotta regain the ability to run chrome desktop apps from Rofi, and I'm at parity with old linux system
2023-06-12 09:04:27 <Fxzxmic> hello, I found an area that needs improvement in Busybox's base64 tool.
2023-06-12 09:06:24 <Fxzxmic> When we performed base64 on "www" string, the tool obtained "d3d3Cg==".
2023-06-12 09:07:04 <Fxzxmic> It contains a line break "\n".
2023-06-12 09:08:42 <Piraty> Fxzxmic: why not post to libera:#busybox
2023-06-12 09:09:05 <ikke> Fxzxmic: how did you test, with echo
2023-06-12 09:09:07 <ikke> My
2023-06-12 09:09:09 <ikke> ?
2023-06-12 09:09:16 <Piraty> yes
2023-06-12 09:09:19 <Piraty> seems echo
2023-06-12 09:09:33 <Piraty> printf '%s' 'www' works as intended
2023-06-12 09:09:50 <Piraty> 
ACTION throws printf at Fxzxmic

2023-06-12 09:11:17 <Fxzxmic> So the echo caused this problem?
2023-06-12 09:11:44 <Fxzxmic> That's why I didn't go to the busybox group to ask this question from the beginning.
2023-06-12 09:14:15 <ikke> Echo by default includes a newline 
2023-06-12 09:14:48 <ikke> echo test | xxd
2023-06-12 09:16:07 <Fxzxmic> I understand, so I should use "echo -n". Thank you all.
2023-06-12 09:16:45 <ikke> Or printf as Piraty suggested 
2023-06-12 09:17:57 <Fxzxmic> Printf is a bit troublesome, I have never really liked it. 😂
2023-06-12 09:19:01 <quinq> But it does what you need
2023-06-12 09:20:04 <aron> or you do not need echo/printf at all, depends on
2023-06-12 09:20:17 <quinq> And it's less of a footgun than random echo implementation
2023-06-12 09:20:21 <quinq> s
2023-06-12 09:20:49 <quinq> Yeah, you can heredoc :>
2023-06-12 09:21:01 <aron> quinq: "echo -n" is posix
2023-06-12 09:21:18 <quinq> More or less :)
2023-06-12 09:21:21 <quinq> It's random
2023-06-12 09:21:24 <quinq> “implementation-defined”
2023-06-12 09:21:50 <aron> I usually read these:
2023-06-12 09:21:51 <aron> https://pubs.opengroup.org/onlinepubs/009695399/utilities/echo.html
2023-06-12 09:21:59 <quinq> If you need to print strings in a script, usually printf is a tad less surprising, printf '%s' "$foo"
2023-06-12 09:22:09 <quinq> Indeed aron
2023-06-12 09:22:38 <quinq> “If the first operand is -n, or if any of the operands contain a <backslash> character, the results are implementation-defined.”
2023-06-12 09:22:42 <quinq> But also “On XSI-conformant systems, if the first operand is -n, it shall be treated as a string, not an option.”
2023-06-12 09:24:26 <aron> I missed the  first paragraph, and I jumped to XSI-conformant
2023-06-12 09:24:42 <aron> sry
2023-06-12 09:25:05 <Fxzxmic> Interesting.
2023-06-12 09:25:35 <quinq> aron, no need :)
2023-06-12 09:25:51 <aron> it seems XSI is really a POSIX extension
2023-06-12 09:25:57 <aron> ah, soooo much standards
2023-06-12 09:27:33 <Fxzxmic> But as long as Alpine is not "random", it's good because I only use it currently. 😂
2023-06-12 09:28:04 <aron> well, it is not just about being Alpine
2023-06-12 09:28:20 <aron> on alpine you have different shells too
2023-06-12 09:28:27 <quinq> Fxzxmic, if it's for your local script yeah, not so much of an issue, as long as it werks I suppose :)
2023-06-12 09:28:29 <aron> ash, dash, bash
2023-06-12 09:29:04 <aron> dash's "echo" behaves very much differently
2023-06-12 09:29:49 <aron> + you can have a 4th "echo" from binutils
2023-06-12 09:30:11 <aron> so yes, if it is a *very* local script, then go for it
2023-06-12 09:31:15 <quinq> Seems like echo -n doesn't print a newline on ash
2023-06-12 09:31:39 <quinq> But yeah, at least you know Fxzxmic :)
2023-06-12 09:31:58 <Fxzxmic> aron: I have recorded this in my memo, thanks.
2023-06-12 09:31:59 <aron> ash is very similar to bash, but without bashism
2023-06-12 09:33:09 <aron> Fxzxmic: it also works if you have a controlled environment, btw
2023-06-12 09:34:16 <Fxzxmic> OK
2023-06-12 09:34:41 <aron> one approach can be:  "busybox echo -n www"  (this is slower, but does not depends on current shell)
2023-06-12 10:07:40 <Piraty> aron: what's up with dash's echo?
2023-06-12 10:08:18 <Piraty> use busybox standalone shell mode to be very sure what will be called
2023-06-12 10:09:45 <Piraty> useful for resqueue shells etc
2023-06-12 10:25:06 <aron> Piraty: nothing special, but for example, "echo -e" is pretty different from the "defacto standard"
2023-06-12 10:26:07 <aron> (yes, I am aware that -e is not even listed in the posix text, but you know how I mean)
2023-06-12 14:25:05 <Ermine> Does apk create apknew files only in /etc ?
2023-06-12 15:18:54 <psykose> in the protected dirs yeah
2023-06-12 15:30:59 <aldcor> in my nativa language `dirsa` means ass
2023-06-12 15:31:07 <aldcor> native*
2023-06-12 15:34:34 <Yasen6275> afternoon
2023-06-12 15:34:50 <Yasen6275> Is mate desktop still supported?
2023-06-12 15:35:03 <Yasen6275> I have problrms installing it
2023-06-12 16:00:44 <Saijin_Naib> What problem, error messages, and are you mixing edge with stable? 
2023-06-12 19:14:27 <acdw> did Syncthing change where it stores its default folder on a recent update?
2023-06-12 19:36:28 <frag_> not to be a smart ass but i think life is better without syncing, so if you can maybe avoid it ...
2023-06-12 19:37:11 <kmmndr> Hi all :-)
2023-06-12 19:38:19 <ikke> o/
2023-06-12 19:38:19 <kmmndr> Is anyone here using power-profiles-daemon ?
2023-06-12 19:40:47 <ikke> what if someone was?
2023-06-12 19:41:42 <psykose> ikke: frustration eh :P
2023-06-12 19:42:08 <ikke> This just landed after you asked about it: http://dup.pw/alpine/aports/2c12a5aa5fef
2023-06-12 19:42:39 <ikke> though nothing consequential
2023-06-12 19:44:09 <psykose> not related to anything no
2023-06-12 19:44:41 <psykose> it's also actually wrong
2023-06-12 19:44:51 <ikke> as it removes python altogether/
2023-06-12 19:44:52 <ikke> ?
2023-06-12 19:47:20 <psykose> glib
2023-06-12 19:47:27 <ikke> oh
2023-06-12 19:47:28 <psykose> but yeah it doesn't work without a bunch of polkit setup
2023-06-12 19:48:22 <psykose> relies on it for any action
2023-06-12 19:48:29 <psykose> not even as root can you use it
2023-06-12 19:51:29 <kmmndr> I was wondering if someone has been successfully used it
2023-06-12 19:51:36 <acdw_> has Syncthing changed its default folder location with an update lately? 
2023-06-12 19:52:50 <psykose> aren't you supposed to manually specify what folders to monitor?
2023-06-12 19:52:52 <abby> pretty sure the default folder location is set in your config file after you create the folder, so it should never change
2023-06-12 19:53:09 <abby> and the default location is a subdir of $HOME, and i don't think that has changed
2023-06-12 19:54:32 <acdw_> weird. when i first installed it i had a folder in /var
2023-06-12 19:54:42 <acdw_> and now the default folder is ~/Sync
2023-06-12 19:55:15 <acdw_> i figured alpine had it packaged differently or something ... this might have nothing to do with alpine of course
2023-06-12 19:55:34 <abby> ~/Sync is the upstream default
2023-06-12 19:57:02 <acdw_> okay something weird is going on then... thanks!
2023-06-12 19:57:37 <ikke> but options is an argument
2023-06-12 19:58:05 <ikke> so it should not complain anyway, but maybe due to the assignment
2023-06-12 19:58:10 <ikke> but I have no idea why they do it like this
2023-06-12 20:01:33 <psykose> options argument?
2023-06-12 20:01:49 <ikke> sorry wrong channel
2023-06-12 20:02:31 <quinq> You can't tease like that about options and not finish
2023-06-12 20:02:41 <ikke> https://github.com/containers/podman/blob/9eef30051c83f62816a1772a743e5f1271b196d7/pkg/bindings/containers/containers.go#LL439C1-L439C13
2023-06-12 20:02:51 <ikke> someone asked about the _ = options
2023-06-12 20:03:45 <quinq> Thanks ;)
2023-06-12 21:47:04 <frag> did you ever check a checksum and it didnt fit?
2023-06-12 21:47:35 <psykose> didn't fit?
2023-06-12 21:47:51 <quinq> In the terminal width
2023-06-12 21:48:15 <frag> didnt match*
2023-06-12 21:48:37 <quinq> frag, some projects just re-tag releases on different commits instead of doing actual fixup releases
2023-06-12 21:49:00 <quinq> (re-tag as in re-using the same tag, for different code)
2023-06-12 21:49:16 <psykose> yeah a bunch of times
2023-06-12 21:49:34 <abby> mfw
2023-06-12 21:49:57 <quinq> yeah ^_^
2023-06-12 21:50:03 <psykose> imagine not fitting the checksums
2023-06-12 21:50:05 <psykose> abby: post fit
2023-06-13 04:56:35 <LACampbell> fixed my chrome web apps issue. my alpine desktop is now at the point my old desktop was. feels good
2023-06-13 04:56:51 <LACampbell> oh wait, I forgot CUPs. but to be fair last time I cheated and just copied the config my wifes mac came up with
2023-06-13 04:57:14 <psykose> i don't think i've ever had a working cups anywhere ever on linux
2023-06-13 04:57:29 <LACampbell> lol. I had one following that method
2023-06-13 04:57:30 <psykose> i heard the generic netprint 'works' these days but..
2023-06-13 04:58:13 <LACampbell> I swear it's true! I was sending documents to a wireless computer from my old desktop like I was in star trek
2023-06-13 04:58:19 <LACampbell> *wireless printer
2023-06-13 04:59:04 <psykose> :)
2023-06-13 05:04:47 <LACampbell> now I kind of want to set up an NFS server
2023-06-13 06:00:24 <tanul> Hello, can anyone please suggest the procedure to use ntlm in alpine
2023-06-13 06:00:37 <tanul> I have tried this but no luck https://github.com/simo5/gssntlmssp
2023-06-13 06:00:43 <tanul> my application is in .net
2023-06-13 06:07:42 <psykose> what is an ntlm
2023-06-13 06:08:12 <psykose> also you tried some 2-year outdated fork of something
2023-06-13 06:08:45 <psykose> ntlm sounds like something you configure samba to do
2023-06-13 06:16:05 <tanul> ntlm is an authentication type.. So we are running our .net app in containers with alpine 3.17 as base image. The target sever is a windows based and the rest api hosted on it expect ntlm authentic client calls
2023-06-13 06:17:14 <tanul> i have checked with curl command curl --ntlm --negotiate and it works fine..But as per my understanding alpine doesn't support ntlm auth calls
2023-06-13 06:17:55 <tanul> I tried this apk get gss-ntlmssp.. But it is unavailable for alpine.. 
2023-06-13 06:18:25 <tanul> Is there any way we can call such kind of rest api. Or changing the base image is the only option
2023-06-13 06:25:44 <grawity> do you *really* need NTLM, as opposed to Kerberos
2023-06-13 06:29:26 <quinq> Seems that NTLM is deprecated from more than 10 years ago, so finding open-source support for such proprietary protocol could be a problem
2023-06-13 06:30:10 <quinq> But as psykose said, samba seems to support it
2023-06-13 06:30:26 <psykose> i googled ntlm
2023-06-13 06:30:28 <psykose> > Since 2010, Microsoft no longer recommends NTLM in applications
2023-06-13 06:30:35 <psykose> yeah no, no idea what package is missing and don't care
2023-06-13 06:30:36 <tanul> Target server is third party... We tried passing kerberos as well but got 401 unauthorized.. We checked with the admin of the rest api and this is his words " In target rest api kerberos authentication falls back to NTLM". I don't know the meaning of this line
2023-06-13 06:31:58 <quinq> Why do you care though what that REST API does behind its API? If it uses kerberos or NTLM internally shouldn't matter to a web client
2023-06-13 06:32:25 <aquarius> Sorry I got logged out.. I dont know what is the meaning of this line..
2023-06-13 06:32:39 <aquarius> kerberos falls back to ntlm
2023-06-13 06:32:48 <quinq> 08:31:57  quinq$ Why do you care though what that REST API does behind its API? If it uses kerberos or NTLM internally shouldn't matter to a web client
2023-06-13 06:33:41 <aquarius> I tried this code https://github.com/dotnet/Docker.DotNet/issues/642#issue-1748456397
2023-06-13 06:34:24 <aquarius> Keep on getting this error
2023-06-13 06:34:25 <aquarius> Flurl.Http.FlurlHttpException: Call failed. NTLM authentication is not possible with default credentials on this platform: GET  ---> System.PlatformNotSupportedException: NTLM authentication is not possible with default credentials on this platform. at
2023-06-13 06:37:19 <aquarius> We have passed the credentials but still while using we alpine we always get this error System.PlatformNotSupportedException: NTLM authentication is not possible with default credentials on this platform
2023-06-13 06:38:57 <psykose> funnily that error string exists in two places but you truncated the error so it's not clear which
2023-06-13 06:39:35 <aquarius> Sorry
2023-06-13 06:39:41 <aquarius> This is the complete error
2023-06-13 06:39:43 <aquarius> Flurl.Http.FlurlHttpException: Call failed. NTLM authentication is not possible with default credentials on this platform: GET  ---> System.PlatformNotSupportedException: NTLM authentication is not possible with default credentials on this platform.
2023-06-13 06:42:28 <psykose> you already posted that 4 times, and it's missing the call stack after it
2023-06-13 06:44:12 <psykose> it does just sound like https://github.com/dotnet/runtime/issues/66633 though
2023-06-13 06:51:33 <aquarius> hmm may be.. 
2023-06-13 06:51:49 <aquarius> But no idea how to resolve it.. Our production environment is impacted beause of this
2023-06-13 07:00:25 <grawity> quinq: samba for server side, gss-ntlmssp is the client side
2023-06-13 07:06:00 <aquarius> gss-ntlmssp this is unavailable in alpine
2023-06-13 07:06:14 <aquarius> I even tried building it but no luck
2023-06-13 07:06:58 <grawity> i'd really focus on getting kerberos to work tbh
2023-06-13 07:07:17 <grawity> i've packaged gss-ntlmssp for arch, it's tiny, but ... never actually used it even once
2023-06-13 07:07:25 <grawity> you said you tried with curl as well?
2023-06-13 07:13:30 <psykose> i built it without issue but i doubt it solves your problem
2023-06-13 07:17:47 <grawity> if you build it locally, make sure the /etc/gss/mech.d/gssntlmssp.conf is installed, to tell libgssapi/libkrb5 where to find the new mech
2023-06-13 07:20:44 <aquarius> I have directly used these files.. This person already build and kept it in the github profile
2023-06-13 07:20:44 <aquarius> https://github.com/mikeTWC1984/gssntlm
2023-06-13 07:22:12 <aquarius> But no luck.. I have made some changes in the code and now getting this error
2023-06-13 07:22:13 <aquarius> GSSAPI operation failed with error - No credentials were supplied, or the credentials were unavailable or inaccessible (SPNEGO cannot find mechanisms to negotiate).
2023-06-13 07:22:55 <aquarius> I'll try to again keep those so file and gssntlmssp.conf 
2023-06-13 07:23:25 <aquarius> Just one question that .gssntlmssp.so should be with which permission
2023-06-13 07:23:28 <aquarius> a+x
2023-06-13 07:23:34 <aquarius> should I use this..
2023-06-13 07:24:49 <aquarius> should I keep it to 666 or 777 permission for this gssntlmssp.so file
2023-06-13 07:29:02 <grawity> I think most .so library files have +x, don't they
2023-06-13 07:29:13 <grawity> generally not world-writable thouh
2023-06-13 07:29:13 <psykose> you can't download random .so's and expect them to work
2023-06-13 07:29:24 <psykose> it sounds like you have no idea what you're doing
2023-06-13 07:29:29 <psykose> if you want to build it, here you go https://img.ayaya.dev/y7Q6HKhcCqMo
2023-06-13 07:38:49 <aquarius> yup correct.. I am not aware of these things
2023-06-13 07:39:49 <aquarius> Thank you so much... So after building this.. That file placement has to be manual
2023-06-13 07:39:51 <aquarius> Am I correct
2023-06-13 07:39:55 <aquarius> Am I correct
2023-06-13 07:40:34 <aquarius> Means i explicitly specify command in docker file for this file /etc/gss/mech.d/gssntlmssp.conf
2023-06-13 07:40:48 <aquarius> Or does building it manages everything
2023-06-13 07:57:20 <aquarius> This error is coming with the docker file
2023-06-13 07:57:23 <aquarius> the --mount option requires BuildKit. Refer to https://docs.docker.com/go/buildkit/ to learn how to build images with BuildKit enabled ##[error]the --mount option requires BuildKit. Refer to https://docs.docker.com/go/buildkit/ to learn how to build images with BuildKit enabled
2023-06-13 08:01:33 <psykose> then simply remove it
2023-06-13 08:02:00 <psykose> or read the docs
2023-06-13 08:02:05 <psykose> looks pretty self explanatory
2023-06-13 09:48:19 <aquarius> @psykose, While running docker file getting this error. gzip: invalid magic tar: Child returned status 1
2023-06-13 09:48:32 <aquarius> Any suggestion please.. Would be grateful
2023-06-13 09:48:55 <Apachez> give up that child for adoption
2023-06-13 09:49:35 <psykose> you're very good at this 'remove the whole error and paste one line of it' thing
2023-06-13 09:49:41 <psykose> the error before it is the actual one
2023-06-13 09:49:57 <psykose> which is going to be 'the http request failed'
2023-06-13 09:50:35 <aquarius> gzip: invalid magic
2023-06-13 10:37:48 <aquarius> I have merged the dockerfile and it is building and placing the so file in /usr/lib/gssntlmssp/gssntlmssp.so
2023-06-13 10:37:52 <aquarius> But the error is coming
2023-06-13 10:38:10 <aquarius> GSSAPI operation failed with error - No credentials were supplied, or the credentials were unavailable or inaccessible (KDC reply did not match expectations).
2023-06-13 10:40:35 <aquarius> After building in docker do we have to still add this file /etc/gss/mech.d/gssntlmssp.conf
2023-06-13 10:41:00 <aquarius> If yes then what should be the content of the file
2023-06-13 16:18:57 <oleovit> hi. i have installed the utmps pkg. i see the services utmp-init, utmp and utmp-prepare. i guess, the utmp serv. must be at runlevel boot, the other ones, if needed (?) at runlevel default. is this correct?
2023-06-13 16:23:41 <oleovit> .. or should i skip this completely, because of security reasons ("because its difficult to implement it in a secure way"/musl)
2023-06-13 16:28:44 <oleovit> .. oh, there is a setup-utmp script. i guess this is the one to use..
2023-06-13 16:40:45 <blt[m]> psykose: would you consider moving tessen from edge/testing to community ? (that's the last package I have on my system which require me to keep edge/testing repo 🙂 )
2023-06-13 17:14:49 <hmmbot> I am trying to try out the alpine standard iso with cmdline qemu using this: `qemu-system-x86_64 -boot d -cdrom ./alpine-standard-3.18.0-x86_64.iso -m 1024 -nographic`
2023-06-13 17:15:25 <hmmbot> alpine starts up, then quickly switches to a blank screen I can't get out of.  What is happening? What is the right way to try out alpine directly from the iso at cmdline?
2023-06-13 17:32:25 <psykose> blt[m]: sure
2023-06-13 17:33:08 <Quillith> Hi, I notice that my audio takes a second or two to kick in when I resume/play a video if there hasn't been any audio playing. It doesn't lag behind, it's just a few seconds of silent video. I have pipewire, wireplumber, pipewire-alsa, and am in the audio and video groups. I'm unsure what to check, any ideas? Thanks
2023-06-13 17:34:28 <psykose> that sounds like pipewire autosuspend of the audio device
2023-06-13 17:34:36 <psykose> and then it takes that long to resume when something plays again
2023-06-13 17:34:55 <psykose> https://unix.stackexchange.com/questions/676846/how-do-i-disable-audio-sink-suspend-on-idle-using-wireplumber-in-fedora-35-so-th is more or less how to change it
2023-06-13 17:35:05 <psykose> one line in wireplumber
2023-06-13 17:35:13 <Quillith> Thank you, I'll do that!
2023-06-13 17:35:14 <psykose> but note the suspend obviously like depowers the device or w/e
2023-06-13 17:35:22 <psykose> it's mostly a power+cpu saving thing
2023-06-13 17:35:27 <Quillith> Oh I see
2023-06-13 17:36:01 <Quillith> I usually have audio playing if my computer is on, so I think I'm okay to turn off the autosuspend if shortening the resume time isn't an option
2023-06-13 17:36:03 <psykose> my bluetooth headphones feel like they do something similar, but the time is like .5s so i don't notice it much
2023-06-13 17:36:25 <psykose> yeah, i don't think the resume time is a tweak, it just takes that long to actually resume
2023-06-13 17:36:36 <Quillith> Yeah if it was small like that I don't think I would mind, but when it's multiple seconds it gets to be annoying, especially with things like YT shorts that are only a few seconds long anyway
2023-06-13 17:36:48 <psykose> indeed
2023-06-13 17:37:14 <psykose> but yeah just copy it into ~/.config for user wireplumber and it should work
2023-06-13 17:37:20 <psykose> if it does not then it was actually something else
2023-06-13 17:40:29 <Guest3000> I am trying to install Whonix on Alpine, and part of the process is extracting the downloaded xz file, but the tar command seems to be having trouble with the qcow2 files. When I run "tar -xf Whonix*" I get the error "tar: unknown typeflag: 0x53". Do I need anything else installed for tar to work correctly?
2023-06-13 17:48:30 <psykose> probably apk add tar
2023-06-13 17:49:55 <Quillith> You can install Whonix on alpine? I didn't know that. I remember making a post on the kicksecure forum saying they should and starting a column for alpine on their usability chart, that's awesome to hear it's coming along. Whonix looks pretty cool
2023-06-13 17:50:00 <quinq> Guest3000, xz -d -c Whonix | tar -x
2023-06-13 17:51:18 <Guest3000> I guess I'm finding out whether I personally am capable of running Whonix on alpine lol. I already got KVM working, and ran a debian-based virtual machine with no problems
2023-06-13 17:51:38 <Guest3000> Thanks quing, I'll try that command
2023-06-13 17:54:37 <quinq> (it just decompressing the xz file to stdout and piping the uncompressed tar archive into tar for extraction)
2023-06-13 17:55:49 <Guest3000> I get the same error with that command, unfortunately, but I did successfully run "tar -xf" to unpack the files on termux for android, so maybe I can just copy the decompressed file over. Maybe I am missing something to fix the tar command? 
2023-06-13 18:00:31 <Quillith> Thanks Psychose, that fix worked great! appreciate the help
2023-06-13 18:01:33 <ikke> Guest3000: maybe try bsdtar
2023-06-13 18:02:41 <quinq> $ xz -d -c linux-6.3.tar.xz | busybox tar -x
2023-06-13 18:02:45 <quinq> worksforme™
2023-06-13 18:03:02 <ikke>  They mentioned qcow2
2023-06-13 18:03:23 <quinq> haha ok.
2023-06-13 18:03:46 <quinq> I read “xz” and “tar”, not “file system”, sorry
2023-06-13 18:04:59 <quinq> Doesn't the qemu-img tool do that?
2023-06-13 18:09:22 <Guest3000> bsdtar worked!
2023-06-13 18:09:40 <Guest3000> Thanks, @ikke
2023-06-13 18:13:07 <quinq> nice :)
2023-06-13 18:30:37 <blt[m]> psykose: so if I get it right this will be part of next major release (3.19) after spending the life cycle in edge/community?
2023-06-13 18:30:43 <psykose> ye
2023-06-13 21:05:47 <LACampbell> I have networking, wpa, and wpa_cli running as boot rc-services. my user is part of the wheel group. but if I run "wpa_cli" at the terminal it says it can't connect to wpa_supplicant
2023-06-13 21:06:03 <LACampbell> which is definitely running, I can see its process ID
2023-06-13 21:08:21 <LACampbell> hmm, even wpa_cli -i wlan0 gives me that
2023-06-13 21:10:39 <psykose> do you have the wpa_supplicant control socket configured
2023-06-13 21:13:29 <LACampbell> that's a good question
2023-06-13 21:21:44 <LACampbell> god damn I get confused by all the rc-* commands. some of them go "service verb", some of them are "verb service"
2023-06-13 21:26:34 <LACampbell> switched to iwd. that was painless
2023-06-13 21:27:53 <LACampbell> everything just works, including iwctl. nice
2023-06-13 21:29:43 <zcrayfish> iwd is amazing. 
2023-06-13 21:30:11 <LACampbell> yeah it's a very nice interface
2023-06-13 21:57:39 <soapman> hello, I have a question about Alpine and starting X without display manager
2023-06-13 21:58:07 <zcrayfish> sure. ask away
2023-06-13 21:58:15 <quinq> LACampbell, first example configuration line of wpa_supplicant.conf(5) shows how to enable the wpa_cli for wheel
2023-06-13 21:58:56 <soapman> I ran the script 'setup-xorg-base', then used apk to add  xfce4 xfce4-terminal dbus dbus-x11 
2023-06-13 21:59:15 <LACampbell> quinq: my mistake, I have to start reading config man pages
2023-06-13 21:59:23 <soapman> I start it from a tty console, and XFCE fires up, but my keyboard & mouse are not working at all.
2023-06-13 21:59:36 <quinq> LACampbell, no mistake, just hinting :)
2023-06-13 22:00:32 <soapman> I start X with 'startx'
2023-06-13 22:00:48 <LACampbell> do you have udev?
2023-06-13 22:01:25 <psykose> are you in the input and video group
2023-06-13 22:03:26 <soapman> udev (yes)
2023-06-13 22:03:37 <soapman> i have not modified any group setting
2023-06-13 22:04:08 <soapman> other users on youtube, started with LDM display manager, and that's the only difference??
2023-06-13 22:05:39 <soapman> i'm in the video group, but 'not' input group
2023-06-13 22:06:05 <LACampbell> that might be it
2023-06-13 22:06:47 <soapman> I'll give that a try.  (good suggestion), didn't think on that one
2023-06-13 22:09:55 <LACampbell> good luck, I just started using alpine as a desktop a couple of weeks ago too
2023-06-13 22:16:30 <soapman> I luv this distro (light), awesome simplicity (server)
2023-06-13 22:16:58 <LACampbell> yeah I've been using it on my server for years. and as my dev environment in windows WSL2. took the plunge on the destkop, nice to have it all matching
2023-06-13 22:18:30 <soapman> i've always disliked systemD
2023-06-13 22:20:39 <LACampbell> oh yeah, why's that? I kind of liked voids one, where you made a symbolic link to start services
2023-06-13 22:20:57 <LACampbell> wasn't alpine considering dropping openRC for something else?
2023-06-13 22:21:12 <abby> systema
2023-06-13 22:21:20 <soapman> it works! (awesome)
2023-06-13 22:21:36 <LACampbell> abby: oh yeah thats right. still happening?
2023-06-13 22:22:08 <abby> the a stands for awesome
2023-06-13 22:22:15 <LACampbell> 
ACTION takes notes

2023-06-13 23:06:59 <helpless> hello everyone, I was trying to install manually alpine because I have a separated home partition, but after following steps in wiki I tried to mount partitions
2023-06-13 23:07:30 <helpless> but I can't because in dmesg it says that utf8 isnot a reccomended IO charset
2023-06-13 23:08:04 <helpless> how can I succesfully mount stuff? it is the first time I experienced such error
2023-06-13 23:10:11 <LACampbell> your partition name has utf8 chars?
2023-06-13 23:11:08 <helpless> I didnt give any name to partitions
2023-06-13 23:11:37 <helpless> shall I format again, giving a simple name?
2023-06-13 23:12:09 <LACampbell> so I did a similar thing a couple of weeks ago, installed alpine on an existing home partition. but I'm not a wizard at this stuff. so take what I say with a grain of salt
2023-06-13 23:12:16 <LACampbell> yeah, what command are you actually typing?
2023-06-13 23:12:36 <helpless> i'm following this https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_options_and_settings
2023-06-13 23:12:50 <helpless> I'm stuck at finalizing part
2023-06-13 23:12:54 <helpless> just mount
2023-06-13 23:13:36 <helpless> for partitions I ran plain mkfs.* /dev/*
2023-06-13 23:13:50 <helpless> replacing * with proper stuff
2023-06-13 23:13:54 <helpless> no flags
2023-06-13 23:14:20 <LACampbell> well you need like 4 partitions, right?
2023-06-13 23:14:25 <LACampbell> boot, swap, home, and root
2023-06-13 23:14:47 <helpless> actually I wanted to have a swap file, so I have boot, efi, root, and home
2023-06-13 23:14:58 <helpless> but home has stuff inside and didn't want to backup
2023-06-13 23:15:42 <LACampbell> I mean backup anyway incase you fat finger something
2023-06-13 23:15:51 <helpless> oh yeah xd
2023-06-13 23:16:17 <helpless> but well, I didn't want to spend time moving files
2023-06-13 23:16:28 <psykose> what is the actual error
2023-06-13 23:17:05 <helpless> can't mount partitions, dmesg says utf8 is not a reccomended IO charset for FAT filesystems
2023-06-13 23:17:15 <helpless> semi-auto install process
2023-06-13 23:17:22 <psykose> that's not an error, it's a warning for the fat partition only
2023-06-13 23:17:35 <psykose> it also doesn't mean or do anything, it continues after that fine to the next thing
2023-06-13 23:17:37 <helpless> LACampbell suggested that maybe is a problem of partition name
2023-06-13 23:18:09 <helpless> well
2023-06-13 23:18:11 <psykose> it's always printed when iocharset=utf8 is in options and can just be ignored
2023-06-13 23:18:16 <psykose> it's not the source of any issues here
2023-06-13 23:18:16 <helpless> mount says Invalid argument
2023-06-13 23:18:27 <psykose> for something else but not that, sure
2023-06-13 23:18:30 <helpless> and I'm not giving any argument
2023-06-13 23:18:39 <psykose> what is your fstab
2023-06-13 23:19:00 <helpless> i have not a fstab yet, I just formatted partitions
2023-06-13 23:19:32 <helpless> just the default one of live iso
2023-06-13 23:19:40 <psykose> what did you run then
2023-06-13 23:20:04 <helpless> mount /dev/partition /mnt
2023-06-13 23:20:11 <helpless> replace partition with partition name
2023-06-13 23:20:20 <helpless> with all partition I had no luck
2023-06-13 23:20:30 <psykose> can you post dmesg
2023-06-13 23:21:02 <helpless> lemme see
2023-06-13 23:25:00 <helpless> ix.io/4yaT
2023-06-13 23:25:43 <psykose> this doesn't look like a normal device
2023-06-13 23:26:10 <helpless> what do you mean
2023-06-13 23:26:10 <psykose> /dev/"partition" is actually some squashfs thing and this is a phone
2023-06-13 23:26:20 <helpless> mh no
2023-06-13 23:26:34 <helpless> i have mmcblk1
2023-06-13 23:27:24 <helpless> after 3649 is because i had to do tethering with my phone
2023-06-13 23:27:34 <helpless> I disconnected it 
2023-06-13 23:27:43 <helpless> and reconnected again to download curl
2023-06-13 23:27:50 <psykose> ah that makes more sense
2023-06-13 23:28:05 <psykose> but yeah this looks fine then
2023-06-13 23:28:08 <psykose> you mounted the same thing 20 times
2023-06-13 23:28:18 <helpless> yeah but without success xd
2023-06-13 23:28:19 <psykose> to see what the einval is you'd have to strace it i guess
2023-06-13 23:28:48 <helpless> i dont know what is this
2023-06-13 23:32:02 <helpless> ok I ran strace mount ... 
2023-06-13 23:32:14 <helpless> and it said exited with error 255
2023-06-13 23:35:55 <helpless> http://ix.io/4yaV
2023-06-13 23:37:39 <psykose> what's the partition type of /dev/mmcblk1p3
2023-06-13 23:37:52 <psykose> you have to to mount -t type ..
2023-06-13 23:38:04 <psykose> it tried vfat and some others but they all fail, that's why it doesn't work
2023-06-13 23:38:13 <psykose> e.g. -t ext4 for ext4
2023-06-13 23:38:31 <helpless> ohhhhhhh
2023-06-13 23:38:34 <helpless> I see
2023-06-13 23:38:38 <helpless> let me try
2023-06-13 23:38:52 <helpless> wow
2023-06-13 23:39:00 <helpless> it was such a stupid error
2023-06-13 23:39:09 <helpless> idk why it wasn't mentioned in the wiki tho
2023-06-13 23:39:14 <helpless> thank you so much!
2023-06-14 02:08:26 <gnu-only-way> does alpine contain proprietary code, binary blobs? or is it open source and libre?
2023-06-14 02:08:58 <gnu-only-way> i know the mainstream linux kernal is loaded with malware, does alpine use the blobs or not?
2023-06-14 02:09:02 <gnu-only-way> *kernel
2023-06-14 02:10:00 <abby> shit you got me, ill remove the malware now
2023-06-14 02:10:17 <gnu-only-way> lol
2023-06-14 02:10:33 <gnu-only-way> does alpine contain non-free binary blobs or is it open source?
2023-06-14 02:10:59 <gnu-only-way> no documentation anywhere i can find answering this
2023-06-14 02:11:19 <gnu-only-way> and yes..... the closed source firmware/drivers _does_ contain spyware :)
2023-06-14 02:11:24 <gnu-only-way> I'm being serious
2023-06-14 02:11:40 <gnu-only-way> linux mainstream kernel is backdoored
2023-06-14 02:16:48 <Someone[m]1> It contain firmware blobs.
2023-06-14 02:17:17 <Someone[m]1> <gnu-only-way> "linux mainstream kernel is..." <- Stop spreading misinformation
2023-06-14 02:17:40 <Someone[m]1> you can use linux-libre and have no microcode and security warning supressed
2023-06-14 02:18:35 <gnu-only-way> Not mis info
2023-06-14 02:18:40 <gnu-only-way> i can back my claim
2023-06-14 02:18:44 <gnu-only-way> but that is different topic
2023-06-14 02:19:01 <gnu-only-way> thanks for answering my question though
2023-06-14 02:19:26 <Someone[m]1> security communities can back their claims
2023-06-14 02:19:32 <gnu-only-way> I work in cybersec
2023-06-14 02:19:52 <Someone[m]1> also I forgot to install microcode for months
2023-06-14 02:20:01 <gnu-only-way> So microcode, even on my intel systems I block microcode, but I am using ARM cpu for most things these days. 
2023-06-14 02:20:21 <gnu-only-way> I'm very new to alpine, possible or easy to build it using linux libre?
2023-06-14 02:20:29 <Someone[m]1> gnu-only-way: in your case, you need full transparency? 
2023-06-14 02:20:32 <gnu-only-way> or way to delete the closed source stuff post install
2023-06-14 02:20:51 <Someone[m]1> you can manually uninstall firmwares lol
2023-06-14 02:21:06 <gnu-only-way> yes I need 100% open source
2023-06-14 02:21:29 <Someone[m]1> also are you a NSA agent
2023-06-14 02:21:55 <gnu-only-way> I am not a criminal
2023-06-14 02:21:56 <gnu-only-way> so no
2023-06-14 02:22:17 <gnu-only-way> is it as easy as rm'ing /lib/firmware? 
2023-06-14 02:22:53 <Someone[m]1> apk del linux-firmware
2023-06-14 02:23:01 <Someone[m]1> if they are backdoored it is impossible to remove lmao
2023-06-14 02:23:19 <gnu-only-way> apk del linux-backdoor
2023-06-14 02:23:27 <gnu-only-way> that won't work?
2023-06-14 02:24:32 <abby> you want apk add linux-frontdoor
2023-06-14 02:24:38 <Someone[m]1> Are you trolling?
2023-06-14 02:25:05 <gnu-only-way> only was trolling on that one comment "linux-backdoor" but was serious
2023-06-14 02:25:11 <gnu-only-way> was referring to linux-firmware
2023-06-14 02:25:37 <gnu-only-way> is apk del linux-firmware same as rm'ing /lib/firmware ?
2023-06-14 02:27:08 <Someone[m]1> I'm using chromium, and I'm having alsa for sound. I'm sandboxing chromium (restricting access to directories with bubblewrap), but it needs some access to /etc for the sound to work. anyone know how to trace the process for its access to /etc?
2023-06-14 02:29:48 <abby> strace -f -e trace=open,read,write or something like that
2023-06-14 02:29:50 <Someone[m]1> it seems that it is accessing a file in /etc
2023-06-14 02:30:07 <Someone[m]1> I allowed /etc/alsa but it doesn't work
2023-06-14 02:30:10 <gnu-only-way> bubblewrap, great choice
2023-06-14 02:31:23 <psykose> styrofoam can be better sometimes
2023-06-14 02:31:54 <abby> i like the packing peanuts made of corn that melt in your mouth
2023-06-14 02:33:20 <gnu-only-way> but with bubble wrap you can twist it and make many bubbles go pop pop pop! in rappid succession, making your neighbor on other side of the wall freak out thinking gun shots went off
2023-06-14 02:33:29 <gnu-only-way> you cant do that with the packing peanuts
2023-06-14 02:35:29 <Someone[m]1> > I work in cybersec
2023-06-14 02:35:29 <Someone[m]1> gnu-only-way also what do you mean in 'cybersec'
2023-06-14 02:43:50 <Someone[m]1> lol
2023-06-14 02:48:07 <abby> "i read people's conspiracy theories online" probably
2023-06-14 02:49:31 <Saijin_Naib> Anyone else having trouble with XFCE Polkit on Edge today after updating? I am having it fail at login across two machines
2023-06-14 02:49:54 <Saijin_Naib> I'm having trouble debugging exactly why, nothing salient seems to be logged in dmesg
2023-06-14 03:04:55 <psykose> hmm
2023-06-14 03:05:01 <psykose> i can think of one thing but not sure
2023-06-14 03:05:48 <psykose> whadya mean by fail
2023-06-14 03:11:24 <ptrc> https://ptrc.gay/xYfMGJWQ
2023-06-14 03:29:19 <psykose> that one i feel like i've seen a bunch of times
2023-06-14 03:29:21 <psykose> no idea what it means tho
2023-06-14 03:59:03 <psykose> also depends on when you last updated
2023-06-14 04:29:13 <Saijin_Naib> Yeah, exactly that, ptrc. Just an empty dialog and nothing printed to dmesg
2023-06-14 04:29:22 <Saijin_Naib> Updated multiple times today, though not in the last few hours
2023-06-14 04:29:32 <Saijin_Naib> Incomplete upgrade of components due to the builders taking a bit on something?
2023-06-14 04:30:31 <Saijin_Naib> In terms of what it causes, on my desktop, issues with login/logout/suspend/etc, on my laptop, all the same, plus I can't interact with DPMS or whatever to change my screen brightness
2023-06-14 04:31:23 <Saijin_Naib> pkexec is broken, so I can't pkexec anything, so adjusting GUFW, for instance, is a no-go
2023-06-14 10:34:43 <maria> pipewire capture on kde seems broken on edge. the popup asking for output selection doesn't open. investigating rn but would like to know if it's a known issue.
2023-06-14 10:36:13 <maria> *screen capture
2023-06-14 10:39:47 <maria> hmm just did an install of 3.18 and it's also broken. was working a few days ago as I was using it to stream with obs :/
2023-06-14 10:51:05 <maria> huh seems kpipewire wasn't installed for some reason. didn't need to install that manually before so I wonder what pulled it in before.
2023-06-14 10:51:59 <maria> Ah, spectacle. sorry for the noise ^^
2023-06-14 12:26:28 <Ins> Привет
2023-06-14 12:26:45 <Ins> Hi
2023-06-14 12:28:21 <Ins> Help me install alpine + labwc for desktop please
2023-06-14 12:35:17 <quinq> Hi Ins, there is this: https://wiki.alpinelinux.org/wiki/LabWC
2023-06-14 12:38:32 <Ins> quinq: I did it according to this instruction. But something doesn't work out. Is there a list of sequential commands to install on a clean system? Maybe I don't understand something and I'm doing something wrong... (yandex translate)
2023-06-14 13:12:20 <Ins> But I put it with a standatd image. I'll try with extended.
2023-06-14 15:31:57 <oleovit> sry. but psykose knows what's going on regarding the switch to 0.35.1_git20230608 . the mpv dev has a patch now, but i have no clue how to do, what he described (never done that): https://github.com/mpv-player/mpv/issues/11773#issuecomment-1591428354
2023-06-14 15:33:10 <psykose> that's not a patch, it's a thing for you to test
2023-06-14 15:33:23 <oleovit> but how..
2023-06-14 15:33:27 <psykose> do you know how to use aports
2023-06-14 15:34:01 <oleovit> .. never done that.. im really sry
2023-06-14 15:34:40 <oleovit> is there a wiki entry?
2023-06-14 15:35:33 <psykose> hmm
2023-06-14 15:35:55 <oleovit> all i know, you know from my git
2023-06-14 15:36:33 <psykose> you can use mrtest from pipeline artifacts i think
2023-06-14 15:36:36 <psykose> i'll build it with that for you
2023-06-14 15:36:44 <oleovit> but i never had bigger issues, so i never worked with some enhanced things like this
2023-06-14 15:37:40 <oleovit> that whould be cool, sry, i know you have a lot of work to do
2023-06-14 15:38:25 <oleovit> maybe, another new apline thing to learn. but not in this context now. i need time
2023-06-14 15:40:24 <psykose> apk add -t .temp mrtest
2023-06-14 15:40:26 <psykose> mrtest add -a 48025
2023-06-14 15:40:30 <psykose> follow that
2023-06-14 15:40:37 <psykose> you'll get an -r1 with the patch applied
2023-06-14 15:40:45 <oleovit> one moment
2023-06-14 15:41:04 <psykose> press `u` on the 'what to install' screen
2023-06-14 15:42:52 <oleovit> your answer: y # i guess *lol*?
2023-06-14 15:44:25 <oleovit>  Otherwise, if you want to proceed, type in the
2023-06-14 15:44:33 <oleovit> MR ID ?
2023-06-14 15:44:45 <oleovit> what id?
2023-06-14 15:45:32 <oleovit> 48025 ?
2023-06-14 15:48:10 <ptrc> yes
2023-06-14 15:51:02 <oleovit> i don t want to mess anything up. which pks to install? 1) mpv-0.35.1_git20230608-r1.apk [installed: 0.35.1_git20230608-r0] only this or the other ones, like  mpv-libs-0.35.1_git20230608-r1.apk ..
2023-06-14 15:51:06 <psykose> press u
2023-06-14 15:51:08 <psykose> then y
2023-06-14 15:53:36 <oleovit> all done! (message) Use 'mrtest zap' to uninstall added packages.
2023-06-14 15:54:43 <oleovit> so, this was all the magic? and if done, how to go back to normal? with the zap command?
2023-06-14 16:00:21 <consus[m]> Hi guys
2023-06-14 16:00:29 <consus[m]> I'm trying to install alpine 3.18 on ppc64le power8 machine 
2023-06-14 16:00:35 <oleovit> indeed 0.35.1_git20230608-r1
2023-06-14 16:00:42 <consus[m]> I have a 3TB nvme disk and I'm choosing sys method
2023-06-14 16:00:55 <consus[m]> All is well until the last stage where packages are installed and I get this:
2023-06-14 16:01:02 <consus[m]> grub-install: error: the PReP partition is not empty. If you are sure you want to use it, run dd to clear it: `dd if=/dev/zero of=/dev/nvme4n1p1'.
2023-06-14 16:01:02 <consus[m]> ERROR: linux-lts-6.1.34-r0: No space left on device
2023-06-14 16:01:02 <consus[m]> ERROR: Failed to create boot/vmlinuz-lts: No space left on device
2023-06-14 16:01:35 <consus[m]> And I have no idea why, /boot is surely big enough:
2023-06-14 16:01:35 <consus[m]> /dev/nvme4n1p2          271.1M      2.5M    249.6M   1% /mnt/boot
2023-06-14 16:02:16 <consus[m]> Or not?
2023-06-14 16:02:45 <consus[m]> Hmm... seems not:
2023-06-14 16:02:45 <consus[m]> linux-lts-6.1.34-r0 installed size:
2023-06-14 16:02:45 <consus[m]> 302 MiB
2023-06-14 16:02:58 <consus[m]> Seems like the default partition size should be increased
2023-06-14 16:03:15 <nangel> uh.. the error message references nve4n1p1, and you mention nve4np2
2023-06-14 16:03:44 <nangel> nve4n1p2...
2023-06-14 16:04:28 <consus[m]> grub-install thing is just a warning
2023-06-14 16:04:29 <consus[m]> > 172.17.144.61/24
2023-06-14 16:04:30 <consus[m]> > ERROR: Failed to create boot/vmlinuz-lts: No space left on device
2023-06-14 16:04:31 <consus[m]> The main problem is here
2023-06-14 16:04:53 <consus[m]> It seems that the default /boot on ppc64le is few mbytes less than the package size
2023-06-14 16:05:24 <Ermine> /boot partition is too small I guess
2023-06-14 16:05:32 <consus[m]> Yes it is
2023-06-14 16:06:00 <consus[m]> I guess I can manually fix this, but I think it would be better to make the defaults work
2023-06-14 16:06:15 <ikke> The default size was already adjusted
2023-06-14 16:06:29 <ikke> but that would not affect already installed systems
2023-06-14 16:06:35 <consus[m]> I'm pretty sure I got the latest installation ISO
2023-06-14 16:06:41 <consus[m]> Just a few hours ago
2023-06-14 16:06:52 <ikke> It should default to 300M
2023-06-14 16:06:59 <consus[m]> The package size is 304M
2023-06-14 16:07:11 <consus[m]> *302M
2023-06-14 16:07:25 <consus[m]> linux-lts-6.1.34-r0 installed size:
2023-06-14 16:07:25 <consus[m]> 302 MiB
2023-06-14 16:09:00 <ikke> But that should not all go to /boot
2023-06-14 16:12:50 <consus[m]> > local boot_size="${BOOT_SIZE:-300}"
2023-06-14 16:12:59 <consus[m]> I have this in setup-disk
2023-06-14 16:13:05 <consus[m]> I'll try with 512
2023-06-14 16:16:06 <consus[m]> Yeah, 512M works
2023-06-14 16:16:30 <consus[m]> > /dev/nvme4n1p2          487.6M    285.2M    166.6M  63% /mnt
2023-06-14 16:17:17 <consus[m]> It seems that 300 is not in 1000 blocks, but in 1000 byte blocks
2023-06-14 16:17:21 <consus[m]> Hence the difference
2023-06-14 16:17:29 <consus[m]> *not in 1024 blocks
2023-06-14 16:18:21 <ikke> I wonder why vmlinux became so big
2023-06-14 16:18:32 <ikke> on the ppc64le server we have (still on 3.16 atm), its 40M
2023-06-14 16:18:46 <ikke> this one is 268.5M
2023-06-14 16:19:26 <consus[m]> Easy
2023-06-14 16:19:29 <consus[m]> It's not stripped
2023-06-14 16:19:46 <consus[m]> > vmlinuz-lts: ELF 64-bit LSB executable, 64-bit PowerPC or cisco 7500, OpenPOWER ELF V2 ABI, version 1 (SYSV), statically linked, BuildID[sha1]=21f2d4a74ec79d390fd43e56439b0d900e5a1084, with debug_info, not stripped
2023-06-14 16:20:53 <psykose> ncopa: ^ what do you think
2023-06-14 16:20:56 <consus[m]> > 38.8M	vmlinuz-lts
2023-06-14 16:20:58 <psykose> looks related to that debug thing that was added
2023-06-14 16:21:00 <consus[m]> After stripping
2023-06-14 16:21:21 <ikke> DEBUG_INFO_BTF
2023-06-14 16:21:25 <ikke> d97c07c67cb9452ec1f7680c5314ea53e622cea5
2023-06-14 16:21:29 <consus[m]> Most likely
2023-06-14 16:21:44 <consus[m]> It should be definitely stripped into a subpackage
2023-06-14 16:22:41 <psykose> should have a special dbg() -dbg yeah
2023-06-14 16:22:47 <psykose> if it's possible at all
2023-06-14 16:24:32 <psykose> seems from https://gitlab.alpinelinux.org/alpine/aports/-/issues/12563 the --strip-debug wasn't added
2023-06-14 16:25:03 <consus[m]> Glad to be of use :)
2023-06-14 16:42:23 <consus[m]> How long do you guys plan to support power8 btw?
2023-06-14 16:47:23 <ikke> consus[m]: There are no plans to deprecated it any time soon afaik
2023-06-14 16:47:56 <ikke> Not sure if there would be a good reason to raise the baseline to power9
2023-06-14 16:48:29 <consus[m]> Nice
2023-06-14 16:48:42 <consus[m]> Because redhat and ubuntu both nuked it lol
2023-06-14 16:48:55 <consus[m]> So my options here are alpine or ause
2023-06-14 16:49:01 <consus[m]> *suse
2023-06-14 17:01:01 <ncopa> i think the DEBUG_INFO_BTF is debug info, so --strip-debug will remove it?
2023-06-14 17:05:07 <oleovit> agaib, thank you for everything. it worked, i guess he found a solution. i also found a wiki on postmarket for mrtest. so i did mrtest zap, tihs removed the virtual pkg, but i still have policy output 0.35.1_git20230608-r1 . do i have to del mpv and add again, or even a apk upgrade -a ? 
2023-06-14 17:18:15 <psykose> latter
2023-06-14 17:18:28 <oleovit> upgrade -a did it. learned something new. thx for building it, and the step-for-step guide!
2023-06-14 17:28:50 <psykose> ncopa: dunno, arch has btf but just does `strip vmlinuz` anyway
2023-06-14 17:29:15 <psykose> and on a bunch of the extra binaries that come with -dev
2023-06-14 17:29:26 <psykose> we don't strip anything except for what mod_install_strip does
2023-06-14 17:36:05 <heat_> psykose, generally you don't strip btf - that's the point
2023-06-14 17:36:46 <psykose> how are you meant to use it without having a 50 million gb kernel then
2023-06-14 17:36:46 <heat_> and it's a custom format anyway, so unless binutils grew knowledge of linux section names, strip isn't kicking BTF away
2023-06-14 17:36:56 <psykose> right
2023-06-14 17:37:02 <psykose> so the strip is still fine to call on the kernel
2023-06-14 17:37:21 <heat_> BTF is a lot smaller than DWARF
2023-06-14 17:37:48 <psykose> for some reason that is also enabled
2023-06-14 17:42:48 <heat_> is the perf package on alpine missing some sort of optional dep? my perf report output on alpine is pretty much $PAGER'd text, while I would expect a curses-like interface
2023-06-14 17:44:54 <mbx> hey everybody. I think I forgot to document something. On a new alpine box, wget complains "wget: can't execute 'ssl_client': No such file or directory", but I already installed openssl and openssh, what am I missing again?
2023-06-14 17:46:28 <ikke> mbx: there is a package called ssl_client
2023-06-14 17:47:33 <ikke> https://pkgs.alpinelinux.org/package/edge/main/x86_64/ssl_client
2023-06-14 17:48:22 <mbx> Thanks ikke, I just found it
2023-06-14 17:48:32 <mbx> you are so fast :D
2023-06-14 17:48:50 <mbx> interestingly, nobody ever posted the exact error message on the internet
2023-06-14 17:50:12 <psykose> heat_: it's probably the slang build-time optdep
2023-06-14 17:53:07 <psykose> added
2023-06-14 18:04:47 <psykose> heat_: are the /usr/libexec/perf-core/tests/ actually used for anything or are they like actual tests
2023-06-14 18:06:36 <heat_> i don't think so, doesn't seem to be present on arch either
2023-06-14 18:07:28 <psykose> ty bb
2023-06-14 18:07:47 <heat_> <3
2023-06-14 18:08:48 <heat_> if you're looking at perf rn, there's also a fun comment at the end of perf report "# (Cannot load tips.txt file, please install perf!)"
2023-06-14 18:08:58 <heat_> i assume you're deleting the tips.txt file too?
2023-06-14 18:12:31 <psykose> hmm
2023-06-14 18:12:58 <psykose> it's in /usr/share/doc/perf-tip/tips.txt with linux-tools-doc
2023-06-14 18:13:46 <psykose> i can patch the message tho
2023-06-14 18:16:25 <psykose> added
2023-06-14 18:18:13 <msiism> So, the sources for the User Handbook are licensed CC-BY-SA 4.0. This also applies to the rendered version of that handbook, right?
2023-06-14 18:18:27 <psykose> maybe
2023-06-14 18:20:44 <msiism> Yeah, I mean, there's no indication of it being under that license at https://docs.alpinelinux.org/.
2023-06-14 18:35:30 <oleovit> sry. me again. much bigger problem. testing my postinstall in a vm, revealed to me, that only the 2 v.18 main/community repos are in /etc/apk/repositorys. no commented edge repos at all. so, everything is breaking my script. is this intended since v.18?
2023-06-14 18:37:33 <psykose> yes
2023-06-14 18:37:44 <oleovit> oh no
2023-06-14 18:38:49 <oleovit> i think this is a very bad idea.
2023-06-14 18:40:01 <oleovit> so i have to append them. had no clue. this is breaking user-scripts and co.
2023-06-14 18:41:17 <psykose> don't think there's any guarantee anywhere that there are magic commented edge repos (though they were there for a while)
2023-06-14 18:41:17 <oleovit> the whoe time since the update to v18..nothing worked and no indo about this massive change at all
2023-06-14 18:41:31 <psykose> 2 lines of text is hardly a massive change
2023-06-14 18:41:37 <psykose> it's similarly easy for you to fix
2023-06-14 18:41:53 <psykose> they were intentionally removed to not give the idea people can just mix the repos i guess
2023-06-14 18:42:35 <consus[m]> Is there an easy way to check for functional mount without modifying the openrc init.d script?
2023-06-14 18:42:48 <psykose> what's a functional mount
2023-06-14 18:42:57 <consus[m]> I wanna make sure that /var/lib/libvirt/qemu is mounted before starting libvirt
2023-06-14 18:43:10 <psykose> you would probably indeed have to modify it then
2023-06-14 18:43:11 <oleovit> yes, but one has to be informed. i really try to follow everything i cab, but this is a massive change, because it never was like that. also nothing in the wiki et cetera.
2023-06-14 18:43:15 <consus[m]> :(
2023-06-14 18:43:17 <consus[m]> Okay
2023-06-14 18:44:16 <oleovit> ahh. ok, people mix, shit happens then, but that's their fault if they dont read the wiki
2023-06-14 18:44:59 <oleovit> anyway, was a shock, will append them. thx anyway
2023-06-14 18:46:38 <psykose> the change was https://gitlab.alpinelinux.org/alpine/alpine-conf/-/merge_requests/155 specifically
2023-06-14 18:46:45 <psykose> though it's actually only in 3.18.2 now
2023-06-14 18:47:02 <psykose> but yeah i don't think anyone predicted this would then break someone's scripts that expected comments to be there
2023-06-14 18:47:09 <psykose> not sure how someone would document that
2023-06-14 18:47:13 <psykose> 'we removed a comment from a file'
2023-06-14 18:48:10 <psykose> ah, no, was in 3.18 proper
2023-06-14 18:48:38 <msiism> "Drop obsolete comment on/in $thing" is what I tend to use.
2023-06-14 18:49:16 <psykose> if there was a changelog with things like that it would be roughly ~40k lines every release
2023-06-14 18:49:20 <psykose> you would not even find such a thing in there
2023-06-14 18:49:30 <msiism> Oh, I thought more like Git log.
2023-06-14 18:49:44 <psykose> well, that's already in the git log of that repo
2023-06-14 18:49:50 <psykose> i guess a log of every log
2023-06-14 18:49:59 <msiism> I see.
2023-06-14 18:52:52 <helpless> hi everyone, today I started having problems with udev-settle, that fails to start and hogs my cpu even after boot. How can I see the logs?
2023-06-15 00:30:57 <zito> if I have an old computer that can't boot from USB and has a broken hard drive, can I use the alpine install cd to install the system onto an usb thumb drive, then use the same install cd every time to boot from that usb?
2023-06-15 01:23:56 <Saijin_Naib> I ran Alpine from USB. It mostly worked okay?
2023-06-15 01:24:09 <Saijin_Naib> Persistent on-drive install, I mean 
2023-06-15 01:32:33 <zito> Saijin_Naib: my question is if I can boot it from usb by using the install cd
2023-06-15 01:39:24 <mindlessjohnny> I'm trying to test Alpine with nginx and php but it appears that php either isn't in the repo or maybe it's not available for aarch64 but I can't seem to find the packages at pkgs.alpinelinux.org regardless of architecture... wanted to make sure I wasn't missing something before I go through the hoops to try and compile php (running aarch64 as I'm running a VM on my M2 mac)
2023-06-15 01:42:00 <psykose> did you forget to enable the /community repo
2023-06-15 01:44:12 <mindlessjohnny> prehaps, new to distro and all
2023-06-15 01:45:30 <mindlessjohnny> That did the trick, appreciate the help for such a silly issue
2023-06-15 02:00:22 <heat> psykose, btw thank you for looking at the perf issues :)
2023-06-15 02:00:37 <psykose> np :)
2023-06-15 02:00:40 <heat> although aarch64 is still stuck on 6.3.4 apparently
2023-06-15 02:00:58 <psykose> yeah that is still like
2023-06-15 02:01:00 <psykose> 2 weeks behind
2023-06-15 02:01:07 <psykose> but Soon it will upload
2023-06-15 02:01:09 <psykose> 200 builds to go
2023-06-15 02:01:49 <psykose> (network went poof for a long time on that builder)
2023-06-15 03:34:17 <Saijin_Naib> hmm... Looks like it was shadow needing a rebuild for my two machines' issues, psykose
2023-06-15 03:34:22 <Saijin_Naib> Is that what you had suspected?
2023-06-15 03:34:26 <psykose> nope
2023-06-15 03:34:29 <psykose> nothing even changed in it
2023-06-15 03:34:33 <Saijin_Naib> hah, wtf
2023-06-15 03:34:45 <Saijin_Naib> All good now... thats the only package that looked salient
2023-06-15 03:34:56 <Saijin_Naib> But like 60 updated today, so it was definitely one of the ones that went by me
2023-06-15 03:34:58 <Saijin_Naib> Ah well
2023-06-15 03:35:02 <Saijin_Naib> Thanks anyway for the help
2023-06-15 03:47:37 <psykose> ~
2023-06-15 05:15:49 <LACampbell> games are a lot slower on alpine than they were on my old glibc system. this is running through flatpak, which the old one did too, so I doubt it's a musl issue
2023-06-15 05:16:51 <aron> what kind of games?
2023-06-15 05:17:16 <LACampbell> just stuff run through steam. tf2, that kind of thing. also noticed it with minecraft
2023-06-15 05:17:27 <LACampbell> currently checking out what graphics card drivers I have
2023-06-15 05:17:32 <aron> so 3D
2023-06-15 05:17:46 <aron> what graphics card?
2023-06-15 05:18:04 <aron> (afaik, official nvidia does not work with musl)
2023-06-15 05:18:51 <LACampbell> 05:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Lucienne (rev c1)
2023-06-15 05:18:56 <LACampbell> it's amd, thank goodness
2023-06-15 05:19:23 <LACampbell> amdgpu is running
2023-06-15 05:20:42 <LACampbell> feel like I needed some other software sitting on top of it
2023-06-15 05:20:56 <LACampbell> though I guess it could be amdgpu not working well with musl? hmm
2023-06-15 05:21:21 <aron> funny thing: I put my nvi card into a vm with pci passthrough, and the overall virtual system runs faster now than it was the host system
2023-06-15 05:22:14 <aron> with less resources
2023-06-15 05:22:18 <aron> weird, huh?
2023-06-15 05:25:08 <aron> host is alpine, ofc
2023-06-15 05:29:26 <psykose> amdgpu is a kernel driver so only the kernel matters there
2023-06-15 05:30:11 <psykose> how much is 'a lot'? did the old system also use flatpak for steam or not?
2023-06-15 06:13:28 <LAC-Laptop> I've messed up my startup services. Networking complains at boot that it can't find any devices, ie "ip: ioctl 0c8193 failed: No such device". there's a nunch of them, I think it's just trying to run all the commands in /etc/network/interfaces
2023-06-15 06:13:48 <LAC-Laptop> if I log in and restart networking manually, it goes ahead and does dhcp and everything is fine
2023-06-15 06:14:22 <LAC-Laptop> even though the only services that run after networking - judging by the logs - are busbybox, chronyd, crond, polkit, rtkit, samba...
2023-06-15 06:18:38 <LAC-Laptop> ok rc-status -a is givng me a more complete picture. I think I accidentally disabled and addded back the networking service
2023-06-15 06:21:16 <LAC-Laptop> can anyone tell me where  'networking' is in there rc-status -a? mine is in both boot and default
2023-06-15 06:22:26 <psykose> generally boot
2023-06-15 06:22:35 <LAC-Laptop> only boot?
2023-06-15 06:23:26 <quinq> Mine is only in boot
2023-06-15 06:23:59 <LAC-Laptop> huh. well I'm stumped
2023-06-15 06:25:20 <psykose> it sounds like a race condition between net and the device being configured
2023-06-15 06:25:26 <psykose> removing it from boot and leaving in default probably works
2023-06-15 06:25:48 <LAC-Laptop> boot: modules hwclock seedrng loadkmap sysctl bootmisc iwd hostname networking syslog
2023-06-15 06:25:57 <LAC-Laptop> ok I'll try that
2023-06-15 06:28:34 <LACampbell> psykose: that did work!! thanks. looks like there is a race condition with that and iwd, since that's all I changed (just following the wiki)
2023-06-15 06:29:40 <psykose> something like that
2023-06-15 07:14:51 <uuidNuniq> what things init the device?
2023-06-15 07:15:17 <uuidNuniq> is it bootmisc?
2023-06-15 07:15:41 <psykose> probably iwd doing something funny i guess
2023-06-15 07:15:44 <psykose> with network config
2023-06-15 07:15:48 <psykose> but if that's not enabled then not sure
2023-06-15 07:16:53 <uuidNuniq> i guess when dev is mounted, network devices will be available?
2023-06-15 07:24:22 <LACampbell> I'm pretty ignorant of init systems - I figured there'd be some dependency graph
2023-06-15 07:26:30 <psykose> there is, in this case iwd is 'before net'
2023-06-15 07:26:40 <psykose> there's probably a specific issue that it's not actually ready
2023-06-15 07:26:47 <psykose> well
2023-06-15 08:28:54 <consus[m]> ifup talks dirty to me, why?
2023-06-15 08:29:07 <consus[m]> # ifup eth1.15
2023-06-15 08:29:07 <consus[m]> ifup: failed to change interface eth1.15 state to 'up'
2023-06-15 08:29:07 <consus[m]> ip: RTNETLINK answers: File exists
2023-06-15 08:29:27 <consus[m]> auto eth1.15... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/wZFRXGUYCZzPScMUUtAceZVh>)
2023-06-15 08:29:33 <consus[m]> vlan config looks decent
2023-06-15 08:29:44 <consus[m]> And it actually brings the interface up
2023-06-15 08:30:14 <consus[m]> But it silently cannot bring it down afterwards
2023-06-15 08:32:27 <consus[m]> Details: https://pastebin.com/fdC249Q3
2023-06-15 09:57:35 <raspbeguy> Hello, I am using LVM2 volume groups. My issue is that the logical volumes block devices in /dev/mapper are not symlinks to ../dm-* and they are owned by group root
2023-06-15 09:58:46 <psykose> are you using eudev or nah
2023-06-15 10:18:28 <consus[m]> Also there seems to be a problem with bridging. Simply adding a new bridge to interfaces and rebooting works, however ifquery -r does not seem to be aware of the interface existing.
2023-06-15 10:18:44 <consus[m]> ifup and ifdown also don't work
2023-06-15 11:43:36 <bbbhltz> Good day/evening all. Easy beginner question: Using Edge, /etc/alpine-release is "3.18.0". With v.3.18.2 out I need to run apk upgrade -a, correct?
2023-06-15 11:46:39 <ikke> bbbhltz: does /etc/alpine-release.apknew exit/
2023-06-15 11:47:14 <ikke> on edge, it's still 3.18.0
2023-06-15 11:47:15 <psykose> edge is 3.18.0 indeed
2023-06-15 11:47:46 <psykose> never understood why that is changed to a release number instead of only in release
2023-06-15 11:48:12 <psykose> edge should always be named some _something suffix
2023-06-15 11:50:44 <bbbhltz> psykose: thanks, I feel like I've already asked that question anyway
2023-06-15 11:51:13 <psykose> your telepathically linked friend must've asked
2023-06-15 12:08:45 <aron> why abuild is adding cmd:XYZ=$pkgver to 'provides' if package contains /usr/bin/XYZ/my_command?
2023-06-15 12:09:28 <psykose> so you can add commands in default path with cmd:xyz
2023-06-15 12:09:45 <aron> but XYZ is a directory
2023-06-15 12:10:03 <psykose> which one
2023-06-15 12:10:12 <aron> in my example the XYZ
2023-06-15 12:10:34 <aron> package contains: /usr/bin/XYZ/my_command
2023-06-15 12:10:37 <psykose> i mean literally which package
2023-06-15 12:10:47 <psykose> i did find the bug already though
2023-06-15 12:10:53 <aron> its a local package of mine
2023-06-15 12:11:23 <aron> s/its/it's/
2023-06-15 12:13:16 <psykose> the check is basically
2023-06-15 12:13:20 <psykose> x in bin/*
2023-06-15 12:13:27 <psykose> -x $x || continue
2023-06-15 12:13:30 <psykose> but you see
2023-06-15 12:13:33 <psykose> that is true for folders
2023-06-15 12:13:38 <aron> yeah
2023-06-15 12:13:41 <aron> it's dumb
2023-06-15 12:13:42 <psykose> needs -f too
2023-06-15 12:25:28 <aron> will be this fix backported?
2023-06-15 12:33:00 <psykose> backported?
2023-06-15 12:33:31 <aron> into v3.15, for instnace
2023-06-15 12:37:12 <psykose> no
2023-06-15 12:38:08 <raspbeguy>  a user who belongs to group disk. It's trying to deal with block devices in /dev/mapper/* so it does not have permission to do anything.
2023-06-15 12:38:08 <raspbeguy> hello, I'm using LVG volume group. Logical volumes block devices are present at 2 locations: in /dev/mapper/* and /dev/dm-*. There are also symlinked in /dev/<vgname>/* to /dev/dm-*. The block devices in  /dev/mapper/* are owned by root:root mode 600 while those in /dev/dm-* are owned by root:disk mode 660. My problem is that I want to use sanlock (called by LVM when managing locks) which is run as
2023-06-15 12:38:32 <raspbeguy> comparatively on debian, /dev/mapper/* are symlinks to /dev/dm-*
2023-06-15 12:38:40 <aron> then I'll patch is myself, I see that line what you were talking about
2023-06-15 12:38:50 <aron> *patch it
2023-06-15 12:42:04 <raspbeguy> that lead me to another question: why are some packaged udev rules deployed in /usr ? and some other in /lib/used/rules.d
2023-06-15 12:42:23 <raspbeguy> that lead me to another question: why are some packaged udev rules deployed in /usr/lib/used/rules.d and some other in /lib/used/rules.d
2023-06-15 12:45:47 <psykose> for the same reason /lib is separate from /usr/lib
2023-06-15 12:45:58 <psykose> no reason at all and it's random stuff in either place and it makes no difference
2023-06-15 12:53:44 <raspbeguy> ok
2023-06-15 12:54:20 <aron> another packaging-related question: can I attach changelog to a package, like "changelog" in arch linux? ( https://wiki.archlinux.org/title/PKGBUILD#changelog )
2023-06-15 12:55:23 <raspbeguy> so about my problem, I'm not an expert in udev and I don't know what I can do
2023-06-15 12:56:11 <psykose> aron: no
2023-06-15 12:56:27 <aron> bummer
2023-06-15 12:56:57 <psykose> raspbeguy: this sounds like some eudev issue
2023-06-15 12:57:01 <psykose> not exactly sure though
2023-06-15 12:57:48 <aron> psykose: do you have any alternative idea to achive this? maybe I should put into -doc pkg?
2023-06-15 12:57:57 <psykose> no
2023-06-15 12:58:02 <psykose> it doesn't exist, you can't do it
2023-06-15 12:58:32 <aron> :(
2023-06-15 12:59:00 <ikke> You can ofcourse in clude it, but someone has to manually look at the changelog file
2023-06-15 13:00:14 <aron> I really liked the changelog option in arch, and I'm missing it here ;)
2023-06-15 13:02:10 <raspbeguy> psykose, yes I think so too, but I cannot find where
2023-06-15 13:02:30 <raspbeguy> I checked the rules on debian and I see no significant difference
2023-06-15 13:04:09 <raspbeguy> https://termbin.com/qirl7
2023-06-15 13:04:57 <psykose> i.e. not a rules-related issue
2023-06-15 13:05:17 <raspbeguy> Oh I see
2023-06-15 13:07:31 <aron> ikke: you're right, a lot of packages put their changelog into -doc pkg, then I'll go with this
2023-06-15 13:07:40 <psykose> please don't
2023-06-15 13:07:59 <psykose> no package does this except some random stuff that autoinstalls a 'changelog' file or w/e
2023-06-15 13:11:46 <aron> well, on my current install the following have changelogs: bash, dos2unix, dosfstool, gnuplot, sshfs
2023-06-15 13:11:51 <aron> other installs probably have more :P
2023-06-15 14:23:32 <raspbeguy> it seems that eudev is deprecated now
2023-06-15 14:25:10 <raspbeguy> it seems that now gentoo stopped developping eudev and use now systemd udev which can now be packaged separately from systemd
2023-06-15 14:26:13 <aron> (it seems like a gateway drug, haha)
2023-06-15 14:27:49 <psykose> says who
2023-06-15 14:27:57 <psykose> ah
2023-06-15 14:28:00 <raspbeguy> says gentoo wiki
2023-06-15 14:28:24 <raspbeguy> but maybe I misinterpreted that
2023-06-15 14:28:33 <psykose> it doesn't say that there
2023-06-15 14:28:54 <raspbeguy> gentoo deprecated it as its mainline udev but the project is still maintained
2023-06-15 14:29:04 <raspbeguy> by some other people
2023-06-15 14:45:22 <invoked> "it's a trap!" -- admiral ackbar
2023-06-15 15:09:32 <Guest3176> I am running alpine with i3 window manager on a laptop, and I the first time I connected a monitor with hdmi, nothing happens, the external monitor just idles
2023-06-15 15:09:48 <Guest3176> xrandr shows hdmi is connected
2023-06-15 15:10:05 <Guest3176> Any packages or configuration that I likely need?
2023-06-15 15:13:13 <artok> did you go through https://wiki.archlinux.org/title/multihead ?
2023-06-15 15:15:39 <Guest3176> No, I'll check it out. Thanks
2023-06-15 15:27:19 <Guest3176> artok, the xrandr command on that page works! Now I have both the builtin monitor and the external one, but I'm not sure the best way to make the settings persist. Any thoughts?
2023-06-15 15:29:17 <Guest3176> Hmm, maybe nothing is needed. When I unplug and plug back in the monitor, it still works. Not sure about after a reboot yet. I'll try that next, I suppose
2023-06-15 15:34:31 <artok> check the autorandr also 
2023-06-15 16:06:48 <martin1> Hi, I was just asking about multiple monitors. Thanks, artok, autorandr seems cool, and the "autorandr horizontal" command does what I want during my session, but when I try to add the xrandr command or the autorandr command to .xinitrc or .profile, the outputs on the monitors are cloned instead of stacked. Maybe I am missing some environment variable or something that gets set
2023-06-15 16:06:50 <martin1> upon login?
2023-06-15 16:30:37 <artok> can't really remember how that was done
2023-06-15 18:00:49 <minimal> raspbeguy: what you are seeing is nothing to do with udev/eudev. Even when udev is enabled Alpine's initramfs' init makes use of nlplug-findfs (which makes use of mdev) to mount the rootfs
2023-06-15 18:01:40 <minimal> so long before udev is started the /dev/mapper/ entries have already been created
2023-06-15 18:01:48 <psykose> can't it remake them after
2023-06-15 18:03:07 <minimal> psykose: well if the /dev/mapper entry for the rootfs has been created and the rootfs is actually mounted via that device then can it change it whilst it is in use?
2023-06-15 18:03:22 <psykose> if it tries really hard maybe
2023-06-15 18:04:45 <minimal> mdev itself doesn't appear to have any specific rules for /dev/dm* or /dev/mapper/ devices, I once thought about adding some but suspected that such an MR would be rejected as it would represent a change in behaviour
2023-06-15 18:07:07 <minimal> basically with Alpine mdev is always used for early (initramfs) stuff regardless of whether mdev / mdevd / udevd is used later
2023-06-15 18:11:15 <khw> I'm looking for documentation of locale handling on Alpine.  There is a man page for locale(1), but that command doesn't seem to exist.  The man page for setlocale(1) mentions the GNU extensions like LC_PAPER, but those don't exist.  Are there supported locales besides C and C.UTF-8?
2023-06-15 18:11:56 <minimal> khw: musl-locales package...
2023-06-15 18:14:50 <khw> thank you
2023-06-15 18:20:23 <raspbeguy> minimal: I will read your answer as soon as my daughter is asleep
2023-06-15 18:25:38 <raspbeguy> minimal: my rootfs isn't on a LV
2023-06-15 18:25:49 <raspbeguy> It's a separate volume
2023-06-15 18:26:39 <minimal> ok. Anyway as mdev is run by initramfs then it is likely the cause of what you see
2023-06-15 18:27:18 <psykose> khw: pretty much all of it is just doing LANG=, the (1) bins don't really do anything
2023-06-15 18:29:35 <raspbeguy> minimal: ok. Do you know any workaround?
2023-06-15 18:29:41 <minimal> nope
2023-06-15 18:29:57 <raspbeguy> That's a shame
2023-06-15 18:31:04 <raspbeguy> That basically means alpine cannot use shared volume groups using sanlock
2023-06-15 18:31:16 <raspbeguy> Which is a huge deal for me
2023-06-15 18:31:16 <minimal> as I never figured out exactly where it is happening. When passing boot param "debug_init" I can see nlplug-findfs dealing with /dev/dm-*, I don't see any reference to /dev/mapper there though, unless vgchange is doing that
2023-06-15 18:32:21 <minimal> as I said earlier I guess adding a mdev rule might resolve it but then that would impact everyone so I doubt it would be accepted
2023-06-15 18:32:52 <raspbeguy> Couldn't one do it locally on it's own machine?
2023-06-15 18:35:28 <raspbeguy> Also maybe that wouldn't hurt to submit the idea anyway. This is only charging the owner group of LV from root to disk, which is already the case for non LVM block devices
2023-06-15 18:36:25 <raspbeguy> In my opinion this is a bug
2023-06-15 18:37:04 <psykose> it pretty much is yeah
2023-06-15 18:38:17 <raspbeguy> minimal: so would it bother you to try a MR with that if you know what to do, please?
2023-06-15 18:39:36 <psykose> hoo boy
2023-06-15 18:39:37 <psykose> history
2023-06-15 18:39:45 <minimal> raspbeguy: I though the "wish" was to change /dev/mapper/WHATEVER to be a softlink to /dev/dm-X
2023-06-15 18:39:57 <raspbeguy> minimal: even better
2023-06-15 18:40:05 <minimal> to match what a typical udev distro does
2023-06-15 18:40:20 <minimal> but, as I said, that would be a change in behaviour for Alpine...
2023-06-15 18:40:20 <raspbeguy> Totally
2023-06-15 18:40:49 <psykose> sounds like a nice improvement in behaviour
2023-06-15 18:41:03 <raspbeguy> minimal: every MR is a change in behaviour 
2023-06-15 18:41:23 <minimal> psykose: depends on how it would be viewed by the "right" person?
2023-06-15 18:41:35 <raspbeguy> Even psykose seems to agree with me, which is quite rare :)
2023-06-15 18:41:45 <psykose> it's pretty common i'd say
2023-06-15 18:42:04 <minimal> just thinking of my past experiences getting some stuff through...
2023-06-15 18:44:23 <minimal> I hit this issue a while ago with cloud-init where its growpart code expects /dev/mapper/* to be a softlink to /dev/dm-X
2023-06-15 18:45:18 <raspbeguy> Also we have quite some time until next release so there is time for this "change in behaviour" to be tested
2023-06-15 18:46:06 <psykose> ah the old quadrifecta
2023-06-15 18:46:28 <psykose> lots of time until release -> submit changes -> every1 ignore until near release -> whoops it's too close until next release -> release
2023-06-15 18:48:22 <raspbeguy> Still, I think (and I hope) this change should be made
2023-06-15 18:48:42 <psykose> yes
2023-06-15 18:48:57 <minimal> raspbeguy: well the first thing is to figure out where/how the /dev/mapper entries are created
2023-06-15 18:50:04 <minimal> nlplug-findfs calls "vgchange --activate ay" which activates all VGs, not just rootfs (if it is one)
2023-06-15 18:50:26 <minimal> I suspect this vgchange call may trigger the /dev/mapper/ entries but not sure
2023-06-15 18:51:09 <raspbeguy> Ok. Let me nuke my test VM and reinstall from scratch without even eudev
2023-06-15 19:21:10 <raspbeguy> minimal, so I installed a fresh VM on edge, I just installed lvm, created a VG from /dev/vdb and added a LV on it.
2023-06-15 19:21:39 <raspbeguy> just after boot there is no /dev/mapper nor /dev/dm-*
2023-06-15 19:22:49 <ikke> raspbeguy: is lvm part of the mkinitfs features?
2023-06-15 19:23:20 <raspbeguy> didn't touch it so if it isn't by default, then no
2023-06-15 19:23:54 <ikke> setup-alpine (setup-bootable) would add it if you used that
2023-06-15 19:24:48 <raspbeguy> didn't use that
2023-06-15 19:24:59 <raspbeguy> I have features="base ext4 scsi virtio"
2023-06-15 19:25:02 <ikke> That makes sure the proper modules are present
2023-06-15 19:25:08 <raspbeguy> I will add lvm to it
2023-06-15 19:25:32 <ikke> and what is your kernel cmdline?
2023-06-15 19:25:49 <ikke> Is this lv meant to be the rootfs?
2023-06-15 19:26:18 <raspbeguy> no it isn't meant to be the rootfs
2023-06-15 19:26:22 <ikke> ok
2023-06-15 19:26:59 <ikke> there is a lvm service that you need to add to boot
2023-06-15 19:27:24 <ikke> (if it's not a rootfs, then adding it to the initfs is not necessary)
2023-06-15 19:28:22 <ikke> the lvm service loads the kernel modules and scans for vgs / lvs
2023-06-15 19:32:27 <raspbeguy> I added lvm in the mkinitfs and now /dev/mapper/* and /dev/dm-* are there
2023-06-15 19:32:37 <raspbeguy> minimal, https://termbin.com/4uoa
2023-06-15 19:34:29 <ikke> raspbeguy: probably because the kernel modules are loaded then
2023-06-15 19:34:46 <raspbeguy> honestly I doubt that it's done in the initfs
2023-06-15 19:35:26 <raspbeguy> I mean, I can remove lvm from the mkinitfs and nothing will be done before that
2023-06-15 19:36:12 <ikke> it would only if it's in the modules cmdline argument
2023-06-15 19:36:13 <raspbeguy> also I didn't install eudev, and now even /dev/dm-* is owned by root:root
2023-06-15 19:37:31 <raspbeguy> I will install eudev, which will make it root:disk
2023-06-15 19:45:05 <raspbeguy> hum, I had to also remove lvm from mkinifs features for that to work
2023-06-15 19:46:49 <minimal> raspbeguy: adding "lvm" to /etc/mkinitfs/mkinitfs.conf results in the "lvm" binary being added to the initramfs, and so nlplug-findfs can call it ("vgchange")
2023-06-15 19:47:23 <minimal> obviously if you don't add the "lvm" feature then nlplug-findfs can't do the vgchange
2023-06-15 19:47:27 <raspbeguy> so not adding it enable eudev to get control on it
2023-06-15 19:47:42 <ikke> minimal: wouldn't nlplug-findfs only be usd for the root fs?
2023-06-15 19:48:18 <minimal> ikke: you mean device (potentially) containing rootfs...
2023-06-15 19:48:20 <raspbeguy> ikke, here it was used also for my non roots lvm
2023-06-15 19:50:22 <minimal> ikke: for LVM device nlplug-findfs runs "vgchange --activate ay" to activate *ALL* VGs, not just a specific one (i.e. containing rootfs)
2023-06-15 19:50:44 <raspbeguy> so there is a problem in the way initfs handles lvm volumes, but there is also definitely a problem in eudev rules too
2023-06-15 19:53:39 <minimal> raspbeguy: what eudev rule problem? I though you said it was set as root:disk when "lvm" feature was not enabled
2023-06-15 19:54:18 <raspbeguy> minimal, well /dev/dm-* are root:dev
2023-06-15 19:54:25 <raspbeguy> but not /dev/mapper/*
2023-06-15 19:54:58 <raspbeguy> and /dev/<vgname>/* are symlinks to /dev/mapper/*
2023-06-15 19:55:14 <minimal> right, that's ok then isn't it, it's a symlink
2023-06-15 19:55:36 <raspbeguy> /dev/mapper/* isn't a symlink
2023-06-15 19:55:37 <minimal> oops, sorry, is /.dev/mapper/XYZ symlinks to /dev/dm-X?
2023-06-15 19:55:50 <raspbeguy> on debian it is
2023-06-15 19:55:54 <raspbeguy> not here
2023-06-15 19:57:19 <minimal> "so there is a problem in the way initfs handles lvm volumes" - I'm not sure, if your rootfs is on a LVM device then you need to add the "lvm" feature, if your rootfs isn't on a LVM device then don't add the "lvm" feature...
2023-06-15 19:57:47 <ikke> and if you want lvm devices to be present at boot, enable the lvm service
2023-06-15 19:57:57 <raspbeguy> minimal, the initfs problem doesn't concern myself
2023-06-15 19:58:15 <raspbeguy> I mean, I'm fine not addinglvm to initfs features
2023-06-15 19:58:33 <raspbeguy> I can always enable lvm service later at boot
2023-06-15 20:01:28 <minimal> raspbeguy: ah, one thought, have you the device-mapper-udev package installed?
2023-06-15 20:01:42 <raspbeguy> yes
2023-06-15 20:02:37 <minimal> so have you checked those rules?
2023-06-15 20:03:57 <raspbeguy> not really an expert on udev
2023-06-15 20:04:11 <raspbeguy> let me try again
2023-06-15 20:04:13 <minimal> well those are where the various LVM / DM stuff should be handled
2023-06-15 20:13:04 <raspbeguy> minimal, here is a diff between dm.rules on alpine and on debian
2023-06-15 20:13:07 <raspbeguy> https://termbin.com/sit7
2023-06-15 20:13:15 <raspbeguy> left is alpine, right is debian
2023-06-15 20:13:46 <ikke> 
ACTION hands raspbeguy a unified diff :P

2023-06-15 20:14:42 <artok> duh.. headless raspberry pi installation seems to be next one I might know how to do
2023-06-15 20:15:14 <raspbeguy> https://termbin.com/qs3mp better ?
2023-06-15 20:17:49 <ikke> raspbeguy: that's readable for me :)
2023-06-15 20:21:53 <minimal> I don't see those lines in the lvm2 source
2023-06-15 20:22:20 <minimal> but also don't see them in the Alpine packaging
2023-06-15 20:23:19 <raspbeguy> minimal, I guess this file is added by device-mapper-udev
2023-06-15 20:24:16 <minimal> eh? device-mapper-udev is a *sub* package of lvm2, I'm looking at that (origin) of that file is the lvm2 source - udev/10-dm.rules.in
2023-06-15 20:24:38 <raspbeguy> oh ok
2023-06-15 20:25:00 <raspbeguy> hum
2023-06-15 20:32:34 <raspbeguy> Anyway I also am very surprised that the block device is present twice (/dev/dm-* and /dev/mapper/*) without symlink
2023-06-15 20:32:56 <raspbeguy> that seems like a dangerous behaviour
2023-06-15 20:33:27 <ikke> how so?
2023-06-15 20:37:29 <raspbeguy> dunno, if there is a filesystem on a LV, you could mount it with /dev/mapper/vgname-lvname and with /dev/dm-x, so is that filesystem isn't made for multiple mounts then you're basically screwing ot
2023-06-15 20:47:25 <raspbeguy> ok, so it's hypothesis time
2023-06-15 20:49:26 <XFaCiEer> <raspbeguy> ok, so it's hypothesis time
2023-06-15 20:49:42 <XFaCiEer> It is possible    to undo coffee peculation
2023-06-15 20:50:37 <raspbeguy> my guess is that somehow the device nodes are created twice instead of once: somehow a device node is also created on /dev/mapper/* after /dev/dm-*. that leads udev to overwrite the correct symlink on /dev/<vgname>/* that was originaly pointing to /dev/dm-* and then to /dev/mapper/*
2023-06-15 20:51:12 <raspbeguy> /dev/mapper/* should definitely be a symlink and i don't know why here it's a full featured device node
2023-06-15 21:03:54 <artok> did someone have apkovl for deploying headless rpi that will do setup-alpine with answer file ?
2023-06-15 21:04:39 <artok> I've seen just this headless.apkovl.tar.gz project that will enable ssh and by that it is still manual things
2023-06-15 21:13:59 <raspbeguy> I guess my only workaround is to run sanlock as root... I dislike that
2023-06-15 22:27:23 <minimal> artok: there is work-in-progress to provide automated installation for alpine
2023-06-16 03:09:25 <Saijin_Naib> Can I e2fsck an LVM volume from Alpine Single-User? 
2023-06-16 03:47:46 <uuidNuniq> Saijin_Naib: why not try it yourself. as long as the drive is mounted RW, you can't 
2023-06-16 04:02:58 <Saijin_Naib> I mean, I did try, but given it's a busybox shell, I don't know the full build and compile flags, I'm on LVM not just a vanilla partition, GPT under UEFI...
2023-06-16 04:03:42 <Saijin_Naib> Looking to hear if other folks have had success, or if I should rather make bootable media and just not even bother with it
2023-06-16 04:05:28 <uuidNuniq> i does not matter single/whatever mode. as long as your whatever drive is mounted `rw` you can't do e2fsck
2023-06-16 04:08:12 <uuidNuniq> man e2fsck | awk '/mounted/{print}'
2023-06-16 07:42:45 <raspbeguy> hello, regarding our chat about LVM device nodes yesterday, I opened an issue https://gitlab.alpinelinux.org/alpine/aports/-/issues/15029
2023-06-16 09:02:33 <LACampbell> having flaky input device behaviour. sometimes the mouse will no longer send xevents by moving, but the buttons still register. hibernating then resuming makes it work again, which suggests it's a software issue. I have a udev service running, do I need anything else?
2023-06-16 10:17:51 <oleovit> hi. short quest regarding intels software guard protection, kernel message, eudev (rule). can i ignore this dmesg output: https://0x0.st/HTKx.txt
2023-06-16 12:57:08 <minimal> Saijin_Naib: you don't e2fsck a LVM volume, you e2fsck a ext2/3/4 filesystem. Now if that ext2/3/4 filesystem is inside a LVM LV then you would need to have the LV available before you try and fsck the filesystem inside it
2023-06-16 12:58:27 <minimal> so from single user mode ("rescue mode") is the LV already available? i.e. does /dev/mapper/<LV name> exist?
2023-06-16 15:28:43 <Saijin_Naib> @minimal: Yeah I can see /dev/vg0/lv_root
2023-06-16 16:06:27 <ikke> l3t me in
2023-06-16 16:27:16 <minimal> Saijin_Naib: so then, assuming it's a ext2/3/4 fs, then you should be able to fsck that. Is there any particular reason you want to do so though? Normally it would be done, if necessary, during boot
2023-06-16 16:36:12 <minimal> raspbeguy: I still think it is a mdev-related issue as /dev/dm-X and /dev/mapper/X entries are for *device mapper*, they're used by both LVM and LUKS encryption and so disabling "lvm" from the mkinitfs.conf only prevents nlplug-findfs running "vgchange" but the device mapper entries are still created by mdev (run by nlplug-findfs)
2023-06-16 16:37:33 <raspbeguy> At boot the device nodes are not created
2023-06-16 16:37:57 <raspbeguy> I have to call LVM with lvs for instance to create them
2023-06-16 16:38:39 <raspbeguy> I do not activate volumes at boot
2023-06-16 16:41:21 <minimal> yeah was checking mkinitfs source, the dm-mod.ko isn't added to the initramfs unless either "lvm" or "cryptsetup" is added to features, I thought it was being added always
2023-06-16 17:40:19 <Saijin_Naib> <minimal> "Saijin_Naib: so then, assuming..." <- Laptop lost power during usage and was acting weird, error looked like commonly FS corruption, which matches my experience with ext4 on laptops, so I wanted to try and fsck it from recovery without having to make media on another machine.
2023-06-16 17:40:45 <Saijin_Naib> Clean installed it again
2023-06-16 17:40:54 <Saijin_Naib> Easier than messing around with it 
2023-06-16 17:41:09 <Saijin_Naib> Does Alpine/APK have anything like debsums? 
2023-06-16 17:41:32 <Saijin_Naib> Or dism/sfc, if you're familiar with that world 
2023-06-16 17:43:20 <mercenary> apk audit ?
2023-06-16 17:44:28 <Saijin_Naib> Hmm 🤔 
2023-06-16 17:45:02 <Saijin_Naib> I'll need to investigate if that does it and gives me actionable results to repair packages that are no longer as-shipped
2023-06-16 17:47:30 <ikke> apk audit --system
2023-06-16 17:53:38 <minimal> Saijin_Naib: why fsck it from recovery? didn't the next boot after the powerloss fsck it automatically and it was uncleanly unmounted?
2023-06-16 17:54:11 <minimal> s/and /as /
2023-06-16 17:55:17 <Saijin_Naib> No, it didn't seem to have automatically checked, and my tune2fs settings were also being ignored 
2023-06-16 17:56:03 <minimal> Saijin: the "dmesg" output should have shown it detected it needed fsck
2023-06-16 17:56:33 <Saijin_Naib> <ikke> "apk audit --system" <- Oh, thank you for that! Lemme investigate that
2023-06-16 17:56:39 <minimal> this is a "Sys mode" install, right?
2023-06-16 17:57:01 <Saijin_Naib> Yeah, sys mode. Nothing too exotic except for being LVM, I guess
2023-06-16 17:57:26 <Saijin_Naib> Second time it got trashed this way during a power loss
2023-06-16 17:57:31 <minimal> I don't see why it would not have been automatically fsck'ed if needed
2023-06-16 17:58:19 <Saijin_Naib> 🤷 It wasn't. Booting to MX LiveUSB showed errors when scanned
2023-06-16 17:58:56 <Saijin_Naib> This happened last time, too. I think maybe because the power loss is so unexpected? My battery is worn and mis-reports remaining life and can just die under load
2023-06-16 17:59:36 <minimal> Saijin_Naib: you *do* have the fsck init.d service enabled, right?
2023-06-16 18:00:00 <Saijin_Naib> Does it come automatically with setup-desktop and setup-Alpine? 
2023-06-16 18:00:47 <minimal> ah, it's "need"ed by the root init.d service - does "rc-status -a" show it as started?
2023-06-16 18:03:53 <minimal> Saijin_Naib: ^^^
2023-06-16 18:04:05 <Saijin_Naib> On this install, yeah, under dynamic run level wanted/needed
2023-06-16 18:04:12 <Saijin_Naib> Sorry, had to go get the computer 
2023-06-16 18:04:47 <Saijin_Naib> Which involves making sure a baby can't get to trouble in the five seconds they're out of view 🤣
2023-06-16 18:05:12 <Saijin_Naib> Challenge level impossible 
2023-06-16 18:05:52 <minimal> and does the /etc/fstab have an entry for the rootfs with its 6th field being "1" ?
2023-06-16 18:13:15 <Saijin_Naib> Yep! 
2023-06-16 18:13:35 <Saijin_Naib> Strange, haha
2023-06-16 18:14:25 <minimal> so everything points to fsck being automatically run during boot if needed
2023-06-16 18:16:05 <Saijin_Naib> Yeah, but it appears that it did not run, or if it did, not aggressive enough, since MX LiveUSB showed the volume as having errors, and shadow failing to modify my groups due to being unable to access the file 
2023-06-16 18:16:10 <minimal> set rc_logger to "YES" in /etc/rc.conf and then you'll have a /var/log/rc.log file to show you the init scripts run on every boot - you should then see it logging "Checking local filesystems"
2023-06-16 18:16:40 <minimal> Saijin_Naib: the *volume* as having errors? or the ext2/3/4 *filesystem* as having errors?
2023-06-16 18:16:42 <Saijin_Naib> Would tune2fs interfere with that autofsck? 
2023-06-16 18:17:06 <Saijin_Naib> FS, I believe. Found and fixed corruption, and said the superblock didn't match 
2023-06-16 18:18:13 <minimal> depends on what you may have set/changed with tune2fs
2023-06-16 18:18:23 <minimal> also which version of Alpine are you using?
2023-06-16 18:21:45 <Saijin_Naib> Nominally 3.18.2 x86_64, but edge, really 
2023-06-16 18:22:14 <minimal> was this installed *originally* using 3.18.x? or an earlier version?
2023-06-16 18:22:23 <Saijin_Naib> tune2fs is just set to -i 1w -c 10 
2023-06-16 18:23:25 <Saijin_Naib> Current install is from 3.18.x media, prior would have been late 3.17
2023-06-16 18:24:10 <minimal> are you saying it was *installed* using 3.18.0, or upgraded from a install of 3.17.x?
2023-06-16 18:24:38 <minimal> I'm asking what version of Alpine was *originally* used to install the currently running system (regardless of any upgrades)
2023-06-16 18:26:02 <Saijin_Naib> Current is from 3.18.2 media 
2023-06-16 18:26:11 <minimal> that likely explains things then
2023-06-16 18:26:36 <Saijin_Naib> ? 
2023-06-16 18:26:55 <Saijin_Naib> Improved defaults on current installer? 
2023-06-16 18:27:09 <minimal> Alpine 3.18.x & Edge ship with e2fsprogs 1.47.x, which contains the mke2fs tool
2023-06-16 18:27:59 <minimal> the 1.47.x mke2fs formats the fs with options that older version of e2fsprogs tools (such as fsck!) DO NO UNDERSTAND
2023-06-16 18:28:29 <minimal> so it is likely that the version of fsck in MX LiveUSB is a version prior to 1.47.x
2023-06-16 18:28:42 <uuidNuniq> that's the feature orphane_file
2023-06-16 18:29:12 <uuidNuniq> orphan_file
2023-06-16 18:29:15 <minimal> and the fsck errors that gives are due to it not understanding some of the options on the 1.47.x-created rootfs
2023-06-16 18:29:46 <minimal> whereas the Alpine install itself has a (1.47.x) version of fsck that does understand those options and so doesn't report any problems
2023-06-16 18:29:50 <uuidNuniq> if this option enabled, then fsck.ext4 1.46.x can't do fsck on it 
2023-06-16 18:30:44 <uuidNuniq> Saijin_Naib: tune2fs -l /dev/sdaX | grep 'Filesystem features' 
2023-06-16 18:30:48 <minimal> I raised this issue prior to the Alpine 3.18.0 release but nothing seems to have made it into the 3.18.0 Release Notes regarding this (I'm sure I added a note to the draft Release Notes in the wiki)
2023-06-16 18:33:06 <Saijin_Naib> Ahh, okay. Fun fun. I'll need to dig and see what MX LiveUSB ships, but it did successfully fsck, which it should not touch if it can't understand a FS feature, right? Just like a kernel won't mount such, right? 
2023-06-16 18:34:07 <minimal> AFAIK yes
2023-06-16 18:34:41 <Saijin_Naib> I love the problems I make for myself by just existing 🤣
2023-06-16 18:35:10 <Saijin_Naib> That's it, I'm going back to JFS2 😠
2023-06-16 18:35:16 <Saijin_Naib> JK
2023-06-16 18:35:34 <minimal> Saijin_Naib: the note I added to the draft Release Notes: https://wiki.alpinelinux.org/wiki/Release_Notes_for_Alpine_3.18.0#e2fsprogs_1.47.0
2023-06-16 18:35:51 <minimal> read the last sentence of it...
2023-06-16 18:36:26 <Saijin_Naib> Yep 😬
2023-06-16 18:37:11 <uuidNuniq> Saijin_Naib: i do btrfs 
2023-06-16 18:37:41 <Saijin_Naib> I need to learn more about it, especially in regards to using it via setup-alpine
2023-06-16 18:37:57 <uuidNuniq> i hate this ext4 change. ext4 should be stable. 
2023-06-16 18:38:34 <uuidNuniq> 1.46.x can't fsck a partition formatted by 1.47.x that's just insane
2023-06-16 18:38:36 <Saijin_Naib> Aren't these FS features important for longterm FS stability and performance? 
2023-06-16 18:38:44 <minimal> the same issue affects Grub in general (but not Alpine's Grub in 3.18.x/Edge) so anyone doing multi-distro-boot using Grub from another distro might have problems booting Alpine 3.18.x/Edge also
2023-06-16 18:39:16 <uuidNuniq> Saijin_Naib: btrfs is the same as ext4 , just mkfs.btrfs and mount it..,
2023-06-16 18:39:20 <minimal> don't understand your question
2023-06-16 18:39:40 <uuidNuniq> the advanced features are not a must 
2023-06-16 18:39:55 <minimal> how can a new feature be important for longterm FS stability?
2023-06-16 18:40:16 <minimal> that fact that is new affects stability lol
2023-06-16 18:42:11 <uuidNuniq> the rule would be : don't use a foreign distro to install/bootstrap another distro
2023-06-16 19:54:13 <somercet1> how do I purge conf files from an old package?
2023-06-16 20:04:15 <somercet1> and by old, I mean, uninstalled a while ago.
2023-06-16 20:56:23 <frag> if i disable wifi/bt/fingerprint in thinkpad bios, it wont boot, seemingly not finding boot medium :S
2023-06-16 21:01:28 <frag> also the 2.5 slot in my thinkpad seem to have a dummy placeholder and no possibility for connecting a hd there (it has m.2 sdd slot another place tho)
2023-06-16 21:02:44 <lopid> you can force it with a hammer and chisel
2023-06-16 21:03:30 <Habbie> frag, but you can re-enable it?
2023-06-16 21:03:34 <Habbie> frag, luckier than most!
2023-06-16 21:03:51 <Habbie> frag, reference: https://nadim.computer/posts/2018-10-26-thinkpad.html
2023-06-16 21:05:25 <frag> Habbie: yes i could, is it that easy to brick it?? :o
2023-06-16 21:05:42 <Habbie> all i know is it happened to one friend
2023-06-16 21:05:51 <Habbie> and he fixed it too, with a lot of effort
2023-06-16 21:07:45 <zcrayfish> 
ACTION wonders if the laptop in question will UEFI stub boot.

2023-06-16 21:21:35 <frag> rpi boot seems cleaner in comparison imo
2023-06-16 21:21:41 <frag> *bios
2023-06-16 21:25:57 <Habbie> boot was the better word, i think
2023-06-16 21:26:12 <frag> hmm when i disable the "wifi wan" in bios is when it doesnt boot ...
2023-06-16 21:26:26 <Habbie> not a lot of legacy, no thunderbolt, many things are simpler for the pi
2023-06-16 21:26:37 <Habbie> (i'm not trying to excuse lenovo one bit, to be clear)
2023-06-16 21:31:58 <frag> simpler is nice :]
2023-06-16 21:32:25 <Habbie> yes
2023-06-16 22:34:56 <minimal> rasbeguy: looking at a strace of "lvcreate" I can see that lvcreate does a "mknod /dev/mapper/vg0-<name>" followed by a "chown root:root /dev/mapper/vg0-<name>" when a LV is created
2023-06-16 22:35:52 <minimal> raspbeguy: ^^^
2023-06-16 22:36:22 <minimal> which explains how /dev/mapper/ entries are not symlinks
2023-06-16 22:36:57 <raspbeguy> But why isn't that the case on other distros ?
2023-06-16 22:37:35 <minimal> something may be "fixing" them afterwards (udev?) or else the way lvm tools are built is having an impact perhaps
2023-06-16 22:37:57 <minimal> haven't yet rebooted this VM to see how the devices are created/recreated upon boot
2023-06-16 22:45:34 <minimal> raspbeguy: a comment in part of LVM code just before a "mknod" call says "# The node may already have been created by udev. So ignore EEXIST."
2023-06-16 22:52:42 <minimal> raspbeguy: BTW this is on a VM *without* the device-mapper-udev package installed as I wanted to see the default lvm behaviour
2023-06-16 23:55:13 <minimal> raspbeguy: interestingly when booting after installing device-mapper-udev package udevd is erroring in "dmesg" output regarding it running "vgchange" (I have NOT enabled the /etc/init.d/lvm service yet) as the dm-mod module is not yet loaded, adding this module to /etc/modules doesn't help as it is loaded *after* udev tries to do the vgchange
2023-06-17 00:13:38 <consus[m]> Which of the firewalls are officially recommended by Alpine: nftables or iptables?
2023-06-17 00:17:31 <psykose> former
2023-06-17 00:18:09 <zcrayfish> iptables is obsolete; nftables is its successor 
2023-06-17 00:23:13 <minimal> minimal: ah, that's due to a packaging problem with lvm2 - the source file udev/69-dm-lvm.rules.in has a rule using the "SYSTEMDRUN" variable and when lvm2 APKBUILD runs "configure" with "--without-systemd-run" then this templated variable is replaced with "no" and so that rule runs the command "no --no-block --property DefaultDependencies=no --unit lvm-activate....." which obviously fails
2023-06-17 00:23:24 <minimal> raspbeguy: ^^^
2023-06-17 00:23:37 <minimal> so that last rule needs to be rewritten for Alpine/OpenRC
2023-06-17 00:26:43 <psykose> minimal: how's https://img.ayaya.dev/lRsqsYcVYvni
2023-06-17 00:27:56 <consus[m]> Okay, so we're currently lacking CONFIG_NF_TABLES_BRIDGE
2023-06-17 00:28:41 <psykose> it's just not set on ppc64le it seems
2023-06-17 00:28:47 <consus[m]> Oh
2023-06-17 00:28:50 <psykose> ah
2023-06-17 00:28:51 <psykose> it is
2023-06-17 00:28:58 <consus[m]> huh?
2023-06-17 00:29:17 <consus[m]> # CONFIG_NF_TABLES_BRIDGE is not set
2023-06-17 00:29:38 <psykose> missing in 3.18
2023-06-17 00:30:15 <consus[m]> I even have a ticket: https://gitlab.alpinelinux.org/alpine/aports/-/issues/15033
2023-06-17 00:30:17 <consus[m]> Just in case
2023-06-17 00:30:55 <psykose> no, even edge too, can't read
2023-06-17 00:30:56 <psykose> yeah sure
2023-06-17 00:30:59 <psykose> anything else missing for now
2023-06-17 00:32:35 <consus[m]> Not sure, currently only stumbling upon this, everything else seem to work
2023-06-17 00:32:39 <minimal> psykose: I guess that would do it - it's later here and I'm off to get some sleep, I'll test that change tomorrow
2023-06-17 00:32:57 <psykose> i'll just commit it since otherwise well, obviously just runs `no` and fails
2023-06-17 00:33:02 <psykose> but it might need more tweaking
2023-06-17 00:33:13 <minimal> it's kinda weird they have a "--without-systemd-run" configure option that doesn't seem to handle things correctly
2023-06-17 00:34:04 <minimal> haven't yet checked upstream (or other non-systemd distros) to see if there are any issues for this
2023-06-17 00:34:32 <psykose> it's mostly correct
2023-06-17 00:34:42 <psykose> in autotools --without-x is identical to --with-x=no
2023-06-17 00:34:45 <psykose> and sets it to that
2023-06-17 00:34:56 <psykose> then they just substitute it in that script instead of actually doing something else
2023-06-17 00:35:08 <psykose> but inside autoconf that's how you use it (x$x =xno or whatever)
2023-06-17 00:35:09 <minimal> well I mean it should replace SYSTEMDRUN *plus* the systemd-run options all with something valid lol
2023-06-17 00:35:16 <psykose> ye
2023-06-17 00:35:25 <psykose> it's obviously wrong, just not unexpected
2023-06-17 00:35:37 <minimal> anyway I'll spend more time tomorrow testing further
2023-06-17 00:35:48 <psykose> ugh
2023-06-17 00:35:49 <psykose> https://gitweb.gentoo.org/repo/gentoo.git/tree/sys-fs/lvm2/files/lvm2-2.03.20-dm_lvm_rules_no_systemd.patch
2023-06-17 00:35:51 <psykose> gentoo has the same fix
2023-06-17 00:35:52 <psykose> xD
2023-06-17 00:35:55 <psykose> so yeah it works fine
2023-06-17 00:35:59 <minimal> ah cool
2023-06-17 00:36:07 <minimal> anyway time for sleep
2023-06-17 00:36:10 <psykose> sleep well dear
2023-06-17 00:36:26 <minimal> I'm cheap at half the price lol
2023-06-17 01:49:26 <idkrn[m]> I haven't used Alpine in a while, do I need elogind and polkit-elogind to allow an unprivileged user to reboot?
2023-06-17 01:49:57 <psykose> or a sudo/doas rule that lets anyone run any command
2023-06-17 02:38:42 <idkrn[m]> psykose: ok, I've gotten used to only needing the polkit itself
2023-06-17 05:27:25 <idkrn[m]> <idkrn[m]> "I haven't used Alpine in a while..." <- Not sure what rule to edit lol
2023-06-17 05:53:49 <idkrn[m]> Probably forgot seatd
2023-06-17 06:55:30 <quinq> idkrn[m], could be something like: permit nopass $username as root cmd poweroff
2023-06-17 07:11:48 <raspbeguy> minimal, psykose: wow that's great ! Thank you
2023-06-17 07:11:53 <psykose> does it work
2023-06-17 07:26:51 <mercenary> always specify as /sbin/poweroff, otherwise the user can copy a shell to 'poweroff' and include it in the path
2023-06-17 07:29:27 <psykose> not without keepenv
2023-06-17 07:30:32 <psykose> default paths are the system ones
2023-06-17 07:36:32 <uuidNuniq> no, does.conf must use full path
2023-06-17 07:36:45 <uuidNuniq> or it won't run without password
2023-06-17 07:38:04 <quinq> mercenary, uniqNuuid, you can check for yourself before stating those :)
2023-06-17 07:38:54 <uuidNuniq> oh, i mean if you set '...nopass .... /sbin/poweroff' , then when you run `doas poweroff` it won't work
2023-06-17 07:39:20 <psykose> and if you nopass poweroff then `doas poweroff` will work
2023-06-17 07:39:21 <psykose> magic
2023-06-17 07:39:31 <uuidNuniq> you have to run doas /sbin/poweroff
2023-06-17 07:39:33 <quinq> Ah, then sure, hence the "poweroff" instead of "/sbin/poweroff"
2023-06-17 07:41:31 <uuidNuniq> well, i played with doas before. but now i don't use doas or sudo
2023-06-17 07:41:49 <uuidNuniq> i just su
2023-06-17 07:42:43 <uuidNuniq> this is due to i need to maintain only one piece of dotfiles for root and myself
2023-06-17 07:43:49 <uuidNuniq> do you guys set `permitrootlogin no` on sshd? 
2023-06-17 07:44:11 <uuidNuniq> i found that there is a security risk if set that to non-no
2023-06-17 07:44:27 <quinq> You should request a CVE then
2023-06-17 07:45:39 <uuidNuniq> in the case of root and normal user uses the same ssh pub, same ssh private keys, this is common right? then when any script/program the user runs can do "ssh root@localhost command"
2023-06-17 07:46:36 <uuidNuniq> you guys get my idea?
2023-06-17 07:46:42 <ptrc> quinq: i'm afraid self-inflicted damage doesn't count when it comes to CVEs
2023-06-17 07:46:52 <psykose> news at 11: if you open the door on your computer people can access it
2023-06-17 07:47:06 <uuidNuniq> especially when `curl | sh`
2023-06-17 07:47:33 <psykose> security reports go to openssh@openssh.com
2023-06-17 07:47:37 <psykose> feel free to send email
2023-06-17 07:48:06 <quinq> uniqNuuid, no that's not common use, an ssh key is private to a user
2023-06-17 07:48:33 <quinq> But yes, if you allow access to root to an ssh key, then the user owning that key can connect to root
2023-06-17 07:48:43 <quinq> That's not a security issue, that's your decision
2023-06-17 07:50:13 <quinq> Note that there might be an easier way to work around that than setting 'permitrootlogin no', that would be to not add the key in the first place
2023-06-17 07:51:20 <uuidNuniq> not put the private key in user's ~/.ssh, right. 
2023-06-17 07:52:18 <uuidNuniq> or use a different ssh pub for root
2023-06-17 07:52:35 <quinq> No, that's about the public key in .ssh/authorized_keys
2023-06-17 07:53:54 <uuidNuniq> not adding the ssh pub for root? then one must enable passwordlogin
2023-06-17 07:54:15 <uuidNuniq> to be able to ssh root@ip
2023-06-17 07:56:07 <uuidNuniq> using a seperate ssh key pair for root plus not putting that root ssh private keys on user's disk
2023-06-17 07:59:46 <uuidNuniq> i decide to open the door psykose for convinience 
2023-06-17 08:00:05 <psykose> >do thing convenience
2023-06-17 08:00:13 <psykose> >me own convenience is a security vuln
2023-06-17 08:00:30 <psykose> abby: how could you do this
2023-06-17 08:00:52 <uuidNuniq> why do i need sudo or doas if i can ctrl+b c and ssh root@localhost to get a shell 
2023-06-17 08:29:05 <uuidNuniq> now i understand why some people use a dedicated hardware for ssh and ssh only. 
2023-06-17 08:29:31 <quinq> uuidNuniq, well that's for you to answer that question, you are complaining about the issue :)
2023-06-17 08:32:09 <uuidNuniq> first, they ssh into that ssh machine, but with ssh passphrase enabled. and that machine has accesss to any of other machine with no passphrase. that's should be secure. 
2023-06-17 08:33:59 <uuidNuniq> or use ssh jump to combine that into onestep?
2023-06-17 08:35:14 <quinq> afaik you can't control if a key was protected by a password or not, and jumping doesn't solve your issue that giving access to root is giving access to root
2023-06-17 08:35:45 <mercenary> that halfway defeats the point of having a jump-box. Generally the jumpbox is locked down tight, and user cannot install anything on it
2023-06-17 08:52:31 <mbx> good morning. This morning I continued the never ending journey with Alpine on T2, the MacBook Pro 2019 16" model I own. I remember that I had trouble installing Alpine on a partition, but now I decided to install Alpine on a 2TB external SSD, because my next goal is to compile a kernel that supports more than just the MacBook and I want to move the OS with me via plug and play.
2023-06-17 08:52:34 <mbx> Like an Alpine for my pocket.
2023-06-17 08:54:13 <uuidNuniq> mbx: well, then you may need to enable all the hooks
2023-06-17 08:54:23 <mbx> I installed Alpine on such drive with "crypt", "lvm" and "sys", but when I attempt to boot it, I get some Hardware Errors (which is obvious), but I'm stuck at "Loading hardware drivers ...". Confused, because the x86_64 standard image booted without problems.
2023-06-17 08:54:24 <uuidNuniq> for initramfs
2023-06-17 08:54:58 <mbx> uuidNuniq: I installed Alpine via setup-alpine, is this still necessary?
2023-06-17 08:55:28 <uuidNuniq> if you want a universal pocket alpine... then yes. 
2023-06-17 08:55:59 <uuidNuniq> setup-alpine detects and just enable the right amount of hooks to the machine 
2023-06-17 08:57:19 <mbx> I don't have to do that prior installation, right? I plan to compile the kernel anyways, to get hardware support for Mac, so then I can enable all hooks during compilation, right?
2023-06-17 08:57:19 <uuidNuniq> for hooks i mean all the /etc/mkinitfs/features.d set in /etc/mkinitfs/mkinitfs.conf
2023-06-17 08:57:31 <mbx> ah, I thought kernel-hooks
2023-06-17 08:58:31 <mbx> uuidNuniq: is there a sample that enables everything?
2023-06-17 08:58:33 <uuidNuniq> oh, build kernel, then that's another thing, you can make the option you want bulitin... then you don't really need most of the features in features.d
2023-06-17 08:59:08 <uuidNuniq> mbx: ls -1 /etc/mkinitfs/features.d 
2023-06-17 08:59:15 <mbx> I feel I should try your route first, though, because as of now, Alpine isn't booting at all
2023-06-17 08:59:57 <uuidNuniq> mbx: turn on openrc debug and read the saved log
2023-06-17 08:59:59 <mbx> Ok, it's time to setup Alpine Linux in Parallels, so I can explore it without reboot and mounting ^^
2023-06-17 09:00:59 <mbx> uuidNuniq: Thanks for the hint, I will do that, not sure how far I come today, but very helpful so far, thanks. :)
2023-06-17 09:01:18 <uuidNuniq> mbx: boot with ISO, mouunt the drive and edit /etc/rc.conf turn on openrc log, boot with hard drive wait until you stuck . and reboot with ISO and view the log in /var/log/openrc.log
2023-06-17 09:02:05 <uuidNuniq> **view the log in /var/log/openrc.log on the hard drive**
2023-06-17 09:03:31 <uuidNuniq> as you said it stucks at Loading hardware drivers, that's a message output by openrc. 
2023-06-17 09:05:12 <uuidNuniq> also you may want to view /var/log/dmesg.log on the hard drive. mbx
2023-06-17 09:05:49 <uuidNuniq> note that these 2 log files are reset everytime you boot with the hard drive. 
2023-06-17 09:06:14 <uuidNuniq> they are not incremental 
2023-06-17 09:07:32 <mbx> uuidNuniq: Thank you so much, will do that now and see what I get, need to wrap up some info about mounting LVM LUKS2 drives before I reboot, tho
2023-06-17 09:14:38 <maakari[m]> If I create a succesful install inside a kvm, is there a documented way to transfer that into a real install?
2023-06-17 09:15:25 <uuidNuniq> maakari[m]: rsync
2023-06-17 09:16:39 <uuidNuniq> maakari[m]: rsync -ahHAX -S --numeric-ids --info=progress2 --partial --delete --exclude={"/dev/*","/proc/*","/sys/*","/tmp/*","/run/*","/mnt/*","/lost+found/*"} / "$mountdir"
2023-06-17 09:16:45 <mbx> uuidNuniq: One more question in regards of /etc/mkinitfs/features.d, once my system boots, I can simply edit it and run `mkinitfs -c /etc/mkinitfs/mkinitfs.conf -b /`, right?
2023-06-17 09:17:06 <uuidNuniq> maakari[m]: and you need to edit grub.cfg and fstab at least after rsync 
2023-06-17 09:17:29 <uuidNuniq> change the uuids to the ones for your disk. 
2023-06-17 09:18:34 <uuidNuniq> mbx: yes, if you meant to regenerate the initramfs
2023-06-17 09:18:58 <maakari[m]> uuidNuniq: thanks. So does that mean I need to create the esp by myself?
2023-06-17 09:19:33 <uuidNuniq> maakari[m]: yes, you need to create fs, format the partition on the target drive.
2023-06-17 09:20:21 <uuidNuniq> you can create same or different partition table and fs, as long as you edit the fstab and grub.cfg after rsync 
2023-06-17 09:20:23 <mbx> exactly, great
2023-06-17 09:21:05 <uuidNuniq> and if you create a different fs, you may need to regenerate the initranfs maakari[m] 
2023-06-17 09:23:06 <maakari[m]> uniqNuuid: So, I mount the intended storage for real hardware into kvm, create esp and fs, and then rsync + gen-fstab + update-grub?
2023-06-17 09:23:07 <uuidNuniq> maakari[m]: safer steps : 1) partition, format, mount 2) rsync 3) chroot into the drive, and edit fstab, grub-mkconfig, mkinitfs 
2023-06-17 09:23:36 <maakari[m]> uuidNuniq: Thanks a lot!
2023-06-17 09:26:28 <uuidNuniq> rsync is able to clone any linux to any vm/vps/real hardware with any partition layout any filesystem as long as you take care of the fstab, grub, initramfs afterwards maakari[m] 
2023-06-17 09:31:10 <uuidNuniq> mbx: how did you install alpine on the hard drive? the LOADING HARDWARE is a fake message. last time i encounter this same issue was because i did not install e2fsprogs and fsck.ext4 can't proceed... 
2023-06-17 09:34:17 <uuidNuniq> maakari[m]: if you learn some basic linux boot process, you'll know. 
2023-06-17 09:37:29 <mbx> uuidNuniq: I just completed setup-alpine and I would expect it to install e2fsprogs for me
2023-06-17 09:39:09 <uuidNuniq> BIOS looks for files installed by grub-install in the drive , UEFI looks for ESP/EFI/bootx64.efi, then grub is run, then grub looks for grub.cfg to load vmlinuz, vmlinuz looks for initramfs which is specified by cmdline "root=", initramfs loads drivers and mounts partition based on info provided by /etc/fstab, and kernel switch to new rootfs mounted by initramfs ... 
2023-06-17 09:39:23 <uuidNuniq> maakari[m]: ^^
2023-06-17 09:40:39 <uuidNuniq> mbx: ok, it should. anyway, take a look at log and then we'll probably find some hints 
2023-06-17 11:12:27 <raspbeguy> psykose: I'll test it as soon as I open my laptop
2023-06-17 11:26:20 <dwfreed> uuidNuniq: initramfs is specified by grub.cfg, not root=
2023-06-17 11:27:51 <raspbeguy> psykose: that works
2023-06-17 11:30:16 <raspbeguy> Many thanks
2023-06-17 11:36:43 <frag> when i use my router to access internet, some domains cant be loaded, is it the DNS? i set it up to get both op/dns "automatically from ISP"
2023-06-17 11:37:30 <dwfreed> how are we supposed to know if you don't provide any details or command output?
2023-06-17 11:37:52 <dwfreed> for all we know, your issues could be because a cow farted 3 towns over
2023-06-17 11:38:31 <frag> doubt its the cows
2023-06-17 11:39:38 <frag> getting  "ERR_NAME_NOT_RESOLVED" in my browser
2023-06-17 11:40:08 <dwfreed> what does dig say
2023-06-17 11:40:22 <dwfreed> does it resolve if you ask one of the popular public resolvers
2023-06-17 11:40:33 <dwfreed> these are some pretty basic troubleshooting steps
2023-06-17 12:38:06 <uuidNuniq> dwfreed: right. i realized the misinfomation after the write up
2023-06-17 12:39:45 <mbx> uuidNuniq: well, that wasn't successful ... let me elaborate: I reinstalled alpine onto my external SSD ("crypt", "lvm", "sys"), then luksOpen'ed, lvchange'd to activate and mounted. updated its mkinitfs.conf to enable all features, but mkinitfs will fail, because it can't find /mnt/lib/modules/6.1.34-0-lts. this bug is pretty old in Alpine, I think I mentioned this a while ago.
2023-06-17 12:39:47 <mbx> I had to `ln -s 6.1.34-2-lts 6.1.34-0-lts` in /mnt/lib/modules to bypass this. Why can the installer run mkinitfs, but I can't? Version should match imo. I then made sure /etc/sysctl.d/local.conf contains `*.*    -/var/log/syslog`  and /etc/rc.conf has `rc_logger="YES"`, unmounted, deactivated the vg, luksClose'd the SSD, tried to boot into the OS (without success). After
2023-06-17 12:39:49 <mbx> mounting the drive again I can see that /var/log/ only contains a chrony log. It doesn't log and I don't know why ... 
2023-06-17 12:41:19 <uuidNuniq> mbx: mkinitfs --help 
2023-06-17 12:41:24 <dwfreed> that line you gave for /etc/sysctl.d/local.conf does not go in that file
2023-06-17 12:41:26 <uuidNuniq> you can specify a kernel version 
2023-06-17 12:41:48 <dwfreed> might go in an syslog.conf or similar, depending on what syslogd you installed
2023-06-17 12:42:14 <mbx> Ok, but why mkinitfs doesn't pick the right kernel version for me? The installer has no problems with it, but it's kinda a low priority problem
2023-06-17 12:42:36 <mbx> dwfreed: I've read /etc/sysctl.d/README which says that it should load
2023-06-17 12:42:38 <dwfreed> it probably picks the running kernel version by default?
2023-06-17 12:42:39 <uuidNuniq> mbx: because you are in chroot
2023-06-17 12:42:55 <dwfreed> mbx: that is not a sysctl configuration
2023-06-17 12:43:01 <dwfreed>  that is a *syslog* configuration
2023-06-17 12:43:07 <mbx> dwfreed: oh yes, that could be the case, indeed
2023-06-17 12:43:09 <dwfreed> there's a *huge*  difference
2023-06-17 12:43:30 <dwfreed> but also if the initramfs fails to decrypt and mount your root filesystem, it's obviously not going to be able to log anything to said filesystem
2023-06-17 12:43:42 <minimal> yes from memory it does a "uname" to get the *running* kernel version unless you specify the kernel version as part of the mkinitfs command line
2023-06-17 12:43:49 <mbx> dwfreed: I mean /mnt/etc/sysctl.d/local.conf, sorry for the confusion, I edited the files on the drive that I will mount at boot
2023-06-17 12:44:13 <mbx> and I wasn't in chroot, it's not mandatory for mkinitfs
2023-06-17 12:44:39 <dwfreed> mbx: it's still not the write place for a *syslog* configuration
2023-06-17 12:44:54 <dwfreed> s/write/right/
2023-06-17 12:45:03 <minimal> mbx: as dwfreed has pointed out several times already, `*.*    -/var/log/syslog` appears to be an entry that should go in the *syslog* config, no in sysctl config
2023-06-17 12:45:06 <uuidNuniq> mbx: you are better run mkinitfs in chroot, or you have to pass all the correct args to it. 
2023-06-17 12:45:49 <mbx> got it, i will read carefully next time, I think I picked something wrong from the net, I think the entry is invalid then
2023-06-17 12:46:35 <minimal> maakari[m]: why are you installing inside a VM and then wanting to transfer to a physical machine? why not install directly on the physical machine?
2023-06-17 12:46:47 <mbx> uuidNuniq: I did call `mkinitfs -c /mnt/etc/mkinitfs/mkinifs.conf -b /mnt` and it worked without chroot
2023-06-17 12:47:09 <uuidNuniq> mbx: i suggest you boot with ISO and mount hard drive and chroot into it (you need to mount sys dev proc run tmp), and then do everything else. 
2023-06-17 12:48:01 <uuidNuniq> mbx: then pass a kernel version too: $(ls /mnt/lib/modules) 
2023-06-17 12:48:12 <uuidNuniq> -k or whatever arg, just read --help 
2023-06-17 12:48:16 <uuidNuniq> i am not on alpine right now. 
2023-06-17 12:49:15 <uuidNuniq> also pass an output option maybe? --output /mnt/boot/initramfs... ? again read --help
2023-06-17 12:49:24 <mbx> uuidNuniq: could you please tell why chroot is mandatory? Alpine docs say it's enough to pass -c and -b, but yes, I will remember to pass the kernel version next time, so thanks for that :)
2023-06-17 12:50:06 <mbx> Alpine docs don't mention an output either, I think that's what -b is for
2023-06-17 12:50:15 <mbx> it defines the root directory
2023-06-17 12:51:20 <mbx> s/root/base/
2023-06-17 12:51:24 <uuidNuniq> i always chroot for that kind of thing... easier, less problem. maybe mkinitfs is special?
2023-06-17 12:52:39 <mbx> It is special in that regard. But I agree that chroot is recommend, i.e. u need to chroot for grub setup anyways, but if you are only doing mkinitfs, it's faster to skip chroot
2023-06-17 12:54:07 <uuidNuniq> you set rc_logger="YES" and you don't find the rc.log afterwards?
2023-06-17 12:54:23 <mbx> correct
2023-06-17 12:55:04 <mbx> also no syslog or dmesg, maybe you have this on top of your head how to enable these logs?
2023-06-17 12:55:54 <uuidNuniq> dmesg log is enabled by /etc/init.d/dmesg? 
2023-06-17 12:56:07 <uuidNuniq> maybe you could try this: rc_interactive="YES"
2023-06-17 12:56:43 <mbx> I also enabled verbose outputs, openrc runs `/lib/rc/sh/openrc-run.sh /lib/rc/sh/openrc-run.sh /etc/init.d/hwdrivers start` before it gets stuck
2023-06-17 12:56:45 <minimal> raspbeguy, psykose: testing further after fixing that udev rule - found some more "issues" - as device-mapper package wasn't installed then udev came errors trying to run dmsetup, I installed device-mapper package and then see in dmesg "udevd[966]: conflicting device node '/dev/mapper/vg0-waffle' found, link to '/dev/dm-0' will not be created"
2023-06-17 12:56:52 <uuidNuniq> also rc_verbose=yes
2023-06-17 12:57:24 <mbx> right after hardware for mdev ...
2023-06-17 12:57:37 <uuidNuniq> mbx: setup-devd mdevd 
2023-06-17 12:57:47 <uuidNuniq> don't use mdev. it is buggy
2023-06-17 12:58:00 <mbx> that was picked for me >.>
2023-06-17 12:58:11 <uuidNuniq> try mdevd
2023-06-17 12:58:32 <minimal> raspbeguy, psykose: as device-mapper-udev has rules that run dmsetup I'd expect device-mapper-udev package to have a dep on device-mapper package but instead it has a install_if="device-mapper eudev" which to be doesn't seem right
2023-06-17 12:58:37 <mbx> do I have to disable mdev or will setup-devd solve it for me?
2023-06-17 12:59:06 <uuidNuniq> just run that, it does everything for you 
2023-06-17 12:59:21 <mbx> thanks, will try that :)
2023-06-17 12:59:26 <uuidNuniq> run in chroot
2023-06-17 12:59:45 <mbx> alright
2023-06-17 13:04:09 <uuidNuniq> mbx: dmesg.log is created by this service https://github.com/OpenRC/openrc/blob/1eccb7f5e6d69cae0fe13b8de1c302fa8d39d472/init.d/bootmisc.in#L223
2023-06-17 13:05:05 <uuidNuniq> mbx: in chroot you may want to run rc-update to check what services are enabled.
2023-06-17 13:41:03 <lopid> how do i see what this `apk upgrade` error is? "1 error; 1500 MiB in 753 packages"
2023-06-17 13:41:32 <ikke> lopid: apk fix
2023-06-17 13:41:44 <lopid> ah, thanks
2023-06-17 13:42:55 <lopid> ERROR: perl-doc-5.36.1-r2: trying to overwrite usr/share/man/man3/Time::HiRes.3pm.gz owned by perl-time-hires-doc-1.9764-r5.
2023-06-17 13:45:25 <uuidNuniq> lopid: apk del perl-time-hires-doc ; apk fix
2023-06-17 13:45:59 <lopid> well yes, i could remove the package
2023-06-17 13:47:16 <ikke> It should probably be removed from perl-doc
2023-06-17 13:53:09 <lopid> i would agree with that probablation
2023-06-17 15:05:21 <maakari[m]> is 326 mb of used ram a bit much or totally normal in tty on real hardware?
2023-06-17 15:08:25 <Apachez> depends
2023-06-17 15:08:40 <Apachez> used ram is so much more than actual being used by applications
2023-06-17 15:08:47 <Apachez> plenty are various caches that can be dropped when needed
2023-06-17 15:10:23 <lopid> 326 millibits is nothing
2023-06-17 15:12:07 <quinq> haha
2023-06-17 15:13:12 <invoked> a good question is whether something needs cache
2023-06-17 15:27:33 <Apachez> everything needs cache
2023-06-17 15:27:36 <Apachez> filecache to say the least
2023-06-17 16:50:52 <minimal> raspbeguy: so what change(s) happened for you after the udev rules was fixed?
2023-06-17 16:52:25 <raspbeguy> minimal: /dev/mapper/* became symlinks to /dev/dm-*
2023-06-17 16:53:03 <minimal> strange, you saw my earlier note about seeing a udev error?
2023-06-17 16:53:18 <minimal> "udevd[966]: conflicting device node '/dev/mapper/vg0-waffle' found, link to '/dev/dm-0' will not be created"
2023-06-17 16:53:34 <raspbeguy> In dmesg ?
2023-06-17 16:53:39 <minimal> yupe
2023-06-17 16:53:59 <raspbeguy> I'll check later when I open my laptop
2023-06-17 16:54:01 <minimal> that's output by udevd code that won't create a symlink if the file already exists
2023-06-17 16:54:28 <minimal> now I'm *not* using the lvm init.d service for these tests, don't know if that will make any difference
2023-06-17 16:54:50 <raspbeguy> I'm not using it either
2023-06-17 16:55:28 <minimal> so the udev rule calls "vgchange" which I believe creates /dev/mapper/vg0-X as a file
2023-06-17 16:56:03 <minimal> if it works for you then sounds like a race condition
2023-06-17 16:57:02 <minimal> I do note that the lvm2 build.log shows "checking whether to enable synchonisation with udev processing... no"
2023-06-17 16:57:19 <raspbeguy> minimal: here is my dmesg https://termbin.com/6612z
2023-06-17 16:57:40 <minimal> however enabling this in the build would likely be problematic and it might cause issues when using lvm with mdev/mdevd systems
2023-06-17 16:58:01 <raspbeguy> Here rae my enabled services https://termbin.com/a0yq
2023-06-17 16:59:18 <minimal> ah, see the errors about dmsetup? I mentioned those earlier today - device-mapper package isn't installed (for dmsetup) but dm/lvm udev rules make use of it
2023-06-17 17:00:31 <minimal> raspbeguy: where are the udev services? they don't show as enabled
2023-06-17 17:01:07 <raspbeguy> I enabled udev
2023-06-17 17:01:08 <minimal> oops, sorry, I see only udev (at sysinit)
2023-06-17 17:01:44 <minimal> what about the other ones? udev-postmount, udev-trigger, udev-settle?
2023-06-17 17:02:14 <raspbeguy> Didn't enabled them. See my second link
2023-06-17 17:02:21 <minimal> why not?
2023-06-17 17:02:36 <minimal> they are typically enabled
2023-06-17 17:02:51 <raspbeguy> Don't know, never heard of them
2023-06-17 17:03:50 <raspbeguy> Actually just enabling udev works for me
2023-06-17 17:03:54 <minimal> never heard of them? I though you were comparing Alpine with Debian? (which will have those as systemd services)
2023-06-17 17:04:29 <raspbeguy> I'm not a udev expert as mentioned earlier :/
2023-06-17 17:04:32 <minimal> and you didn't run setup-devd udev, which will enable these?
2023-06-17 17:04:46 <raspbeguy> Hum no
2023-06-17 17:05:05 <raspbeguy> Didn't know it existed
2023-06-17 17:05:14 <raspbeguy> You want me to run it?
2023-06-17 17:05:17 <minimal> so you have a partial udev system configured and have dmsetup missing which some of the lvm dm/lvm rules require
2023-06-17 17:05:26 <minimal> which is why you see those errors in dmesg
2023-06-17 17:06:12 <minimal> I'd recommend: "apk add device-mapper; setup-devd udev; reboot"
2023-06-17 17:08:41 <raspbeguy> minimal: here is my new dmesg https://termbin.com/1qid
2023-06-17 17:10:34 <minimal> ok, so no more dmsetup errors as expected, however also no symlink error like the one I get
2023-06-17 17:10:51 <minimal> and what about the contents of /dev/mapper ? files or symlinks?
2023-06-17 17:11:21 <raspbeguy> Symlinks
2023-06-17 17:11:57 <minimal> very strange. so what is different between your config and mine I wonder
2023-06-17 17:13:25 <raspbeguy> Here is my apk world file https://termbin.com/gkez
2023-06-17 17:16:24 <raspbeguy> My only config modification is enabling use of lvmlockd in LVM config
2023-06-17 17:18:11 <minimal> I have some addition stuff installed such as blkid & lsblk (rather than using the Busybox versions)
2023-06-17 17:18:53 <minimal> though I don't think udev uses those (it used libblk from memory) whereas mdev does use those commands
2023-06-17 17:19:34 <minimal> I'm currently testing on VirtualBox with Grub, let me build a Qemu Syslinux VM just to narrow the differences
2023-06-17 17:22:51 <mbx> uuidNuniq: hey :) had a little break. So I tried to `setup-devd mdevd` while being in chroot of my drive. The package `mdevd` is installed, rc-update reports that bootmisc runs at boot, whereas mdevd and mdevd-init runs at sysinit. It was looking got for me, but `setup-devd mdevd` as well as trying to boot reports: start-stop-daemon: fopen `/run mdevd.pid': No such file or
2023-06-17 17:22:54 <mbx> directory
2023-06-17 17:23:22 <mbx> s/got/good/
2023-06-17 17:23:32 <raspbeguy> minimal: If you wish I can share my libvirt domain definition xml
2023-06-17 17:25:20 <minimal> I'm just using qemu-system on CLI rather than via libvirt
2023-06-17 17:32:34 <d__> Hey, im having trouble with a bluetooth mouse. it is showing as connected but im unable to move the cursor at all
2023-06-17 18:18:24 <d__> nvm solved by running modprobe uhid
2023-06-17 18:18:55 <minimal> raspbeguy: I'm seeing that udevd symlink error on QEMU also
2023-06-17 18:19:13 <raspbeguy> Hum
2023-06-17 18:19:35 <raspbeguy> No clue where that comes from
2023-06-17 18:20:05 <codedroid> Is there a spamass-milter package or alternative for alpine?
2023-06-17 18:21:49 <raspbeguy> minimal: I'm using that to create my alpine images https://github.com/alpinelinux/alpine-make-vm-image
2023-06-17 18:31:28 <minimal> raspbeguy: I use something similar (though more comprehensive) that I wrote: https://github.com/dermotbradley/create-alpine-disk-image
2023-06-17 18:33:24 <raspbeguy> I could send my exact image if you wish
2023-06-17 18:33:53 <minimal> that might help
2023-06-17 18:34:55 <minimal> apart from lvm2lockd I'm not sure what is different
2023-06-17 19:11:28 <uuidNuniq> which package owns etc/securetty ?
2023-06-17 19:11:43 <uuidNuniq> that thing is not needed at all imo
2023-06-17 19:16:35 <uuidNuniq> oh it is owned by busybox. 
2023-06-17 19:48:17 <psykose> minimal: no, that seems correct, they'll just be there with device-mapper and udev
2023-06-17 19:55:37 <WhyNotHugo> I'd like to mount USB thumb drives and similar devices as non-root. My first though was to write a u-dev rule to grant users in the 'seat' group access to the device blocks themselves and mount them with fuse... but it looks like fuse+ext4 is readonly.
2023-06-17 19:55:54 <WhyNotHugo> Anyone else tackled this kind of issue? Anyone tried this particular approach?
2023-06-17 19:59:20 <Habbie> i don't know much, but i'm pretty sure you don't need fuse here
2023-06-17 20:14:27 <ikke> uuidNuniq: why do you think it's unnecessary?
2023-06-17 20:25:18 <uuidNuniq> ikke: yeah, when i start alpine with systemd-nspawn it won't login. because pts/0 is not in the list
2023-06-17 20:25:33 <uuidNuniq> also alpine's inittab does not have "console" ?
2023-06-17 20:25:45 <uuidNuniq> /dev/console
2023-06-17 20:27:31 <uuidNuniq> anyways. i need to edit /etc/inittab remove tty* add console, rm /etc/securetty, that way i am able to boot alpine with systemd-nspawn --boot
2023-06-17 20:30:30 <mbx> uuidNuniq: I was reading setup-devd.in as well as following some information by skarnet. seems I forgot to coldplug via -C flag which seems to be mandatory for mdevd setup?
2023-06-17 20:33:02 <uuidNuniq> mbx: https://git.alpinelinux.org/aports/tree/main/mdevd/mdevd-init.initd
2023-06-17 20:34:06 <uuidNuniq> enable that to the same runlevel as mdevd 
2023-06-17 20:35:45 <mbx> uuidNuniq: I had mdevd and mdevd-init in the sysinit runlevel, that's what setup-devd did for me ...
2023-06-17 20:36:09 <uuidNuniq> then it is coldpluged 
2023-06-17 20:36:15 <mbx> still it wouldn't find the pid, so something goes wrong, but I can't get my head around
2023-06-17 20:36:34 <mbx> fopen /run/mdevd.pid No such file or directory
2023-06-17 20:37:55 <uuidNuniq> because /run is not ready?
2023-06-17 20:38:59 <uuidNuniq> but can you check that file after booted
2023-06-17 20:39:03 <mbx> idk, I got the error message right in setup-devd and when I try to boot
2023-06-17 20:39:22 <mbx> file was missing
2023-06-17 20:40:16 <uuidNuniq> mbx: in chroot?
2023-06-17 20:40:24 <mbx> yes
2023-06-17 20:40:38 <uuidNuniq> did you mount /run before chroot?
2023-06-17 20:40:39 <mbx> I also bound /run to /mnt/run
2023-06-17 20:40:45 <mbx> yes
2023-06-17 20:41:12 <uuidNuniq> ohh. that's not an important messgae. ignore if you are in chroot
2023-06-17 20:41:23 <uuidNuniq> see if you can see that file after a real boot
2023-06-17 20:41:31 <mbx> I can't boot though
2023-06-17 20:41:53 <mbx> It's still stuck reading hardware and I see the error on boot as well
2023-06-17 20:42:10 <mbx> looks like it can't start mdevd
2023-06-17 20:43:00 <uuidNuniq> looks like your /etc/init.d/devfs is not started
2023-06-17 20:44:02 <mbx> rc-update reports devfs is at runlevel sysinit as well
2023-06-17 20:44:55 <raspbeguy> minimal, I'll send a private message containing a link to the image
2023-06-17 20:45:46 <mbx> uuidNuniq: shouldn't https://git.alpinelinux.org/aports/tree/main/mdevd/mdevd-init.initd declare an `after devfs`? could it load in the wrong order?
2023-06-17 20:46:04 <raspbeguy> did you get it? (not sure it works since I'm using a gateway to IRC)
2023-06-17 20:46:22 <uuidNuniq> maybe your initramfs can't decrypt the drive mbx 
2023-06-17 20:46:54 <uuidNuniq> those dirs /proc /sys /dev /run are created by initramfs 
2023-06-17 20:50:46 <mbx> uuidNuniq: as far as my understanding goes, luks decrypts it first. I asks me for the passphrase before boot, so it should be mounted. But I didn't see mdevd-init be launched before attempting to start mdevd, so I'm a bit skeptical in regards of the coldplug.
2023-06-17 20:51:40 <mbx> would I have to re-run the coldplug from a recovery when bringing alpine to new hardware anyways?
2023-06-17 20:56:08 <minimal> psykose: the point is what causes device-mapper package to be installed?
2023-06-17 20:56:21 <psykose> you would already have it installed
2023-06-17 20:56:33 <minimal> if you installdevice-mapper-udev it won't work right without device-mapper
2023-06-17 20:56:40 <minimal> psykose: I do
2023-06-17 20:56:54 <minimal> doh, I didn't see any deps to cause device-mapper to be installed though
2023-06-17 20:57:12 <psykose> you do manually install it yet
2023-06-17 20:57:13 <psykose> yes
2023-06-17 20:57:32 <psykose> idk who is manually installing device-mapper-udev and then expecting it to install device-mapper instead of Simply doing it the other way around
2023-06-17 20:57:34 <psykose> which works just fine
2023-06-17 20:57:34 <minimal> but if you manually install device-mapper-udev you get nonfunctional rules in place for udev
2023-06-17 20:57:38 <psykose> they should probably just stop doing that
2023-06-17 20:58:39 <minimal> well someone with lvm2 installed may install device-mapper-udev to get rules for lvm not realising that device-mapper is also required for those rules to functional
2023-06-17 20:59:33 <psykose> and someone might also install any -openrc package which then also does not actually pull in the parent package either
2023-06-17 20:59:37 <psykose> just don't do it
2023-06-17 21:00:49 <minimal> anyway that's a side-issue unrelated to the main issue of /dev/mapper entries as files versus softlinks
2023-06-17 21:02:43 <uuidNuniq> mbx: what filesystem do you have on your fstab
2023-06-17 21:06:33 <mbx> uuidNuniq: ext4 I think, I have to check later, need to go now, thanks for all your help. good night :)
2023-06-17 21:10:24 <raspbeguy> minimal, I was thinking that device-mapper was a dependency of LVM
2023-06-17 21:35:36 <WhyNotHugo> Habbie: it's either fuse or some hack to allow priviledge escalation for regular users to mount as root. fuse sounds cleaner.
2023-06-17 23:43:30 <consus[m]> Sorry, me again with weird powerpc issues :(
2023-06-17 23:43:38 <consus[m]> Please consider enabling VFIO like on x86_64
2023-06-17 23:47:46 <consus[m]> And VFIO_PCI
2023-06-18 00:07:55 <minimal> raspbeguy: nope, lvm2 does not depend on device-mapper, it depends on device-mapper-event-libs and lvm2-libs (and indirectly on device-mapper-libs)
2023-06-18 01:22:06 <orowith2os> Hi, currently trying to set up Alpine on a headless Pi and to put it simply it's a major pain.... not gonna finish that sentence.
2023-06-18 01:22:06 <orowith2os> I can set it up just fine if I connect a keyboard and display to the Pi, but nothing if I try going in through USB.
2023-06-18 01:22:22 <orowith2os> I've tried the headless config here: https://github.com/macmpi/alpine-linux-headless-bootstrap
2023-06-18 01:22:23 <orowith2os> but nothing
2023-06-18 01:22:55 <orowith2os> Generally I'm unsure how to handle this. it all feels so convoluted and undocumented.
2023-06-18 01:27:30 <orowith2os> I really don't want to use another distro, since I want something really small so I have as much free as possible.
2023-06-18 01:28:28 <orowith2os> maybe worth mentioning that this is a Pi 02W
2023-06-18 01:34:01 <psykose> not sure if that works at all or not or when the last time someone tested it was
2023-06-18 01:34:18 <psykose> that said that headless repo there is for ssh and not usb setup
2023-06-18 01:34:19 <psykose> not sure though
2023-06-18 01:34:41 <orowith2os> the repo has some instructions for ssh over usb, which I have followed
2023-06-18 01:34:46 <orowith2os> but still didn't work
2023-06-18 01:35:37 <orowith2os> so either the 02w doesn't support it, or there's something else going on?
2023-06-18 01:35:54 <orowith2os> Maybe I need to follow steps for the Pi3, since they seem to be pretty similar?
2023-06-18 01:36:28 <psykose> maybe, i only have an rpi4 so i can't test that myself
2023-06-18 01:37:22 <orowith2os> if you can try it there I would appreciate it too!
2023-06-18 01:37:36 <orowith2os> I have a Pi4 also, but it's in a case that doesn't expose the main USB connector.
2023-06-18 01:43:13 <psykose> will take a while but lets see if i figure it out
2023-06-18 01:52:56 <heat_> i did a zero 2w setup a few days ago, but on wifi
2023-06-18 01:54:34 <orowith2os> any and all info is good, I want to write a little guide on this for others in the future
2023-06-18 01:55:43 <heat_> getting to the setup for me was pretty smooth, just used headless.apkovl.... and I got ssh
2023-06-18 01:56:20 <psykose> orowith2os: seems to work on 4b
2023-06-18 01:56:34 <psykose> just writing the default 3.18 tar.gz and the headless apkovl thing
2023-06-18 01:56:41 <heat_> the setup was a bit screwed up, it did goofed when (not) enabling udhcpcd so i took like 30 minutes to muck around in weird filesystem post-mortems, hell, even chrooted using qemu-aarch64 to debug it even better
2023-06-18 01:56:41 <orowith2os> interesting. Can you walk me through it, step by step, explicitly?
2023-06-18 01:56:44 <psykose> and adding the wpa_supplicant and dwc2 thing
2023-06-18 01:56:48 <psykose> mm
2023-06-18 01:57:28 <orowith2os> aaa what I would give for my old Pi0...
2023-06-18 01:59:47 <ptrc> orowith2os: it's basically: create a fat32 partition on some sd card, download latest raspberry pi image from https://alpinelinux.org/downloads/, unpack it onto said fat32 partition; download headless.tar.gz from the github repo, place it in the topmost directory besides all the files, optionally add some other files to the same directory (like the mentioned wpa_supplicant.conf or usercfg.txt with `dtoverlay=dwc2`)
2023-06-18 02:00:27 <psykose> for 02w it seems wifi specifically might be missing some stuff (see all of https://gitlab.alpinelinux.org/alpine/aports/-/issues/13302) but that shouldn't block the regular usb gadget parts from this afaict
2023-06-18 02:00:36 <ptrc> also, it's important that you don't unpack the headless.apkovl.tar.gz file
2023-06-18 02:00:38 <psykose> (also ptrc did 90% of the work here and is sitting next to me)
2023-06-18 02:00:47 <ptrc> just place it next to the files
2023-06-18 02:00:59 <orowith2os> I'm calling it now, my SD card is going to die the moment I delete the partitions again, lmao
2023-06-18 02:01:05 <orowith2os> will get back to you momentarily 
2023-06-18 02:01:09 <ptrc> it's possible that the dtoverlay= value is different for the other pi models?
2023-06-18 02:01:31 <orowith2os> I'll try with a Pi3 and see if it works there too
2023-06-18 02:01:35 <orowith2os> see if that does anything
2023-06-18 02:01:42 <ptrc> pi 3 doesn't have otg, as far as i'm aware
2023-06-18 02:01:54 <ptrc> but it *should* work via wifi or ethernet
2023-06-18 02:01:57 <orowith2os> ....dammit.
2023-06-18 02:02:19 <orowith2os> yeah, it seems like it can't do the USB stuff other models can
2023-06-18 02:02:30 <orowith2os> ANYWAYS
2023-06-18 02:05:31 <heat_> psykose, 02w had wifi here, out of the box
2023-06-18 02:06:34 <psykose> heat_: that's interesting
2023-06-18 02:06:39 <psykose> wonder what the others are doing wrong
2023-06-18 02:06:45 <psykose> it could also be the separate model
2023-06-18 02:06:54 <psykose> since there's 2 02w's that i know of with separate chips/firmware
2023-06-18 02:08:01 <orowith2os> I can confirm that on my two pis they have WiFi ootb, though I do need to install an extra firmware package after installation to keep it
2023-06-18 02:08:42 <psykose> which one?
2023-06-18 02:08:51 <orowith2os> 02W
2023-06-18 02:08:57 <orowith2os> (2 of em)
2023-06-18 02:08:59 <psykose> i mean which firmware packies
2023-06-18 02:09:13 <orowith2os> Cypress, iirc? Something along those lines
2023-06-18 02:09:17 <orowith2os> It's documented in the wiki
2023-06-18 02:09:18 <psykose> all the linux-rpi* ones already depend on -cypress -brcm
2023-06-18 02:09:23 <orowith2os> Something about a firmware split
2023-06-18 02:09:25 <orowith2os> Yeah
2023-06-18 02:09:26 <psykose> though i think that was somewhat more recent
2023-06-18 02:10:08 <heat_> https://gist.github.com/heatd/41e2bde75f980e3a140f31464ea37387 <-- zero2w dmesg, if this helps
2023-06-18 02:11:48 <psykose> heat_: that looks like what i would expect, yeah
2023-06-18 02:12:00 <psykose> wonder why the people with non-working wifi in that thread have a completely barren one
2023-06-18 02:12:37 <orowith2os> ptrc just to confirm, extract to fat32, dtoverlay=dwc2, and USB login should work?
2023-06-18 02:13:10 <orowith2os> Nothing else?
2023-06-18 02:17:00 <ptrc> extract alpine image to fat32, place headless.apkovl.tar.gz onto that same fat32 partition *unextracted*
2023-06-18 02:17:05 <ptrc> besides that, yes
2023-06-18 02:34:57 <orowith2os> What cursed black magic is this, now it starts without me moving the files in boot/ one directory up?
2023-06-18 02:35:25 <orowith2os> still no connection though
2023-06-18 02:35:46 <orowith2os> maybe something with my USB cable
2023-06-18 02:37:31 <orowith2os> nope
2023-06-18 02:37:49 <orowith2os> maybe there's something else I'm missing on the desktop side...
2023-06-18 02:39:59 <ptrc> orowith2os: you have to set the interface up manually on the desktop side
2023-06-18 02:40:08 <ptrc> i did it with `doas ip link set usb0 up; doas ip addr add 10.42.0.1/24 dev usb0`
2023-06-18 02:40:21 <ptrc> you can check in dmesg if it shows up as a usb gadget device
2023-06-18 02:42:41 <orowith2os> yeah, I'm not seeing it in dmesg at all
2023-06-18 02:44:00 <ptrc> and did you put the `dtoverlay=dwc2` in usercfg.txt or config.txt? 
2023-06-18 02:44:35 <orowith2os> usercfg
2023-06-18 02:45:50 <ptrc> i'm out of ideas then
2023-06-18 02:48:26 <orowith2os> tried with the other pi02w, still nada
2023-06-18 02:48:51 <orowith2os> I'll try the official flasher tools and other distros to see what goes on there
2023-06-18 02:49:13 <orowith2os> OH I can try an armhf image too, I've only been doing aarch64...
2023-06-18 02:50:56 <psykose> aarch64 should be fine, prefered if anything
2023-06-18 02:51:28 <psykose> is the usercfg in the same folder as the config.txt
2023-06-18 02:54:13 <orowith2os> yep
2023-06-18 03:13:40 <psykose> can't think of much else then except otg not working for 02w with that setup
2023-06-18 03:28:07 <idkrn[m]> <psykose> "does it work" <- Not but I'm not running a graphical session
2023-06-18 03:28:45 <idkrn[m]> <quinq> "idkrn, could be something like..." <- On my other systems I only needed polkit related things
2023-06-18 03:31:14 <idkrn[m]> idkrn[m]: I think this is why I never had this problem before. I was in a Sway session
2023-06-18 05:20:04 <uuidNuniq> mbx: i run out ideas then, maybe post your fstab 
2023-06-18 08:10:56 <kitzman> It seems as if it cannot find the firmware image. Hmm
2023-06-18 08:40:00 <consus[m]> psykose: Please, be my PowerPC guarding angel once again and enable CONFIG_VFIO and CONFIG_VFIO_PCI for ppc64le
2023-06-18 08:46:37 <Someone[m]1> idkrn: you may remove the 'poweroff --> busybox' symlink and replace it with a setuid shell script; and chattr +i it
2023-06-18 08:48:04 <Someone[m]1> <idkrn[m]> "On my other systems I only..." <- systemd systems?
2023-06-18 08:49:26 <Someone[m]1> unprivileged user doesn't have the privileges to perform acpi-related tasks on system with real init
2023-06-18 08:49:48 <Someone[m]1> Someone[m]1: this is safe and used on artix linux
2023-06-18 08:50:10 <Someone[m]1> on alpine linux the 'poweroff' is the symlink to busybox so it can't be setuid-ed 
2023-06-18 08:54:42 <quinq> What's a fake init?
2023-06-18 08:55:20 <Someone[m]1> quinq: systemd
2023-06-18 08:55:44 <psykose> 2016 called they want their memes back
2023-06-18 08:57:55 <quinq> oh ok, I don't think that systemd is a fake init, the criticism there is that it couldn't stick to be an init
2023-06-18 08:58:18 <quinq> psykose, whatte happened in 2016?
2023-06-18 08:58:26 <quinq> s/whatte/what/
2023-06-18 08:58:55 <Lantizia> Hey is there some docs some place about the minimum required x86 cpu?  e.g. what it must support... whether it goes as far back as 386,486,586,686, etc.. ?
2023-06-18 08:59:22 <Lantizia> I'm thinking of running it on a Via C3
2023-06-18 09:00:23 <Lantizia> i mean... wiki.alpinelinux.org/wiki/Custom_Kernel says the x86 target includes 5x86/6x86 so it's probably fine?
2023-06-18 09:02:21 <quinq> Years ago it would run on my i586 at least
2023-06-18 09:04:35 <quinq> https://wiki.alpinelinux.org/wiki/Requirements seems to hint that the x86 target is i386, not sure how that's accurate though, would have to look at the builders
2023-06-18 09:11:47 <psykose> via c3 is i486 and not supported
2023-06-18 09:12:35 <consus[m]> quinq: Its not trying to
2023-06-18 09:12:42 <psykose> actual target is.. i586 but mostly between i586/i686 (some stuff has sse2)
2023-06-18 09:12:54 <psykose> and since 3.18 pae since for some reason the x86 kernel got pae enabled
2023-06-18 09:13:01 <psykose> but regardless of any of that via c3 was never supported
2023-06-18 09:14:54 <quinq> psykose, is there somewhere where this is documented? (for future reference)
2023-06-18 09:15:58 <psykose> or was it something else i was thinking of
2023-06-18 09:16:27 <psykose> ah, i was thinking of some vortex thing
2023-06-18 09:19:54 <quinq> psykose, and about the doc?
2023-06-18 09:20:20 <quinq> Would that be in the CI somewhere otherwise?
2023-06-18 09:32:14 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues/14789
2023-06-18 09:45:31 <psykose> quinq: the wiki page 'i386' is just the generic name like how random things or docker call it
2023-06-18 09:45:46 <psykose> as for documented it would mostly just be by what it is and not by what it isn't i guess
2023-06-18 09:45:53 <psykose> the gcc default target is i586- etc
2023-06-18 09:57:37 <quinq> Thank you, ikke and psykose 
2023-06-18 10:43:30 <leaix> Hi. I get "_nc_safe_fopen: symbol not found" when running "tput", what's the cause? I've already run "apk add ncurses" (Alpine v3.17). I haven't found anything online.
2023-06-18 10:46:51 <quinq> worksforme (musl/edge)
2023-06-18 10:48:06 <quinq> And the symbol is provided by /usr/lib/libncursesw.so.6.4
2023-06-18 10:48:52 <quinq> Sorry, I missed that you said it's on v3.17
2023-06-18 10:49:20 <quinq> leaix, what arch?
2023-06-18 10:50:32 <leaix> I enabled "edge" repositories (main/community/testing) too, but no luck
2023-06-18 10:50:49 <leaix> I'm using Alpine on a Raspberry Pi 4b
2023-06-18 10:51:04 <quinq> Yeah, apparently /usr/lib/libncursesw.so.6.3 doesn't provide that symbol
2023-06-18 10:51:19 <quinq> Maybe you mixed alpine versions?
2023-06-18 10:52:15 <leaix> well I'm quite new to Alpine and yesterday I upgraded from 3.15 to 3.17 (I followed the wiki). I don't know if tput worked before
2023-06-18 10:52:30 <quinq> Yeah, tput from ncurses 6.3 doesn't require _nc_safe_fopen
2023-06-18 10:53:33 <leaix> should I try to downgrade ncurses? Or should I disable the edge repositories?
2023-06-18 10:54:06 <quinq> You shouldn't mix release and edge (AFAIU)
2023-06-18 10:54:24 <quinq> Main reason is to avoid this kind of issue :)
2023-06-18 10:54:48 <leaix> Yes, I didn't want to, but let me explain
2023-06-18 10:55:25 <quinq> You don't have to justify yourself, but first make sure everything is upgraded to 3.17 (if you wish to follow that release)
2023-06-18 10:57:59 <leaix> I was trying to install pi-hole, but it isn't officially supported on Alpine. I didn't want to use docker. So I found an unofficial repository on gitlab which provided the installation script for Alpine. But it required to enable edge repositories... so first I tried to upgrade to 3.17, it didn't work, and I enabled the edge repositories
2023-06-18 10:58:19 <leaix> Ok let's switch to edge disabling 3.17, thank you, hope nothing breaks again
2023-06-18 10:59:14 <quinq> Edge is teh bleeding edge, but I haven't got any issues so far, but then the narrative is that if something breaks that's part of the game
2023-06-18 11:01:00 <leaix> I was trying to avoid it for this reason, seems like I can't heh
2023-06-18 11:14:41 <leaix> Now it works. Thanks again
2023-06-18 11:16:14 <quinq> nice
2023-06-18 13:50:59 <raspbeguy> minimal: I reinstalled from scratch manually with iso, I still don't have any dmesg error
2023-06-18 13:54:49 <idkrn[m]> <Someone[m]1> "idkrn: you may remove the '..." <- This is too hacky
2023-06-18 13:55:21 <minimal> raspbeguy: I tested your VM image and see the *same* udevd dmesg error about symlinks
2023-06-18 13:55:33 <raspbeguy> Hum
2023-06-18 13:55:49 <minimal> so the only difference between you and me must be the way we're creating the LVs
2023-06-18 13:56:34 <raspbeguy> The virtual hardware too
2023-06-18 13:58:54 <minimal> I'm using "qemu-img create", then "parted -a opt /dev/vdb" with "mklabel; mkpart primary ext2 2048s 100%", then "vgcreate /dev/vdb1; vgcreate vg0 /dev/vdb1; lvcreate -L 30m vg0 -n waffle"
2023-06-18 13:59:01 <minimal> we're both using virtio-blk
2023-06-18 14:00:21 <minimal> are you partitioning the device or creating the PV straight on the device?
2023-06-18 14:01:02 <minimal> oops, first "vgcreate" above should be "pvcreate"
2023-06-18 14:05:54 <raspbeguy> minimal: I don't partition the device
2023-06-18 14:06:06 <minimal> maybe that's a factor then
2023-06-18 14:06:13 <minimal> I'll test without partitioning
2023-06-18 14:06:14 <raspbeguy> Also the pvcreate step isn't necessary 
2023-06-18 14:06:52 <minimal> why not? pvcreate initialises a PV
2023-06-18 14:07:14 <raspbeguy> Vgcreate does it automatically 
2023-06-18 14:07:44 <minimal> ah right, force of habit to run pvcreate
2023-06-18 14:14:46 <dwfreed> it's still useful to know that it needs doing, especially if you plan to do non-standard things with vgs
2023-06-18 14:18:24 <minimal> raspbeguy: trying "vgcreate" gave error "Global lock failed: check that global lockspace is started."
2023-06-18 14:19:21 <raspbeguy> minimal: because I activated lock management in LVM settings
2023-06-18 14:19:33 <raspbeguy> use_lvmlock
2023-06-18 14:19:41 <minimal> how/where ?
2023-06-18 14:19:41 <raspbeguy> You can disable it
2023-06-18 14:20:02 <raspbeguy> /etc/lvm/lvm.conf
2023-06-18 14:20:26 <minimal> I mean what do YOU do in order to run vgcreate? I'm trying do test things in a similar manner to you
2023-06-18 14:21:24 <raspbeguy> vgcreate --shared myVG /dev/vdb
2023-06-18 14:22:10 <raspbeguy> And after that, after each reboot : vgchange --lock-start
2023-06-18 14:23:52 <minimal> but isn't udev running "vgchange" to activate it when it sees the VG/LV?
2023-06-18 14:24:01 <minimal> that's what the rule we fixes does
2023-06-18 14:24:41 <raspbeguy> No. Nothing gets created in /dev until I run lvs
2023-06-18 14:25:51 <minimal> with which options?
2023-06-18 14:26:04 <raspbeguy> Just lvs
2023-06-18 14:26:34 <raspbeguy> I guess vgscan could do that too
2023-06-18 14:26:35 <minimal> lvs just displays info, why would running it make a difference?
2023-06-18 14:26:53 <raspbeguy> It must trigger a scan or something 
2023-06-18 14:27:11 <minimal> so what is the EXACT sequence of commands you run every boot?
2023-06-18 14:27:31 <raspbeguy> To activate a LV ?
2023-06-18 14:27:55 <minimal> do "make things work"
2023-06-18 14:28:00 <minimal> s/do/to/
2023-06-18 14:28:17 <minimal> I'm trying to test your scenario but I can't do that if I don't know what you're doing
2023-06-18 14:28:41 <raspbeguy> This is a test VM. Nothing works per se
2023-06-18 14:28:52 <minimal> the softlinks work...
2023-06-18 14:29:00 <raspbeguy> Ok
2023-06-18 14:29:18 <minimal> so I need to know exactly what you're doing
2023-06-18 14:29:26 <raspbeguy> Then I run vgchange --lock-start, then lvs
2023-06-18 14:29:54 <raspbeguy> I guess activating the lvm service at boot should also work
2023-06-18 14:30:16 <raspbeguy> Problem is that it tries to activate all my lv, which I don't want
2023-06-18 14:31:12 <minimal> vgcreate is failing for me with "VG vg0 init failed: -19"
2023-06-18 14:32:25 <raspbeguy> Look at /var/log/sanlock.log
2023-06-18 14:32:34 <raspbeguy> Your answer lies there
2023-06-18 14:32:55 <minimal> it created LV "vg0/lvmlock" ok
2023-06-18 14:33:14 <raspbeguy> But its owned by root:root
2023-06-18 14:33:33 <raspbeguy> So sanlock cannot do anything 
2023-06-18 14:34:04 <minimal> "no permission to open /dev/mapper/vg0-lvmlock"
2023-06-18 14:34:12 <raspbeguy> Bingo
2023-06-18 14:34:41 <minimal> so you haven't told me how YOU managed to run vgcreate successfully using this image yet...
2023-06-18 14:34:43 <raspbeguy> That's the error I had before udev rules fix
2023-06-18 14:35:08 <raspbeguy> Well it works on my machine
2023-06-18 14:35:23 <minimal> this IS your (VM) machine I'm using
2023-06-18 14:41:36 <raspbeguy> Yes
2023-06-18 14:41:49 <raspbeguy> For some reason it doesn't work when you run it
2023-06-18 14:46:21 <raspbeguy> Actually I'm not sure what you're looking for. Since udev rules have been fixed I don't have any problem 
2023-06-18 14:46:41 <minimal> you may not have a problem but I still do
2023-06-18 14:47:02 <minimal> so your scenario appears to be the only one where the softlinks actually work
2023-06-18 14:47:33 <raspbeguy> I'll make a screen record for you
2023-06-18 14:48:08 <minimal> I just disabled lockd, did a vgcreate and a lvcreate and STILL see the udevd error in dmesg and no softlink
2023-06-18 14:48:25 <minimal> no it is not a partitioned vs unpartitioned issue
2023-06-18 14:50:13 <raspbeguy> I'll do that later when I open my laptop
2023-06-18 14:51:55 <minimal> basically if I can't reproduce your scenario then there's no real way for me to figure what is different and therefore no likely way to fix this issue in general
2023-06-18 14:58:48 <raspbeguy> I'll try in a virtual box so that I can send you the EXACT situation, including virtual hardware
2023-06-18 15:00:04 <minimal> I don't believe this is related to "exact" virtual hardware, we are both using virtio-blk for the storage device
2023-06-18 15:01:25 <minimal> I don't understand how I get sanlock related permissions errors that you don't get when using the same VM image
2023-06-18 15:01:42 <OdaNobunaga[m]> Hi! I'm trying to build a NAS/server with Alpine Linux. So far, I managed to get a diskless version running from my usb stick, but it looks like the diskless kernel is lacking the zfs module. :(
2023-06-18 15:02:34 <raspbeguy> minimal: or more precisely why /dev/mapper/* are not symlinks when you run it
2023-06-18 15:03:15 <minimal> raspbeguy: well I didn't get as far as testing that as the vgcreate failed with the sanlock errors
2023-06-18 15:03:44 <raspbeguy> That's not difficult to verify
2023-06-18 15:04:00 <minimal> what's not difficult to verify?
2023-06-18 15:04:28 <raspbeguy> Just disable use_lvmlock in LVM settings, create a vg and then a LV on it
2023-06-18 15:04:44 <raspbeguy> You'll see that /dev/mapper/* is a file
2023-06-18 15:04:46 <minimal> I told you about 10-15mins ago I just did that and saw the same udevd error
2023-06-18 15:05:20 <raspbeguy> Well that's why sanlock fails then
2023-06-18 15:05:30 <minimal> but I'm trying to test your *exact* scenario which you say works, which means using lvmlock
2023-06-18 15:06:27 <raspbeguy> Well, if /dev/mapper/* isn't a symlink or do not belong to group 'disk' then sanlock will fail
2023-06-18 15:07:28 <raspbeguy> So if you can't manage to get symlinks when NOT using lvmlock then it's useless to try further 
2023-06-18 15:08:34 <minimal> well I'm trying to see symlinks being created in some way as ALL the various testing I've done so far does not result in that yet you tell me it works perfectly for you
2023-06-18 15:09:07 <minimal> I'm suspecting a race condition of some sort
2023-06-18 15:09:38 <raspbeguy> That's why I'll do a working vbox proof then send it to you
2023-06-18 15:10:17 <minimal> but you already sent me a "working" qemu "proof" which didn't work for me...
2023-06-18 15:10:19 <raspbeguy> Because for now we do not know what exact virtual hardware we are using
2023-06-18 15:10:25 <minimal> how will VB be any different?
2023-06-18 15:10:55 <raspbeguy> With vbox I can send to you the machine exact definition
2023-06-18 15:11:08 <minimal> I know exactly which virtual hardware I'm using and as I pointed our we are both using virtio-blk, not emulating physical hardware, for storage...
2023-06-18 15:11:22 <raspbeguy> That's why I suggested sending the libvirt xml but you're not using that
2023-06-18 15:11:52 <raspbeguy> That's not precise enough
2023-06-18 15:11:56 <minimal> I don't use libvirt, I use qemu-system directly (which libvirtd does behind the scenes)
2023-06-18 15:12:07 <raspbeguy> I know
2023-06-18 15:19:24 <minimal> raspbeguy: what CPU are you running libvirtd/VB on?
2023-06-18 15:22:07 <raspbeguy> minimal: https://termbin.com/jqmp
2023-06-18 15:22:59 <minimal> a Xeon, ok I'm using a Celeron - so if it is race condition that would make sense
2023-06-18 15:25:20 <minimal> basically I'm guessing I always see the problem of the race condition as I have a slow CPU, whereas you never see it as you have a fast CPU
2023-06-18 15:26:12 <raspbeguy> I also could give you access to a VM hosted by me
2023-06-18 15:44:10 <minimal> raspbeguy: ran it in a i7 CPU, made no difference
2023-06-18 16:11:47 <radiomark> Trying to upgrade 3.17 to 3.18, it's failing to resolve dependencies ... https://pastebin.com/5K88q53T
2023-06-18 16:12:05 <radiomark> I assume I've got something installed that's not from the core repos ... how can I hunt this down?
2023-06-18 16:16:43 <ikke> radiomark: add --available, which you should provide when switching repos, see if that helps
2023-06-18 16:16:56 <radiomark> Ah, yes, already doing that as it's in the upgrade guide
2023-06-18 16:17:28 <radiomark> For now I've uninstalled a bunch of stuff -- chromium, firefox, libreoffice, inskscape, mpv ... it seems to have satistied the dependency
2023-06-18 16:17:41 <radiomark> Doing an upgrade now of 786 packages so let's see
2023-06-18 16:19:01 <uuidNuniq> poor guy
2023-06-18 16:19:41 <uuidNuniq> apk upgrade -aU is all you need
2023-06-18 16:20:35 <radiomark> Apparently not
2023-06-18 16:23:58 <uuidNuniq> plus a apk fix
2023-06-18 16:24:33 <radiomark> I had done "apk fix" but it did not seem to find anything it wanted to resolve
2023-06-18 16:25:16 <radiomark> Like I said, I assumed some package was not from the core repos. But I went through the list and most of them had upgraded versions incoming in the new 3.18 repo
2023-06-18 16:25:31 <radiomark> (all of the ones I went through)
2023-06-18 16:25:47 <radiomark> Hence asking if there was an easy way to find these packages
2023-06-18 16:27:10 <uuidNuniq> what is core repo
2023-06-18 16:27:43 <radiomark> "main" and "community"
2023-06-18 16:29:10 <radiomark> I did a recent 3.16 -> 3.17 upgrade without a hitch, and with just "main" and "community"
2023-06-18 16:30:47 <uuidNuniq> you have manual built pkgs?
2023-06-18 16:31:48 <radiomark> No
2023-06-18 16:32:08 <radiomark> But the pastebin shows the packages it's having trouble with
2023-06-18 16:33:17 <uuidNuniq> you said some packages was not from CORE repo
2023-06-18 16:33:39 <radiomark> Some packages not from "core" repos (ie. main or community)
2023-06-18 16:33:48 <radiomark> It's a possibility, if I once tried some other packages
2023-06-18 16:34:13 <radiomark> Hence, is there a way to list the installed packages, by repo?
2023-06-18 16:34:36 <uuidNuniq> check apk list --help?
2023-06-18 16:37:14 <ikke> apk does not keep track what repo a package came from
2023-06-18 16:37:45 <radiomark> ikke: ok, thanks, so i'm left doing it by hand
2023-06-18 16:38:23 <radiomark> The upgrade was 'successful' but now I can't install firefox ... so it's helping to narrow down the problem
2023-06-18 16:38:40 <radiomark> Damn, I have no browser now to do pastebin
2023-06-18 16:39:15 <ikke>  | curl https://tpaste.us -F tpaste=\<-
2023-06-18 16:40:31 <radiomark> Helpful, thanks. Let me try that
2023-06-18 16:40:31 <georgelefeu_> Hi there i got aproblem using gnome-software under xfce in Alpine Linux 3.18.2. I installed gnome-software (doas apk add gnome-software) and started apk-polkit-server service as recommended. When i start gnome-software from the commandline, i got the following messages: ~ $ gnome-software1
2023-06-18 16:40:31 <georgelefeu_> 16:18:38:855 Gs  failed to get featured apps: no apps to show
2023-06-18 16:40:31 <georgelefeu_> 16:18:38:858 Gs  Only 0 apps for curated list, hiding
2023-06-18 16:40:31 <georgelefeu_> 6:18:38:731 Gs  Only 0 apps for recent list, hiding
2023-06-18 16:40:32 <georgelefeu_> 16:18:38:956 GsPluginApk app */*/*/org.alpinelinux.alpine-3.18.2/* has missing sources; skipping
2023-06-18 16:41:00 <georgelefeu_> What is going wrong?
2023-06-18 16:42:07 <radiomark> ikke: https://tpaste.us/b1p1
2023-06-18 16:42:43 <radiomark> So I'm adding firefox, and it's complaining it can't upgrade packages. But all those packages also have new versions available in the repos
2023-06-18 16:44:12 <radiomark> "apk upgrade --available" shows nothing ot do
2023-06-18 16:44:12 <ikke> radiomark: does `apk list --orphaned` return something?
2023-06-18 16:44:26 <radiomark> Yes, a few packages
2023-06-18 16:44:39 <ikke> which?
2023-06-18 16:44:49 <ikke> they might be the cause
2023-06-18 16:46:03 <radiomark> https://tpaste.us/1XBV
2023-06-18 16:46:37 <radiomark> The wx and sdl ones sound like they could be old. But wouldn't they appear in the dependency report when eg. installing firefox or doing the upgrade?
2023-06-18 16:46:46 <radiomark> Part of the "satisfies" list
2023-06-18 16:48:41 <ikke> radiomark: what happens if you do: `apk add darktable\>=4.2.1
2023-06-18 16:48:44 <ikke> `
2023-06-18 16:49:55 <radiomark> https://tpaste.us/pa6x  
2023-06-18 16:50:00 <radiomark> Leads back to wxgtk3
2023-06-18 16:50:30 <ikke> wxgtk3 is the culprit indeed
2023-06-18 16:50:33 <kwadroke[m]> georgelefeu_: maybe add gnome-software-plugin-apk ?
2023-06-18 16:51:26 <radiomark> ikke: am I being stupid not noticing that ... how would I tell from the "updgrade --available" or "add firefox" 
2023-06-18 16:51:41 <radiomark> Is it because it's only printing one level of dependency?
2023-06-18 16:51:55 <radiomark> I'll remove wxgtk3 and see
2023-06-18 16:53:23 <radiomark> So that worked. Now the firefox install is upgrading a load of other (seeminly unrelated) packages ... presumably because of the underlying dependency
2023-06-18 16:53:51 <ikke> you should do an upgrade now
2023-06-18 16:53:58 <ikke> to make sure everything is upgraded
2023-06-18 16:54:14 <radiomark> Well, don't forget "apk upgrade --available" was reporting nothing to do
2023-06-18 16:54:34 <ikke> radiomark: it would also say that if something was holding that back
2023-06-18 16:54:37 <radiomark> Even though newer packaes in the repo, it was prioritising the dependcies from wxgtk3
2023-06-18 16:54:40 <ikke> yes
2023-06-18 16:54:51 <raspbeguy> minimal, you won't believe it
2023-06-18 16:54:54 <ikke> so now that you removed that, all packages should be able to be upgraded
2023-06-18 16:55:11 <radiomark> ikke: ah, I see yes. Ok when this firefox is done, let's see if there's any residual
2023-06-18 16:55:14 <raspbeguy> I tried to reinstall VMs. Like 3 or 4 times.
2023-06-18 16:55:22 <raspbeguy> Now it hit the same error
2023-06-18 16:55:50 <georgelefeu_> kwadroke[m]: gnome-software-plugin-apk is added.
2023-06-18 16:57:07 <radiomark> ikke: https://tpaste.us/Wxm0
2023-06-18 16:57:28 <ikke> so it did upgrade a few packages
2023-06-18 16:57:52 <radiomark> Thanks for the help, at least I can progress
2023-06-18 16:58:05 <ikke> keep in mind the other packages in that orphaned list can still cause issues
2023-06-18 16:58:12 <radiomark> Yeah, that's what I'm thinking 
2023-06-18 16:58:17 <radiomark> What does 'orphaned' mean in this context?
2023-06-18 16:58:31 <ikke> no repo providing that package anymore
2023-06-18 16:58:35 <radiomark> Some of them -- like SDL -- I can imagine I went looking for old package
2023-06-18 16:58:52 <radiomark> But others, like xf86-input-keyboard sound pretty important and not somtehing I've messed with!
2023-06-18 16:59:05 <ikke> they are no longer relevant
2023-06-18 16:59:16 <radiomark> To be safe, I should probably remove all those packages
2023-06-18 16:59:22 <radiomark> I'll have to hope it doesn't break some functionality though
2023-06-18 16:59:26 <ikke> you just need xf86-input-libinput
2023-06-18 17:01:16 <radiomark> ikke: Ok I've left the fonts for now, going to restart X just to be sure
2023-06-18 17:02:06 <raspbeguy> minimal, I got it
2023-06-18 17:03:02 <minimal> raspbeguy: believe what?
2023-06-18 17:03:10 <radiomark> ikke: well at least I've got keyboard + mouse after a reboot :)
2023-06-18 17:03:14 <raspbeguy> I was working with existing VG before. When I recreated the VM, I never bothered to erase the secondary test volume
2023-06-18 17:04:05 <raspbeguy> this is what I tried
2023-06-18 17:04:56 <radiomark> ikke: thanks for the help
2023-06-18 17:05:01 <raspbeguy> I temporarly ran sanlock by root user by modifying the /etc/conf.d/sanlock file
2023-06-18 17:05:01 <ikke> radiomark: no problem
2023-06-18 17:05:41 <raspbeguy> then I created the VG by running vgcreate --shared testVG /dev/sdb
2023-06-18 17:06:17 <raspbeguy> I then restored the original user in /etc/conf.d/sanlock
2023-06-18 17:06:20 <raspbeguy> then reboot
2023-06-18 17:08:37 <raspbeguy> then run vgchange --lock-start
2023-06-18 17:08:47 <raspbeguy> (notice that it's a success)
2023-06-18 17:09:05 <raspbeguy> also I can now create lv
2023-06-18 17:09:30 <raspbeguy> and most importantly, enable them
2023-06-18 17:11:18 <minimal> raspbeguy: am confused, you're saying you fixed something? you'd previously said it worked fine so what was there to fix if it worked fine?
2023-06-18 17:14:40 <raspbeguy> didn't fix anything
2023-06-18 17:15:20 <raspbeguy> I just discovered why I thought that was working but yet there were still issues
2023-06-18 17:17:09 <raspbeguy> also, only the "lvmlock" lv is a symlink. Didn't noticed that the other LV (the real useful ones) are still files
2023-06-18 17:17:53 <raspbeguy> but if the lvmlovk lv is read/write for sanlock then everything works fine
2023-06-18 17:18:03 <kwadroke[m]> georgelefeu_: might be an issue with gnome-software-apk then - https://github.com/Cogitri/gnome-software-plugin-apk/issues/38 . on my Pinephone with PostmarketOS I have 2 Alpine repos disabled which I can't enable . Might be a similar issue
2023-06-18 17:18:08 <minimal> ah ha, so you're still seeing the problem - yet the "dmesg" doesn't show a udevd error about the other symlinks failing?
2023-06-18 17:18:18 <raspbeguy> by "everything works fine" I mean that locks are correctly managed
2023-06-18 17:18:30 <kwadroke[m]> honestly I just use the commandline to install/update
2023-06-18 17:20:48 <minimal> raspbeguy: as you say the non-lvmlock entries in /dev/mapper/ are *not* symlinks does "dmesg" show any udevd errors?
2023-06-18 17:22:25 <raspbeguy> https://termbin.com/04ckf
2023-06-18 17:23:07 <minimal> raspbeguy: right, the last line shows the udevd error I've been talking about
2023-06-18 17:23:50 <minimal> so udevd wanted to create /dev/mapper/testVG-tintin as a symlink pointing to /dev/dm-1 but could not as /dev/mapper/testVG-tintin already existed as a file
2023-06-18 17:24:03 <minimal> so you ARE seeing the issue I've been talking about
2023-06-18 17:25:02 <georgelefeu_> kwadroke[m]: I'm quite shure there's an issue with the apk-polkit-server service. In Alpine Linux 3.17 apk-polkit-rs has also to be installed, but no service had to be started. And gnome-software worked without problems.
2023-06-18 17:28:33 <leaix> Why does "awall translate --verify" say "awall: ssh: filter.1: Invalid zone: WAN"? I already configured /etc/awall/private/base.json, and yes "WAN" (= eth0) exists and is up on the device (as per ip link).
2023-06-18 17:30:07 <ikke> Did you Import that policy in the policy you refer to it?
2023-06-18 17:32:53 <raspbeguy> minimal: yes. Sorry apparently I was too negligent 
2023-06-18 17:33:58 <leaix> ikke: How do I "import" base.json? I just refer to "WAN" in /etc/awall/optional/ssh.json. And yes I've run "awall enable base". The ssh.json's content is exactly this: https://wiki.alpinelinux.org/wiki/Zero-To-Awall#SSH_service
2023-06-18 17:34:16 <minimal> raspbeguy: I might try and build lvm2 with udev sync to see if that makes any difference
2023-06-18 17:36:52 <leaix> "awall enable ssh"* since "awall enable base" doesn't work ("Not an optional policy"). 
2023-06-18 17:42:24 <raspbeguy> minimal: there is an option udev_sync in the lvm config file
2023-06-18 17:42:45 <raspbeguy> Lead me to think this is a runtime option, not a build option
2023-06-18 17:42:59 <minimal> raspbeguy: but the build.log for lvm2 indicated it was disabled, so the config option should have no effect
2023-06-18 17:43:59 <minimal> "hecking whether to enable synchronisation with udev processing... no"
2023-06-18 17:44:21 <raspbeguy> Ok then
2023-06-18 17:45:52 <minimal> that has to be turned on during build when "configure" is called with "--enable-udev_sync"
2023-06-18 17:46:59 <ikke> leaix: "import": "base"
2023-06-18 17:47:08 <ikke> in the file that you use it
2023-06-18 17:47:35 <ikke> /etc/awall/private is not consulted for policy files and meant for configuration that you can include
2023-06-18 17:48:12 <ikke> "Sometimes a policy file depends on other policy files. In this case, the policy file must have a top-level attribute import, the value of which is a list of policy names, which correspond to the file names without the .json suffix. The imported policies may be either optional policies or private policies, located in /usr/share/awall/private or /etc/awall/private."
2023-06-18 18:14:07 <leaix> ikke: solved, thanks
2023-06-18 18:33:41 <leaix> Another issue: seems like '/usr/libexec/command-not-found' takes ~3 seconds to run on my Raspberry 4B (I know it isn't very powerful, but 3s? for this?), even on ash. Where do I start to troubleshoot this? Also there are just 300 pkgs
2023-06-18 18:38:45 <quinq> leaix, that would be apk-search
2023-06-18 18:39:46 <ikke> leaix: run strace to see what takes so long 
2023-06-18 18:44:36 <uuidNuniq> strace?
2023-06-18 18:44:40 <uuidNuniq> is that a shell script
2023-06-18 18:44:57 <uuidNuniq> what about sh -x /usr/libexec/command-not-found
2023-06-18 18:46:07 <quinq> strace on apk search
2023-06-18 18:46:55 <leaix> quinq: the issue is, I'm not searching anything, it runs automatically if I type the wrong command
2023-06-18 18:47:22 <quinq> Remove the package :)
2023-06-18 18:50:53 <uuidNuniq> yeah. del it, that thing is mostly useless
2023-06-18 18:54:35 <leaix> yes "fixed" this way. Seems like apt-cache isn't able to cache the packages list
2023-06-18 19:03:23 <netstickbat> anyone using lxd? when trying to launch a vm i am getting "QEMU failed to run feature checks / qemu: could not load PC BIOS '/usr/share/OVMF/OVMF_CODE.fd'" even though the ovmf package is installed and that file exists
2023-06-18 19:04:02 <uuidNuniq> leaix: you need setup-apkcache and setup-lbu
2023-06-18 19:08:15 <leaix> uuidNuniq: I'm using Alpine in system mode, not diskless, so I don't use lbu. I tried again, and setup-apkcache doesn't change anything unfortunately
2023-06-18 19:22:03 <uuidNuniq> leaix: https://wiki.alpinelinux.org/wiki/Local_APK_cache
2023-06-18 20:01:39 <leaix> 13,99uuidNuniq99,99: maybe it's me, maybe the problem is elsewhere (i/o-bound and not cpu-bound?), but even after trying again I had no luck so far. I guess it can't be fixed
2023-06-18 20:04:16 <ikke> it would probably faster if it used apk search cmd:<command> instead
2023-06-18 20:06:02 <quinq> That's what this command-not-found does' so I doubt it'd be faster
2023-06-18 20:09:05 <leaix> apk search -xv cmd:ls still takes 3s
2023-06-18 20:09:15 <ikke> that's not normal
2023-06-18 20:10:35 <quinq> What the storage media?
2023-06-18 20:11:46 <ikke> yeah, might be a storage issue
2023-06-18 20:15:11 <leaix> it's a 32GB sdcard, using dd seems like the write speed is something around ~20MB/s. Trying to test the read speed now
2023-06-18 20:16:00 <quinq> Erf, SDCard, most likely the issue :)
2023-06-18 20:16:49 <ikke> leaix: do you write directly to the card? filesystem cache might be skewing the result
2023-06-18 20:20:11 <leaix> ikke: /dev/mmcblk0: Timing cached reads:   1406 MB in  2.00 seconds = 704.24 MB/sec; Timing buffered disk reads: 102 MB in  4.08 seconds =  24.98 MB/sec
2023-06-18 20:20:21 <leaix> quinq: Yeah I didn't expect it to be soo slow
2023-06-18 20:21:18 <leaix> so it was using the cache from the beginning... anyway, thanks
2023-06-18 20:21:58 <ikke> those are 2 different caches
2023-06-18 20:22:29 <ikke> setup-apkcache sets up package caching, while apk search uses the index cache
2023-06-18 20:23:40 <leaix> understood
2023-06-18 20:24:22 <ikke> so if you enabled that, you should be aware that /var/cache/apk might grow over time
2023-06-18 20:30:12 <leaix> thanks
2023-06-18 20:32:39 <quinq> Hummm, why does plocate provide *upgradedb* but not *locate* (it provides *plocate*)?
2023-06-18 20:33:31 <Habbie> (fwiw 'plocate' registers itself as an alternative for 'locate' on debian - but i don't know if alpine/apk has an alternatives mechanism)
2023-06-18 20:34:01 <quinq> :)
2023-06-18 21:07:22 <consus[m]> Default sysctl on both ppc64le and x86_64 has this:
2023-06-18 21:07:27 <consus[m]>  * Configuring kernel parameters ...
2023-06-18 21:07:27 <consus[m]> sysctl: cannot stat /proc/sys/vm/unprivileged_userfaultfd: No such file or directory
2023-06-18 21:11:37 <psykose> that hasn't existed in a million years
2023-06-18 21:12:31 <psykose> nothing even tries to touch that anywhere that i can see
2023-06-18 21:14:34 <consus[m]> Fresh 3.18
2023-06-18 21:15:51 <consus[m]> aha
2023-06-18 21:15:55 <consus[m]> I see who
2023-06-18 21:16:26 <consus[m]> /usr/lib/sysctl.d/60-qemu-postcopy-migration.conf is owned by libvirt-qemu-9.3.0-r0
2023-06-18 21:18:24 <psykose> yeah
2023-06-18 21:18:25 <psykose> shrug
2023-06-18 21:18:28 <psykose> doesn't matter very much
2023-06-18 21:19:10 <consus[m]> Yeah, I'm just triggered by warnings
2023-06-18 21:19:21 <consus[m]> Will overwrite it in /etc
2023-06-19 04:18:00 <novorol[m]> what do I need to install to be able to fully inspect C++ core dumps on Alppine with gdb?
2023-06-19 04:34:51 <ikke> apk add musl-dbg, build the relevant project with debug symbols enabled, potentially relevent libraries as well
2023-06-19 12:00:01 <raspbeguy> minimal, did you manage to build lvm2 with udev dync feature ?
2023-06-19 12:00:30 <raspbeguy> I did try but it adds a dependancy to libudev
2023-06-19 12:00:41 <raspbeguy> so I'n not sure how to handle this
2023-06-19 12:02:08 <minimal> yes I've build it, am currently building VM with it for testing
2023-06-19 12:05:21 <raspbeguy> nice
2023-06-19 12:09:20 <psykose> what does it allow doing
2023-06-19 12:12:35 <CcxWrk> Is it just the part of libudev to access /sys/* or actually one that requires udev to communicate with via dbus (gudev)?
2023-06-19 12:12:37 <minimal> psykose: it makes lvm tools "sync" with eudev regarding /dev/mapper/ device creation - I just want to see if this resolves the file vs symlink situation
2023-06-19 12:12:52 <psykose> CcxWrk: former
2023-06-19 12:13:39 <CcxWrk> Ah, then it should work with libudev-zero
2023-06-19 12:14:15 <minimal> not sure, haven't dug too deep into it, just want to see what behavioural change occurs with it enabled
2023-06-19 12:15:24 <psykose> minimal: the only part i'm interested in is what happens when eudev isn't in use but the option is enabled buildtime i guess
2023-06-19 12:15:30 <minimal> currently udevd doesn't create /dev/mapper/vg-XYZ as a symlink as it already exists as a file, which this feature enabled I'm hoping that "vgchange --activate" won't create /dev/mapp/vg0-XYZ itself but rather will let udev do it
2023-06-19 12:16:33 <minimal> psykose: well it is a buildtime option but there's also a lvm.conf setting, from a quick glance at the code it handles the situation where it cannot talk to udev (i.e. in a mdev-only system)
2023-06-19 12:17:47 <psykose> sure, enabled
2023-06-19 12:17:50 <psykose> feel free to just test from edge
2023-06-19 12:18:18 <minimal> psykose: you've enabled it? I have a locally built lvm2 that I'#m testing for it
2023-06-19 12:18:40 <psykose> ye
2023-06-19 12:18:46 <minimal> there was no need to enable it in Edge yet
2023-06-19 12:19:47 <psykose> it's not a very meaningful toggle
2023-06-19 12:31:36 <minimal> raspbeguy: and................./dev/mapper/vg0-waffle *IS* now created as a symlink to /dev/dm-0 ;-)
2023-06-19 12:31:56 <raspbeguy> minimal: 🎉️
2023-06-19 12:32:00 <raspbeguy> good job
2023-06-19 12:32:04 <minimal> no to do some testing with mdev-only to see that there's no ill effects
2023-06-19 12:32:08 <minimal> s/no/now/
2023-06-19 12:33:23 <minimal> CcxWrk: enabling this feature results in tools such as vgcreate being linked with libudev
2023-06-19 12:36:22 <psykose> (libudev is a 64kb sized binary that doesn't do anything)
2023-06-19 12:36:42 <psykose> just before someone crawls out of the woodwork
2023-06-19 12:37:27 <minimal> psykose: and with this enabled in lvm2 then if you're using LVM for rootfs then initramfs has libudev libs added to it
2023-06-19 12:37:43 <psykose> indeed
2023-06-19 12:44:34 <grawity> "via dbus" udev doesn't use dbus...
2023-06-19 12:44:37 <CcxWrk> It does too many things, most of them only tangentially related. Mostly wrapping access to various devices and hardware databases.
2023-06-19 12:45:21 <CcxWrk> grawity: It at least used to. Also several netlink interfaces.
2023-06-19 12:45:48 <CcxWrk> For device hotplug notifications and such.
2023-06-19 12:46:18 <grawity> oh, netlink is also bad now? missed the memo
2023-06-19 12:46:45 <minimal> psykose: the lvm.conf entry for udev_sync explains better what it does
2023-06-19 12:47:02 <psykose> grawity: everything is bad except what i like
2023-06-19 12:47:15 <grawity> hald used to use dbus, I literally can't remember when udev had any interaction with it, was that pre-systemd or post-systemd?
2023-06-19 12:47:29 <CcxWrk> I'm not saying something is good or bad. I'm asking whether the functionality requires communication with running userspace daemon.
2023-06-19 12:47:36 <grawity> oh huh, that was pre-hald even
2023-06-19 12:47:55 <psykose> minimal: hmm, does it default to =1?
2023-06-19 12:48:15 <psykose> it says if udev is not running and it's enabled (=1) then it will just wait forever
2023-06-19 12:48:31 <minimal> yes it defaults to "1" (aka enabled) which previously did nothing as the functionality wasn't built
2023-06-19 12:48:46 <psykose> get to testing then :-)
2023-06-19 12:48:57 <psykose> if it actually breaks then i assume it's a small patch to just default =0
2023-06-19 12:49:30 <minimal> yes I'm in the midst of testing different scenarios
2023-06-19 12:50:49 <psykose> there's a DEFAULT_UDEV_SYNC for it so i know where to toggle it if something doesn't work
2023-06-19 12:51:13 <WhyNotHugo> Would be nice to allow comments in /etc/apk/world. E.g.: ignore # and anything after it.
2023-06-19 12:53:53 <psykose> i think someone opened an issue for that before but i can't find it
2023-06-19 12:54:03 <psykose> it's not that simple because world is re-written after whatever transaction
2023-06-19 12:54:05 <psykose> and sorted
2023-06-19 12:54:17 <psykose> would require changing that entire thing
2023-06-19 13:02:35 <CcxWrk> Probably better to put comments in a configuration management system you use, if you have one.
2023-06-19 13:03:42 <WhyNotHugo> Oh, retaining them would be non-trivial, right
2023-06-19 13:06:01 <psykose> for typical short-comments virtual works ok
2023-06-19 13:06:06 <psykose> apk add -t .deps-for-xyz ..
2023-06-19 13:28:07 <WhyNotHugo> Smart, I will do that :)
2023-06-19 13:28:40 <WhyNotHugo> Installing 'docs' and 'libcap-getcap' doesn't install relevant man pages (from libcap-doc). I feel like it should.
2023-06-19 13:30:37 <OdaNobunaga[m]> The diskless Alpine Linux kernel doesn't contain the zfs module. What is the best way to get zfs support? Compile a custom kernel myself (there is a warning in the Alpine wiki though) or can I somehow add the zfs module via "update-kernel -p (add module or firmware package)" into the kernel? I'm rather a beginner, so please bear that in mind.
2023-06-19 13:33:14 <minimal> or install Alpine in Sys rather than diskless mode
2023-06-19 13:40:42 <raspbeguy> couldn't you install zfs and then use lbu to persist the change?
2023-06-19 13:41:09 <raspbeguy> (never used diskless mode other than install iso)
2023-06-19 13:41:28 <OdaNobunaga[m]> I have installed zfs, but the kernel module seems to be missing.
2023-06-19 13:41:33 <psykose> kernel module is zfs-lts
2023-06-19 13:41:48 <psykose> but i'm not sure how that works with diskless and updating aside from the update-kernel thing
2023-06-19 13:42:02 <psykose> (zfs-lts is for linux-lts)
2023-06-19 13:47:01 <OdaNobunaga[m]> I might have an idea I need to try out later. Thanks for the help!
2023-06-19 13:50:15 <zito> I am reading the docs and the wiki and I don't fully understand how do setup-alpine and setup-disk interact. Where can I read more about this? I am happy with setup-alpine except the partitioning part. I want just /boot on the hard drive, but swap and root on an USB thumb drive. And I want to use LUKS encryption (so I assume this means both / and swap have to be encrypted, not sure about /boot)
2023-06-19 13:51:04 <minimal> zito: the Alpine setup tools don't support that layout
2023-06-19 13:53:19 <zito> minimal: let me try asking differently, maybe this one will be better. if I do the manual partitioning and creation of filesystems using setup-disk, cryptsetup and friends, is it possible to call setup-alpine after that and tell it to work with that already prepared disk configuration?
2023-06-19 13:54:59 <minimal> zito: you can manually partition, format, and mount before running setup-disk, however I've never seen/tried a layout where /boot and / are no 2 different devices, really don't understand why you'd chose to do that
2023-06-19 13:55:42 <zito> minimal: my hard drive is old and has bad sectors, my goal is to get a bootable system by minimal interaction with the hard drive
2023-06-19 13:56:08 <zito> I also can't just install everything onto an USB thumb drive, because the computer is too old and BIOS doesn't support boot from USB
2023-06-19 13:56:19 <zito> didn't have success making it boot from USB using Plop either
2023-06-19 13:56:54 <zito> so I thought to try having just /boot on the hard drive and /, swap on an external USB thumb drive
2023-06-19 13:57:19 <humm> zito: try it
2023-06-19 13:57:28 <humm> and read the scripts
2023-06-19 13:58:12 <zito> couple of days ago I also asked here a related by different question, but got no answer unfortunately
2023-06-19 13:58:57 <zito> I asked if I can just install everything onto the external USB thumb drive (LUKS encrypted) then use the alpine install cd to boot the system from that USB thumb drive
2023-06-19 13:59:07 <minimal> zito: if your disk has bad sectors then things could get corrupted in future and stop working (e.g. updated kernel and initramfs files is /boot)
2023-06-19 13:59:07 <zito> I'm willing to try this option as well if it's possible
2023-06-19 13:59:31 <zito> minimal: true but the chances are smaller than using it for everything
2023-06-19 14:00:05 <minimal> smaller, but not close to zero, kernel updates occur semi-frequently
2023-06-19 14:00:14 <zito> I just have no idea where to get a replacement, it's an old IDE drive that's like 20 years old :D
2023-06-19 14:00:23 <zito> and I live in a shithole 
2023-06-19 14:06:10 <OdaNobunaga[m]> Would this work? I'm writing it down from memory: setup-alpine -> abort at setup-disk with ctrl+c -> mount -t FILESYSTEM /dev/YOURUSB /mnt/ -> mkdir /mnt/boot -> mount -t FILESYSTEM /dev/YOURHDD /mnt/boot -> mkdir /mnt/boot/efi -> setup-disk -m sys /mnt/
2023-06-19 14:06:44 <OdaNobunaga[m]> I forgot the swap. Nowadays, wouldn't a swapfile or zram suffice?
2023-06-19 14:07:39 <zito> I'm afraid I'd have to use swap
2023-06-19 14:07:47 <zito> only got like 400MB RAM there
2023-06-19 14:07:49 <minimal> you said your laptop is too old to boot off USB? but it has UEFI? UEFI should check all devices looking for ESP partitions
2023-06-19 14:07:57 <zito> it has no UEFI
2023-06-19 14:08:14 <minimal> oops, got you mixed up with the other person
2023-06-19 14:08:17 <zito> it's like 20 years old
2023-06-19 14:08:54 <zito> I don't think it even has SSE2 because I get "Invalid instruction" errors a lot when I try to start misc complex software
2023-06-19 14:09:41 <zito> alpine is the only thing that booted on that hardware
2023-06-19 14:54:13 <eletrotupi> this dumbass right here made his /boot too smol
2023-06-19 14:54:29 <eletrotupi> now I can't upgrade my linux kernel
2023-06-19 14:54:43 <eletrotupi> >ERROR: grub-2.06-r12.trigger: script exited with error 1
2023-06-19 14:54:50 <eletrotupi> >cat: write error: No space left on device
2023-06-19 14:54:58 <eletrotupi> any tips?
2023-06-19 14:55:19 <eletrotupi> hm, maybe I can force apk to not upgrade the kernel?
2023-06-19 14:55:58 <heat> you can either repartition your drive and enlargen /boot or attempt to free up some space manually in /boot (while not screwing up)
2023-06-19 14:56:20 <heat> or never ever upgrade the kernel but that's obviously not advisable
2023-06-19 14:56:55 <eletrotupi> damn
2023-06-19 14:58:12 <raspbeguy> fun fact, my OVH server ran something like 5 years without downtime because I accidentally screw the /boot partition and was too ignorant and lazy to learn how to fix it
2023-06-19 15:00:07 <eletrotupi> yeah, for some reason I've reserved 34mb for it
2023-06-19 15:00:12 <eletrotupi> and it apparently wasn't enough
2023-06-19 15:00:29 <eletrotupi> apk doesn't store old kernels there or something, right?
2023-06-19 15:00:56 <eletrotupi> at least, by looking at it it only has vmlinuz-lts, System.map-lts, config-lts
2023-06-19 15:01:05 <eletrotupi> doesn't appear to have anything versioned here
2023-06-19 15:01:33 <Ermine> old kernels are not stored
2023-06-19 15:01:43 <eletrotupi> well, that sucks
2023-06-19 15:01:48 <eletrotupi> I mean, for my case 
2023-06-19 17:08:06 <kitzman> Did anyone have trouble running alpine as a dom0 with linux-lts? I can't start uefi domUs
2023-06-19 17:13:25 <minimal> psykose: that sync_udev change is causing problems for mdev-only systems - trying to boot a VM with a LVM rootfs fails in the initramfs with "/dev/mapper/vg0-root: Can't open blockdev" and drops to emergency shell. I see /dev/dm-* do exist, but /dev/mapper/ only has the control file and /dev/vg0 doesn't exist. This mirrors what I saw when trying to "lvcreate" on a non-LVM-rootfs system with only mdev, the difference being in the 1st case 
2023-06-19 17:13:25 <minimal> there are no lvm tools (or /etc/lvm/lvm.conf) in the initramfs, so nlplug-findfs is using libdevicemapper directly
2023-06-19 20:32:15 <kitzman> Why isn't xen built with ovmf support on alpine?
2023-06-19 20:49:33 <minimal> "because" (insert random reason)
2023-06-19 20:50:00 <minimal> "because no-one decided to built it with OVMF support"
2023-06-19 20:52:10 <minimal> also #14970 refers to "Booting Xen via UEFI is broken on multiple Linux distributions and FreeBSD"
2023-06-19 21:48:47 <consus[m]> psykose: Also CONFIG_FUNCTION_TRACER is missing on ppc64le
2023-06-19 21:56:55 <psykose> minimal: does defaulting it to 0 fix it or nah
2023-06-19 22:00:44 <minimal> psykose: am just testing that currently
2023-06-19 22:02:28 <minimal> appears to have made no difference
2023-06-19 22:13:43 <psykose> hmm
2023-06-19 22:14:12 <minimal> I need to track down why the /dev/vg0/LVNAME creation is failing
2023-06-20 01:34:15 <baalajimaestro> what does it indicate when a file named "oom" gets created
2023-06-20 01:37:20 <psykose> that an application made a file named oom
2023-06-20 02:39:17 <baalajimaestro> Nope. I run just docker containers. And this is getting created on home directory, which isn't mounted to any of the containers. I'm seeing a file named oom get created every once in a while. Either it should be zfs crying about something, or idk if it's something from the os
2023-06-20 02:40:49 <psykose> 98% sure nothing on linux makes a file named oom in the home directory
2023-06-20 02:41:02 <psykose> so, some random application, ...
2023-06-20 02:41:45 <psykose> don't think zfs does either
2023-06-20 02:43:57 <baalajimaestro> That's bizarre af now. I switched oses as well, and replicated it yet :|
2023-06-20 02:43:58 <baalajimaestro> Besides all this, I am getting panics or idk what that is, on random times which neither logged on previous boot logs nor show up at a replica table time
2023-06-20 03:12:07 <abby> did you `touch oom`
2023-06-20 03:53:03 <Saijin_Naib[m]> minimal: any luck with this? My desktop just failed to boot with that error after upgrading today
2023-06-20 03:53:21 <psykose> how recently did you upgrade
2023-06-20 03:53:34 <psykose> was it to -r8 or -r9 lvm
2023-06-20 03:53:40 <Saijin_Naib[m]> About 30min ago
2023-06-20 03:54:05 <psykose> hm
2023-06-20 03:54:07 <psykose> okie
2023-06-20 03:54:10 <Saijin_Naib[m]> Cant recall exactly which revision 
2023-06-20 03:54:24 <Saijin_Naib[m]> Going to do it to my laptop and see what happens
2023-06-20 03:54:30 <Saijin_Naib[m]> Is this recoverable? 
2023-06-20 03:55:06 <psykose> r10 should be fine after a chrooted upgrade + mkinitfs regeneration but that will take a while to build
2023-06-20 03:55:42 <Saijin_Naib[m]> Mm, alright, thank you 
2023-06-20 03:56:05 <Saijin_Naib[m]> Almost had another heart attack when it dropped me to the recovery shell
2023-06-20 03:57:26 <psykose> actually my fault this time
2023-06-20 03:58:07 <Saijin_Naib[m]> Hey, it wasn't me? Alright! 
2023-06-20 04:58:59 <tm512> trying to get umurmur running in this Alpine container, but trying to start it through openrc prints out the error "start-stop-daemon: did not create a valid pid in `/run/umurmurd/umurmurd.pid'"
2023-06-20 04:59:20 <tm512> though the daemon does start up just fine, openrc still thinks it's stopped
2023-06-20 04:59:37 <ikke> what kind of container?
2023-06-20 04:59:52 <tm512> it's an LXC container
2023-06-20 04:59:54 <ikke> ok
2023-06-20 05:01:08 <tm512> http://sprunge.us/yqHaDq
2023-06-20 05:02:20 <ikke> tm512: might be that it's a timing issue
2023-06-20 05:02:37 <psykose> the default thing does wait 100ms only
2023-06-20 05:03:35 <tm512> is that configurable? this is a pretty fast system, so it's a bit weird, unless umurmur just has a particularly slow startup
2023-06-20 05:03:55 <psykose> fc810262dc290a3917d8afefc6e5c2ecc6c148ef
2023-06-20 05:04:02 <psykose> just delete that and should be fine if you don't want to wait for the build
2023-06-20 05:07:02 <tm512> I'm not really that familiar with openrc. where would I find that? if the builds are quick I don't really mind waiting though, especially since the daemon itself is running fine, just in a way that confused openrc
2023-06-20 05:08:51 <psykose> etc/init.d/umurmur
2023-06-20 05:10:53 <tm512> ah I see, thanks
2023-06-20 05:21:50 <tm512> the updated binary got pushed out pretty pretty quick, seemingly works fine now
2023-06-20 08:50:52 <malmi70[m]> I ecrypted my whole disk during the install process. I want to shrink it but I can't figure how. I booted live Linux and tried to use gparted and cgdisk. They can decrypt the Alpine disk but doesn't allow to shrink/resize. What I am missing here?
2023-06-20 08:53:01 <lopid> luks?
2023-06-20 08:56:13 <malmi70[m]> Lvm
2023-06-20 08:58:07 <lopid> luks encrypted?
2023-06-20 09:07:16 <malmi70[m]> The default in Alpine. I chose encrypted, lvm, sys
2023-06-20 09:08:37 <malmi70[m]> At least luksOpen opens it in terminal
2023-06-20 09:11:39 <lopid> there're a few guides here: https://search.brave.com/search?q=shrink+luks+partition
2023-06-20 09:12:47 <malmi70[m]> Thanks a lot for nothing. The exact search parameters I used in the exactly the same search engine. You're the mvp of the day!
2023-06-20 09:14:40 <lopid> perhaps if you were a bit more explicit about what the problem is. "doesn't allow" is quite vague
2023-06-20 09:25:31 <quinq> That's not really a good attitude for getting help :)
2023-06-20 12:31:02 <raspbeguy> wow that escalated quickly
2023-06-20 12:44:18 <minimal> 60% of the time it works every time
2023-06-20 13:24:51 <raspbeguy> hi minimal. Di you have any idea on how to progress on the LVM issue?
2023-06-20 13:59:59 <shrizza> Hm, was there a change recently that would cause boot to fail after an upgrade?
2023-06-20 14:04:59 <ikke> That's too vague to answer 
2023-06-20 14:05:09 <ikke> Fail in what way?
2023-06-20 14:05:41 <minimal> shrizza: on Edge? with LVM?
2023-06-20 14:06:08 <raspbeguy> > Hm, was there a change recently that would cause boot to fail after an upgrade?
2023-06-20 14:06:08 <raspbeguy> are you using LVM on your root volume? Are you usind eudev or mdev?
2023-06-20 14:06:10 <minimal> raspbeguy: not so far, am trying to think what to test next
2023-06-20 14:07:26 <shrizza> Yeah, I just noticed a couple others mentioning it in my scrollback. Seems that LVM is failing to mount my root/swap volumes.
2023-06-20 14:07:47 <shrizza> And yes, edge.
2023-06-20 14:08:33 <raspbeguy> shrizza, is you manage to update now that should be fixed now
2023-06-20 14:08:53 <shrizza> Not sure about the mdev/eudev thing.
2023-06-20 14:10:09 <raspbeguy> if you don't know then you're proaby using mdev
2023-06-20 14:13:05 <minimal> shrizza: so yes it seems you updated to a "broken" lvm2 package that has since been fixed
2023-06-20 14:19:12 <shrizza> Someone paste me The Ideal mkinitfs incantation while I USB recover.
2023-06-20 14:21:17 <psykose> apk fix mkinitfs
2023-06-20 14:21:53 <minimal> they'd have to update the lvm2/device-mapper packages first
2023-06-20 14:31:39 <shrizza> OK, did the upgrade (pulled down new lvm2), and apk fix mkinitfs. But now I just boot to a blinking cursor. I try typing something to see if I'm just blind at console, and it beeps on keypress after like 10 silent key inputs.
2023-06-20 14:32:32 <minimal> shrizza: which bootloader?
2023-06-20 14:33:01 <psykose> that sounds like something completely different broken now
2023-06-20 14:33:04 <shrizza> Grub I think.
2023-06-20 14:33:24 <psykose> did you mount all the usual mounts for usb recovery
2023-06-20 14:33:30 <minimal> so when Grub menu appears edit the boot option and remove "quiet" from the command-line
2023-06-20 14:33:31 <psykose> /dev etc
2023-06-20 14:35:04 <minimal> shrizza: did you only update lvm2 package or device-mapper/device-mapper-libs/device-mapper-event-libs also?
2023-06-20 14:35:46 <shrizza> I apk added lvm2 so I could vgchange -a y my volumes, then mounted that to /mnt/ and /dev/sda1 (my boot part) to /mnt/boot, chrooted into /mnt, did the apk upgrade (new lvm2 got downloaded), and finally did apk fix mkinitfs.
2023-06-20 14:36:29 <minimal> shrizza: and the "apk fix" also updating device-mapper packages?
2023-06-20 14:36:45 <psykose> you forgot all the mounts then
2023-06-20 14:37:34 <minimal> yeah the bind-mounts for /dev, /sys, /proc etc before chrooting
2023-06-20 14:38:02 <psykose> mkinitfs does some stuff based on detected devices and w/e
2023-06-20 14:38:11 <psykose> so it will skip a ton of stuff without a full chroot
2023-06-20 14:39:06 <psykose> mostly just need these 3 things https://wiki.archlinux.org/title/chroot#Using_chroot
2023-06-20 14:43:35 <shrizza> I have those three. I'm using a script on the Alpine Linux Chroot wiki page.
2023-06-20 14:44:14 <psykose> not that then
2023-06-20 14:45:49 <shrizza> Removing quiet from boot args in grub conf did not appear to do anything.
2023-06-20 14:46:19 <minimal> shrizza: from grub conf? or from entry in menu you see during boot?
2023-06-20 14:46:49 <shrizza> I actually don't even see the menu at boot.
2023-06-20 14:47:16 <minimal> ah, you have it hidden then
2023-06-20 14:48:50 <shrizza> I'm not particularly married to grub if switching to syslinux would move me further along this problem.
2023-06-20 14:49:00 <shrizza> I could try that.
2023-06-20 14:50:03 <minimal> switching bootloader is unlikely to have an effect on your problem
2023-06-20 14:51:05 <minimal> is your /boot inside LVM?
2023-06-20 14:51:30 <shrizza> No.
2023-06-20 14:51:57 <minimal> ok. If it was then you coludn't switch as Grub supports LVM directly but Syslinux (from memory) does not
2023-06-20 14:53:50 <shrizza> In times like this I'm wishing none of my crap is in LVM.
2023-06-20 14:54:26 <minimal> there's nothing wrong/bad about LVM in general
2023-06-20 14:54:55 <shrizza> I don't know why I setup my daily driver like this. All the Alpine VMs I setup are always non-LVM.
2023-06-20 15:12:45 <eletrotupi> hmm, I believe that my alpine box is broken now 😞
2023-06-20 15:12:59 <eletrotupi> it's stuck on loading initial ramdisk
2023-06-20 15:13:25 <eletrotupi> is there's a way to debug it further or should I just reinstall it?
2023-06-20 15:15:47 <shrizza> So I notice when I do "apk fix mkinitfs" there's a warning: depmod: WARNING: could not open modules.builtin.modinfo at /tmp/mkinitfs.d1FzMi/lib/modules/4.19.176-0-vanilla: No such file or...
2023-06-20 15:16:03 <eletrotupi> I think it's related to that failed Linux kernel upgrade from yesterday
2023-06-20 15:17:05 <psykose> ... 4.19.176-0-vanilla?
2023-06-20 15:34:42 <minimal> eletrotupi: linux-vanilla was removed from Alpine 3.11.0 onwards. Are you the same person that was in the Libera alpine-linux channel recently?
2023-06-20 15:38:34 <psykose> i think you read the wrong message
2023-06-20 15:40:55 <minimal> ah, confused by the mixed posts of eletrotupi and shrizza
2023-06-20 15:41:19 <shrizza> So I removed linux-vanilla and installed linux-edge, and did apk fix mkinitfs again but still same result...
2023-06-20 15:41:41 <minimal> skrizza: linux-vanilla does not exist since Alpine 3.11.0 so how did you have it on an Alpine Edge system?
2023-06-20 15:42:24 <shrizza> I don't know... This laptop's been through a lot.
2023-06-20 15:43:50 <shrizza> I also switched to syslinux, removed quiet from boot args, and commented out "menu hidden". Still just a blinking cursor though. Is this just hosed?
2023-06-20 15:44:30 <minimal> psykose: there was someone on Libera alpine-linux recently claiming there were running Alpine 3.18.2 but said they had a initramfs-vanilla file - seemed they'd originally installed 3 years ago and upgraded since (but never indicated how/how many times they upgraded)
2023-06-20 15:45:54 <shrizza> I have 3.15 main/community pinned for like 2 packages but everything else is edge.
2023-06-20 15:46:05 <minimal> shrizza: why did you install linux-edge rather than linux-lts? why did you have linux-vanilla installing  on an Alpine Edge system in the first place? The Alpine 3.11.0 Release Notes indicate to install linux-lts as part of upgrading to 3.11.0
2023-06-20 15:46:28 <minimal> shrizza: that is NOT supported, you cannot mix edge repos with another release
2023-06-20 15:47:39 <minimal> shrizza: sounds like you have some sort of Frankenstein installation
2023-06-20 15:47:50 <shrizza> No clue, I just apk upgrade the things.
2023-06-20 15:48:07 <shrizza> So I'll try linux-lts instead.
2023-06-20 15:48:34 <minimal> just trying "random" things is not a good way to debug a problem
2023-06-20 15:50:00 <shrizza> OK but wait I'm getting further after removing USB HDD from my BIOSs boot devices. Not sure why that would matter when I have the boot USB removed...
2023-06-20 15:50:26 <minimal> as mkinitfs failed with a reference to a 4.19.176-0-vanilla kernel then it seems you no longer have a valid initramfs file
2023-06-20 15:50:51 <minimal> and you've installed linux-edge (which is not 4.19.176)
2023-06-20 15:51:08 <shrizza> OK, everything works again.
2023-06-20 15:51:39 <shrizza> I alreaded apk del'd linux-vanilla before adding linux-edge.
2023-06-20 15:52:02 <shrizza> And did mkinitfs after that.
2023-06-20 15:52:37 <minimal> and mkinitfs gave that error about 4.19.176-0-vanilla then?
2023-06-20 15:53:13 <shrizza> That was before I removed linux-vanilla.
2023-06-20 15:55:15 <shrizza> I agree that my kernel situation was frankensteiny but I don't know man like I said I just apk upgrade things and they tend to just work.
2023-06-20 15:55:33 <minimal> what is on "USB HDD"? could there be an Alpine install on there?
2023-06-20 15:56:08 <minimal> BIOS or UEFI?
2023-06-20 15:56:16 <shrizza> Yeah it was the USB boot stick I was using.
2023-06-20 15:56:21 <shrizza> BIOS.
2023-06-20 15:56:58 <shrizza> Having it unplugged with the USB HDD prioritized was causing problems.
2023-06-20 15:57:22 <shrizza> That was a large source of my confusion while troubleshooting this.
2023-06-20 15:58:38 <shrizza> Everything appears to be working beautifully again so thanks!
2023-06-20 16:01:01 <shrizza> eletrotupi: If our issues are similar, you gotta boot into Alpine Linux install USB, mount your LVM volumes and /boot partition (into say /mnt), then chroot into /mnt, apk upgrade the fixed lvm2 packages, and run apk fix mkinitfs.
2023-06-20 16:18:19 <eletrotupi> shrizza: yeah, but I can't seem to mount the decrypted LUKS partition
2023-06-20 16:18:43 <eletrotupi> mount complains about an invalid argument 
2023-06-20 16:18:57 <eletrotupi> mount /dev/mapper/partition /mnt
2023-06-20 16:19:58 <shrizza> Oh, I just went through scrollback and I see more context for your case. Limited space on /boot is compounding?
2023-06-20 16:20:46 <eletrotupi> yeah, I think that during the upgrade 
2023-06-20 16:21:12 <eletrotupi> it failed to upgrade and run mkinitfs
2023-06-20 16:21:35 <eletrotupi> then I rebooted now and it got stuck on loading initial ramdisk 
2023-06-20 16:22:00 <eletrotupi> but now I'm confused, since appears that I can mount the drive anymore
2023-06-20 16:22:24 <eletrotupi> I mean, luksOpen worked fine
2023-06-20 16:22:28 <shrizza> So /dev/mapper/partition contains your luks open'd partition?
2023-06-20 16:22:39 <eletrotupi> but mounting seems to be problematic as well
2023-06-20 16:22:42 <eletrotupi> yeah
2023-06-20 16:23:01 <shrizza> What's the exact complaint from mount?
2023-06-20 16:23:09 <eletrotupi> Invalid argument 
2023-06-20 16:23:14 <eletrotupi> tried with -v as well
2023-06-20 16:23:19 <minimal> eletrotupi: "mount -t ext4 /.dev/mapper/partition /mnt" ?
2023-06-20 16:23:25 <eletrotupi> but it outputed the same thing
2023-06-20 16:23:30 <eletrotupi> hm lemme try
2023-06-20 16:23:51 <eletrotupi> oh, there we go
2023-06-20 16:23:52 <eletrotupi> nice
2023-06-20 16:24:10 <eletrotupi> ok, now for the boot part
2023-06-20 16:24:39 <shrizza> Yeah, you gotta copypasta the script on Alpine Linux's Chroot wiki to chroot the right way.
2023-06-20 16:24:42 <minimal> Busybox mount tends to need you to tell it the type of filesystem
2023-06-20 16:25:26 <shrizza> The script is named 'start-chroot'.
2023-06-20 16:25:55 <eletrotupi> ok, thanks shrizza 
2023-06-20 16:26:04 <eletrotupi> I'll eat something then I'll try it again
2023-06-20 16:26:20 <eletrotupi> I saw also a guy saying that he just wiped his /boot partition 
2023-06-20 16:26:23 <eletrotupi> recreated 
2023-06-20 16:26:31 <eletrotupi> and moved on 
2023-06-20 16:27:45 <shrizza> I had the mindset of as long as my .ash_history survives I don't care about anything else.
2023-06-20 16:28:08 <shrizza> Bit of an exaggeration but not far-fetched.
2023-06-20 16:30:52 <minimal> eletrotupi: where did you see someone say he just wiped his /boot etc?
2023-06-20 16:35:39 <shrizza> I mean, in a sense I mostly wiped my /boot by doing apk del linux-vanilla, apk add linux-edge, and apk fix mkinitfs.
2023-06-20 16:39:32 <minimal> shizza: I assume the guy he was telling you about wasn't you ;-)
2023-06-20 17:07:35 <eletrotupi> ok, I'm back
2023-06-20 17:07:51 <eletrotupi> minimal: on some other forum
2023-06-20 17:08:22 <eletrotupi> hm, I wonder if one could use arch-chroots scripts to chroot
2023-06-20 17:08:30 <eletrotupi> I think I actually did that last time
2023-06-20 17:16:37 <uuidNuniq> eletrotupi: you just create your own chroot
2023-06-20 17:17:05 <uuidNuniq> learn from here: https://wiki.gentoo.org/wiki/Chroot
2023-06-20 17:17:25 <uuidNuniq> and then, you make a simple script ... and use that to do the mounting preparation. 
2023-06-20 17:18:29 <eletrotupi> ok, chrooted
2023-06-20 17:18:33 <eletrotupi> apk del linux-lts
2023-06-20 17:19:44 <eletrotupi> which left 5.4MB on /boot then apk add linux-lts 
2023-06-20 17:19:56 <eletrotupi> which fails when executing mkinitfs 
2023-06-20 17:19:59 <eletrotupi> fucking hell
2023-06-20 17:20:07 <eletrotupi> I think I might unmount everything 
2023-06-20 17:20:37 <eletrotupi> then free, I dunno, 100MB from my /home partition 
2023-06-20 17:20:48 <eletrotupi> so I can shove it unto the /boot partition again 
2023-06-20 17:36:52 <uuidNuniq> eletrotupi: what's in your /boot? ls -lh /boot
2023-06-20 17:52:22 <eletrotupi> nothing much, efi and grub
2023-06-20 17:54:41 <eletrotupi> but I'm going to do a full reinstall anyway
2023-06-20 17:58:55 <eletrotupi> dunno why the /boot partition was so smol
2023-06-20 18:00:02 <ikke> Because it could for a long time 
2023-06-20 18:02:25 <singham> I installed alpine on lime2
2023-06-20 18:02:33 <singham> With a single FAT32 partition
2023-06-20 18:03:29 <singham> During alpine-setup, I enabled partition to be mounted at /media/mmcblk0p1 which is not where I want it
2023-06-20 18:03:42 <singham> I want the 14.8GB partition as root
2023-06-20 18:04:32 <singham> When I took out the sd card and mounted the partition at sda1 (14.8GB) one, it doesn't show any root directories
2023-06-20 18:04:50 <singham> But the normal contents of tarball of apk boot, etc.
2023-06-20 18:04:58 <singham> I need your help
2023-06-20 18:05:12 <singham> Why can't I see the filesystems in the mounted partition?
2023-06-20 18:05:20 <singham> on another system?
2023-06-20 18:06:03 <singham> In fstab I changed the partition mounting from /media/mmcblk0p1 to / and I couldn't boot anymore!
2023-06-20 18:57:02 <singham> Please post the answers. I'll check the logs.
2023-06-20 19:22:24 <eletrotupi> aight, managed to get back where I was
2023-06-20 19:22:33 <eletrotupi> only issue (as usual) is sound
2023-06-20 19:22:46 <eletrotupi> apparently aplay -l doesn't detect any soundcards
2023-06-20 19:23:02 <eletrotupi> (and I thought I would have to battle again with nvidia)
2023-06-20 19:27:39 <bl4ckb0ne> just use pipewire
2023-06-20 19:27:45 <eletrotupi> yeah
2023-06-20 19:27:49 <eletrotupi> I have that installed
2023-06-20 19:28:03 <eletrotupi> but wpctl status lists only a Dummy Output
2023-06-20 19:28:17 <eletrotupi> since it prolly can't find a sound card (?)
2023-06-20 19:28:29 <eletrotupi> might be a kernel module missing, I guess?
2023-06-20 19:29:57 <bl4ckb0ne> not likely
2023-06-20 19:30:07 <bl4ckb0ne> is your user in the audio group
2023-06-20 19:30:12 <eletrotupi> yes
2023-06-20 19:30:21 <eletrotupi> same with root (since the alsa wiki page says it needs to)
2023-06-20 19:31:54 <eletrotupi> bah, I kinda remember having to tinker with SOF firmware binaries on this lappy
2023-06-20 19:36:08 <eletrotupi> that was it
2023-06-20 19:48:05 <psykose> what is the tinkering
2023-06-21 00:28:46 <shrizza> eletrotupi: Oh no, did you end up reinstalling?
2023-06-21 00:28:57 <eletrotupi> yeah
2023-06-21 00:29:31 <eletrotupi> only considered that due to the fact that this is a work machine & we use docker for all development
2023-06-21 00:29:40 <eletrotupi> so it was 'cheaper' to recreate everything
2023-06-21 00:30:07 <eletrotupi> but while I had it mounted I took the opportunity to double check if anything was missing (config files, pending git changes, etc)
2023-06-21 00:30:29 <eletrotupi> I'm using it right now normally
2023-06-21 00:30:45 <eletrotupi> also, setup-disks slapped a nice 300mb partition for /boot now
2023-06-21 00:30:47 <eletrotupi> :D
2023-06-21 00:31:50 <shrizza> Ah, well at least you seem to be on top of it. What laptop BTW?
2023-06-21 00:34:26 <shrizza> My daily driver was installed in 2018 I think and this is my /boot situation:
2023-06-21 00:34:32 <shrizza> /dev/sda1          92M   34M   52M  40% /boot
2023-06-21 00:35:19 <minimal> shrizza: yes setup-disk was changed in past to size /boot differently
2023-06-21 00:35:22 <eletrotupi> it's a dell vostro 15 5510
2023-06-21 00:44:33 <tm512> seems like Alpine is a little bit weird with urxvt? I installed rxvt-unicode-terminfo which got things mostly working, but if I open up tmux, "10;rgb:d900/d900/d900]11;rgb:0d00/0d00/0d00" is spat out to stdin
2023-06-21 00:44:58 <tm512> my TERM is set to rxvt-unicode-256color
2023-06-21 01:07:39 <tm512> hmm, one thing I just noticed is that inside tmux, on Alpine TERM gets set to tmux-256color, whereas on Ubuntu, Void, and DragonflyBSD, where tmux does not have this issue, TERM gets set to screen on Ubuntu and screen-256color on the others
2023-06-21 01:10:17 <tm512> adding this to ~/.tmux.conf does not change things though: set -g default-terminal "screen-256color"
2023-06-21 01:17:35 <psykose> tmux having a term of tmux sounds correct
2023-06-21 01:23:15 <tm512> yeah, but having escape codes input onto the command line isn't correct
2023-06-21 01:23:53 <psykose> certainly not
2023-06-21 01:25:43 <tm512> doesn't happen if I nest tmux within a tmux session. such a weird issue, never seen this happen before and I've been using tmux for a while
2023-06-21 01:27:05 <tm512> gonna see if this happens in xterm or just urxvt
2023-06-21 01:27:20 <tm512> 
ACTION apparently doesn't even have xterm installed

2023-06-21 01:27:29 <psykose> doesn't seem to
2023-06-21 01:27:33 <psykose> classic urxvt moment
2023-06-21 01:27:50 <psykose> i mistakenly actually went to check if there was something interesting in the git history
2023-06-21 01:27:52 <psykose> but you see
2023-06-21 01:27:54 <psykose> it's urxv
2023-06-21 01:27:57 <psykose> so it's cvs
2023-06-21 01:30:39 <tm512> indeed, launching this in actual xterm doesn't cause this issue, but TERM=xterm tmux still has the issue, so I guess it's not terminfo-related
2023-06-21 01:31:07 <tm512> if it's urxvt though I'm not sure why it behaves on every other system I can try
2023-06-21 01:33:43 <psykose> it's probably just ncurses
2023-06-21 01:36:00 <psykose> hm, nah prolly not
2023-06-21 01:36:08 <psykose> there's also https://gitlab.alpinelinux.org/alpine/aports/-/issues/14525 as a separate issue where it has a bunch of newlines
2023-06-21 01:38:11 <tm512> oh yeah, that issue. I dunno that it's related though
2023-06-21 01:40:42 <psykose> not very
2023-06-21 01:42:25 <tm512> well, I guess in the meantime I'll just keep using tmux on the host rather than in the Alpine container, since that works fine still
2023-06-21 01:43:35 <tm512> I just kinda wanted to make an openrc service script to launch this minecraft server inside of a tmux session, keep it entirely within the container and have it autostart
2023-06-21 01:44:19 <tm512> I kinda need it to be in tmux so that I can attach to the server console whenever I need to
2023-06-21 02:00:58 <shrizza> Whoa, just noticed all the xterm talk here as I'm getting ready to paste a screenshot of some weirdness I'm seeing in xterm after upgrading yesterday.
2023-06-21 02:01:02 <shrizza> https://imgur.com/a/8Uyfwz1
2023-06-21 02:01:44 <shrizza> As is mentioned in the description: Some Unicode glyphs disappear after screen redraw. Ctrl-L makes everything appear properly again, but the same glyphs disappear when the window is hidden and redisplayed (such as through change of tile/workspace focus).
2023-06-21 02:26:47 <Saijin_Naib[m]> Minimal and psykose, you beautiful people, thank you for helping me get my systems running again after the LVM incident
2023-06-21 02:27:04 <Saijin_Naib[m]> Now polkit is broken at XFCE login, but that I believe to be unrelated 
2023-06-21 02:27:41 <psykose> classic
2023-06-21 02:27:57 <shrizza> Yeah, thanks dudes.
2023-06-21 02:28:13 <Saijin_Naib[m]> Is polkit normally a pain in the booty? 
2023-06-21 02:45:20 <Saijin_Naib[m]> Do I need shadow-login or until-Linux-login if I'm running elogind OpenRC service? 
2023-06-21 04:19:24 <shrizza> I discovered that toggling on "Line-Drawing Characters" in xterm's right-button menu fixes the above glitch. Now trying to figure out how to have that toggled on by default.
2023-06-21 12:55:04 <minimal> raspbeguy, psykose: regarding /dev/mapper/ symlinks, I'm starting to wonder the point of trying to "fix" this as it can only be "fixed" in very specific situations in Alpine
2023-06-21 13:15:47 <txnintn> I have been trying different ways to install alpine on a Raspberry Pi with lvm over raid1, with this last method, I lose kernal message during boot and shut down after the initial install.  Google was no help as usual. I will be out of office until 4pm cst  Thanks
2023-06-21 14:41:13 <raspbeguy> minimal, well for me the only way to fix it is to patch busybox mdev I guess
2023-06-21 14:43:01 <minimal> raspbeguy: well first it depends on exactly *what* is to be fixed - /dev/mapper softlinks when using LVM with udev? or also softlinks when using LUKS with udev? softlinks in general whether using udev or mdev?
2023-06-21 14:43:43 <minimal> raspbeguy: mdev isn't creating the files in /dev/mapper, lvm/devmapper and cryptsetup are so there's nothing really to fix in mdev
2023-06-21 14:44:56 <raspbeguy> I mean, implement udev_sync feature in mdev
2023-06-21 14:45:12 <raspbeguy> don't know if that's feasible but this is my only idea
2023-06-21 14:45:24 <minimal> raspbeguy: you mean turn mdev into a clone of udev? mdev does not work the same way as udev
2023-06-21 14:45:40 <raspbeguy> I see
2023-06-21 14:46:40 <raspbeguy> or somehow disable lvm udev_sync feature in initfs
2023-06-21 14:47:26 <raspbeguy> I guess this bug should be reported to busybox devs, maybe they'll know how to address it
2023-06-21 14:47:39 <minimal> that doesn't "solve" the problem. I have udev_sync disabled when the initramfs runs, which means that /dev/mapper/ entries are created as *files* by libdevmapper (via nlplug-findfs)
2023-06-21 14:47:58 <minimal> raspbeguy: why busybox? it has nothing to do with busybox/mdev
2023-06-21 14:48:13 <raspbeguy> oh I thought mdev was part of busybox
2023-06-21 14:48:30 <minimal> yes mdev is part of busybox but there is no problem with mdev
2023-06-21 14:48:47 <raspbeguy> then report the issue to lvm devs
2023-06-21 14:49:04 <minimal> what "bug" exactly to report for mdev? It does NOT create /dev/mapper entries at all, other programs do
2023-06-21 14:49:23 <raspbeguy> so what is exactly creating the bug?
2023-06-21 14:49:49 <maria> hi friends, I'm setting up my laptop with iwd for wifi by following the wiki entry. unfortunately, whether my wifi works or not seems to be a race condition as it changes each boot. dbus and iwd start before networking but, for example, I get output "ip: ioctl xxxxxx failed: no such device\n udhcp: ioctl xxxxxx failed: no such device\n ifup: failed to change interface WLAN state to 'up'". 
2023-06-21 14:49:49 <maria> restarting the service post boot fixes it but is rather annoying. should I have iwd handle the networking itself like is noted after the /etc/network/interfaces file contents section?
2023-06-21 14:49:49 <minimal> raspbeguy: report an issue to lvm devs that they don't offer functionality to "sync" with mdev which doesn't offer such sync functionality in the 1st place?
2023-06-21 14:50:16 <minimal> raspbeguy: what buy? there is a difference in *behaviour* between mdev and udev, that's not a bug
2023-06-21 14:50:53 <raspbeguy> all I know is that there is a fatal issue when using lvm with udev_sync with mdev
2023-06-21 14:51:01 <raspbeguy> I know literally nothing else
2023-06-21 14:51:33 <minimal> mdev as set up in Alpine does not have a daemon (whereas udev does). Now there is a daemon for mdev that can be enabled (if Busybox package is rebuilt) but I don't know if it would "handle" any sync interface
2023-06-21 14:51:48 <raspbeguy> so tell LVM people that when running lvm with udev_sync build option, without even using the ferature, that creates a bug
2023-06-21 14:52:32 <minimal> what "fatal issue"? there is a difference in behaviour between Alpine and many/most other distros which is due, in part, to Alpine using mdev in the initramfs even if udev is used later in boot
2023-06-21 14:53:41 <minimal> raspbeguy: what bug? I don't see any bug with lvm udev_sync, I have it "working" on a mdev-only VM (as in the system boots) but the /dev/mapper entries are files not symlinks
2023-06-21 14:54:37 <raspbeguy> The bug in my opinion is that lvm behaviour when built without the feature should be the same as lvm build with it, with the feature disabled in the config
2023-06-21 14:55:04 <minimal> and the behaviour is the same
2023-06-21 14:55:30 <minimal> but that doesn't solve the "issue" of file vs symlink
2023-06-21 14:57:24 <raspbeguy> In my case, /dev/mapper is not populated at initfs step, because I don't activates the volume at boot
2023-06-21 14:57:40 <raspbeguy> so I need the udev_sync option
2023-06-21 14:58:12 <minimal> aargh, sorry this conversation seems to be going nowhere
2023-06-21 14:58:37 <minimal> yes you need the udev_sync option for your specific situation where you are using udev and it works as expect in that very specific situation
2023-06-21 14:59:17 <raspbeguy> another idea would be to have another piece of software that to the symlinks
2023-06-21 14:59:37 <raspbeguy> so we can let lvm built without the feature
2023-06-21 14:59:39 <minimal> however it does not work as "expected" (where this means /dev/mapper entries should be symlinks) in other situations
2023-06-21 15:00:49 <minimal> if you build lvm2 without the udev_sync feature then we are back to the start - that lvm2/device-mapper will create *FILES* in /dev/mapper so unless this "another piece of software" removes those files and replaces them with symlinks will resolve nothing
2023-06-21 15:01:22 <raspbeguy> true
2023-06-21 15:01:57 <minimal> lvm2/device-mapper creates FILES at all times *unless* the udev_sync feature is enabled and it is able to talk to udev to arrange for udev to create the entries as symlinks instead
2023-06-21 15:01:58 <raspbeguy> so build it WITH the feature and let the magic piece of software handle the sync stuff
2023-06-21 15:03:19 <minimal> how? lvm2/device-mapper "discovers" udev to talk to it so the magic software would have to somehow ensure they talk to it instead of udev and then udev events probably come into the picture as well
2023-06-21 15:03:49 <minimal> and that doesn't handle the mdev-only scenarios
2023-06-21 15:04:14 <minimal> and doesn't handle udev scenarios where either LVM or LUKS is being used for the device containing the rootfs
2023-06-21 15:05:13 <raspbeguy> it would if that program was part of the initfs I guess
2023-06-21 15:05:30 <minimal> nope as initramfs only uses mdev not udev
2023-06-21 15:05:36 <raspbeguy> but again, I speak about hypothetic stuff that I don't know
2023-06-21 15:06:34 <raspbeguy> not taking about udev right now
2023-06-21 15:06:57 <raspbeguy> my fancy idea would be to have something between lvm and mdev
2023-06-21 15:07:22 <minimal> as I said mdev (as used by Alpine) has no "sync" interface
2023-06-21 15:08:11 <raspbeguy> I don't even know if that something have to speak to mdev
2023-06-21 15:09:22 <minimal> as I explained the lvm buildin functional is basically "either (a) if udev_sync is enabled and I can talk to udev I ask it to create mapper entries (which is will do as symlinks), otherwise (b) I create files in mapper directory
2023-06-21 15:09:38 <raspbeguy> correct
2023-06-21 15:09:54 <raspbeguy> that something would make believe to lvm that it's udev
2023-06-21 15:10:36 <minimal> mdev (a) currently has no rules to create mapper entries, and (b) lvm sees either udev_sync is disabled or it cannot contact udev and goes ahead and creates files itself
2023-06-21 15:11:15 <minimal> raspbeguy: which as i said before means rewriting/reconfiguring mdev to look/act like udev
2023-06-21 15:11:26 <minimal> mdev is NOT intended to be a clone of udev
2023-06-21 15:11:28 <raspbeguy> also, another idea would be to fill a feature request to lvm for an option to create symlinks in the case it doesn't find udev
2023-06-21 15:12:29 <raspbeguy> (or patch it ourself)
2023-06-21 15:12:32 <minimal> and a similar feature request for cryptsetup as well I guess
2023-06-21 15:12:46 <raspbeguy> yes
2023-06-21 15:13:40 <minimal> which all goes back to my original point several hours ago, what *exactly* is the extent of the intention/wish of what to "fix"?
2023-06-21 15:14:28 <minimal> to modify lvm2, cryptsetup, perhaps mkinitfs and other packages, to have both mdev and udev situations create symlinks for LVM and LUKS devices?
2023-06-21 15:15:46 <raspbeguy> that seems correct
2023-06-21 15:16:34 <minimal> well apart from yourself who else want this "fix"? Would such a broad change be accepted for merging in Alpine?
2023-06-21 15:16:58 <minimal> I've had changes refused in the past as people didn't agree with them
2023-06-21 15:21:53 <minimal> so these seems like the sort of thing that the TSC would have to decide upon
2023-06-21 15:22:32 <raspbeguy> I'm just giving my opinion since you asked for it 🙂
2023-06-21 15:24:23 <raspbeguy> I'll manage to do what I want one way or another. But as I noticed I am not the only one that the lack of symlinks bothers, you are also
2023-06-21 15:25:01 <minimal> I can just code-around the mapper entries being files rather than symlinks
2023-06-21 15:25:28 <raspbeguy> I can too
2023-06-21 15:27:31 <raspbeguy> In my sanlock-using case, the udev fix that submitted psykose solved my issue **for existing volume groups only** which is something I can live with
2023-06-21 15:29:46 <minimal> raspbeguy: did you *fully* test it to be sure? including a reboot test?
2023-06-21 15:29:55 <raspbeguy> but I still think that's a bug that should be addressed by lvm and cryptsetup and whatever
2023-06-21 15:30:09 <raspbeguy> minimal, even a reinstall test yes
2023-06-21 15:30:27 <minimal> you keep saying "bug" when it appears to be a *difference in behaviour*
2023-06-21 15:30:42 <minimal> between mdev systems and udev systems
2023-06-21 15:30:56 <invoked> raspbeguy: can you explain your sanlock-using case and what problem that solves for you?  just curious (and sorry if this is already deep in the scroll somewhere)
2023-06-21 15:30:58 <raspbeguy> well it's a bug that LVM doesn't address well to this difference in behaviour
2023-06-21 15:31:40 <raspbeguy> invoked, well I use a shared volume group between my hypervisors
2023-06-21 15:31:45 <minimal> raspbeguy: says who? lvm2 only has support for udev, I assume they'd tell you they don't support mdev (or at least didn't design for mdev)
2023-06-21 15:32:23 <raspbeguy> minimal, if they do, that would at least be clear
2023-06-21 15:32:53 <minimal> raspbeguy: is it a "bug" that a Ford Focus doesn't work well when you remove the differential and replace it with a Mitsubishi differently?
2023-06-21 15:32:59 <raspbeguy> and they do support systems withour udev, it's just they do do it properly
2023-06-21 15:33:30 <raspbeguy> that's why you're supposed to disable udev features in the config
2023-06-21 15:33:43 <raspbeguy> if those options exists, they support the case
2023-06-21 15:34:27 <minimal> eh? with udev disabled the mapper entries are files not symlinks
2023-06-21 15:34:37 <raspbeguy> corect
2023-06-21 15:34:47 <minimal> so how is that "supporting the case"?
2023-06-21 15:34:49 <raspbeguy> because lvm created those files
2023-06-21 15:35:11 <raspbeguy> lvm should have created symlinks instead
2023-06-21 15:35:19 <minimal> says who?
2023-06-21 15:35:24 <raspbeguy> that's why I say : they support the case, but not properly
2023-06-21 15:35:53 <minimal> who says lvm2 should create symlinks?
2023-06-21 15:35:59 <minimal> the udev devs?
2023-06-21 15:36:04 <minimal> someone else?
2023-06-21 15:37:36 <raspbeguy> back to sanlock : when using lvm shared features (using sanlock), lvmlock (which is part of lvm code) ask to sanlock to manage its locks on a special lv that lvm created just for it
2023-06-21 15:37:39 <minimal> lvm2 is doing things "differently", that does not necessarily mean it is doing things wrong
2023-06-21 15:37:57 <raspbeguy> if mermissions are not right on this lv device node, that's lvm fault
2023-06-21 15:38:27 <minimal> what is not right about permissions?
2023-06-21 15:38:53 <raspbeguy> the file in /dev/mapper is owned by root:root mode 600
2023-06-21 15:39:04 <minimal> ok, and?
2023-06-21 15:39:24 <raspbeguy> it should either be a symlink, or have correct permissions, which are root:disks mode 660
2023-06-21 15:39:45 <minimal> who said that all distros have a group called "disks"?
2023-06-21 15:40:08 <raspbeguy> they sure don't, but lvm should have a config option to manage it
2023-06-21 15:40:23 <minimal> perhaps they do, I don't know
2023-06-21 15:40:59 <raspbeguy> s/disks/disk
2023-06-21 15:41:59 <raspbeguy> anyway, I'm not revendicating anything, just speaking because you asked for my opinion
2023-06-21 17:36:48 <quinq> Hola
2023-06-21 17:37:10 <quinq> Anybody around here is using unfs3 successfuly?
2023-06-21 17:38:08 <quinq> When running: doas unfsd -d -l my_LAN_IPv4
2023-06-21 17:38:16 <quinq> It tells me: unable to register (NFS3_PROGRAM, NFS_V3, udp).
2023-06-21 17:38:48 <quinq> But when strace-ing it, I can see that it tries to bind to ::
2023-06-21 17:38:59 <quinq> And fails
2023-06-21 17:53:44 <zcrayfish> does it magically work if you pass the -p option as well?
2023-06-21 17:54:48 <quinq> Well, it doesn't complain, but then nfs doesn't mount ;p
2023-06-21 20:02:29 <leitz> Newbie question, can Alpine run on older (x86_64) Dell and HP laptops?
2023-06-21 20:04:36 <lopid> you could try the extended iso
2023-06-21 20:05:17 <lopid> or maybe just the standard
2023-06-21 20:06:58 <leitz> lopid, will do. I wasn't sure if there was a limited hardware list, or what. Haven't looked at Alpine much, but what I see looks good.
2023-06-21 20:07:29 <lopid> not that i know of
2023-06-21 20:09:21 <leitz> Is Alpine more like BusyBox, where there's one big binary that does lots of Unix-y stuff? Or is it more a "just the facts" with actual software?
2023-06-21 20:10:00 <abby> alpine uses busybox
2023-06-21 20:10:01 <lopid> it's quite minimal, comes with busybox as default, though you can extended it with other gnu type stuff
2023-06-21 20:10:12 <lopid> -ed
2023-06-21 20:10:56 <lopid> there are those who run it as their daily desktop
2023-06-21 20:11:44 <leitz> I'm happy to give it a go, anything without systemd is a win.
2023-06-21 20:11:50 <lopid> ;)
2023-06-21 20:14:45 <leitz> I used to do NetBSD, and loved the small kernel and the zippy fast response.
2023-06-21 20:20:03 <quinq> And now you have to deal back with Linux
2023-06-21 23:31:25 <Saijin_Naib[m]> > there are those who run it as their daily desktop
2023-06-21 23:31:25 <Saijin_Naib[m]> Raises hand
2023-06-21 23:32:18 <elly> same here :)
2023-06-21 23:34:21 <zcrayfish> desktop? what's that? Alpine is clearly the superior OS for daily laptops!
2023-06-21 23:34:35 <Saijin_Naib[m]> Minus some Edge hiccups that I obviously signed myself up for, absolutely solid and brilliant on latest-stable always
2023-06-21 23:35:08 <Saijin_Naib[m]> zcrayfish: Truth! That's also where it is! Last thing is my NAS, once I figure out if I can force it on there somehow 
2023-06-21 23:35:17 <elly> laptop is what I meant for myself anyway
2023-06-21 23:35:20 <elly> it's on my Framework
2023-06-21 23:35:54 <Saijin_Naib[m]> elly: Yoooooooo. How is it on there? ALS working? Suspend/resume? Brightness and other function keys? 
2023-06-21 23:36:41 <Saijin_Naib[m]> I might go Star Labs anyway because $$$, but the FrameWork is so cool
2023-06-21 23:41:42 <luks2[m]> framework is that modular laptop brand right?
2023-06-21 23:42:26 <abby> yeah
2023-06-21 23:49:02 <elly> yep!
2023-06-21 23:49:19 <elly> Saijin_Naib[m]: not sure what ALS is, I do not have suspend/resume working, and I do have the function keys working :)
2023-06-21 23:49:35 <elly> https://elly.town/d/blog/2022-10-20-alpine-framework.txt I should write an update to this
2023-06-21 23:53:59 <zcrayfish> ambient light sensor??
2023-06-21 23:54:08 <Saijin_Naib[m]> ALS is the Ambient Light Sensor, yeah
2023-06-21 23:54:24 <Saijin_Naib[m]> Miss that from my HP EliteBook 2740p 😭
2023-06-21 23:54:36 <Saijin_Naib[m]> My GeoBook has nothing 🤣
2023-06-21 23:55:06 <zcrayfish> my latitude 5430 rugged doesn't have one either, but I wondered about using the webcam to simulate one. oh well.
2023-06-22 00:00:32 <Saijin_Naib[m]> <zcrayfish> "my latitude 5430 rugged doesn'..." <- Should be doable, but I feel like the power consumption would be nuts, even if you just poll once every few seconds 
2023-06-22 05:29:06 <shrizza> Daily on my laptop (ThinkPad X200), desktop (Fujitsu Primergy), and a buncha servers and VMs.
2023-06-22 05:30:48 <psykose> don't think i've ever even heard of the primergy series
2023-06-22 05:37:41 <tm512> I'm personally appreciating just how tiny Alpine is, seems great for containers
2023-06-22 05:38:24 <zbrown> IIRC that was the original point (great for containers)
2023-06-22 05:38:48 <tm512> the Alpine image from linuxcontainers.org was just 14MB, whereas even Void was over 200MB, before installing anything at all
2023-06-22 05:40:43 <psykose> alpine existed almost a decade before docker so that was very much not the original point
2023-06-22 05:41:14 <tm512> whereas Ubuntu 23.04 was over 700MB. feels like for a large number of containers you'd end up wanting to do something where every container shares that original rootfs and has all of their specific stuff in a separate volume, maybe mounted as overlayfs or something
2023-06-22 05:42:59 <ikke> that's how docker works in the first place
2023-06-22 05:43:58 <ikke> but not somehting like lxc
2023-06-22 05:44:10 <zbrown> containers are just processes with pieces of flare, it's important to remember that
2023-06-22 05:44:35 <zbrown> psykose: I stand corrected. Or sit I guess :P
2023-06-22 05:45:18 <abby> looks like lxc just downloads void's regular rootfses, which are more intended for a bare-metal system
2023-06-22 05:45:37 <tm512> ikke: yeah I'm talking about lxc, though that use case would probably strongly point towards docker
2023-06-22 05:45:57 <abby> if you made an lxc container using base-minimal (which is meant for containers), it'd probably be a lot smaller
2023-06-22 05:46:45 <tm512> for my recent (and first) foray into containers I had assumed lxc would be simpler to get working even though docker would've fit the task pretty well
2023-06-22 05:47:09 <tm512> seems making assumptions is dangerous, because it was a bit of an ordeal
2023-06-22 05:48:16 <abby> XBPS_ARCH=x86_64-musl xbps-install -R https://repo-fastly.voidlinux.org/current/musl -r test-void -S base-minimal && du -sh test-void -> 77MB
2023-06-22 05:48:31 <abby> which is quite a bit smaller than lxc's image
2023-06-22 05:49:06 <psykose> imagine small image when big image
2023-06-22 05:49:17 <abby> love to make big image
2023-06-22 05:50:11 <tm512> abby: that seems more reasonable. of course Alpine still has it beat but I think busybox is a compromise personally, generally having coreutils is more comfortable
2023-06-22 05:50:27 <abby> yeah void isn't really focused on size
2023-06-22 05:51:16 <abby> maybe we could build a container rootfs tho
2023-06-22 05:52:19 <Saijin_Naib[m]> Is XFS over LVM (in?) doable with setup-alpine/setup-disk using arguments and/or environment variables, or is this in the do it myself offline and then mount and point Alpine to the mount camp? 
2023-06-22 05:52:38 <tm512> I wonder how tiny a Chimera Linux container would end up being, I think it ports its userland from various BSDs
2023-06-22 05:52:53 <Saijin_Naib[m]> I think I need to reinstall my desktop and might as well use XFS, I guess
2023-06-22 05:52:58 <abby> just freebsd
2023-06-22 05:53:08 <abby> i think
2023-06-22 05:53:35 <Saijin_Naib[m]> Is Chimera the Alpine variant for retrogaming? 
2023-06-22 05:53:46 <abby> chimera has nothing to do with alpine
2023-06-22 05:53:51 <abby> you're thinking of GamerOS
2023-06-22 05:54:08 <abby> tm512: smallest rootfs on chimera's download page is 5MBish https://repo.chimera-linux.org/live/latest/
2023-06-22 05:54:16 <Saijin_Naib[m]> Ah, sorry
2023-06-22 05:56:40 <psykose> tm512: you can just apk add coreutils and get them
2023-06-22 05:57:00 <tm512> abby: pretty tiny, though it probably decompresses a bit larger than Alpine if the image sizes displayed through `lxc images` indicate anything. the alpine/3.18 image is 2.75MB compressed (for aarch64)
2023-06-22 05:57:43 <psykose> Saijin_Naib[m]: ROOTFS=xfs iirc
2023-06-22 05:59:11 <tm512> psykose: yeah, I noticed coreutils were available, just haven't had a pressing need to grab them so far
2023-06-22 06:00:35 <tm512> I'm not doing a whole lot in this container, only got a Minecraft server and umurmur running. I'd probably end up wanting GNU or BSD userland if this were more of a daily driver
2023-06-22 06:06:08 <Saijin_Naib[m]> psykose: it's that simple? Damn, these setup scripts are amazing 
2023-06-22 06:06:22 <psykose> dunno if i'd call that amazing
2023-06-22 06:06:22 <Saijin_Naib[m]> 100% chance I fuck it up if I do it manually 
2023-06-22 06:07:04 <Saijin_Naib[m]> Is it as cozy as Calamares or whatever? No, but it does what is needed fairly friendly
2023-06-22 06:07:16 <Saijin_Naib[m]> Asking for fstype would be much better, but this works
2023-06-22 15:07:20 <eletrotupi> hm, I wonder how (and if) should we package stuff that should be runned under the local user 
2023-06-22 15:07:53 <eletrotupi> for instance, I was looking at packaging a LSP, which by default installs on $HOME/.local/bin 
2023-06-22 15:08:29 <ikke> If you package it, you would install it globally
2023-06-22 15:08:32 <ikke> not in the users home
2023-06-22 15:08:32 <eletrotupi> but I dunno know, installing globally (or at /usr/local/bin) sounds... weird?
2023-06-22 15:08:40 <ikke> why?
2023-06-22 15:08:48 <ikke> packages should never install in /usr/local
2023-06-22 15:08:56 <eletrotupi> well, /usr/bin*
2023-06-22 15:09:03 <ikke> why is it weird to install there?
2023-06-22 15:10:24 <eletrotupi> well, mostly because the package will default to somewhere else
2023-06-22 15:11:26 <psykose> default in what sense?
2023-06-22 15:11:48 <psykose> it starts, checks its own location, and tells you to move it?
2023-06-22 15:12:42 <eletrotupi> but I dunno, might send a patch upstream to allow configure the make install rule
2023-06-22 15:13:03 <eletrotupi> in the sense that it's literally typed to install to .local/bin
2023-06-22 15:13:14 <psykose> i don't think it's relevant what some line in a makefile does
2023-06-22 15:13:23 <psykose> which project
2023-06-22 15:13:24 <ikke> either patch it, or manually install it
2023-06-22 15:13:38 <eletrotupi> also, saw that there was a couple of issues with dotnet+musl on the repo
2023-06-22 15:13:47 <psykose> ah, dotnet
2023-06-22 15:13:50 <psykose> just don't package it then
2023-06-22 15:13:51 <psykose> ez
2023-06-22 15:13:51 <eletrotupi> so I'll take a look at it after work
2023-06-22 15:13:56 <eletrotupi> ahahahaha
2023-06-22 15:14:01 <eletrotupi> yeah, that's what I thought too
2023-06-22 15:14:44 <psykose> past me was smoking some crack and actually merged it for some reason
2023-06-22 15:15:19 <eletrotupi> kek
2023-06-22 17:53:07 <thresh> hi.  I'm working on a software package on alpine where I need to define specific user:group for a catalog we ship.  The problem is, the user is not created at the time package is built.  Is my best option to create the directory in pre-install scripts?
2023-06-22 17:54:08 <psykose> pkgusers=user install -o user
2023-06-22 17:54:41 <ikke> the pre-install script is to make sure the user exists in the target system
2023-06-22 17:54:48 <ikke> not during build (package)
2023-06-22 17:56:54 <thresh> sure
2023-06-22 18:06:24 <thresh> hmm, I've added both pkgusers="nginx" to my APKBUILD, and using pkgusers=nginx install -d -m750 -o nginx -g root ./var/lib/nginx/state;  and it's still install: invalid user ‘nginx’ under fakeroot
2023-06-22 18:08:08 <lopid> thresh! :-o
2023-06-22 18:09:31 <psykose> you don't need it again i just condensed it, it's just `install -d -o nginx`...
2023-06-22 18:09:34 <psykose> how are you running abuild
2023-06-22 18:11:20 <thresh> cd abuild-base/ && abuild checksum && abuild unpack && abuild prepare && abuild -K build && abuild rootpkg
2023-06-22 18:13:57 <psykose> if you just use `abuild -r` does it work
2023-06-22 18:14:43 <psykose> `mkusers` is the part you are missing
2023-06-22 18:15:35 <thresh> mhm, thanks
2023-06-22 18:16:34 <thresh> so through abuild -r it will create system users on an actual build host?
2023-06-22 18:16:53 <psykose> ye
2023-06-22 18:17:06 <thresh> not fun :/  but that helps
2023-06-22 18:18:03 <thresh> don't really want to give build user that kinda privileges
2023-06-22 18:19:00 <psykose> what is the actual way to create users/groups that can be used for files/dirs that then get put into tar that works inside fakeroot/userns
2023-06-22 18:19:15 <psykose> not aware of one that is not 794 steps
2023-06-22 18:20:19 <thresh> I assume rpm/deb just have a clever hack to chown to whatever on some internal install phase
2023-06-22 18:21:09 <psykose> the install part is the easy one
2023-06-22 18:21:12 <psykose> there's no hacks there
2023-06-22 18:21:26 <psykose> the info to do it with that comes from somewhere is the one in question
2023-06-22 18:21:40 <thresh> yeah I understand the problem
2023-06-22 18:22:04 <psykose> seems like some subuid thing inside 'user name spaces' is the webscale solution of 2037 to use
2023-06-22 18:22:08 <psykose> maybe i'll look at it eventually
2023-06-22 18:22:08 <thresh> I'll just hack it in postinst..
2023-06-22 18:28:43 <papertigers> I don't know if this is the best place to ask but I have a question about cloud-init. My config (NoCloud) provides a new name for an ethernet device.  However things seem to happen out of order.  /etc/init.d/networking fails to bring the interface up because it doesn't yet exist. However when I login to the box I am able to see the new name in /sys/class/net/ -- So if I run
2023-06-22 18:28:44 <papertigers> /etc/init.d/networking restart it then brings the interface up properly.  Does anyone know where I would look to fix this? Perhaps I need to change the order things are depended upon?
2023-06-22 18:41:27 <papertigers> seems like there needs to be a dependency on devfs somewhere?
2023-06-22 18:51:44 <thresh> heh, chown in postinst seems to get reverted
2023-06-22 18:52:15 <ikke> There is nothing that would do thaat
2023-06-22 19:01:21 <psykose> indeed, nothing that runs post post-install that would undo your chowns normally
2023-06-22 19:09:35 <thresh> https://gist.github.com/thresheek/bcafd907c63e58d66a3565b8cf244140
2023-06-22 19:09:47 <thresh> looking at strace apk add now
2023-06-22 19:13:17 <thresh> there seem to be some chown/chmod calls after postinst is finished though: https://gist.github.com/thresheek/bcafd907c63e58d66a3565b8cf244140?permalink_comment_id=4607623#gistcomment-4607623
2023-06-22 19:37:21 <thresh> looks like it happens ins apk_db_update_directory_permissions in apk_solver_commit_changeset
2023-06-22 19:38:24 <ikke> Most likely apk setting the permissions as they are in the package
2023-06-22 19:38:32 <ikke> which is where you should manage them
2023-06-22 19:39:18 <psykose> i guess indeed apk does it once more after post-install
2023-06-22 19:39:19 <thresh> yeah
2023-06-22 19:44:39 <thresh> not sure if I'm at ease giving sudo apk privileges to a non-privileged user
2023-06-22 19:45:07 <thresh> precreating a user before the build will probably work in my build environment, but it's a meh option for external users
2023-06-22 19:45:33 <ikke> thresh: there is apk rootbld
2023-06-22 19:45:46 <ikke> which uses bwrap to create an isolated build environment
2023-06-22 19:48:28 <psykose> which also makes users on the host the same way using abuild-addgroup/adduser
2023-06-22 19:48:39 <psykose> same issue
2023-06-22 19:48:54 <psykose> i thought of a fix for it but it would take forever to make it
2023-06-22 19:49:44 <thresh> I'll work it around now by precreating a user I guess, and since this is only for closed-source software for which I don't need to ship build instructions it's probablty fine for now
2023-06-22 19:50:12 <thresh> but when this feature hits mainline nginx, it's going to be somewhat of an issue
2023-06-22 19:50:45 <ikke> what feature?
2023-06-22 19:51:05 <thresh> the one I need to create a directory for
2023-06-22 19:51:40 <ikke> note that the aports nginx package already creates a user / group for nginx
2023-06-22 19:52:11 <thresh> sure, that's where I got the inspiration for the packages we ship from nginx.org/nginx.com :) much appreciated
2023-06-22 19:52:30 <thresh> but we need to ship our own builds, so the need of non-aports packages
2023-06-22 19:52:35 <ikke> right
2023-06-22 19:53:47 <ikke> Was not aware nginx was providing their own packages
2023-06-22 20:02:22 <thresh> oh, I just though that I can remove install -d from package() and just create the dir in the post-install
2023-06-22 20:02:37 <thresh> which is horrible admittedly from package manager point of view
2023-06-22 22:17:29 <LACampbell> is there some musl based reason why x crashes more on alpine?
2023-06-22 22:18:15 <LACampbell> never had this on glibc linux. so not sure if it's related to that, or some package that is needed for X but doesn't exist
2023-06-22 22:19:11 <psykose> maybe
2023-06-22 22:19:17 <psykose> nobody is going to fix it probably tho
2023-06-22 22:19:33 <LACampbell> lol. ok so it is a known issue? 
2023-06-22 22:19:51 <LACampbell> I'm fully prepared to blame X11 here :P
2023-06-22 22:20:07 <psykose> 'x crashes sometimes' isn't much of an issue (in terms of that doesn't establish any cause and isn't debuggable at all) but i don't generally hear much about that now that i think of it
2023-06-22 22:20:12 <psykose> sometimes maybe but in general not 'often'
2023-06-22 22:20:32 <LACampbell> Once a week or so. compared to my old glibc linux which was... never
2023-06-22 22:20:34 <psykose> for light contrast i haven't seen xwayland crash in months
2023-06-22 22:20:50 <zcrayfish> I'm good at crashing stuff, and I can't recall crashing X11 on any non-glibc distros...
2023-06-22 22:20:54 <LACampbell> it crashed with error messages, I search them and they're chrome error messages. so chromium is crashing X? this shit is complicated
2023-06-22 22:20:55 <zcrayfish> are you using a weird GPU?
2023-06-22 22:20:55 <psykose> it's more likely the wm you're using manages to crash x in some way and changed (which is still an x bug but how you see it is how you see it)
2023-06-22 22:21:07 <psykose> well
2023-06-22 22:21:14 <psykose> start with the xorg log and also what you're seeing pasted somewhere
2023-06-22 22:21:16 <LACampbell> zcrayfish: nah, one of the amdgpu ones
2023-06-22 22:21:17 <psykose> i can have a light look
2023-06-22 22:21:26 <psykose> amdgpu has good and bad stuff more or less
2023-06-22 22:21:56 <LACampbell> where's the Xorg log
2023-06-22 22:22:07 <psykose> var/log/X* or somesuch
2023-06-22 22:22:07 <LACampbell> hang on let me use the man page
2023-06-22 22:22:14 <psykose> maybe something in .local/
2023-06-22 22:22:18 <psykose> i don't remember
2023-06-22 22:22:43 <zcrayfish> /var/log/Xorg.*.log*
2023-06-22 22:22:44 <LACampbell> yeah I don't have any of these :/
2023-06-22 22:23:07 <LACampbell> which suggests to me my system is screwed at that level
2023-06-22 22:23:19 <psykose> could be /tmp/X*
2023-06-22 22:23:33 <psykose> just find / -name "Xorg*" -type f
2023-06-22 22:23:34 <psykose> hehe
2023-06-22 22:23:59 <psykose> (that will take forever so constrain it a bit)
2023-06-22 22:23:59 <zcrayfish> are you rebooting when X crashes? :)
2023-06-22 22:24:12 <LACampbell> no. just doing startx again
2023-06-22 22:24:17 <psykose> it deletes the old one on start
2023-06-22 22:24:20 <LACampbell> the fact I have no X logs makes me think X is screwed up itself
2023-06-22 22:24:23 <psykose> though you would still find the new one
2023-06-22 22:25:05 <LACampbell> "sched_getscheduler: Function not implemented (38)". then it X disconnects
2023-06-22 22:25:12 <LACampbell> disconnects cleanly it seems
2023-06-22 22:25:27 <LACampbell> I logged this using my pen + notepad service
2023-06-22 22:25:33 <psykose> that's just the chromium stderr and not that related
2023-06-22 22:26:01 <psykose> well, it says 'chromium logs something or crashes which then crashes x'
2023-06-22 22:26:03 <psykose> but that's about it
2023-06-22 22:26:05 <LACampbell> ok so X crashes, and I have no logs. I guess I'll try and reinstall X, and read the wiki more carefully?
2023-06-22 22:26:49 <psykose> i don't think "reinstalling" anything is going to make the logs appear
2023-06-22 22:26:56 <psykose> just run the find command or similar and you'll find something
2023-06-22 22:27:16 <LACampbell> I did run the find command, I found nothing
2023-06-22 22:27:22 <LACampbell> just links to the binaries and man pages
2023-06-22 22:28:15 <psykose> can't think of why that would happen presently
2023-06-22 22:28:42 <LACampbell> idk. maybe the fact I ran alpine succesfully on a VPS for years doesn't really translate to desktop usage
2023-06-22 22:29:06 <LACampbell> I tried running setup-xorg-base again
2023-06-22 22:29:08 <LACampbell>  * rc-update: service `hwdrivers' is not in the runlevel `sysinit'
2023-06-22 22:29:11 <LACampbell>  * rc-update: service `mdev' is not in the runlevel `sysinit'
2023-06-22 22:29:13 <LACampbell> so I'll add those
2023-06-22 22:31:21 <psykose> that's not relevant
2023-06-22 22:31:26 <psykose> it's just debug output
2023-06-22 22:31:39 <psykose> you should actually not add those
2023-06-22 22:31:43 <LACampbell> oh ffs
2023-06-22 22:32:03 <psykose> just run it again
2023-06-22 22:32:17 <psykose> don't do anything special though
2023-06-22 22:33:17 <LACampbell> I dunno my whole system just seems screwed. randomly typing commands in that alpine programs mention to me like a monkey. no x logs. maybe a wipe is in order
2023-06-23 02:32:25 <shrizza> LACampbell: apk add mlocate && updatedb && locate -i xorg | fgrep log
2023-06-23 08:18:33 <LACampbell> shrizza: appreciate it
2023-06-23 08:18:36 <LACampbell> but: "updatedb: can not open a temporary file for `/var/lib/mlocate/mlocate.db'"
2023-06-23 08:18:50 <LACampbell> I think my conclusion is that alpine on the desktop is beyond my skill
2023-06-23 08:18:52 <shrizza> You need to be root.
2023-06-23 08:19:25 <shrizza> It's probably on par with most other distros once you get the ball rolling.
2023-06-23 08:20:45 <shrizza> With the benefit of less default mysterious layers to complicate interaction between you and the OS.
2023-06-23 08:20:54 <LACampbell> nah I'm really struggling with it in a way I didn't with other stuff
2023-06-23 08:20:57 <LACampbell> great on the server though
2023-06-23 08:21:12 <LACampbell> thanks that gets me the log. TODO: learn to read these logs. where is the time. and where are the crashes
2023-06-23 08:22:29 <shrizza> If you're already familiar with Alpine on the server side, I'd imagine insight you glean from using it on the desktop could tend to reciprocate with your server experience.
2023-06-23 08:23:59 <LACampbell> I'd hope so
2023-06-23 08:24:05 <LACampbell> but like on the server - don't need x
2023-06-23 08:25:00 <LACampbell> basically X sometimes just crashes. can't find a reason in the log files. don't know why. didn't happen in other distros. I'm sure it's solvable but server stuff is way easier than X11 lol
2023-06-23 08:25:37 <aron> nope, X tends to randomly crash on other distros as well, just wait for it, hehe
2023-06-23 08:26:02 <LACampbell> no I was using another distro on this machine for a year or so
2023-06-23 08:26:03 <aron> none of my X sessions lasted more that 1 year
2023-06-23 08:26:18 <aron> s/that/than/
2023-06-23 08:26:20 <LACampbell> like I'm sure it's my fault
2023-06-23 08:26:29 <shrizza> Is it reproducible.
2023-06-23 08:26:33 <aron> X is buggy as hell
2023-06-23 08:26:36 <LACampbell> no. the best kind
2023-06-23 08:26:44 <LACampbell> it happened when I chatted on discord
2023-06-23 08:26:54 <LACampbell> and then it printed chrome debug errors when I quit
2023-06-23 08:27:03 <LACampbell> maybe it's pipewire? that's a shitshow on this install
2023-06-23 08:27:05 <LACampbell> really who knows
2023-06-23 08:27:22 <LACampbell> maybe x didn't crash. maybe it just decided to quit
2023-06-23 08:27:26 <LACampbell> for its own reasons
2023-06-23 08:27:28 <PureTryOut> I would recommend Wayland 😉
2023-06-23 08:27:35 <LACampbell> oh yeah
2023-06-23 08:27:40 <LACampbell> maybe it's time
2023-06-23 08:27:45 <shrizza> Also, could it be an emoji displayed in a window title. I've noticed many posts about X crashes that end with the acknowledgment that it's the window manager displaying an emoji and triggering a bug in some X library or something.
2023-06-23 08:28:11 <aron> haha, sounds good
2023-06-23 08:29:04 <shrizza> What window manager do you use.
2023-06-23 08:29:17 <LACampbell> me? openbox
2023-06-23 08:29:21 <LACampbell> which is ancient at this point
2023-06-23 08:29:25 <LACampbell> last release was 2016 IIRC
2023-06-23 08:29:56 <aron> 
ACTION uses xfce4 (on armv7 and x86_64 as well)

2023-06-23 08:30:13 <LACampbell> x or wayland?
2023-06-23 08:30:19 <LACampbell> I guess if I go wayland I'll need a new WM
2023-06-23 08:30:29 <aron> X
2023-06-23 08:30:53 <LACampbell> i have amd graphics thankfully
2023-06-23 08:30:53 <PureTryOut> XFCE4 doesn't have Wayland support
2023-06-23 08:31:01 <LACampbell> oh thats a shame
2023-06-23 08:31:08 <aron> I'm fine with X
2023-06-23 08:31:22 <PureTryOut> You can use either GNOME or Plasma, or several Wayland-only window managers
2023-06-23 08:32:34 <aron> I have phosh on the pinephone tho, but it is a mess -- regular crashes, weird glitches
2023-06-23 08:32:44 <aron> actuall, the whole pmos seems to be a mess
2023-06-23 08:32:59 <LACampbell> trying to find something similar to openbox. just a stacking thing that's highly customisable and lightweight
2023-06-23 08:33:23 <aron> (I have pinephone, I really, really liked the idea, but it is miserable, and I put it away)
2023-06-23 08:34:02 <PureTryOut> aron: as a pmOS dev, I'm sad to hear that
2023-06-23 08:34:13 <PureTryOut> But the PinePhone isn't a great device tbh, terrible specs for one
2023-06-23 08:34:54 <PureTryOut> Please do report the crashes and glitches you encounter, otherwise we can't do much about it
2023-06-23 08:35:40 <aron> true, I have to blame the pp hw
2023-06-23 08:35:53 <aron> it is terrible indeed
2023-06-23 08:36:24 <aron> maybe my impressions are heavily influenced by the pp hw
2023-06-23 08:39:03 <PureTryOut> Well things still shouldn't crash, but performance issues are definitely caused by the hardware
2023-06-23 08:39:06 <tm512> LACampbell: apparently part of the reason why Openbox seems abandoned is that the dev basically considers it complete
2023-06-23 08:39:20 <LACampbell> understandable
2023-06-23 08:39:29 <tm512> LACampbell: if you compile it from source, there has been a bit of development in another repo that goes up to 2018 http://git.openbox.org/?p=mikachu/openbox.git;a=summary
2023-06-23 08:40:06 <LACampbell> toyed with the idea of writing an openbox like WM recently. not using XML, or any gtk decorations
2023-06-23 08:40:20 <LACampbell> wonder how you test window managers. logout, recompile, login? or I guess a vm
2023-06-23 08:40:34 <PureTryOut> launch it in a window probably
2023-06-23 08:41:07 <LACampbell> launch a window manager in a window?
2023-06-23 08:41:33 <aron> PureTryOut: for instance, latest release made LLV player defunctional; moved to mpv again, but it sometimes deadlocks the WM when it exits -- that's why I put it away, for good
2023-06-23 08:41:55 <PureTryOut> I... Don't know what LLV player is 🤔
2023-06-23 08:42:09 <PureTryOut> what's the package name?
2023-06-23 08:42:11 <tm512> LACampbell: in xinitrc you can have the WM launched in a loop, so that when it quits a new instance gets launched
2023-06-23 08:42:18 <aron> it's a movie player on pmos
2023-06-23 08:42:18 <LACampbell> oh
2023-06-23 08:42:35 <PureTryOut> LACampbell: yes, KWin from KDE can do that. You can even test a Wayland session on X11 that way
2023-06-23 08:42:35 <tm512> 
ACTION does not know if that's how people tend to test in-development WMs though

2023-06-23 08:42:41 <LACampbell> tm512: so you play around with it, recompile, quit.
2023-06-23 08:42:43 <aron> dunno the pkg, it is preinstalled
2023-06-23 08:43:06 <PureTryOut> 
ACTION is not a Phosh/GNOME/GTK person

2023-06-23 08:43:24 <aron> sure, I just wanted to share some "fun"
2023-06-23 08:43:58 <PureTryOut> sure but I want to see if we can figure out what makes the LLV Player non-functional. But I don't even know what it is lol, and I see no package named anything like that as a dep of Phosh
2023-06-23 08:44:04 <tm512> LACampbell: as mentioned nested X is also a possibility, especially early on in development, or when dealing with some kind of crashing bug
2023-06-23 08:44:17 <LACampbell> huh I had no idea you could nest x
2023-06-23 08:45:00 <tm512> https://en.wikipedia.org/wiki/Xephyr
2023-06-23 08:45:33 <shrizza> LACampbell: I'm assuming you've tried i3?
2023-06-23 08:45:35 <aron> no reason to figure out anything on it, I'm not intended to debug pp anymore
2023-06-23 08:45:54 <LACampbell> shrizza: yeah i used to use it back in the day. not so into strict tiling anymore
2023-06-23 08:45:57 <LACampbell> whys that?
2023-06-23 08:46:02 <tm512> apparently it tends to get packaged as xorg-server-xephyr. at least this is how Arch and Void both package it
2023-06-23 08:47:28 <tm512> I've only ever used nested X over the network, almost a VNC-like setup, and I don't even remember what the use-case was
2023-06-23 08:47:29 <aron> (according to wiki, pp has one of the best support, so I'll not try others either -- the smartphone concept is dead to me)
2023-06-23 09:01:36 <PureTryOut> aron: you'd be helping others at least reporting your crashes though
2023-06-23 09:58:17 <shrizza> GTK2+Clearlooks 4 Lyfe
2023-06-23 11:57:00 <mebeim> Hello people :) just wanted to ask a quick question about the latest Alpine ISOs. I've set up the arm64 "virt" ISO in a qemu/libvirt VM and I can't seem to get SSH with PAM working. The openssh-server-pam package does not seem to have done anything because I have some limits in /etc/security/limits.conf that are not being enforced by PAM. UsePAM yes is set in sshd_config. Is
2023-06-23 11:57:01 <mebeim> this a known issue? Also, if this is not the righe place to ask could someone please point me in the right direction? Thanks!
2023-06-23 12:02:07 <ikke> This is the correct place to asj
2023-06-23 12:02:09 <ikke> Ask
2023-06-23 12:02:30 <minimal> mebeim: which settings in limits.conf are you trying to use?
2023-06-23 12:15:55 <mebeim> minimal: uid hard maxlogins N, to set the max number of concurrent logins through SSH for some user
2023-06-23 12:19:04 <mebeim> AFAIK it should be enforced by PAM, but it is not. SSH auth does not even seem to go through PAM.
2023-06-23 12:19:42 <mebeim> I can see the following in /var/log/messages: auth.info sshd[2114]: rexec line 82: Unsupported option UsePAM <LF> auth.info sshd[2114]: Accepted password for user from [...]
2023-06-23 12:21:01 <minimal> do you have the openssh-server package installed as well as openssh-server-pam package?
2023-06-23 12:21:17 <minimal> that sounds like it is using the wrong binary for ssh server
2023-06-23 12:21:18 <mebeim> That "Unsupported option UsePAM" is suspicious, makes me think something's wrong with my openssh installation. It's just a fresh Alpine install with alpine-virt-3.18.2-aarch64.iso, where I then added a new user and the openss-server-pam and shadow-login 
2023-06-23 12:21:47 <mebeim> minimal: yep, I have openssh-server-pam, is there some config I need to do after installing it with apk add maybe?
2023-06-23 12:22:08 <minimal> no, I asked if you ALSO had openssh-server package installed
2023-06-23 12:22:17 <mebeim> oh sorry my bad
2023-06-23 12:23:18 <mebeim> yeah I do
2023-06-23 12:23:29 <mebeim> so I need to uninstall that I guess
2023-06-23 12:23:34 <minimal> that's likely your problem then
2023-06-23 12:28:57 <mebeim> minimal: ok I can now see that PAM is being used in the logs, but still it does not seem to enforce the limit
2023-06-23 12:28:59 <mebeim> 
ACTION thinking

2023-06-23 12:30:40 <mebeim> The limit I have now is `user hard maxlogins 1` in /etc/security/limits.conf, but I am able to open 2 SSH sessions
2023-06-23 12:34:41 <minimal> type "-" rather than "hard"
2023-06-23 12:35:25 <mebeim> minimal: same behavior
2023-06-23 12:35:27 <minimal> oops, s/type/try/
2023-06-23 12:35:56 <mebeim> I can confirm it works in another VM with Debian 12 + PAM, where I have the exact same line in limits.conf
2023-06-23 12:36:12 <minimal> ok, saw that suggested in a serverfault question similar to yours from 12 years ago
2023-06-23 12:37:00 <minimal> ah, pam_limits uses UTMP - that's probably the issue
2023-06-23 12:37:48 <mebeim> Yeah I came across quite the number of Q&As while trying to fix this (I had a setup with SSHD and PAM working before the current one where I forgot about the package conflicts)
2023-06-23 12:38:02 <minimal> try installing the utmps package and running setup-utmp
2023-06-23 12:38:12 <mebeim> Ah I see!
2023-06-23 12:38:20 <minimal> musl by default has NO utmp support
2023-06-23 12:38:28 <minimal> utmps adds UTMP support
2023-06-23 12:38:41 <mebeim> I was expecting some kind of error logs in /var/log/messages
2023-06-23 12:38:48 <minimal> what sort of errors?
2023-06-23 12:38:49 <mebeim> Thanks I'll try that 
2023-06-23 12:39:25 <mebeim> Idk something like "pam session: cannot enforce this limit because I am missing this functionality/binary"
2023-06-23 12:39:35 <minimal> it doesn't know utmps is missing
2023-06-23 12:39:47 <minimal> so it cannot error
2023-06-23 12:40:21 <mebeim> Got it, makes more sense now
2023-06-23 12:40:53 <minimal> out-of-the-box typical tools like "w", "who", "last" which rely on utmp do nothing
2023-06-23 12:41:02 <minimal> until utmps is installed and activated
2023-06-23 12:41:07 <mebeim> ah, that's why then
2023-06-23 12:41:21 <mebeim> I could notice `who` not working, I had no idea about UTMP 
2023-06-23 12:41:37 <minimal> who reads data from utmp/wtmp files
2023-06-23 12:41:49 <minimal> which don't exist without utmps
2023-06-23 12:42:04 <mebeim> yep, today I learned
2023-06-23 12:43:27 <minimal> mebeim: somewhat out-of-date: https://github.com/dermotbradley/alpine-misc-notes/blob/main/utmp-status.md
2023-06-23 12:43:34 <minimal> I must get around to updating it
2023-06-23 12:44:02 <mebeim> minimal: hmmm... now `who` works as expected, but still the limit is not enforced
2023-06-23 12:44:13 <mebeim> 
ACTION is confused

2023-06-23 12:45:03 <mebeim> (tried both `hard` and `-`)
2023-06-23 12:45:32 <minimal> ah, seems linux-pam was never linked against utmps libs
2023-06-23 12:47:09 <mebeim> oh snap
2023-06-23 12:47:28 <mebeim> so it cannot support this functionality at all?
2023-06-23 12:47:34 <minimal> I'd not tested pam limits together with utmp in the past
2023-06-23 12:47:55 <minimal> it just needs to be built/linked with it AFAIK
2023-06-23 12:48:17 <mebeim> so I could theoretically get it working if I build and install the package manually
2023-06-23 12:51:20 <minimal> trying a local build here currently
2023-06-23 12:51:59 <mebeim> minimal: nice, thanks a lot for your time and support
2023-06-23 13:22:53 <kwadroke[m]> Does anyone have a link to documentation on how to get Active Directory authenication working? I'm able to join the domain and query users via id and getent groups username. I Just need to get logins working
2023-06-23 13:23:18 <kwadroke[m]> course this is via Winbind (Samba)
2023-06-23 13:25:00 <kwadroke[m]> I normally use SSSD on RHEL/Alma systems, but SSSD doesn't have Active Directory support compiled in. Spent most of the afternoon trying to get that working, but ran into some gettext issues. I have a APKBUILD for Realm and ADCLI that appears to work.
2023-06-23 14:00:20 <consus[m]> I have musl-nslcd config
2023-06-23 14:01:01 <consus[m]> Works for me
2023-06-23 14:07:34 <singham> Has anyone hosted a webserver with minihttpd?
2023-06-23 14:07:51 <singham> I need help with setting up tls with minihttpd.
2023-06-23 14:09:52 <kwadroke[m]> consus: can you sanitize and post your config?
2023-06-23 14:14:21 <consus[m]> yep
2023-06-23 14:14:22 <consus[m]> sec
2023-06-23 14:14:26 <kwadroke[m]> thanks
2023-06-23 14:17:26 <consus[m]> This is /etc/nslcd.conf: https://pastebin.com/dvBUjDUs
2023-06-23 14:20:51 <kwadroke[m]> thanks
2023-06-23 14:26:59 <kwadroke[m]> hmmm ldap bind. trying to stay away from that
2023-06-23 14:34:14 <consus[m]> hm?
2023-06-23 14:34:16 <consus[m]> why?
2023-06-23 14:36:16 <kwadroke[m]> trying to keep the password off the systems
2023-06-23 14:36:39 <kwadroke[m]> on most of our VMs the users have sudo access
2023-06-23 14:36:56 <Lantizia> Has Alpine ever been in the SPI like Debian and Arch?
2023-06-23 14:37:32 <kwadroke[m]> and they can get access to the password that way
2023-06-23 14:38:11 <kwadroke[m]> and I don't want to have to create a user for every system to query ldap
2023-06-23 14:38:39 <Lantizia> or SFC
2023-06-23 14:38:58 <kwadroke[m]> and the AD admins really want to disable LDAP as well - probably will never be able to do that though
2023-06-23 14:39:34 <consus[m]> oh
2023-06-23 14:41:06 <ikke> Lantizia: no, there has never been a legal entity, nor has Alpine Linux been represented by one.
2023-06-23 14:42:59 <Lantizia> does Alpine still work on a Via C3 cpu?  
2023-06-23 14:47:28 <ikke> Depends on what instructionset it supports 
2023-06-23 14:48:07 <Lantizia> well is there an alpine doc saying what instruction sets at a minimum are required?
2023-06-23 14:48:12 <Lantizia> I'm having trouble locating one
2023-06-23 14:48:44 <Lantizia> I mean 'Via C3' was going to be 'Cyrix III' until Via bought Cyrix... so essentially 586-like
2023-06-23 14:49:10 <ikke> It needs to be at least full 586
2023-06-23 14:49:21 <Lantizia> "full" :)
2023-06-23 14:49:27 <ikke> But some applications may require 686
2023-06-23 14:50:56 <Lantizia> Some things (e.g. https://gcc.gnu.org/legacy-ml/gcc-patches/2004-07/msg02718.html ) claim it's essentially a i686 without the 'cmov' instruction set... if that helps?
2023-06-23 14:51:19 <ikke> https://gitlab.alpinelinux.org/alpine/tsc/-/issues/20
2023-06-23 15:08:24 <Lantizia> is that you saying.... "We know" :P
2023-06-23 15:11:07 <ikke> No, just giving context on what we expect for x86
2023-06-23 17:24:07 <raspbeguy> Hello, is there any people having 3TB or 4TB HDDs ? Could you do me a favor and give me the exact size in bytes of those disks please ?
2023-06-23 17:24:52 <ikke> as reported by what?
2023-06-23 17:25:08 <minimal> doesn't that vary slightly by vendor/model?
2023-06-23 17:25:56 <minimal> also when drive sizes are quoted they are in TB (multiples of 1000), not TiB (multiples of 1024)
2023-06-23 17:26:47 <minimal> so 3TB in theory is 3 x 1000 x 1000 x 1000 x 1000 bytes
2023-06-23 17:27:58 <ikke> sda           8:0    0 4000787030016
2023-06-23 17:28:58 <minimal> 4000787030016 = 4.00078703TB
2023-06-23 17:29:31 <raspbeguy> minimal: that's what I want to verify
2023-06-23 17:29:40 <ikke> and 3.6TiB
2023-06-23 17:29:51 <minimal> verify what exactly?
2023-06-23 17:30:40 <raspbeguy> I thought that the exact size could vary but I found the exact same size of different brand
2023-06-23 17:31:43 <minimal> does it matter? (apart from RAID1 if you've got 2 drives of varying size lol)
2023-06-23 17:31:45 <raspbeguy> Mine 4TB disks are also 4000787030016 bytes
2023-06-23 17:32:07 <raspbeguy> minimal: that's exactly because of a raid array I'm asking
2023-06-23 17:33:00 <minimal> raspbeguy: I suspected as much. I used to leave a small bit of unused space at end of drives to deal with variations in disk size ;-)
2023-06-23 17:33:56 <minimal> plus over time drives can have reallocated sectors (which is normally fine as they have "spare" sectors....until the run out of those lol)
2023-06-23 17:34:56 <zcrayfish> Any reallocated sectors are a very very bad sign on modern mechanical harddrives.
2023-06-23 17:36:03 <raspbeguy> ikke: would you mind telling me the model please?
2023-06-23 17:40:37 <minimal> raspbeguy: so what do you do in the future if 1 of the drives dies and the replacement you get is slightly smaller? ;-)
2023-06-23 17:43:10 <raspbeguy> minimal: that's precisely what I want to avoid 
2023-06-23 17:43:37 <minimal> raspbeguy: which is why I talked about slightly undersizing usage
2023-06-23 17:46:51 <minimal> benn there, software RAID-1 2 disk of different makes & models with slightly different sizes ;-)
2023-06-23 17:48:11 <minimal> always a good idea not to use the same make/model of drive from the same manufacturing batch for RAID1
2023-06-23 17:52:20 <ikke> raspbeguy: it's called G-Drive (it's an external disk), not sure what model
2023-06-23 17:52:32 <raspbeguy> Ok thanks 
2023-06-23 17:52:51 <ikke> https://www.techradar.com/reviews/g-technology-g-drive-4tb
2023-06-23 17:53:58 <raspbeguy> If more people could share the exact size of their 3 or 4 TB HDD that would help me
2023-06-23 17:54:33 <minimal> many of those non-big-name external drives tend to use multiple of the (2? 3? drive manufacturers left) makes as and when they buy batches of a particular make&model
2023-06-23 17:55:16 <minimal> so a G-Drive could be Seagate inside this month and WD inside next month
2023-06-23 17:56:40 <minimal> raspbeguy: what are you trying to achieve? you've been told that drives of theoretically the same capacity may vary in actual bytes size
2023-06-23 17:57:16 <raspbeguy> Ideally I would like to build a list of models that have the exact same size
2023-06-23 17:57:44 <raspbeguy> So I know what to buy to replace one of my dying disk
2023-06-23 17:57:47 <minimal> a list of models *at certain points in time*?
2023-06-23 17:58:33 <minimal> a model today may have a specific size but that is no guarantee the same make/model will always have the same size or that that particular make/model will still be for sale when you go looking for it
2023-06-23 17:58:52 <raspbeguy> From experience same models have the exact same size no matter the time 
2023-06-23 17:59:24 <minimal> and models are always sold rather than being replaced by similar models?
2023-06-23 17:59:47 <minimal> I've got (from memory) at least 3 different models of WD Red drives of the same capacity
2023-06-23 17:59:57 <minimal> as they were bought at different times
2023-06-23 18:00:13 <minimal> the model numbers are *slightly* different for each of them
2023-06-23 18:00:55 <minimal> yet they're all WD Red (not RED+ / Red Pro) drives of the same size
2023-06-23 18:11:31 <minimal> raspbeguy: also not some vendors have been quietly replacing some drive models with SMR versions without telling people
2023-06-23 18:11:41 <minimal> s/not/note/
2023-06-23 18:12:29 <zcrayfish> (WDC)
2023-06-23 18:18:32 <psykose> all of them do that
2023-06-23 19:51:16 <frag_> is there a default binding for toggle or changing between 1 or 2 pages pr row in zathura?
2023-06-23 19:52:38 <rob[m]> I don't remember by heart, but it's in the man page
2023-06-23 19:55:48 <frag_> indeed, thought i had checked the man page :> (it was 'd')
2023-06-23 20:40:53 <linuxsquare> Hey guys, I have a quick question for you. I'm trying to create my own alpine-iso. I followed the official guide and it's working so far. What I'm confused with, is with the $apks variable in the profile script. Is there a way, that those programs are installed into the iso, without mentioning them inside the world-file in the genapkovl? When generating the iso, I've noticed the packages are getting downloaded but not installed. How do I solve 
2023-06-23 20:40:55 <linuxsquare> this issue? Thank you in advance.
2023-06-23 21:06:27 <rob[m]> <frag_> "indeed, thought i had checked..." <- It's ok. It happens. I got used to checking man pages now that I'm on OpenBSD and with AIX
2023-06-23 21:06:35 <rob[m]> *and on AIX
2023-06-24 00:38:52 <latex_> Anyone know how to get wifi working on raspberry pi 2?
2023-06-24 01:24:24 <imprevisto[m]> 🔨
2023-06-24 09:04:03 <oleovit> hi. regarding abort: https://git.alpinelinux.org/aports/commit/?id=457c8966262864a43a4a9ba8ed60d5efd6447aae resp. https://gitlab.alpinelinux.org/alpine/aports/-/commit/457c8966262864a43a4a9ba8ed60d5efd6447aae , i guess i have to execute "setup-devd udev" again, in order to "see" the change in /lib/udev/rules.d/50-udev-default.rules + dmesg (even if this is only a cosmetic thing)?
2023-06-24 09:12:26 <psykose> no, you have to do nothing except reboot
2023-06-24 09:13:05 <psykose> (but i forgot to actually apply the patch)
2023-06-24 09:13:09 <oleovit> i did psykose.
2023-06-24 09:14:00 <oleovit> maybe 6 reboots since the update. yep, maybe something went wrong applying it
2023-06-24 09:23:24 <oleovit> ..... or did i not do a specific _reboot_, if this is required. i only do a reboot after a kernel update, and that was the day before..
2023-06-24 10:22:34 <ovf> recently (maybe with the 6.3.9 update) wg-quick started to miscalculate the mtu if none is set in the config file. the symptom is that the sequence wg-quick up wg0;wg-quick down wg0;wg-quick up wg0 results in 'RTNETLINK answers: No such device'. anyone seen that yet?
2023-06-24 10:26:48 <psykose> doesn't ring a bell
2023-06-24 10:27:59 <psykose> regular wg stuff seems ok and i don't put mtu
2023-06-24 10:28:03 <psykose> (not wg-quick)
2023-06-24 10:29:39 <satmd> ovf: do you mind finding the exact version where this starts to happen?
2023-06-24 10:31:19 <satmd> I have just opened a(n unrelated) bug with gentoo's 6.3.x series where I get rcu stalls with wireguard+bonding+igb. This is a long shot, but *if* your problem also starts with the switch of 6.2 to 6.3, it might be worth mentioning. But it's also just a gut feeling.
2023-06-24 10:36:13 <ovf> psykose: i'm pretty sure it's specifically the delicate dance wg-quick does with guessing the mtu (that i don't yet understand). it feels like after 'down' the old mtu is still out there somewhere and wg-quick then subtracts from it again, resulting in a value too low
2023-06-24 10:37:01 <ovf> satmd: i will try a bit later, thanks
2023-06-24 10:37:02 <psykose> perhaps
2023-06-24 10:46:06 <satmd> oh, and I noticed that the wireguard kernel module has not changed between 6.2.8 and 6.3.8. I believe it's some other part of networking code.
2023-06-24 11:36:48 <frag_> can thinkpad/alpine be on for a monthish time in suspension?
2023-06-24 11:37:33 <lopid> why not?
2023-06-24 11:38:15 <ikke> With or without powersupply?
2023-06-24 11:38:44 <frag_> i dunno, just have an asus/void laptop that can do that, buy this thinkpad eats quite a lot of battery with the lid closed
2023-06-24 11:38:50 <frag_> ikke: without powersupply
2023-06-24 11:39:13 <frag_> it wouldnt last 24h me thinks
2023-06-24 11:39:31 <frag_> maybe its just setup tho, set that one up 6ish years ago ..
2023-06-24 11:39:52 <frag_> hmm might have changed distro come to think of it... anyways..
2023-06-24 11:40:51 <psykose> the distro has no relevance when you put the device into some sleep mode
2023-06-24 11:40:59 <psykose> and the answer is no for any device i can think of
2023-06-24 11:44:00 <frag_> oh you need swap to hibernate? damn ..
2023-06-24 11:44:15 <ikke> There are different suspension levels
2023-06-24 11:44:34 <ikke> hibernate should work indefinely, as the power is completely off
2023-06-24 11:45:11 <psykose> swapfiles also work
2023-06-24 12:33:58 <msiism> So, is there any merit in adding a normal user to the cdrom group these days?
2023-06-24 12:34:05 <msiism> Doesn't seem like it.
2023-06-24 12:34:22 <msiism> I mean, I can access contents of a CD just fine without being in that group.
2023-06-24 12:36:51 <msiism> I would prefer being able to mount it as a normal user. But adding myself to that group doesn't have that effect.
2023-06-24 12:45:08 <ikke> mounting as user requires CAP_SYS_ADMIN
2023-06-24 12:45:33 <ikke> A tool like pmount (afaik not available for alpine) is suid, so it can defer it to normal users
2023-06-24 12:46:11 <Hello71> being in the "disk" group wouldn't let you mount disks either
2023-06-24 12:46:29 <Hello71> you can read and write the devices, for example you could do mpv /dev/sr0
2023-06-24 12:46:59 <Hello71> or cdrecord /dev/sr0
2023-06-24 12:47:51 <Hello71> if you want to mount without root, you can use the "user" fstab option, pmount, udisks, etc
2023-06-24 12:48:07 <msiism> Oh, okay. Good stuff.
2023-06-24 12:49:15 <msiism> Hello71: But doesn't uisng cdrecord like this require me to be in the cdrw group too?
2023-06-24 12:49:51 <ikke> cdrom has rw on /dev/sr0 for me
2023-06-24 12:49:55 <ikke> so that would suffice
2023-06-24 12:50:09 <msiism> Yeah, that's the default, apparently.
2023-06-24 12:50:12 <Hello71> I don't think there's any such group. in the traditional unix permission model, each device can only be owned by one group, so it doesn't make sense
2023-06-24 12:50:16 <ikke> brw-rw---- 1 root cdrom 11, 0 Jun  9 18:59 /dev/sr0
2023-06-24 12:50:34 <msiism> But Alpine has cdrw as a group.
2023-06-24 12:51:06 <Hello71> hm, it does
2023-06-24 12:51:12 <ikke> may be legacy
2023-06-24 12:51:24 <ikke> Some groups were checked by grsec
2023-06-24 12:52:56 <ikke> Hmm, grsec website theme is quite similar to alpine linux
2023-06-24 12:53:15 <ikke> Including an image of the alps (or some outher snowy mountain)
2023-06-24 12:53:33 <psykose> theirs has better marketing
2023-06-24 12:53:43 <psykose> (i don't like marketing)
2023-06-24 12:55:16 <Apachez> cdrw but not dvdrw?
2023-06-24 12:55:24 <Apachez> what about bluray as a group? ;)
2023-06-24 12:55:46 <msiism> It should be "od…" then, for "optical disc".
2023-06-24 12:56:05 <msiism> I'm okay with "cd…" though.
2023-06-24 12:56:38 <psykose> cdrom sounds fine shrug
2023-06-24 12:57:20 <msiism> It's a bit like using a floppy disk icon for "Save" in a GUI.
2023-06-24 12:57:24 <msiism> But that's okay.
2023-06-24 12:57:32 <ikke> msiism: I was thinking exactly the same :D
2023-06-24 12:57:41 <grawity> you are not complaining about /dev/tty* being owned by 'uucp' even though it's neither a teletypewriter nor being used for UUCP
2023-06-24 12:59:06 <grawity> ...huh, I didn't realize alpine miniroot was *that* tiny
2023-06-24 12:59:18 <psykose> lots of guis do have a floppy icon for save
2023-06-24 12:59:23 <psykose> 
ACTION is confused

2023-06-24 13:00:15 <msiism> Sure, but there are more abstract icons these days, too.
2023-06-24 13:00:18 <ikke> grawity: that's the reason why alpine docker images are that small
2023-06-24 13:00:46 <msiism> psykose: Like, a downward arrow pointing into an open drawer.
2023-06-24 13:01:08 <msiism> Or pointing onto something resempling a HDD.
2023-06-24 13:01:20 <psykose> indeed, there's hundreds
2023-06-24 13:01:21 <msiism> s/resem/resam/
2023-06-24 13:01:29 <psykose> 
ACTION is still confused what the point is

2023-06-24 13:02:21 <msiism> Well, the point of these newer icons is to generalize the visual representation of the operation and abstract it aways from a distinct storage medium, I guess.
2023-06-24 13:02:22 <ikke> psykose: That it's ok to use anachronisms, so kind reverting their initial 'complaint'
2023-06-24 13:02:45 <psykose> ah
2023-06-24 13:03:09 <msiism> Okay, right.
2023-06-24 13:04:33 <msiism> Being in the cdrom group also allows me to use `eject`. That's nice.
2023-06-24 13:40:14 <rob[m]> I don't blame anybody, but I remember when I used to care about icons when I was doing mobile app development. When google didn't even follow their design guidelines... funny
2023-06-24 13:41:10 <psykose> they change them yearly
2023-06-24 13:41:54 <rob[m]> I'm glad that I left that world
2023-06-24 13:43:13 <ikke> Nobody gets a promotion by keeping things the same
2023-06-24 13:44:02 <psykose> mobile app world seemed like a slightly roided version of web app world to me every time i looked at it
2023-06-24 13:44:04 <psykose> but never made anything so
2023-06-24 13:52:59 <rob[m]> ikke: yeah, but at one point, I'm like, who cares. I just want a cli. I despise android now
2023-06-24 13:54:56 <msiism> GUIs have their place, though.
2023-06-24 13:55:12 <msiism> For example in CAD or audio/video production.
2023-06-24 13:55:32 <rob[m]> psykose: mobile apps are a mess imo. You have to rewrite everything if you don't use something cross platform and stable design guidelines are prohibited from what I saw 
2023-06-24 13:55:32 <rob[m]> mobile apps are doomed, I don't see them lasting more than 10 years. web is the way to go
2023-06-24 13:55:49 <msiism> I think most GUIs would benefit a lot from integratin a command line.
2023-06-24 13:56:06 <rob[m]> msiism: I agree as well, I personally don't need much of a gui, but they are very useful when needed
2023-06-24 14:50:40 <msiism> Is there an official description of what the "main", "community", and "testing" repositories are and how they differ?
2023-06-24 14:53:11 <msiism> And, what are these called, actually. "Branches?"
2023-06-24 14:53:40 <msiism> Okay, they are apparently just called repositories.
2023-06-24 14:53:40 <ikke> They are repositories
2023-06-24 14:53:51 <msiism> I see.
2023-06-24 14:54:23 <ikke> https://docs.alpinelinux.org/user-handbook/0.1a/Working/apk.html#_repositories_releases_and_mirrors
2023-06-24 14:55:04 <msiism> Oh, totally forgot about the Handbook. I was looking through git repos and man pages.
2023-06-24 15:11:38 <modlooper> hey everyone!
2023-06-24 15:12:03 <modlooper> does anyone know by any chance whether it is possible to build a kernel with all modules built in to avoid the modloop?
2023-06-24 15:12:13 <psykose> yourself you can build anything
2023-06-24 15:12:31 <modlooper> I am running a diskless setup, and would like to unmount and remove the USB boot stick after boot
2023-06-24 15:13:00 <modlooper> which is currently not possible because the modloop is mounted from that through th loop dev
2023-06-24 15:13:09 <ikke> You can unmount modloop 
2023-06-24 15:13:13 <ikke> by stopping the modloop service
2023-06-24 15:13:24 <ikke> then it's safe to unmount the usb disk
2023-06-24 15:13:24 <modlooper> strangely it refuses to stop for me...
2023-06-24 15:13:27 <modlooper> tried that first
2023-06-24 15:13:37 <ikke> What happesn when you try to stop it?
2023-06-24 15:13:59 <modlooper> my initial idea was to copy the modloop file from the usb stick to ram (I have plenty), unmount the modloop, remount if from ram, remove stick
2023-06-24 15:14:08 <modlooper> will check again and post the exact error
2023-06-24 15:14:11 <modlooper> give me 5 min
2023-06-24 15:31:01 <modlooper> ikke: of course it works perfectly... not sure what caused it before that it refused...
2023-06-24 15:31:13 <modlooper> thanks for assuring me that I should try again lol :D
2023-06-24 15:32:05 <ikke> modlooper: no problem :-0
2023-06-24 15:32:07 <ikke> :-)
2023-06-24 15:37:18 <modlooper> have a good one y'all
2023-06-24 20:55:25 <frag> is there something in busybox (or posix) that can make folder B the same as folder A? (been using rsync for that, wondering if there something more basic ans simpler for local use)
2023-06-24 20:56:13 <Habbie> rm -rf + cp?
2023-06-24 20:56:39 <frag> Habbie: yeah, mostly use it for larger backups tho :/ (like 400G)
2023-06-24 20:57:00 <Habbie> you want something roughly as smart as rsync?
2023-06-24 20:58:28 <frag> well would be nice if it didnt use too much resources :] im on rpi and similar
2023-06-24 20:58:32 <quinq> There is pax \o/
2023-06-24 20:59:21 <quinq> Though tar should work too
2023-06-24 20:59:46 <frag> pax? tar?
2023-06-24 21:00:05 <frag> taring 400G is a lot eh
2023-06-24 21:00:22 <quinq> Well, I didn't put the requirement of 400G
2023-06-24 21:00:47 <quinq> But that has nothing to do with the tool you'll be using to transfer data
2023-06-24 21:00:54 <quinq> It'll still be 400G
2023-06-24 21:01:04 <Habbie> except when rsync decides it only has to transfer 20G
2023-06-24 21:01:09 <Habbie> which is a feature almost unique to rsync
2023-06-24 21:01:13 <quinq> haha
2023-06-24 21:01:27 <quinq> It's not a bug, it's a feature!
2023-06-24 21:01:36 <Habbie> it.. was not a joke
2023-06-24 21:01:56 <Habbie> these are the choices - transfer 400G, or map the differences so you only transfer the differences
2023-06-24 21:01:56 <quinq> Yeah I guess it's not really funny when that happens :/
2023-06-24 21:02:05 <Habbie> i wasn't talking about a bug either..
2023-06-24 21:02:09 <quinq> Well
2023-06-24 21:02:45 <quinq> copying 20G out of 400GB when you ask it to copy 400GB sounds like qualifying as a bug :)
2023-06-24 21:02:48 <Habbie> but, it's unclear to me whether frag is relying on that rsync feature
2023-06-24 21:02:58 <Habbie> quinq, of course, but it's not what i meant :)
2023-06-24 21:03:16 <Habbie> i don't know if frag already has a B that has diverged from A, or no B at all yet
2023-06-24 21:03:41 <frag> Habbie: yes thats often the case
2023-06-24 21:03:50 <quinq> I think that the target might not be very clear
2023-06-24 21:03:54 <Habbie> quinq, i agree
2023-06-24 21:04:20 <quinq> Are you actually asking for a tool to replicate a hierarchy of files, frag, or are you asking for keeping two file hierarchies in sync (rsync)?
2023-06-24 21:04:21 <Habbie> frag, and is the target on the same system, or somewhere else?
2023-06-24 21:05:12 <frag> Habbie: same system
2023-06-24 21:05:51 <frag> quinq: rsync can do both eh... i guess the latter 
2023-06-24 21:06:13 <Habbie> rsync can do both, but picking the -right- tool does depend on this question
2023-06-24 21:06:22 <Habbie> 21:01Z <Habbie> these are the choices - transfer 400G, or map the differences so you only transfer the differences
2023-06-24 21:07:34 <Habbie> are you running into trouble with rsync using resources?
2023-06-24 21:07:35 <quinq> Yeah, and/or actively keep track of unsynced files
2023-06-24 21:07:43 <Habbie> that's a good one
2023-06-24 21:08:43 <quinq> But maybe that's also already an rsync feature ^^
2023-06-24 21:08:59 <quinq> (kind of)
2023-06-24 21:09:06 <Habbie> well, not actively
2023-06-24 21:09:11 <Habbie> lsync can do that, and can then call rsync for you
2023-06-24 21:09:15 <Habbie> but inotify over 400GB will also hurt
2023-06-24 21:09:34 <quinq> Yeah, I meant just keeping state of the source, and assuming you only have to update the target
2023-06-24 21:09:53 <quinq> Yeah
2023-06-24 21:10:06 <quinq> frag, use better hardware :
2023-06-24 21:10:08 <quinq> :)
2023-06-24 21:10:16 <quinq> Or sync when nobody's using the rpi
2023-06-24 21:10:56 <frag> just wanted something simpler :] ive been using a script that uses rsync, i think rsync is very picky(?), eg having to have / after dirs or somthing iirc
2023-06-24 21:15:14 <ikke> If you append / to the source, it copies the content (skips parent directory)
2023-06-24 21:19:57 <Habbie> it's not specifically picky; it gives you choices, and those choices are encoded by that single / being there or not
2023-06-24 21:20:05 <Habbie> if you're not used to it it will often do the exact thing you did not want, though
2023-06-25 00:19:30 <powers> I went on a spiritual journey, and I'm back on OFTC where I started
2023-06-25 00:24:27 <Habbie> better than ending up on freenode
2023-06-25 00:32:27 <powers> lmaooo true
2023-06-25 00:33:45 <powers> being honest I'm just really mad. at both the current hardware/software-landscape, and my own incompetence
2023-06-25 00:34:41 <powers> but I'll just have to do my due diligence and grow, just like everyone else
2023-06-25 03:19:20 <LACampbell> a bit off topic, but does anyone know anything about terminal emulators? the only that has been able to handle vim colours properly has been xfce4-terminal. tried alacritty, xterm, urxvt. they all fall apart when trying to display the cursor - they can only do it in one colour
2023-06-25 03:37:58 <raven523> xfce4-terminal uses vte, but xterm and urxvt don't. I don't think alacritty does
2023-06-25 05:50:53 <LACampbell> ah. so does lxterminal actullay
2023-06-25 05:50:59 <LACampbell> and lxterminal doesn't have an issue with it
2023-06-25 08:21:41 <psykose> lxterm is also vte
2023-06-25 14:37:41 <Ceterisparibus0> Hi Everyone.
2023-06-25 14:37:55 <Ceterisparibus0> I have Windows 11 installed and I am trying to install Alpine as dual boot
2023-06-25 14:38:03 <Ceterisparibus0> I could not manage to do this. I need your help !
2023-06-25 14:38:08 <Ceterisparibus0> Can sombeody guide me ?
2023-06-25 14:38:30 <Ceterisparibus0> I have already some knowledge about linux & alpine as well (I am running alpine as server in Raspberry Pi4 already)
2023-06-25 14:41:33 <txnintn> Same hd or a 2nd one?
2023-06-25 14:42:50 <Ceterisparibus0> Same HDD
2023-06-25 14:45:15 <txnintn> Personally I wouldn't do it that way, If your just looking to play with it go with a usb.  
2023-06-25 14:45:34 <txnintn> UEFI system or legacy boot
2023-06-25 14:47:05 <txnintn> I'm setting up my rpi4 with lvn over raid1 with two usb and a sd spare for a docker server.
2023-06-25 14:48:21 <Ceterisparibus0> I already did an installation to USB and it is really performing well. I want to install it to HDD as daily use. Please help. Are there any documentation
2023-06-25 14:49:06 <txnintn> UEFI boot or Legacy?
2023-06-25 15:24:58 <Ceterisparibus0> It is LEGACY
2023-06-25 15:25:32 <quinq> of kain
2023-06-25 15:27:14 <txnintn> Yeah Uefi would been a lot easier..lol
2023-06-25 17:36:13 <txnintn> ping
2023-06-25 18:17:28 <raspbeguy> Hello, just noticed that my server freshly installed with Alpine has constant available entropy, even when using `cat /dev/random` to try to drain it.
2023-06-25 18:18:46 <raspbeguy> Also I installed two servers (bare metal, not VM) and I noticed some that should be random, that were identical on both servers
2023-06-25 18:22:11 <txnintn> I hope not RPI4's <- just got my docker server up and running
2023-06-25 18:39:50 <psykose> `cat /dev/random` doesn't "drain entropy"
2023-06-25 18:40:06 <psykose> what was 'should be random, that were identical on both servers'
2023-06-25 18:44:46 <Apachez> random vs urandom ?
2023-06-25 18:45:31 <psykose> those are the same
2023-06-25 18:47:38 <psykose> (since 5.6 after init)
2023-06-25 18:47:55 <psykose> before init /dev/random will block, after that they are the same output with no difference
2023-06-25 18:49:30 <raspbeguy> psykose: when you install open-iscsi, an initiator name is generated and put in /etc/iscsi/initiatorname.iscsi
2023-06-25 18:50:21 <Kazu[m]> didn't know element allowed you to connect bridge to IRC, that's cool😸
2023-06-25 18:50:27 <raspbeguy> Also I didn't know that /dev/urandom was the same as random now
2023-06-25 18:51:50 <psykose> what generates /etc/iscsi/initiatorname.iscsi
2023-06-25 18:52:12 <psykose> i assume you got the same :xxxx on both machines or w/e
2023-06-25 18:55:20 <raspbeguy> That's right
2023-06-25 18:56:41 <raspbeguy> The program the generate the name is iscsi-iname
2023-06-25 18:59:43 <psykose> if you run this at early boot before rng init this is likely to give you the same result but no other time
2023-06-25 18:59:45 <psykose> when does it run
2023-06-25 19:00:25 <psykose> that said i don't think this is meant to be unique?
2023-06-25 19:03:21 <raspbeguy> psykose: this is meant to be run once, when running the service for the first time. That was not during boot time
2023-06-25 19:03:31 <psykose> this code opens /dev/urandom to read some bytes to feed into md5 and then mixes in gettimeofday() and then opens it again to pick which bytes and then outputs that, and also doesn't fail if it couldn't open /dev/urandom
2023-06-25 19:03:41 <psykose> i don't know who wrote this but if this is expected to be random i would tell you to find something else lol
2023-06-25 19:04:11 <psykose> that said, getting the same result assuming a different time is weird
2023-06-25 19:05:08 <psykose> if you can consistently reproduce it you can go from there
2023-06-25 19:05:27 <raspbeguy> I discovered that because this created a conflict that was messing with the LUN target and getting very bad I/O
2023-06-25 19:06:02 <raspbeguy> I will try on VMs tomorrow 
2023-06-25 19:06:15 <psykose> it also mixes in uname
2023-06-25 19:06:52 <psykose> idk i can't tell if i'm being trolled or not by this code
2023-06-25 19:07:41 <raspbeguy> psykose: I guess it's not intended to be rock-solid collision free generation.
2023-06-25 19:08:07 <psykose> yes, it should be 'good enough', especially because it passes time.tv_usec + time.tv_sec into md5
2023-06-25 19:08:10 <raspbeguy> That's not a cryptography feature
2023-06-25 19:08:37 <psykose> i don't actually see how you get a dupe here unless you get the same uname(3) output + the same time_t struct + /dev/urandom failed to open + it was before rng init (early boot)
2023-06-25 19:09:02 <psykose> so my guess is i am either missing some bug, it's miscompiled due to some bug, or you actually just got the same value via md5 +shuffle luck
2023-06-25 19:09:28 <zcrayfish> Doesn't the kernel print a message when urandom is used before its ready? 
2023-06-25 19:09:40 <raspbeguy> I'll do more comprehensive tests tomorrow 
2023-06-25 19:10:06 <psykose> no idea but even if that was the case it wouldn't be an exact dupe
2023-06-25 19:10:28 <raspbeguy> Anyway, I am very surprised that my available entropy is constant at 256
2023-06-25 19:10:49 <psykose> that said it's 6 bytes, so a dupe is 1 in 255**6 with perfect entropy of everything
2023-06-25 19:11:09 <psykose> reducing that a lot makes it even more likely
2023-06-25 19:11:30 <psykose> 256**6 *
2023-06-25 19:12:04 <psykose> raspbeguy: it always returns that
2023-06-25 19:12:24 <raspbeguy> Why ?
2023-06-25 19:13:04 <raspbeguy> Is that specific to Alpine or is that since kernel 5.6 ?
2023-06-25 19:13:24 <zcrayfish> Since kernel 5.6
2023-06-25 19:13:26 <zcrayfish> AFAIK
2023-06-25 19:13:51 <raspbeguy> Is that a bug or a feature ?
2023-06-25 19:15:54 <psykose> latter
2023-06-25 19:16:53 <psykose> 256 is 'full', i.e. the init is done
2023-06-25 19:17:04 <psykose> from there you can't run out more or less
2023-06-25 19:17:34 <psykose> the exact specific answer is hard to find
2023-06-25 19:19:42 <psykose> somewhat mentioned in https://www.zx2c4.com/projects/linux-rng-5.17-5.18/ i guess
2023-06-25 19:19:51 <psykose> https://www.zx2c4.com/projects/linux-rng-5.17-5.18/inside-linux-kernel-rng-presentation-sept-13-2022.pdf for more context
2023-06-25 19:20:03 <psykose> but the tl;dr is that if it was >256 it wouldn't make any difference
2023-06-25 19:22:31 <psykose> what i don't understand is why this iscsi thing doesn't just do char buf[6]; getrandom(&buf, 6, 0); and delete every other line of code
2023-06-25 19:22:59 <psykose> i guess i don't get paid enough :-)
2023-06-25 19:23:31 <raspbeguy> ¯ \ _ (ツ) _ / ¯
2023-06-25 19:24:33 <raspbeguy> so, are stuffs like haveged still useful?
2023-06-25 19:24:45 <psykose> afaict no
2023-06-25 19:24:55 <psykose> not after 256 init
2023-06-25 19:25:00 <psykose> or.. before, idk
2023-06-25 19:25:14 <psykose> i forget the rules for crediting with the special syscall
2023-06-25 19:39:48 <minimal> raspbeguy: haveged hasn't really been useful for some time. With the recent changes in the kernel's entropy behaviour rngd and jitterentropy-rngd have also become not realy useful (I'm the Alpine maintainer for rng-tools and I've discussed this with the rng-tools author, see https://github.com/nhorman/rng-tools/issues/195) 
2023-06-25 20:07:54 <raspbeguy> good to know
2023-06-25 20:08:25 <Saijin_Naib[m]> Good to know... I've been using it on my Alpine installs 
2023-06-25 20:08:39 <Saijin_Naib[m]> Is it harmful or just... Useless? 
2023-06-25 20:08:46 <raspbeguy> I guess useless
2023-06-25 20:12:07 <VidarReturns[m]> When I change to root shell with `sudo -i` Why is ~/.bashrc Not running? Is this a bug?
2023-06-25 20:12:07 <VidarReturns[m]> I have sudo installed
2023-06-25 20:12:32 <VidarReturns[m]> * have sudo and bash installed
2023-06-25 20:12:39 <lopid> no
2023-06-25 20:13:56 <abby> it would run /root/.bashrc, if root's shell is bash
2023-06-25 20:14:03 <lopid> yes
2023-06-25 20:15:02 <lopid> well, .bash_profile, which may, and some would say should, source .bashrc
2023-06-25 20:16:01 <VidarReturns[m]> Root's shell is bash. When I run echo $SHELL it says /bin/bash, but it did not run .bashrc because nothing from .bashrc was sourced
2023-06-25 20:16:37 <psykose> which bashrc did you expect it to source
2023-06-25 20:16:47 <psykose> full path
2023-06-25 20:17:34 <VidarReturns[m]> After I run sudo -i, then have a bash shell with .bashrc unsourced, Then run bash it will source .bashrc
2023-06-25 20:17:43 <VidarReturns[m]> I don't understand why
2023-06-25 20:17:57 <lopid> are you sure it's bash and not ash?
2023-06-25 20:18:54 <VidarReturns[m]> After running sudo -i, I run echo $SHELL and get /bin/bash
2023-06-25 20:19:12 <lopid> i don't think alpine sets up bash to source .bashrc from .bash_profile, you have to do that yourself
2023-06-25 20:20:18 <psykose> it works fine though https://img.ayaya.dev/Aq0yJV2JytuQ
2023-06-25 20:20:26 <psykose> what did you expect to get sourced
2023-06-25 20:20:47 <psykose> well
2023-06-25 20:20:51 <psykose> just `sudo -i` doesn't i guess
2023-06-25 20:21:14 <psykose> your issue is https://unix.stackexchange.com/questions/228314/sudo-command-doesnt-source-root-bashrc
2023-06-25 20:21:57 <psykose> sudo -i runs login bash and login bash doesn't source bashrc
2023-06-25 20:21:58 <VidarReturns[m]> Thank you
2023-06-25 20:23:45 <VidarReturns[m]> That's unlike other distributions I've used. Hopefully Alpine not be a outsider. Thank you very much for the help ❤️
2023-06-25 20:23:58 <VidarReturns[m]> * Hopefully Alpine will not be
2023-06-25 20:24:19 <lopid> the behaviour is as it should be
2023-06-25 20:26:18 <psykose> i'm sure every other distro has a 500 line /etc/bash/bashrc that does magic in any scenario to source every config file on the system
2023-06-25 20:26:41 <psykose> personally i always found that somewhat annoying
2023-06-25 20:26:41 <lopid> 
ACTION eyes red hat

2023-06-25 20:28:56 <minimal> Saijin_Naib[m]: depends on which you're referring to
2023-06-25 20:30:03 <minimal> if you're referring to haveged then it is harmless as the kernel has had its own jitter equivalent for some time
2023-06-25 20:32:07 <minimal> if you're referring to rngd then then, due to the recent kernel changes, rngd doesn't act as originally designed for reseeding due to the kernel changes in urandom_min_reseed_secs and write_wakeup_threshold and so it can "waste" CPU unnecessarily
2023-06-25 20:46:37 <Saijin_Naib[m]> Ah, excellent, thanks for the summary. I'm using haveged after having read some blogs about it being useful and potentially fixing a papercut I had with lightdm taking a long time to "wake" up from suspend/hibernate 
2023-06-25 20:46:41 <Saijin_Naib[m]> I'll drop it now 😎 Less is morr
2023-06-26 01:07:14 <livy> Hello there. I am using Google Chrome downloaded from Flathub. I know that it is running inside a container, not my host root filesystem.
2023-06-26 01:07:39 <livy> I download a file to ~/Downloads directory (inside the container). Does anyone know where to locate it on the host?
2023-06-26 01:09:02 <livy> I mean can someone instruct me where to get the downloaded file? I've check ~/.var/app/com.google.Chrome but found nothing.
2023-06-26 01:09:28 <livy> Chrome was installed by Flatpak in user mode.
2023-06-26 01:10:11 <Someone[m]1> docker can copy file inside containers to host
2023-06-26 01:10:34 <zcrayfish> Google Chrome can be set to do the same in Flatseal
2023-06-26 01:15:22 <livy> Yes I also know that I can allow it to access host filesystem using Flatseal. But for now, I want to locate the ~/Downloads directory inside its container.
2023-06-26 01:15:34 <psykose> flatpak chrome should put it in downloads already assuming the portals are running
2023-06-26 01:15:36 <psykose> iirc
2023-06-26 01:15:45 <psykose> xdg-desktop-portal-gtk and some xdg-user-dirs configuration
2023-06-26 08:41:33 <tarahb> I'm trying to setup a local repository migrating from pacman. I'm using this tool(https://gitlab.com/Durrendal/toAPK) to convert from PKGBUILD to APKBUILD. Some packages are not getting converted properly, so I'm using `abulid lint .` in my convertion script. It always prints the help message. What should we give to the '<project directories>' argument?
2023-06-26 08:45:05 <psykose> 'abuild lint' is not a thing
2023-06-26 08:48:20 <ikke> There is a seperate tool
2023-06-26 12:13:34 <tarahb> Yes found the tool by Ikke. Thanks people. The default help message is misleading though since it prints, `abuild lint [flags] <project directories>` when I typed `abuild lint`
2023-06-26 12:15:07 <psykose> https://img.ayaya.dev/KOMWjNCiJw1O
2023-06-26 12:21:22 <tarahb> Must be due to different versions I guess. I'm still using the 2.12. https://pastebin.com/Fw5axBKu
2023-06-26 12:21:38 <psykose> that's not abuild
2023-06-26 12:21:57 <psykose> it's something completely different, named abuild
2023-06-26 12:23:09 <psykose> that looks like something for android
2023-06-26 12:26:01 <tarahb> lmao I'm sorry. It seems so. Old abuild seems to run the $1 in shell. I tried `abuild checksum` and it works as expected but `abuild ls` prints the contents of the directory. I guess it's a bug in the old version.
2023-06-26 12:27:51 <psykose> that's not an "old version" it's something completely different
2023-06-26 12:28:24 <psykose> abuild ls running ls is expected because all it does is execute a shell command
2023-06-26 12:28:35 <psykose> +- some wrapping
2023-06-26 12:35:45 <tarahb> `abuild help` gives me the proper abuild help messages. And the `abuild checksum` works properly. https://pastebin.com/SeVfgYHm
2023-06-26 12:43:12 <psykose> that is ancient but yeah it's at least real abuild
2023-06-26 13:20:31 <tarahb> psykose: kk
2023-06-26 14:01:43 <tarahb> Is there any option like `apk add --root` in abuild. I tried the following https://pastebin.com/AFAJsZQ8, seems like it is trying to find the database in / but I don't have any. Though I have the alpine root structure in some other folder. Can I point abuild to that folder?
2023-06-26 14:02:17 <psykose> not really
2023-06-26 14:02:45 <psykose> what you're building doesn't make sense
2023-06-26 14:03:54 <tarahb> I'm just using it for experiment. ninja-bin has the most minimal of dependencies.
2023-06-26 14:05:18 <tarahb> psykose: Having --root feature in abuild is very desirable. Would you mind if I that functionality? Or does that goes againt the philosophy of alpine?
2023-06-26 14:06:29 <psykose> this sounds like a useless feature useful to 1 person with a specific setup
2023-06-26 14:21:29 <tarahb> Can I convince you otherwise? I think this will be helpful for a lot of people. Having --root option in abuild makes it possible to use it with large cross platform native projects like browser. I work on a rendering engine and managing those big dependencies is an utter pain. The pain keeps increasing for every platform we start to support. If abuild is able to make packgages
2023-06-26 14:21:31 <tarahb> locally, people can basically use this as an cpp package manager, bascially freeing themselves from vcpkg hell. I guess we just have to modify the /etc/apk/world? given the proper repositories are configured. Also homebrew is pain to use in mac because of ruby implementation. Mac doesn't allow to change root without disabling SIP, so building using abuild there is extremely
2023-06-26 14:21:33 <tarahb> desirable. Another use case is using apk in windows, windows doesn't have chroot of any kind, with abulid the packages can be made easily available there without much effort.
2023-06-26 14:25:27 <psykose> abuild isn't a package manager
2023-06-26 14:26:01 <psykose> i have no idea what you're imagining but i don't think you know what you're talking about
2023-06-26 14:26:13 <psykose> and there are already like 50 projects that are 'build system for xyz to manage dependencies'
2023-06-26 14:26:16 <psykose> maybe more than 50
2023-06-26 14:26:29 <psykose> that do what you are describing
2023-06-26 14:33:50 <tarahb> psykose: I understand that abuild isn't a packgae manager but there is virtually no difference between apk+abuild combo and other package managers for native projects. Good c++ package managers makes sysroot locally without affecting the / so the build system can find pkg-config files inside the sysroot. So they basically have the UNIX filesystem structure like /bin /lib /share
2023-06-26 14:33:53 <tarahb> /etc. There is literally no difference in functionality. We pretty much use pacman to handle cpp dependencies in our org. But pacman has bash as the base system which doesn't work properly in native windows. And is really hard to port pacman to new systems.
2023-06-26 14:34:07 <psykose> yes, 'Good c++ package managers makes sysroot locally without affecting the /'
2023-06-26 14:34:34 <psykose> abuild is not a 'good c++ package manager' for 'windows to build web browser without chroot'
2023-06-26 14:34:46 <psykose> it's merely a small build tool for some alpine packages +- some adjacent third party trees
2023-06-26 14:39:46 <psykose> that doesn't mean you can't just fork it to do whatever you want, obviously
2023-06-26 14:40:06 <psykose> for ""your org"" assuming it ""literally does what you want"" this is by far the easiest option and would take you 5 minutes to do anything with it
2023-06-26 14:41:09 <psykose> it itself though, is an internal alpine tool for packaging, and i really don't care if someone in some business somewhere builds some web browsers or whatever or what the usecase is
2023-06-26 14:41:17 <psykose> it's not the audience and i don't care what's useful for that target
2023-06-26 14:41:57 <psykose> doesn't stop you from just forking it, which takes 5 minutes :)
2023-06-26 14:45:02 <tarahb> The thing is there is no good c++ package management tool, and it's not possible to have one. And most c++ projects or big projects just don't use the same languages. There is atleast a mix of two or three languages and no package managers that I know of handles this well. For example we use a tool called djinni to convert IDL files into cpp and hpp files for binding with other
2023-06-26 14:45:05 <tarahb> languages, but djinni has to installed in the system and it's dependency java has to be installed. We can handle the invocation in a simple Makefile but the distribution part is really messy. That's where I think abuild+apk can be useful.
2023-06-26 14:45:29 <psykose> you already have an answer
2023-06-26 14:48:25 <aron> "c++ package management tool" ugh
2023-06-26 14:48:50 <aron> what's wrong with the distro's package management tool?
2023-06-26 14:49:12 <tarahb> psykose: hahaha okay, I had to try. Thanks for answering though.
2023-06-26 14:52:22 <tarahb> aron: By c++ I really meant native projects. There is disparity between the available package in distro and the ones used by projects. Say if I need protobuf 2.19, the distro might only have 22.1 or something like that.
2023-06-26 14:53:25 <aron> just let distros package your stuff, and set proper version requirements if any
2023-06-26 14:58:04 <tarahb> aron: In cases where the dependencies are old we can't expect the distros to support them, only option is using static linking in that case.
2023-06-26 14:59:38 <aron> indeed, but what's the problem with that?
2023-06-26 15:00:05 <aron> add that project's specific version as a source dependency
2023-06-26 15:00:15 <aron> download it, configure it, and link against it
2023-06-26 15:00:22 <aron> nothing fancy
2023-06-26 15:01:06 <aron> also, keep in mind, every situation is different, you probably can't make a tool that is suitable for all kind of situations
2023-06-26 15:02:10 <aron> and it might be more future-proof, if you keep it as simple as possible, and not depend from another tool
2023-06-26 15:03:33 <aron> building against very old libs is always a pain, especially if the lib is not designed carefully
2023-06-26 15:03:50 <aron> maintaining libs in long term is hard!
2023-06-26 15:09:15 <tarahb> aron: It doesn't have to be if maintaining a repository with all the dependencies. The problem with adding another project as soure dependency is that there is no agreement to use the same build system like nodes packgaes.json. Some people use cargo, some use make, some use cmake, etc. Having an APKBUILD for every dependency and installing it in a common sysroot and adding a
2023-06-26 15:09:18 <tarahb> -Isysroot/include is the easiest and less messiest and ultimate perfect way of managing dependencies I can think of.
2023-06-26 15:12:24 <aron> I disagree
2023-06-26 15:12:35 <aron> anyway, you can always fork abuild if you want
2023-06-26 15:18:58 <tarahb> aron: okay cool
2023-06-26 16:05:17 <Saijin_Naib[m]> How can I troubleshoot what appears to be a bubblejail issue with Evolution on edge?
2023-06-26 16:05:44 <Saijin_Naib[m]> It never loads, though the processes are running. If I launch via terminal, I get this wholly unhelpful print
2023-06-26 16:06:20 <Saijin_Naib[m]> bwrap: Can't make symlink at /var/run: File exists
2023-06-26 16:06:20 <Saijin_Naib[m]> evolution 
2023-06-26 16:07:41 <psykose> that sounds like something being confused by /var/run -> /run
2023-06-26 16:07:41 <psykose> hmm
2023-06-26 16:07:56 <psykose> and by something i mean webkit
2023-06-26 16:08:25 <psykose> i can't reproduce it for some reason
2023-06-26 16:08:53 <Saijin_Naib[m]> Fresh install of 3.18.2 upgraded to edge
2023-06-26 16:09:05 <psykose> if you start e.g. epiphany do you get the same
2023-06-26 16:09:08 <Saijin_Naib[m]> evolution, evolution-on, evolution-ews, evolution-data-server
2023-06-26 16:09:11 <Saijin_Naib[m]> let me check
2023-06-26 16:09:48 <Saijin_Naib[m]> bwrap: Can't make symlink at /var/run: File exists
2023-06-26 16:09:48 <Saijin_Naib[m]> epiphany 
2023-06-26 16:10:08 <psykose> this sounds like https://github.com/flatpak/flatpak/issues/5255#issuecomment-1380445755
2023-06-26 16:10:22 <Saijin_Naib[m]> it spawns a zombie Epiphany that shows up in my XFCE task list, but doesn't actually draw and just blocks the screen
2023-06-26 16:10:30 <Saijin_Naib[m]> Yeah, but neither are via flatpak
2023-06-26 16:10:35 <psykose> not related to flatpak
2023-06-26 16:10:38 <psykose> read closer
2023-06-26 16:10:39 <Saijin_Naib[m]> and my flatpaks run fine, of course
2023-06-26 16:11:01 <psykose> it's some magic media symlink stuff
2023-06-26 16:11:02 <Saijin_Naib[m]> hmmm
2023-06-26 16:11:06 <psykose> or other symlink stuff
2023-06-26 16:11:08 <Saijin_Naib[m]> Is this safe to mess with?
2023-06-26 16:11:19 <Saijin_Naib[m]> I assume not, otherwise Alpine would have done it by default, right?
2023-06-26 16:11:36 <psykose> you should "look" not "mess" and see what's where
2023-06-26 16:11:37 <psykose> idk
2023-06-26 16:11:52 <psykose> i don't have any symlinks for instance
2023-06-26 16:12:13 <Saijin_Naib[m]> Wha, how?
2023-06-26 16:12:31 <Saijin_Naib[m]> Yeah, I have a symlink in /var/run that points to /run
2023-06-26 16:12:36 <psykose> no, that one is fine
2023-06-26 16:12:40 <psykose> i mean /media
2023-06-26 16:14:27 <Saijin_Naib[m]> /media does not appear to be a symlink, but rather just a directory at the root
2023-06-26 16:14:44 <Saijin_Naib[m]> same goes for /var/run/media
2023-06-26 16:14:51 <Saijin_Naib[m]> that's just a directory in run
2023-06-26 16:14:57 <psykose> `ls -l /; ls -l /media; ls -l /run; ls -l /var`
2023-06-26 16:15:35 <Saijin_Naib[m]> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/CpdSXyLccyzaICPmgpuegYBI

2023-06-26 16:16:32 <pj> preferably as pastebin next time
2023-06-26 16:16:38 <psykose> that is a pastebin
2023-06-26 16:16:42 <psykose> Saijin_Naib[m]: looks normal
2023-06-26 16:17:03 <pj> psykose: wrong
2023-06-26 16:17:23 <Saijin_Naib[m]> psykose: is gdb the correct next step?
2023-06-26 16:20:29 <psykose> more so `strace -f -s 999 epiphany` but it will have a ton of logs
2023-06-26 16:22:31 <psykose> (add -o file too and paste that)
2023-06-26 16:24:31 <Saijin_Naib[m]> Shit, 6.4MB
2023-06-26 16:24:42 <psykose> sounds fine
2023-06-26 16:24:50 <Saijin_Naib[m]> Can I upload that, or actually copy/paste the log so the IRC software pastebins it for me>?
2023-06-26 16:24:58 <psykose> it's the same shit
2023-06-26 16:25:04 <psykose> it's a url and i get it
2023-06-26 16:25:22 <Saijin_Naib[m]> 
ACTION posted a file: saijin_naib-epiphany-strace (6278KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/svEarbJFlGFSxLiwNYUzoFds >

2023-06-26 16:25:23 <Saijin_Naib[m]> Ah, awesome, thanks
2023-06-26 16:25:27 <psykose> thanks
2023-06-26 16:27:23 <psykose> so yeah it seems like the same immediate symptoms that it's trying to add newroot/var/run -> ../run and it already exists...
2023-06-26 16:27:25 <psykose> not sure why
2023-06-26 16:28:17 <Saijin_Naib[m]> hmm... I don't have anything custom in my profile.d
2023-06-26 16:28:28 <psykose> not really related
2023-06-26 16:28:32 <Saijin_Naib[m]> I did restore my /home from my previous install on this machine (that worked fine minus polkit issues)
2023-06-26 16:28:48 <Saijin_Naib[m]> So this is lower-level than profile/home?
2023-06-26 16:29:10 <Saijin_Naib[m]> As far as I know, I've done nothing funky aside from switch to edge and restore a backed-up apk world and restoring my /home folder
2023-06-26 16:33:18 <psykose> what does `doas tree /run | grep var` give for you
2023-06-26 16:36:55 <Saijin_Naib[m]> Empty output
2023-06-26 16:37:17 <Saijin_Naib[m]> Null? It ran successfully, it just didn't find anything
2023-06-26 16:37:52 <psykose> makes sense
2023-06-26 16:38:18 <psykose> no good ideas i guess
2023-06-26 16:43:43 <Saijin_Naib[m]> Fuuuuuck 
2023-06-26 16:43:47 <Saijin_Naib[m]> Thank you, though 
2023-06-26 16:44:07 <Saijin_Naib[m]> Brain is itching that it might be something broken with XFCE and AT-SPI
2023-06-26 16:44:28 <Saijin_Naib[m]> At least Builder works, so half my work environment is good 
2023-06-26 16:44:30 <Saijin_Naib[m]> 🤣
2023-06-26 18:27:18 <uuidNuniq> the zstd compress level 19 needs to go away.
2023-06-26 18:27:37 <uuidNuniq> that's crazy for initramfs. 
2023-06-26 18:28:05 <uuidNuniq> just use the default 3 
2023-06-26 18:48:16 <Saijin_Naib[m]> Decompress should be mostly transparent even at that level, no? 
2023-06-26 18:56:22 <uuidNuniq> Saijin_Naib[m]: decompress has nothing to do with compress level. 
2023-06-26 18:59:19 <Saijin_Naib[m]> I'm aware. I'm just confused about why you would care about compression level used when it really has no bearing on decompress, which is going to be the more frequent interaction with a compressed file 
2023-06-26 18:59:20 <Apachez> not entirely true
2023-06-26 19:00:29 <Saijin_Naib[m]> For ZSTD, the variance in decompress is pretty insignificant, though for other compression methods it isn't always, you're right
2023-06-26 19:28:27 <quinq> uuidNuniq, going away from where?
2023-06-26 19:28:34 <quinq> It seems that the default is still gzip
2023-06-26 19:29:22 <quinq> I was just about to ask how to modify the default mkinitfs compression scheme ^^
2023-06-26 20:24:19 <quinq> uuidNuniq, found it, zstd) cmd_exists zstd; comp="zstd -T0 -19" ;;
2023-06-26 20:33:20 <uuidNuniq> quinq: you can set a varaible in /etc/mkinitfs/mkinitfs.conf
2023-06-26 20:33:53 <uuidNuniq> to use zstd
2023-06-26 20:33:54 <Apachez> there are some not neglible differences when you choose different compression level and blocksize
2023-06-26 20:34:28 <uuidNuniq> quinq: the script also detect pigz. so installing pigz reduces time to create gzip initramfs 
2023-06-26 20:35:12 <Apachez> depending on cpu the sweatspot often is at arond 64-128kbyte blocksize
2023-06-26 20:35:37 <Apachez> anything higher than that (which higher compression by default often utlizes) will bring more cachemisses so you end up with lesser performance
2023-06-26 20:35:50 <Apachez> its like you have two graphs crossing each other
2023-06-26 20:36:59 <Apachez> sure for pure storage I also prefer maximum compression but for lets say a ramdisk using maximum compression is often a bad thing... often -c6 or -c7 is the sweetspot where -c9 is maximum
2023-06-26 20:38:16 <quinq> uuidNuniq, I couldn't find anything about that in the man-page
2023-06-26 20:38:35 <Apachez> for example with xz compression and ramdisks using 256k as blocksize instead of 1M would increase the readspeed by 14% (compared to using 1M) with the cost of about 5% larger image
2023-06-26 20:38:40 <quinq> It only says “Default configuration file. It contains the list of default features in the format of features="...".”
2023-06-26 20:39:07 <quinq> And I couldn't find any module about that
2023-06-26 20:39:17 <Apachez> https://jonathancarter.org/2015/04/06/squashfs-performance-testing/
2023-06-26 20:39:54 <uuidNuniq> quinq: initfscomp=zstd 
2023-06-26 20:40:43 <quinq> I guess that calls for a documentation patch :D
2023-06-26 20:40:44 <quinq> Thanks
2023-06-26 20:40:55 <uuidNuniq> Line 243: . $(readlink -f "$config")
2023-06-26 20:41:21 <uuidNuniq> that will replace the default initfscomp=gzip
2023-06-26 20:41:48 <uuidNuniq> Line 301: gzip) comp="$(command -v pigz 2>/dev/null || echo gzip) -9" ;;
2023-06-26 20:41:57 <quinq> I meant documentation in the man-page :)
2023-06-26 20:42:00 <uuidNuniq> that will use pigz instead of gzip if available
2023-06-26 20:42:28 <quinq> yep
2023-06-26 20:43:51 <uuidNuniq> but they need to drop the "-19" that is too slow. 
2023-06-26 20:45:24 <uuidNuniq> i mean alpine is a rolling release, there will be only one initramfs in /boot, the size reduction that -19 brings is merely significant
2023-06-26 20:46:27 <abby> alpine is not rolling
2023-06-26 20:47:07 <uuidNuniq> oh, right. i mean the kernel is rolling 
2023-06-26 20:48:07 <uuidNuniq> e.g. one kernel image, one initramfs img at any given time.
2023-06-26 20:48:23 <uuidNuniq> not like you got 10 initramfs where size matters 
2023-06-26 20:49:00 <ikke> well, size does matter, as people are running into space issues for their /boot partitions lately
2023-06-26 20:49:11 <ikke> (I've already run into that on one system)
2023-06-26 20:49:42 <uuidNuniq> hmm... too many modules enabled?
2023-06-26 20:50:17 <uuidNuniq> or not planned well in the beginning? or use /boot as ESP? 
2023-06-26 20:50:58 <ikke> /boot was 100M by default until not long ago
2023-06-26 20:51:02 <quinq> how much space does -19 actually save compared to default level?
2023-06-26 20:51:06 <zcrayfish> The default list of mkinitfs features that are enabled always loads more than you need to boot your system.
2023-06-26 20:51:22 <zcrayfish> Not always a bad though though, if you get dumped to an emergency shell, some of those things are nice to have
2023-06-26 20:51:24 <uuidNuniq> quinq: depends on how many modules you enable
2023-06-26 20:51:37 <ikke> In this case, it's vmlinuz that takes most space, though
2023-06-26 20:52:04 <quinq> uuidNuniq, with the same modules.
2023-06-26 20:52:28 <uuidNuniq> quinq: you can do a test :)
2023-06-26 20:53:09 <uuidNuniq> modify the mkinitfs script remove -19 and genereate new initramfs. and compare with previous size
2023-06-26 20:53:17 <quinq> I was hoping that people claming it matters already did that so that I wouldn't have to do it again
2023-06-26 20:53:49 <uuidNuniq> zcrayfish: i always trim it to the minimal. 
2023-06-26 20:54:28 <uuidNuniq> i have 2 modules enabled actually : the base, and my custom .module in /etc/mkinitfs/feature.d/
2023-06-26 20:54:31 <zcrayfish> same. if something fails at this point I just use a rescue disk (well USB flash drive, but still)
2023-06-26 20:55:41 <uuidNuniq> i use zstd level 1 where possible 
2023-06-26 20:56:03 <uuidNuniq> e.g. i use btrfs mount option : -o compress=zstd:1
2023-06-26 20:56:29 <uuidNuniq> that is a great level. it makes my rootfs 50% size of ext4
2023-06-26 20:57:12 <uuidNuniq> and i don't see any speed difference compared to ext4
2023-06-26 20:58:25 <uuidNuniq> btw: btrfs is really great :) although i don't use the features like subvolume or snapshot
2023-06-26 20:58:34 <uuidNuniq> too complex to me. 
2023-06-26 20:59:25 <uuidNuniq> i just use it like it is ext4. 
2023-06-26 20:59:50 <uuidNuniq> but with meta+data csum + compress
2023-06-26 21:02:07 <uuidNuniq> oh btw, btrfs does not need fsck :) so in fstab it is  0 0 
2023-06-26 21:02:22 <uuidNuniq> but i do monthly scrub
2023-06-26 21:02:28 <uuidNuniq> with a cron
2023-06-26 21:03:22 <zcrayfish> yeah, I switched to btrfs just for compression and deduplication 
2023-06-27 06:46:25 <Forza> Hello :)
2023-06-27 06:49:17 <OdaNobunaga[m]> <uuidNuniq> "btw: btrfs is really great..." <- Just bear that in mind: https://lists.alpinelinux.org/~alpine/infra/%3C20201018103924.GC17790%40home.c-otto.de%3E
2023-06-27 06:50:32 <Forza> OdaNobunaga[m]: doesn't say why the filesystem failed
2023-06-27 06:50:54 <Forza> Any filesystem can get damaged for various reasons 
2023-06-27 06:53:10 <Forza> I wanted to suggest that Alpine provides a virtiofs install image in addition to the standard kvm image on the download page 
2023-06-27 06:59:57 <Forza> It's basically just a rootfs in a directory. Then provide libvirt XML domain config or a qemu command line wrapper script to launch virtiofsd and enable direct kernel boot. 
2023-06-27 07:00:58 <LACampbell> am I supposed to start wireplumber is an openRC service? the wiki isn't clear on this. also it doesn't seem to install an openRC service
2023-06-27 07:02:03 <psykose> no
2023-06-27 07:02:53 <Forza> psykose: what do you think of my proposal? 
2023-06-27 07:03:05 <psykose> what proposal
2023-06-27 07:03:33 <Forza> I suppose there isn't much interest in virtiofs as rootfs? 
2023-06-27 07:03:41 <psykose> i have no idea what that is
2023-06-27 07:03:46 <Forza> Ah
2023-06-27 07:03:57 <Forza> Then it's hard to know :) 
2023-06-27 07:04:25 <Forza> Virtiofs is a shared filesystem between hosr/hypervisor and the guest 
2023-06-27 07:04:37 <Forza> Like nfs, Samba or plan9 
2023-06-27 07:04:42 <psykose> i know that part
2023-06-27 07:04:48 <psykose> i don't see what's stopping you from doing it
2023-06-27 07:04:54 <Forza> But is built on virtio instead of a network filesystem 
2023-06-27 07:05:16 <Forza> Nothing. I use it with alpine and it works very well 
2023-06-27 07:05:34 <Forza> In fact, much better than with Debian and Ubuntu 
2023-06-27 07:06:51 <Forza> I guess the normal installer would work fine. The only change I think needs to be done is excluding a boot loader
2023-06-27 07:07:36 <Forza> I might try documenting the process for others to use
2023-06-27 07:11:23 <Forza> psykose: example https://gist.tnonline.net/PQ 
2023-06-27 07:11:54 <psykose> looks normal
2023-06-27 07:12:53 <Forza> https://gist.tnonline.net/2E mount output 
2023-06-27 07:13:05 <Forza> But yes. That's the point. It's normal 
2023-06-27 07:13:36 <Forza> Just a shared dir instead of dedicating a virtual block device 
2023-06-27 07:15:02 <Forza> I need to check if the installer has the fuse.ko and virtiofsd.ko modules 
2023-06-27 07:16:31 <Forza> IMHO it makes real good sense to use alpine on virtiofs. It minimise the storage footprint and memory usage because the guest doesn't have to load filesystem drivers etc. 
2023-06-27 08:01:57 <LACampbell> what's the big attraction of pipewire over pulse? you need a lot less packages installed to get pulse working
2023-06-27 08:03:19 <abby> 1. it was designed to work with flatpak, 2. you can use it to do screen recording on wayland, 3. it can integrate with alsa, pulse, jack, probably others
2023-06-27 08:05:37 <luks2[m]> whats the consensus on flatpak ive heard a lot of negative things
2023-06-27 08:05:58 <abby> i don't think there is consensus on flatpak
2023-06-27 08:06:27 <abby> but it's just a fancy wrapper around running a program in a container (in this case, using bubblewrap)
2023-06-27 08:09:04 <psykose> 4. it works more gooder
2023-06-27 08:10:22 <abby> that too
2023-06-27 08:12:29 <pj> 5. it's not snap
2023-06-27 08:13:12 <psykose> 6. it got pipes and it got wires, great for electrical plumbers
2023-06-27 08:14:07 <ikke> 😨
2023-06-27 08:14:08 <pj> 7. nostalgic reference to windows xp screensaver
2023-06-27 08:18:23 <psykose> 8. i put on my robe and wizard hat
2023-06-27 08:29:45 <pj> 9. Are you going to ponder your orb?
2023-06-27 08:39:41 <teapot9> podman doesn't work, i get `OCI runtime error: crun: invalid file system type on '/sys/fs/cgroup'`, even if cgroups service is started
2023-06-27 08:41:06 <psykose> what's the output of `mount`
2023-06-27 08:45:11 <ikke> teapot9: did you reboot after a kernel upgrade?
2023-06-27 08:51:00 <teapot9> psykose: mount: https://bpa.st/4CO7W
2023-06-27 08:51:41 <teapot9> ikke: rebooted not long ago, no recent kernel update
2023-06-27 08:52:03 <psykose> hmm
2023-06-27 08:52:09 <psykose> does it not work anymore with hybrid mode
2023-06-27 08:52:33 <teapot9> switched to hybrid after getting the error with unified
2023-06-27 08:52:58 <psykose> which one is the mount output from
2023-06-27 08:53:10 <teapot9> wait, rebooted again and it works
2023-06-27 08:53:13 <teapot9> i dont understant
2023-06-27 08:56:17 <teapot9> switched back to unified and it also works, thanks for the help
2023-06-27 09:56:39 <doodoo> Hello, I've just installed Alpine Linux on my PC, however there is no eth0 network interface showing up. It was present during the install. Might anyone know what to do to fix it ?  ifconfig -a outputs "lo" only.
2023-06-27 10:11:11 <Forza> doodoo: did you configure networking properly? 
2023-06-27 10:11:47 <Forza> What network card is it? 
2023-06-27 10:12:28 <doodoo> When it asked for what interface to use I selected eth0, and dhcp for the next question
2023-06-27 10:12:34 <doodoo> I'm not too sure, how do I check exactly ?
2023-06-27 10:14:45 <Forza> doodoo: what's the output of lspci -k (use a pastebin service) 
2023-06-27 10:16:07 <doodoo> let me boot into my other os as alpine doesnt have lspci and pciutils does not seem to  be a package
2023-06-27 10:17:53 <Forza> Don't remember what the package is called 
2023-06-27 10:18:19 <aron> pciutils
2023-06-27 10:18:42 <doodoo> https://pastebin.com/ahgwjJVX
2023-06-27 10:18:48 <Forza> https://pkgs.alpinelinux.org/package/v3.18/main/x86_64/pciutils
2023-06-27 10:18:50 <doodoo> not sure but it was telling me pciutils is not found for some reason
2023-06-27 10:19:18 <Forza> r8169
2023-06-27 10:19:27 <Forza> Seems you have a realtek card 
2023-06-27 10:20:40 <doodoo> does that imply something ?
2023-06-27 10:21:06 <Forza> Yes that it is found by the kernel and a driver was loaded 
2023-06-27 10:21:23 <doodoo> then why is it not showing up in alpine ?
2023-06-27 10:21:39 <Forza> Check /etc/network/interfaces 
2023-06-27 10:21:45 <doodoo> it contains eth0
2023-06-27 10:21:57 <doodoo> (and lo for that matter)
2023-06-27 10:22:25 <ikke> doodoo: but `ip link show` does not return it?
2023-06-27 10:22:28 <aron> Forza: "let me boot into my other os" <-- doodoo
2023-06-27 10:22:57 <doodoo> I dont have lspci on my alpine, so I booted into my nix to check
2023-06-27 10:23:12 <Forza> Ah.. 
2023-06-27 10:23:13 <aron> doodoo: check   ip link show
2023-06-27 10:23:19 <aron> that's a good idea
2023-06-27 10:23:33 <doodoo> alright 1 sec, lemem boot back into alpine
2023-06-27 10:23:50 <Forza> But you can also check in dmesg or /proc/interrupts
2023-06-27 10:24:40 <aron> also: lsmod | grep r8169
2023-06-27 10:25:34 <psykose> might also be missing firmware (iirc that one is linux-firmware-rtl_nic)
2023-06-27 10:25:41 <psykose> though by default all the firmware is installed post-install
2023-06-27 10:26:37 <doodoo> well my kb is not inputting into alpine rn.... out of nowhere
2023-06-27 10:26:41 <doodoo> so ill be a hot minute
2023-06-27 10:31:51 <doodoo> ok on boot I see this error "ip ioctl 0x8913 failed No such device", followed with "udhcpc ioctl 0x8933 failed: No such device". I am still not able to login as my kb is not working (i just reinstalled alpine too...)
2023-06-27 10:32:47 <psykose> the ioctl thing is some race condition somewhere that i didn't figure out
2023-06-27 10:33:35 <Forza> :/
2023-06-27 10:33:47 <doodoo> ok what is "/proc/sys/kernel/hotplug" as it is not creating it
2023-06-27 10:33:52 <doodoo> and my keyboard isnt working
2023-06-27 10:33:56 <doodoo> i am guessing it might be related
2023-06-27 10:34:11 <doodoo> how tf did this break, even after a fresh install. I am so confused.
2023-06-27 10:35:44 <psykose> how you installed is relevant since it affects the mkinitfs features etc
2023-06-27 10:36:10 <doodoo> I see, I am installing it onto a partition as I am dualbooting. 
2023-06-27 10:36:28 <doodoo> mount -t ext4 /dev/sdXY /mntsetup-disk -m sys /mnt
2023-06-27 10:36:51 <doodoo> mount -t ext4 /dev/sdXY /mnt                              setup-disk -m sys /mnt
2023-06-27 10:37:15 <doodoo> i am using a partition that was formatted to ext4 before tho, i didnt use mkfs.ext /dev/sdXY
2023-06-27 10:37:19 <doodoo> do i need to use mkfs ?
2023-06-27 10:37:57 <psykose> wouldn't matter no
2023-06-27 10:38:41 <doodoo> ill just reformat to hopefully get that kb working...
2023-06-27 10:38:44 <doodoo> and reinstall
2023-06-27 10:39:02 <psykose> don't think reinstalling and doing the same thing would change anything
2023-06-27 10:39:30 <doodoo> im really confused as the kb was working when i first joined the IRC
2023-06-27 10:40:14 <psykose> what's in features= in /etc/mkinitfs/mkinitfs.conf after install
2023-06-27 10:47:22 <doodoo> well 1 sec as now when i try boot into it i am gteting filesystem errors
2023-06-27 10:50:51 <doodoo> « features="ata base ide scsi usb virtio ext4" »
2023-06-27 10:55:56 <psykose> seems ok
2023-06-27 10:58:42 <doodoo> ok i rly dont know what to do without my keyboard working in it now
2023-06-27 10:58:53 <doodoo> it literally broke from one both to another, reinstalling makes no difference
2023-06-27 10:59:42 <doodoo> boot*
2023-06-27 10:59:49 <aron> doodoo: usb keyb?
2023-06-27 10:59:54 <doodoo> yeah
2023-06-27 11:00:05 <aron> tried to unplug/replug?
2023-06-27 11:00:16 <doodoo> ok i put it in another port and it works
2023-06-27 11:00:33 <doodoo> unplug/replug didnt
2023-06-27 11:00:45 <aron> different port, works too
2023-06-27 11:00:53 <doodoo> ip link show only shows "lo"
2023-06-27 11:01:01 <aron> at least you have keyb now :)
2023-06-27 11:01:04 <doodoo> yes haha
2023-06-27 11:01:12 <doodoo> now i need networking
2023-06-27 11:01:31 <aron> what is this yield?   lsmod | grep r8169
2023-06-27 11:01:49 <doodoo> nothing
2023-06-27 11:02:09 <aron> try this:  modprobe r8169
2023-06-27 11:02:21 <aron> it should print it's problems, if any
2023-06-27 11:02:27 <doodoo> "Module not found in directory /lib/modules/5.18.6-arch1-1
2023-06-27 11:02:34 <aron> ooooh
2023-06-27 11:03:30 <doodoo> well i understand whats wrong
2023-06-27 11:03:39 <doodoo> but im not too sure why its not there
2023-06-27 11:03:46 <aron> yeah, it's wrong
2023-06-27 11:04:17 <doodoo> i suppose its an alpine issue ?
2023-06-27 11:05:20 <ikke> doodoo: could you try the live iso again?
2023-06-27 11:05:27 <aron> or: it is not a *module* module anymore
2023-06-27 11:05:29 <doodoo> the live iso works
2023-06-27 11:05:29 <ikke> and see what the situation is there, given that it did work there
2023-06-27 11:05:33 <doodoo> sure
2023-06-27 11:05:36 <doodoo> ill reboot into it
2023-06-27 11:06:29 <aron> hmm, what kind of kernel is that "5.18.6-arch1"?
2023-06-27 11:06:38 <ikke> that's not alpinelinux
2023-06-27 11:06:41 <ikke> that's archlinux
2023-06-27 11:06:54 <doodoo> but i installed alpine linux haha
2023-06-27 11:06:56 <doodoo> idk...
2023-06-27 11:07:08 <doodoo> wait does it have something to do with booting maybe ?
2023-06-27 11:07:20 <doodoo> coz i copied the boot config from my archlinux
2023-06-27 11:07:27 <doodoo> (systemd-boot)
2023-06-27 11:07:36 <aron> sounds like bad rootfs is mounted
2023-06-27 11:07:54 <doodoo> oh, how do i go about that ?
2023-06-27 11:08:13 <aron> I checked for r8169.ko.gz, and it is a thing in all alpine -lts kernels
2023-06-27 11:08:13 <doodoo> (i suppose its not particularly alpine linux related then but id still appreciate any help if you know)
2023-06-27 11:08:39 <aron> first:   uname -a
2023-06-27 11:08:46 <aron> second:    cat /etc/alpine-release
2023-06-27 11:09:02 <Forza> doodoo: you probably booted wrong kernel, then the alpine modules won't work 
2023-06-27 11:09:24 <aron> (uname "comes from" the initramfs, and alpine-release is just a regular file)
2023-06-27 11:09:35 <ikke> uname comes from the running kernel
2023-06-27 11:09:38 <ikke> (it's a syscall)
2023-06-27 11:09:42 <doodoo> alpine-release is 3.18.2
2023-06-27 11:09:46 <ikke> and the kerken comes from the boot partition
2023-06-27 11:09:49 <ikke> kernel*
2023-06-27 11:09:55 <aron> hm, sure?
2023-06-27 11:10:05 <doodoo> uname -a is "Linux swejqa 5.18.6-arch1-1 #1..... x86_64 Linux"
2023-06-27 11:10:08 <doodoo> haha yeah im running arch kernel
2023-06-27 11:10:16 <doodoo> idek where it is, I uninstalled arch
2023-06-27 11:10:28 <doodoo> I see boot partition
2023-06-27 11:10:37 <aron> doodoo: so definitely different kernel
2023-06-27 11:10:54 <Forza> dmesg |less will tell you among the first few lines what kernel and command line arguments and rootfs you have 
2023-06-27 11:11:10 <doodoo> i i didnt realise the kernel is in the boot partition 
2023-06-27 11:11:25 <Forza> It is 
2023-06-27 11:11:40 <Forza> But you don't have to have a boot partition 
2023-06-27 11:11:48 <aron> ikke: right, my bad
2023-06-27 11:11:54 <doodoo> I mean since im dualbooting tho, I think I need one
2023-06-27 11:13:14 <doodoo> i suppose i update the vmlinuz-linux, initramfs-linux.img and initramfs-linux-fallback.img insied the boot partition ?
2023-06-27 11:13:33 <ikke> alpine does not have a fallback image
2023-06-27 11:13:36 <ikke> that's archlinux
2023-06-27 11:13:41 <doodoo> ah ok
2023-06-27 11:13:45 <doodoo> well the first 2 are what matter anyway
2023-06-27 11:13:49 <aron> (at least we know that the kernel version is incorrect)
2023-06-27 11:14:42 <ikke> it's initramfs-lts for alpine
2023-06-27 11:14:58 <ikke> and vmlinuz-lts
2023-06-27 11:15:06 <doodoo> yeah makes sense
2023-06-27 11:15:12 <doodoo> ok ill do it and get back
2023-06-27 11:23:50 <psykose> yeah you booted some other kernel but everything else is from alpine
2023-06-27 11:24:00 <psykose> so there are no modules available for it
2023-06-27 11:25:20 <doodoo> just to confirm i copy the initramfs-lts and vmlinuz-lts  from /boot in the alpine partition and use those to boot correct ?
2023-06-27 11:25:41 <psykose> no idea
2023-06-27 11:26:01 <ikke> What bootloader are you using?
2023-06-27 11:26:04 <doodoo> systemd-boot
2023-06-27 11:26:27 <psykose> i haven't done dualbooting where i had to copy random files around before
2023-06-27 11:26:35 <doodoo> im getting an issue mounting the alpine partition "/sysroot: No such file or directory"
2023-06-27 11:26:49 <doodoo> haha I dont wanna fuck up my windows boot tbh, its such a pain in the arse
2023-06-27 11:27:16 <ikke> Can you see the boot entry for alpinelinux?
2023-06-27 11:27:21 <ikke> what arguments etc
2023-06-27 11:27:37 <ikke> (I'm not that familiar with systemd-boot)
2023-06-27 11:27:38 <aron> doodoo: I shoved my old windoz system into a VM, and no need for dual boot :)
2023-06-27 11:27:52 <doodoo> sadly I only use windows for gaming with AC like valorant
2023-06-27 11:28:05 <doodoo> wait let me open up the entry for alpine boot
2023-06-27 11:29:11 <aron> passing vga with pci passthrough works like a charm
2023-06-27 11:30:12 <doodoo> i think AC is finicky in VMs tho
2023-06-27 11:30:29 <doodoo> anyway, arguments are "root=PARTUUID=.... rw"
2023-06-27 11:30:39 <doodoo> the issue is with the /sysroot file tho
2023-06-27 11:31:06 <aron> (AC stands for... alternate current? :P)
2023-06-27 11:31:07 <ikke> /sysroot is a directory
2023-06-27 11:31:38 <doodoo> anti-cheat
2023-06-27 11:31:50 <doodoo> yeah it is, i suppose i gotta make it lol
2023-06-27 11:31:51 <doodoo> idk where tho
2023-06-27 11:32:06 <ikke> doodoo: no
2023-06-27 11:32:26 <doodoo> or something with boot options?
2023-06-27 11:32:28 <doodoo> fstab?
2023-06-27 11:32:28 <doodoo> idk
2023-06-27 11:33:11 <ikke> you're missing modules=sd-mod,usb-storage,ext4 modules=sd-mod,usb-storage,ext4
2023-06-27 11:33:19 <ikke> assuming you use ext4
2023-06-27 11:33:31 <ikke> sorry, modules was duplicated
2023-06-27 11:33:41 <doodoo> yes i was about to say, found this on the wiki "* incorrect device UUID * missing module in /mnt/etc/update-extlinux.conf or /mnt/etc/mkinitfs/mkinitfs.conf"
2023-06-27 11:33:51 <doodoo> ill add 
2023-06-27 11:34:03 <ikke> rootfstype=ext4
2023-06-27 11:34:33 <doodoo> i add rootfstype=ext4 + modules=sd-mod,usb-storage,ext4 ?
2023-06-27 11:34:38 <ikke> yezs
2023-06-27 11:40:23 <doodoo> i still get the same issue
2023-06-27 11:40:38 <doodoo> which is weird coz when it puts me into that safety shell i do "ls -a" and I see /sysboot
2023-06-27 11:40:53 <doodoo> could it be related to "Can't open blockdev"
2023-06-27 11:41:34 <aron> what kind of disk do you have? nvme?
2023-06-27 11:42:00 <ikke> https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L593
2023-06-27 11:42:21 <ikke> aron: yeah, good question
2023-06-27 11:42:52 <doodoo> nope, sata ssd
2023-06-27 11:43:18 <ikke> lvm?
2023-06-27 11:43:53 <ikke> in the emergency shell, do you see the blockdevices for your disk?
2023-06-27 11:43:53 <aron> raid, maybe?
2023-06-27 11:43:58 <markand> o/. I'm dd'ing the alpine-standard iso (x86_64) into a USB key, booting from it finds grub etc, then I get a recovery shell saying that mount failed 
2023-06-27 11:43:59 <doodoo> whats lvm ?
2023-06-27 11:44:07 <ikke> logical volume manager
2023-06-27 11:44:46 <doodoo> how do i check that
2023-06-27 11:45:11 <ikke> If you did not explicitly configured that, then most likely you are not
2023-06-27 11:45:27 <doodoo> im rebooting in to check blockdevices
2023-06-27 11:46:14 <doodoo> if u want to know if blkid writes anything it does, everything
2023-06-27 11:47:00 <ikke> Does the UUID match what's the cmdline?
2023-06-27 11:47:06 <doodoo> is there no systemd boot example for alpine ? 
2023-06-27 11:47:14 <doodoo> i am using the partuuid to boot
2023-06-27 11:47:16 <doodoo> does it need uuid ?
2023-06-27 11:47:28 <Forza> doodoo: doodoo for dual boot you just need to make sure that the boot loader knows of the kernel and initramfs for both OS 
2023-06-27 11:47:50 <doodoo> yes for windows I really dont understand that mess tho
2023-06-27 11:50:21 <doodoo> ill try use uuid not partuuid
2023-06-27 11:51:47 <doodoo> oh it works !!
2023-06-27 11:52:08 <doodoo> with internet
2023-06-27 11:52:09 <doodoo> hahaha nice!!
2023-06-27 11:52:41 <aron> cool
2023-06-27 11:52:47 <doodoo> finally, took a few hours haha
2023-06-27 11:53:10 <doodoo> thanks a lot guys, i managed to fix it and learnt a fair bit
2023-06-27 11:53:21 <doodoo> now i can go have lunch in peace
2023-06-27 12:04:20 <paper__> hi, linuxcontainers stopped building lxc containers for 32bit architectures, does alpine have their own builds for armv7 that we could use?
2023-06-27 12:14:18 <psykose> i don't think we maintain anything 'lxc image' related but i imagine you could create it from the minirootfs or the like
2023-06-27 12:14:47 <psykose> ikke: what did we use for the armv7 builder in lxc if not this same thing?
2023-06-27 12:15:39 <psykose> iirc you just use some template and it pulls the rootfs and makes it, or something
2023-06-27 12:17:01 <paper__> there is distrobuilder that prepares the container that lxc can then pull from, but that would require some infra we currently don't have, so for now I am just asking about whether there aren't pre-prepared containers somewhere
2023-06-27 12:17:38 <paper__> also, I think we would have to maintain our own template for distrobuilder since they dropped the armv7 variant
2023-06-27 12:18:09 <ikke> We use the legacy templates 
2023-06-27 12:18:27 <ikke> It doesn't require anything prebuilt/hosted 
2023-06-27 12:20:14 <ikke> It uses apk.static to install into a chroot 
2023-06-27 12:37:20 <psykose> ah, i guess those are basically just a setup script
2023-06-27 12:37:31 <psykose> and we have them patched for new keys and stuff
2023-06-27 12:37:36 <ikke> Yes
2023-06-27 15:02:52 <maria> hi i'm wanting to make some custom alpine packages for my network and am unsure of the best way to do so. is there a way to build packages for other versions of alpine than what is installed on my system? should i use a vm or chroot instead of my primary install? what's the recommended way to go about this?
2023-06-27 15:05:10 <psykose> rootbld slightly lets you do that if you start from e.g. 3.18-stable but overall there's a lot of caveats and messes there
2023-06-27 15:06:03 <psykose> having a vm or some container stuff with a repo to run buildrepo in is probably easiest, but it does take a bunch of setup
2023-06-27 15:14:43 <pj> eventually using docker-abuild aka dabuild
2023-06-27 15:14:58 <pj> if you want to just build stuff manually
2023-06-27 15:20:32 <Saijin_Naib[m]> @psykose: My second Alpine XFCE edge x86_64 machine (also clean install from 3.18.2 upgraded to edge this week) is experiencing the same issues with Evolution and Epiphany not working
2023-06-27 15:20:45 <psykose> idk what it could be
2023-06-27 16:02:04 <Saijin_Naib[m]> Hm, alright. Any pointers of where I should look next to file an issue?
2023-06-27 16:02:06 <Saijin_Naib[m]> I have no STR other than install Evolution under edge x86_64 with XFCE as the DE
2023-06-27 16:02:24 <Saijin_Naib[m]> Where to look next meaning info to collect, sorry
2023-06-27 17:29:11 <uuidNuniq> OdaNobunaga[m]: i guess that btrfs was converted from ext4
2023-06-27 21:03:36 <Forza> Maybe LTO builds are causing issues? I saw something like this on #gentoo earlier. 
2023-06-27 21:04:28 <psykose> not in this class of error
2023-06-27 21:04:45 <Forza> Saijin_Naib[m]: ^^
2023-06-27 21:04:53 <Forza> OK 
2023-06-27 21:04:59 <Forza> Sorry 
2023-06-27 21:07:45 <psykose> what is "something like this on #gentoo earlier"
2023-06-27 22:56:44 <zito> are there any third party alpine package repositories for VERY old hardware? I have a CPU that doesn't even support sse2, so a lot of software fails with Invalid Instruction: neovim, mpv, midori, firefox to name a few. also I am afraid I can't handle rebuilding packages myself and managing my own versions of everything. are there repos I can just plug and install from them?
2023-06-27 23:02:20 <psykose> don't think so
2023-06-27 23:02:43 <psykose> neovim probably fails in luajit which is meant to check for sse but it's broken somewhere
2023-06-27 23:03:11 <psykose> mpv i'm not sure but if you get a `bt` and `x/5i $pc` from gdb i can look at it
2023-06-27 23:03:19 <psykose> firefox is probably unfixable
2023-06-27 23:04:01 <psykose> that said i think even the kernel from 3.18 won't work anymore with such hardware if it doesn't support pae so idk
2023-06-27 23:08:07 <zito> psykose: there's pae in the output of /proc/cpuinfo
2023-06-27 23:08:13 <psykose> works then
2023-06-27 23:08:30 <zito> it's a mobile AMD Athlon 1400+
2023-06-27 23:08:39 <zito> 800 MHz
2023-06-27 23:09:42 <zito> I found some post of a dev in the issues tracker saying that very few people use such hardware, hence staying on sse2 is fine sinec it doesn't affect many users
2023-06-27 23:09:46 <zito> well it affects me
2023-06-27 23:10:35 <zito> and I'm not looking for getting babysitted with a particular single package, I am trying to find options so I could service myself
2023-06-27 23:10:50 <zito> so I thought such a third party repo could be an option (if it existed)
2023-06-27 23:11:51 <zito> hence I find pretty much pointless to start spamming the issue tracker with complaints for every package that does not work for me
2023-06-27 23:12:13 <zito> such hardware is too old to be officially supported? fair, I can totally understand that
2023-06-27 23:12:37 <zito> but alpine happens to be the only distro that boots and installs on this thing, so I don't have other options
2023-06-27 23:14:01 <zito> I can even get Xorg to work, with a lot of manual voodoo config patching, it would be great if I find a way to get past this nvalid Instruction limitation on my own, every time it happens for a package I may need
2023-06-27 23:14:32 <psykose> yeah, i got that part
2023-06-27 23:14:36 <psykose> but no such repo exists that i know of
2023-06-27 23:14:58 <abby> what's debian i386's minsupport?
2023-06-27 23:15:19 <psykose> 686
2023-06-27 23:15:23 <zito> I wasted a whole day with debian 12 i386, just yesterday
2023-06-27 23:15:53 <zito> couldn't even beat the manual partitioner to get a LUKS encrypted filesystem without LVM
2023-06-27 23:15:54 <abby> what about arch linux 32? they have both 686 and 486 repos
2023-06-27 23:16:08 <abby> https://archlinux32.org/architecture/
2023-06-27 23:16:40 <zito> the automated partitioner worked, then after many hours of installing packages it tried to set up the console and garbled it
2023-06-27 23:16:50 <zito> then froze at the grub install step
2023-06-27 23:16:52 <zito> so no debian
2023-06-27 23:17:06 <zito> owait, I thought arch abandoned the x86 branch years ago
2023-06-27 23:17:17 <zito> I did not know there's still arch for 32bit
2023-06-27 23:17:31 <raven523> archlinux32 is a separate distro
2023-06-27 23:17:39 <abby> ^ same with arch linux arm
2023-06-27 23:19:43 <zito> the alpine install script is so easy to use, does archlinux32 have something similar? or does it require a lot of manual intervention copypasting commands?
2023-06-27 23:20:18 <raven523> arch does have a guided installer now
2023-06-27 23:20:41 <raven523> though it's not recommended
2023-06-27 23:22:34 <tm512> I think gentoo supports 32-bit x86 still and for packages you could set up distcc to get packages compiled on more capable hardware, and there's first-class support for having control over the build flags that get used. might be a bit annoying overall though
2023-06-27 23:24:57 <zito> installing gentoo is even worse than installing arch. especially while being biased by the simplicity of installing a;pine
2023-06-27 23:25:06 <tm512> this one video I saw recently on Arch 32 it seemed like the distro has (or had) a lot of issues with outdated keyrings screwing up pacman installations. not sure if that's been sorted out by now
2023-06-27 23:25:46 <zito> I just went through the bugs they mention and a random one is about qt making use of sse2
2023-06-27 23:26:01 <abby> outdated keyring issues is just arch in general
2023-06-27 23:26:26 <ilyaigpetrov> Hi. I try to launch a game xonotic from a VT (not from X11). `$ xonotic-sdl` throws the following error: http://ix.io/4zht.
2023-06-27 23:26:49 <tm512> abby: in this video the keyring issues seemed so much worse than I ever remembered dealing with in the 5 years I used Arch
2023-06-27 23:28:24 <zito> I could just use vim (which works) instead of neovim, but not having mpv and a single functional mainstream web browser is just too limiting
2023-06-27 23:28:38 <zito> most sites don't work in dillo unfortunately
2023-06-27 23:28:50 <zito> or in text mode links
2023-06-27 23:29:34 <tm512> zito: unless Qt is making use of intrinsics and doesn't offer fallback code paths, it'd be GCC emitting SSE2 instructions, which could be worked around by just compiling for a different -march or -mcpu or whatever the right incantation is
2023-06-27 23:30:57 <zito> https://archlinux32.org/packages/i686/extra/dosbox/ NOOO :D
2023-06-27 23:31:06 <zito> doesn't work in 32bit alpine either
2023-06-27 23:31:13 <zito> now this makes me sad for real
2023-06-27 23:31:45 <psykose> ilyaigpetrov: `error: XDG_RUNTIME_DIR is invalid or not set in the environment.` seems pretty self explanatory
2023-06-27 23:31:53 <tm512> coincidentally I've been thinking of getting this old 32-bit laptop I have laying around up and running, I think it lacks SSE2 as well, so now I'm wondering what kinds of issues I'm gonna run into
2023-06-27 23:32:20 <tm512> SSE2 was introduced with the Pentium 4, right? this laptop has a Pentium M which I think is based on the P3 architecture
2023-06-27 23:32:33 <psykose> although not sure why that's printed at all
2023-06-27 23:32:39 <tm512> definitely an improved P3 though so who knows
2023-06-27 23:32:48 <psykose> you probably need SDL_VIDEODRIVER=kmsdrm or w/e
2023-06-27 23:34:02 <psykose> tm512: pentium M has sse2
2023-06-27 23:34:04 <tm512> oh I guess PM does have SSE2, so maybe things won't be quite as painful
2023-06-27 23:34:33 <psykose> zito: if you get the mpv bt i can look at it
2023-06-27 23:35:11 <zito> psykose: like I said above, that's very kind of you, but I don't know how much would help me having the issue with a single package eventually solved
2023-06-27 23:35:12 <tm512> I'm suspecting the main trouble with this thing is going to be graphics acceleration. it has Intel Extreme Graphics 2, an i855 chipset
2023-06-27 23:35:29 <zito> until the next app I want to use will give me the same error, thus forcing me to come here and whine again?
2023-06-27 23:35:34 <psykose> that has no gpu support
2023-06-27 23:35:38 <psykose> zito: isn't that everything in life
2023-06-27 23:36:39 <zito> psykose: tbh no. actually I expected that slackware (which I tried first) would just run on this hardware
2023-06-27 23:36:49 <zito> well, the installer did not work, which was quite surprising
2023-06-27 23:36:58 <tm512> psykose: the pre-gallium i915 driver in mesa seems to support it, though someone in #dri-devel said something about it not really being tested for years
2023-06-27 23:37:05 <zito> maybe Pat doesn't pay as much attention to x86 nowadays...
2023-06-27 23:37:05 <tm512> nowadays it's relegated to the mesa-amber branch
2023-06-27 23:37:52 <psykose> the truth is that fewer and fewer people care about it
2023-06-27 23:37:56 <psykose> though that was already obvious
2023-06-27 23:38:29 <zito> I still got a few exotic distros to try, maybe one of them will just work. for example I did not get to try tinycore yet
2023-06-27 23:39:42 <tm512> zito: maybe BSDs? I don't expect an OS like NetBSD to give up on 32-bit x86 anytime soon, though I think x86_64 and ARM are the main development targets
2023-06-27 23:39:52 <zito> psykose: you're correct of course, maybe I should just see what can I sell for cheap (because a kidney is too expensive) and try getting a refurbished 64bit CPU laptop for $100-200 :D
2023-06-27 23:39:52 <tm512> for understandable reasons
2023-06-27 23:40:14 <psykose> depending on where you live used stuff can be a huge bargain
2023-06-27 23:40:38 <psykose> locally there's always random zen1/2 cpus for like <$100 etc as a great start to build something
2023-06-27 23:40:41 <psykose> random laptops and whatever
2023-06-27 23:40:43 <zito> not the best place 
2023-06-27 23:41:02 <tm512> zito: stuff from like 2012-2013 should be really cheap at this point (exempting Apple hardware) and is still pretty usable
2023-06-27 23:41:23 <tm512> if you're trying to go for something very budget
2023-06-27 23:41:49 <zito> I'll just say that if the Russians blow up that nuclear plant, issues like mpv not working on this 20 years old laptop won't bother me anymore, I won't even have enough time to run tfo :D
2023-06-27 23:42:02 <tm512> if craigslist is a thing in your area you might be able to bargain with people for cheap hardware
2023-06-27 23:43:08 <tm512> a while back I saw one of those tiny Dell Optiplex systems with a 6th-gen i3 for like $100 with an SSD and even wifi included
2023-06-27 23:43:50 <psykose> maybe not in that place, heh
2023-06-27 23:43:53 <psykose> not super fortunate
2023-06-27 23:45:26 <ilyaigpetrov> psykose: `$ SDL_VIDEODRIVER=kmsdrm xonotic-sdl` starts execution but fails complaining about opengl: http://ix.io/4zhF. From X11 the game works fine.
2023-06-27 23:47:25 <psykose> sounds like https://github.com/libsdl-org/SDL/issues/5399 which would be a xonotic bug
2023-06-27 23:47:34 <psykose> when using kmsdrm anyway
2023-06-27 23:47:57 <psykose> you can try =directfb instead
2023-06-27 23:49:06 <ilyaigpetrov> directfb fails the same way as with undefined SDL_VIDEODRIVER like in my first message here.
2023-06-27 23:50:34 <psykose> unfortunate
2023-06-27 23:50:46 <ilyaigpetrov> Thanks, you response was helpful. By what keywords did you find the 5399 bug?
2023-06-27 23:51:05 <psykose> glGetString in search on sdl repo
2023-06-27 23:51:47 <psykose> the directfb failure is probably directfb being broken (not unexpected) and kmsdrm being broken is most likely an application issue and not sdl
2023-06-27 23:51:51 <psykose> that about summarises it
2023-06-27 23:52:58 <ilyaigpetrov> thanks
2023-06-27 23:57:00 <ilyaigpetrov> Why directfb is broken?
2023-06-28 00:00:56 <psykose> it's a library for outputting to fb and fb is not exactly maintained in linux itself and nobody cares about it in general and as a result nobody uses directfb in any normal setting so nothing broken in it gets fixed, ...
2023-06-28 00:03:08 <zito> they should put these AI things to start working on fixing old software
2023-06-28 00:04:07 <psykose> drm existed since 1999 as the successor to fb for actual rendering
2023-06-28 00:04:10 <psykose> idk dawg
2023-06-28 00:04:31 <psykose> of course the actual 'everything is drm' stuff was way later than that
2023-06-28 00:04:39 <psykose> maybe just 10 years ago
2023-06-28 02:01:24 <livy> Does anyone use the software RAID install of Alpine? I mean giving multiple devices 'sda sdb sdc sdd' when answering the installer question where to put the OS.
2023-06-28 02:03:05 <livy> It creates md0 as RAID1 (metadata 0.9) for '/boot' and md1 as RAID5 (metadata 1.2) for '/'.
2023-06-28 02:04:01 <livy> If any Alpine developer see this, please take a look at how broken it is. It is very unstable.
2023-06-28 02:04:31 <livy> When booting, I encounter the error: "alpine mounting /dev/vg0/lv_root on /sysroot failed"
2023-06-28 02:05:16 <livy> And I am forced to reboot. I would say it happens 4 out of 5 times. I have to reboot until it works.
2023-06-28 02:06:11 <livy> And when it successfully boots up, the last drive (sdd) was kicked out of the RAID arrays (both md0 and md1).
2023-06-28 02:07:17 <livy> I tried 'ls -l /dev/sd*', and found that there were [sda + sda1 + sda2], same for sdb, same for sdc, but there was only sdd with no sdd1 or sdd2.
2023-06-28 02:07:39 <livy> 'parted /dev/sdd print' showed that there were 2 partitions on sdd.
2023-06-28 02:08:06 <livy> I tried to swap my 4 drives around, and re-installed Alpine, and ended up with the same issue.
2023-06-28 02:08:40 <livy> Whichever drive becomes 'sdd', is kicked out of the mdraid arrays for no reason.
2023-06-28 02:09:54 <livy> I tried to boot to live Alpine USB, and I can see sdd + sdd1 + sdd2 just fine. So it cannot be a hardware issue.
2023-06-28 02:10:11 <Saijin_Naib[m]> <psykose> "not in this class of error" <- Alright, I'll standby and make due. It must be some quirk. Back to Thunderbird (shudder)
2023-06-28 04:04:40 <jonno> hello, how can I find out why a process was killed?
2023-06-28 04:10:05 <zcrayfish> any messages like "kernel: Out of memory: Kill" in dmesg?
2023-06-28 04:10:24 <zcrayfish> otherwise there's hundreds of reasons a process would be killed.
2023-06-28 04:12:03 <jonno> it just says "killed" at the end
2023-06-28 04:12:34 <jonno> there is no dmesg, it says dmesg: klogctl: Operation not permitted
2023-06-28 04:15:06 <abby> run it with superuser
2023-06-28 04:15:15 <abby> s/it/dmesg/
2023-06-28 04:15:46 <jonno> I am root
2023-06-28 04:15:58 <jonno> this is alpine running in docker
2023-06-28 04:27:02 <aron> jonno: I think your question is too general; to see the "Killed" message at the end means (a) the OOM reaper came for it (see above); or (b) it received a SIGKILL from *a process*
2023-06-28 04:28:32 <aron> for case (a), dmesg will help
2023-06-28 04:29:13 <aron> for case (b), you should share more details
2023-06-28 04:38:26 <jonno> how do I get dmesg enabled in alpine?
2023-06-28 04:38:50 <jonno> I don't think it was OOM because the system has a lot of ram, I think it's because some files were being accessed at the same time
2023-06-28 04:39:15 <zcrayfish> sysctl kernel.dmesg_restrict may need to be changed.
2023-06-28 04:39:30 <jonno> it's read only
2023-06-28 04:39:59 <jonno> it's also set to 0
2023-06-28 04:40:36 <abby> <jonno> how do I get dmesg enabled in alpine? <- you're running alpine in a container, so you aren't running an actual alpine kernel unless your host OS is alpine too
2023-06-28 04:40:49 <abby> you'd want to be doing dmesg on the host, outside of the container
2023-06-28 04:40:51 <jonno> good point
2023-06-28 04:43:14 <jonno> looks like it was the oom
2023-06-28 04:43:38 <zcrayfish> you've been reaped! :o
2023-06-28 04:43:52 <jonno> maybe docker needs more ram
2023-06-28 04:43:56 <jonno> where can I download some?
2023-06-28 04:44:36 <jonno> I'm only trying to generate a pdf with a font loaded
2023-06-28 04:45:17 <abby> rams'r'us
2023-06-28 04:48:08 <zcrayfish> If the system is memory constrained enough, but you don't mind a performance loss, maybe look at zswap
2023-06-28 06:35:54 <livy> After manually formatting /dev/sda as ext4 and make a proper fstab automount (to /media/savestates), I ran 'setup-alpine' and made a diskless install.
2023-06-28 06:36:28 <livy> I select the 'savestates' directory as the location to store .apkovl, then issued an 'lbu commit' command.
2023-06-28 06:36:54 <livy> How do I suppose reboot and load the .apkovl file?
2023-06-28 06:37:19 <livy> I mean if I reboot, and boot to the live USB drive again, how can it know where to load the .apkovl?
2023-06-28 06:44:27 <livy> The Wiki has a section "Saving and loading ISO image customizations
2023-06-28 06:44:49 <livy> but it only instructs how to save states, and doesn't mention how to load them.
2023-06-28 06:46:02 <ikke> livy: it has some predefined locations it looks for it, but you can also specify the location with a kernel cmdline argument 
2023-06-28 06:46:43 <OdaNobunaga[m]> lbu.conf?
2023-06-28 06:47:30 <livy> @OdaNobunaga[m] I am booting from a live USB which was created using dd command (meaning read-only).
2023-06-28 06:47:45 <livy> I do not know what lbu.conf is, but it is clear that the USB drive content cannot be changed.
2023-06-28 06:51:09 <livy> @ikke What are the default locations? It is inconvenient to halt the bootloader and specify the kernel boot arguments every time.
2023-06-28 06:52:46 <ikke> Yeah, the iso is read only. You'd use setup-bootable to create a writable usb disk, or do it manually 
2023-06-28 06:56:16 <livy> Um, so as I understand, to run diskless mode, I still need some kind of bootable (and writable) USB drive.
2023-06-28 06:56:20 <ikke> Apparently it uses nlplug-findfs to discover apkovls, most likely in the root of partitions 
2023-06-28 06:56:39 <ikke> livy: the ovl has to be stored somewhere 
2023-06-28 06:57:18 <livy> I know that the .apkovl files need to be stored somewhere, and I want to store them on my internal disks (/dev/sda).
2023-06-28 06:57:19 <ikke> The manual argument is called apkovl 
2023-06-28 06:58:18 <livy> But even if I use 'apkovl=' as kernel argument, how can I point it to /dev/sda, which is currently not mounted?
2023-06-28 06:58:29 <livy> I really do not understand the idea how it works.
2023-06-28 06:59:17 <livy> What I currently have: a Alpine CD, and an internal HDD.
2023-06-28 07:01:35 <ikke> How is the ovl file called?
2023-06-28 07:02:26 <OdaNobunaga[m]> Create two partitions:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/LdzYMcrtSRIczNNmLZRVyitY>)
2023-06-28 07:02:33 <ikke> nlplug-findfs will look for any file that matches *.apkovl.tar.gz*
2023-06-28 07:03:10 <ikke> And it does not require the device to be mounted 
2023-06-28 07:04:42 <ikke> maxdepth=1 so it will only look in the root
2023-06-28 07:06:20 <ikke> So, put the file in the root of /dev/sda, with an appropriate name, and I think it should find it
2023-06-28 07:06:34 <livy> I just tried to reboot, and yes it found the .apkovl and loaded it.
2023-06-28 07:06:51 <livy> I just do not understand why and how Alpine CD can do that.
2023-06-28 07:08:06 <livy> And what if I have another partition (e.g: /dev/sdb2) which is formatted and has .apkovl.tar.gz files?
2023-06-28 07:09:56 <OdaNobunaga[m]> It will look in lbu.conf where the apk cache is located at.
2023-06-28 07:10:06 <OdaNobunaga[m]> * look in /etc/lbu, * lbu/lbu.conf where
2023-06-28 07:10:41 <OdaNobunaga[m]> Diskless Alpine is complicated though. For example, if you need a new kernel module, it gets ugly. "Apk add" can't install them and diskless is neither using linux-lts nor linux-virt. Instead, you update the kernel via "update-kernel".
2023-06-28 07:11:01 <livy> As the Alpine bootable media is an ISO9660 filesystem, it cannot be modified. So the content of that file is always default.
2023-06-28 07:11:41 <livy> After the reboot, my live Alpine USB now automatically loads the previous .apkovl.tar.gz file which is on /dev/sda.
2023-06-28 07:12:03 <livy> I do not know how to boot without loading it. Maybe unplugging the SATA cable? :D
2023-06-28 07:14:41 <ikke> edit the cmdline on boot 
2023-06-28 07:15:31 <ikke> Set apkovl to something non-existing 
2023-06-28 07:17:04 <livy> Which key should I use to stop the ISO Linux 6.04 screen?
2023-06-28 07:17:17 <livy> To edit the boot command line?
2023-06-28 07:18:43 <ikke> Tab
2023-06-28 07:18:57 <livy> Thanks. And I think I can just boot into the OS, and delete the .apkovl.tar.gz file.
2023-06-28 07:19:16 <ikke> Yes, that would also work
2023-06-28 12:00:30 <quinq> Hummm, I have a strange thing happening, could somebody explain it to me? :)
2023-06-28 12:00:33 <quinq> $apk info -r ssl_client
2023-06-28 12:00:35 <quinq>  
2023-06-28 12:00:37 <quinq> ssl_client-1.36.1-r2 is required by:
2023-06-28 12:00:48 <quinq> $ doas apk del ssl_client
2023-06-28 12:00:54 <quinq> World updated, but the following packages are not removed due to: ssl_client: grub-bios linux-lts zfs […]
2023-06-28 12:10:12 <ikke> quinq: apk info -P ssl_client
2023-06-28 12:21:42 <quinq> Thank ikke 
2023-06-28 12:21:53 <quinq> But apk info -r cmd:ssl_client still doesn't return anything
2023-06-28 12:22:27 <quinq> (apk info -P ssl_client returns "ssl_client-1.36.1-r2 provides: cmd:ssl_client=1.36.1-r2"
2023-06-28 12:23:22 <quinq> Question is, is there a way which packages depends on a another package of this type, thas provides something?
2023-06-28 12:23:50 <quinq> Or not, and that would be an apk improvement?
2023-06-28 12:29:58 <zcrayfish> are the D: lines in /lib/apk/db/installed what you are asking about?
2023-06-28 12:41:40 <quinq> Let-me have a look
2023-06-28 12:42:01 <quinq> P:ssl_client
2023-06-28 12:42:07 <quinq> D:so:libc.musl-x86_64.so.1 so:libcrypto.so.3 so:libssl.so.3
2023-06-28 12:42:34 <quinq> I don't think so, what does D: mean? That P(ackage?): D(epends?) on libc?
2023-06-28 12:44:09 <quinq> I'm asking how to see with pkg info -r which packages depend on the said package (apparently it's kind of a virtual one?)
2023-06-28 12:45:08 <quinq> Well, at this point the question is rather to confirm that it's currently not possible :)
2023-06-28 12:49:54 <ikke> Hmm, grub-bios only depends on grub
2023-06-28 13:05:08 <ikke> Hmm, I'm not sure why apk would mention grub-bios or linux-lts
2023-06-28 13:08:09 <ikke> Or the others for that matter 
2023-06-28 13:10:52 <ikke> ssl_client is installed due to an install_if
2023-06-28 14:45:35 <Ty3r0X> hello, the netavark package in the latest stable does not seem to install a binary and my container manager (podman) cannot start any of my containers.
2023-06-28 15:04:56 <z4w3p> hallo alpine root
2023-06-28 16:04:21 <psykose> Ty3r0X: it does, it's in /usr/libexec/podman/netavark
2023-06-28 16:04:27 <psykose> and it works fine so the issue must be something else
2023-06-28 16:13:00 <Ty3r0X> psykose: is this dir included in the podman $PATH?
2023-06-28 16:13:34 <ikke> libexec is rarely in PATH, rather it's executed by absolute path
2023-06-28 16:16:50 <Ty3r0X> yes I can find netvark there and it can execute just fine, I wonder why is podman complaining No such file or directory
2023-06-28 16:17:44 <ikke> Ty3r0X: can you try to strace podman?
2023-06-28 16:17:50 <ikke> see what it tries to execute?
2023-06-28 16:24:19 <Ty3r0X> ikke: I cannot do grep or tee with strace
2023-06-28 16:24:39 <ikke> Ty3r0X: you can write the strace output to a file
2023-06-28 16:24:44 <ikke> -o podman.strace
2023-06-28 16:25:00 <ikke> you can filter specific syscalls with -e
2023-06-28 16:25:05 <Ty3r0X> oh nice!
2023-06-28 16:25:05 <ikke> for example -e execve
2023-06-28 16:36:04 <Ty3r0X> that's weird, I switched to cni and then back to netavark and the problem dissapeared..?
2023-06-28 19:55:09 <RoyalTiger[m]> Short question: Is there any good guide for Alpine Linux with full disk encryption at all (found some, but they are not very enlighting in the sense of what exactly happens (https://wiki.alpinelinux.org/wiki/Full_disk_encryption_secure_boot)) - so maybe a good guide for luks would also be good) and my further question: RAM based Alpine and Disk Encryption (+dropbear/ssh unlock) - is that possible, does this somehow integrate in the
2023-06-28 19:55:10 <RoyalTiger[m]> sense of usability and does this make sense at all? I have a Virtualisiation Server that has much RAM, but limited hard disk R/W, so I would prefer spin all my machines from RAM by one iso (already achived that), but encryption of the machines would be great.
2023-06-28 19:59:34 <JohannesJ[m]> oh!
2023-06-28 19:59:50 <JohannesJ[m]> that guide is what5 i used as well
2023-06-28 20:13:36 <RoyalTiger[m]> <RoyalTiger[m]> "Short question: Is there any..." <- I also don't fear to head into Terraform/Vagrant idk for that, if that's more like a "one script for everything  following" thing, cause I would do that for all my alpine machines. One of my first problems with alpine was for example, that the reduced image (virtual), doesn't support qemu-guest-agent, cause the serial vport was cut out of the image.
2023-06-28 20:14:51 <psykose> there's no dropbear/ssh unlock integration no
2023-06-28 20:17:20 <minimal> Ro
2023-06-28 20:17:50 <minimal> RoyalTiger[m]: I wrote a MR to add dropbear unlock integration but it wasn't merged
2023-06-28 21:53:30 <abk> Greetings humans and bots!
2023-06-28 21:56:04 <abk> I just installed sddm and, according to the screen at boot time, it loads after everything else. Yet by some reason sddm «freezes» while loading, and just appears a lonely frozen prompt in a black screen. I need to stop and re-start the corresponding service for sddm to function.
2023-06-28 21:56:45 <abk> There appear no errors in the log. Does anyone have a suggestion on how to fix it?
2023-06-28 22:36:55 <tokyovigilante> Morning, can I ask about Wifi during install? I have a thinkpad x230 with a 6000 series intel chipset, firmware in intel-firmware-other but not in install image. If I run the diskless install as a first step, can I then install the firmware into the live environment before installing to disk, or is it easier to just install without networking and set up once the system is on disk? Thanks
2023-06-28 23:43:04 <Saijin_Naib[m]> What image are you using? Does Alpine Extended not have Linux-firmware-other? 
2023-06-29 00:38:27 <tokyovigilante> it does I think as an installable package but it doesn't seem to be included in the boot "live" environment from the ISO
2023-06-29 00:39:57 <tokyovigilante> tried with the standard then with the extended, I haven't had a chance to experiment too much with the extended image, need to figure out how to replace my LUKS / partition with Alpine and preserve my /home, on EFI/GPT etc, so haven't wanted to experiment on my internal disk too much
2023-06-29 00:50:16 <nektro> is there any way to use myhostname on alpine? not seeing a package for it on the website
2023-06-29 00:51:01 <psykose> musl doesn't have nss modules
2023-06-29 02:15:37 <LACampbell> does anyone understand elogind? I read the man page and don't feel any closer to getting "loginctl hibernate" to work (the command just exits silently). "loginctl session" lists a session which is nice
2023-06-29 02:29:49 <Saijin_Naib[m]> No 😭
2023-06-29 02:30:16 <Saijin_Naib[m]> My lightdm can't suspend, restart, shutdown... 
2023-06-29 02:30:32 <Saijin_Naib[m]> I had it working before, too 
2023-06-29 02:30:51 <Saijin_Naib[m]> Are you using a whole DE with session manager? 
2023-06-29 02:41:59 <LACampbell> Saijin_Naib[m]: no, just dbus session
2023-06-29 02:42:29 <LACampbell> does 'loginctl reboot' work for you? I got that working. issue seems to be that waking from hibernating wipes the swap partition
2023-06-29 02:43:52 <LACampbell> (for that to work you need a PAM login)
2023-06-29 02:56:23 <LACampbell> Saijin_Naib[m]: I fixed it. if you want to see if I can help you ping me quickly before I forget :)
2023-06-29 02:58:07 <LACampbell> (TL;DR elogind service, pam login, kernel resume parameter)
2023-06-29 02:58:12 <Saijin_Naib[m]> Nice! 
2023-06-29 02:58:19 <Saijin_Naib[m]> I think I'm missing Pam 
2023-06-29 02:58:27 <Saijin_Naib[m]> Thank you for the pointer 
2023-06-29 02:58:36 <LACampbell> ah! so as I unserstand the default alpine linux busybox login isn't a pam login
2023-06-29 02:58:55 <LACampbell> is elogind running?
2023-06-29 03:15:34 <Saijin_Naib[m]> Supposedly, alongside dbus and polkit 
2023-06-29 03:15:57 <Saijin_Naib[m]> XFCE polkit dies at login, and polkit service fails during boot
2023-06-29 03:16:06 <LACampbell> ok, so if you just type "loginctl" are any sessions listed?
2023-06-29 03:18:45 <Saijin_Naib[m]> Nope, no sessions 
2023-06-29 03:18:59 <Saijin_Naib[m]> XFCE polkit broke after adding PAM stuff
2023-06-29 03:19:55 <psykose> no sessions means you had no pam login
2023-06-29 03:20:06 <psykose> you do have to log in again after adding util-linux-login or shadow-login
2023-06-29 03:20:53 <Saijin_Naib[m]> Rebooted after adding útil-Linux-login days ago 
2023-06-29 03:21:27 <Saijin_Naib[m]> Is there a Pam OpenRC service? 
2023-06-29 03:21:59 <psykose> that's not how pam works so no
2023-06-29 03:22:08 <psykose> if you have it then that's not the issue
2023-06-29 03:22:08 <Saijin_Naib[m]> pam-rundir package seems to be what breaks XFCE polkit
2023-06-29 03:22:21 <psykose> if you have that at the same time as elogind then yeah
2023-06-29 03:22:29 <Saijin_Naib[m]> Oooh 
2023-06-29 03:22:37 <Saijin_Naib[m]> Should they not conflict then? 
2023-06-29 03:22:51 <psykose> not really no
2023-06-29 03:23:07 <Saijin_Naib[m]> I have a session now, but lightdm still can't suspend/restart/shutdown/logout 
2023-06-29 03:23:14 <Saijin_Naib[m]> Does that need Pam? 
2023-06-29 03:24:02 <psykose> you already have that
2023-06-29 03:24:14 <Saijin_Naib[m]> Ah, right, sorry 
2023-06-29 03:24:48 <Saijin_Naib[m]> Does lightdm need to be in any groups like plugdev or something? 
2023-06-29 03:24:52 <Saijin_Naib[m]> Wheel? 
2023-06-29 03:25:58 <LACampbell> Saijin_Naib[m]: curious, can you do "loginctl hibernate" ? now that you have a session
2023-06-29 03:26:18 <LACampbell> https://wiki.alpinelinux.org/wiki/Elogind#Hibernating I did a brief write up, but I feel not a good one :/
2023-06-29 03:26:19 <Saijin_Naib[m]> 😬
2023-06-29 03:26:40 <Saijin_Naib[m]> That never worked reliably for me on this crap top, but let me try
2023-06-29 03:26:49 <Saijin_Naib[m]> Good test of how robust XFS is 🤣
2023-06-29 03:26:53 <LACampbell> lol. a crap top with windows is a viable linux machine
2023-06-29 03:27:29 <Saijin_Naib[m]> Eeh... This worked great under 10s 🤷
2023-06-29 03:27:48 <LACampbell> under 10 seconds?
2023-06-29 03:27:51 <psykose> a swapfile also works so that's not necessarily accurate
2023-06-29 03:27:56 <Saijin_Naib[m]> Took me about 6mo to figure out what kernel tweaks it needed and then I made a bunch of really bad PRs haha
2023-06-29 03:28:05 <Saijin_Naib[m]> Windows 10s
2023-06-29 03:28:33 <Saijin_Naib[m]> Command executed, resuming now
2023-06-29 03:28:40 <LACampbell> oh. yeah I made the mistake of upgrading to 11 on my win 10 laptop. considering putting alpine on there too, but linux + laptop batteries... eh
2023-06-29 03:28:59 <Saijin_Naib[m]> 11 was lighter for me
2023-06-29 03:29:11 <LACampbell> they destroyed my beloved old school windows panel
2023-06-29 03:29:11 <Saijin_Naib[m]> Should uptime be non-zero from hibernate? 
2023-06-29 03:29:34 <psykose> i found windows to be quite useless for power management these days because some random idle service jumps to 50% cpu all the time doing nothing
2023-06-29 03:29:42 <LACampbell> you can't make the taskbar or whatever look "classic" anymore. it's that tool 'modern' look with the big fat icons. it's like a bezel
2023-06-29 03:29:47 <Saijin_Naib[m]> I like the new tiling they added to DWM. It's my favorite WM
2023-06-29 03:29:54 <psykose> linux instead takes the opposite approach of having 7850 knobs for tuning power stuff that nobody knows about
2023-06-29 03:30:01 <LACampbell> but yeah how did logicntl hibernate go?
2023-06-29 03:30:10 <LACampbell> psykose: lol, so battery time may not be any worse? hmm
2023-06-29 03:30:20 <LACampbell> maybe I will put alpine on there too. getting sick to death of windows
2023-06-29 03:30:28 <psykose> on linux it's better if you set up some half decent power management but it's not that easy
2023-06-29 03:30:29 <Saijin_Naib[m]> It looks like it just executed a shutdown. Nothing resumed, and my uptime is just from this boot
2023-06-29 03:30:41 <psykose> depends on if all the drivers support all the low power states and whatever, bunch of knobs
2023-06-29 03:30:59 <LACampbell> Saijin_Naib[m]: aha, so - do you have a resume kernel flag? like in /etc/default/grub
2023-06-29 03:31:14 <LACampbell> https://wiki.alpinelinux.org/wiki/Elogind#Hibernating actually follow this. and if it's confusing I'll edit it and then other people can benefit in the future lol
2023-06-29 03:31:17 <Saijin_Naib[m]> I have a 4GB SWAP volume over LVM from setup-alpine 
2023-06-29 03:32:23 <Saijin_Naib[m]> So it would be lv_swap or vg0-lv_swap
2023-06-29 03:32:36 <psykose> you need swap>ram to hibernate
2023-06-29 03:32:36 <Saijin_Naib[m]> Or I do the path to the swap volume? 
2023-06-29 03:33:14 <LACampbell> oh it would be partition
2023-06-29 03:33:40 <LACampbell> like if you do lsblk, do you see a swap partition? /dev/something
2023-06-29 03:33:59 <Saijin_Naib[m]> Does the chunk taken out for Intel iGPU and other shit matter? RAM shows to be max 3.7GB
2023-06-29 03:34:47 <Saijin_Naib[m]> Swap is under mmcblk0p2, name of vg0-lv_swap
2023-06-29 03:35:10 <LACampbell> I'm a bit lost with those names. the partition is /dev/mmcblk0p2 ?
2023-06-29 03:35:35 <LACampbell> if so, add 'resume=/dev/mmcblk0p2" to the kernel flags. then do grub-mkconfig again
2023-06-29 03:36:00 <Saijin_Naib[m]> I'm bad with LVM but I believe the partition is mmcblk0p2, which holds two LVM volumes 
2023-06-29 03:37:10 <LACampbell> can you post the output of lsblk?
2023-06-29 03:38:53 <Saijin_Naib[m]> Yeah, in a year when Thunderbird connects to Matrix
2023-06-29 03:38:59 <LACampbell> lol
2023-06-29 03:39:06 <Saijin_Naib[m]> This is a very painful implementation of Matrix
2023-06-29 03:39:25 <LACampbell> basically do you see that wiki article? does it look anything like that, like is there a line with [SWAP] on it
2023-06-29 03:39:39 <Saijin_Naib[m]> Yep, on my LVM volume 
2023-06-29 03:39:49 <LACampbell> can you just paste that line?
2023-06-29 03:39:59 <Saijin_Naib[m]> Once it connects 😭
2023-06-29 03:40:30 <LACampbell> lol. I assume it's just /dev/mmcblk0p2 though
2023-06-29 03:40:32 <Saijin_Naib[m]> It loads each channel one by one and fetches messages one by one, so it takes forever (and it doesn't cache anything between sessions ) 
2023-06-29 03:40:56 <LACampbell> like if you do dev/mmcblk0p2 I assume that returns something
2023-06-29 03:41:03 <LACampbell> NAME      MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
2023-06-29 03:41:06 <LACampbell> nvme0n1p4 259:4    0 48.7G  0 part [SWAP]
2023-06-29 03:42:02 <Saijin_Naib[m]> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/JjIPhtyLeEXxEdQDvNdJWRvK

2023-06-29 03:42:51 <Saijin_Naib[m]> It's not on a simple partition, but one of two LVM volumes under that partition 
2023-06-29 03:44:26 <LACampbell> oh. no idea how to do that
2023-06-29 03:45:17 <LACampbell> would that be /dev/mmcblk0p2/vgo-lv_swap ? :/
2023-06-29 03:45:33 <LACampbell> lsblk /dev/mmcblk0p2/vgo-lv_swap
2023-06-29 03:46:31 <Saijin_Naib[m]> ```
2023-06-29 03:46:31 <Saijin_Naib[m]> /dev/vg0/lv_swap
2023-06-29 03:46:31 <Saijin_Naib[m]> ```
2023-06-29 03:46:35 <Saijin_Naib[m]> seems to be the direct path
2023-06-29 03:46:49 <LACampbell> ah! ok can you add that to the grub kernel flags then rebuild grub?
2023-06-29 03:47:07 <Saijin_Naib[m]> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/pXFbPXUgMRKpUeTpOMfiUsVe

2023-06-29 03:47:30 <LACampbell> can't you just paste it into the chat window? :P
2023-06-29 03:48:37 <Saijin_Naib[m]> Done, trying now
2023-06-29 03:49:07 <Saijin_Naib[m]> I was told to always put it in code blocks if it goes over a line so it gets pastebinned
2023-06-29 03:49:27 <LACampbell> IDK, I'm not an old hand here
2023-06-29 03:49:35 <LACampbell> but I reckon if the traffic is low and it's like 6 lines, who really cares
2023-06-29 03:50:09 <Saijin_Naib[m]> If there is anything I've learned about IRC, it is that someone cares
2023-06-29 03:50:25 <Saijin_Naib[m]> Hibernate worked a treat 
2023-06-29 03:50:34 <LACampbell> LOL
2023-06-29 03:50:38 <Saijin_Naib[m]> I didn't do the steps below the grub editing yet 
2023-06-29 03:50:46 <LACampbell> yeah IIRC it will work once
2023-06-29 03:50:50 <LACampbell> but then if you check lsblk again
2023-06-29 03:50:59 <LACampbell> your SWAP is gone. or the partition is ther ebut its not marked as swap
2023-06-29 03:51:21 <Saijin_Naib[m]> Nah, still shows the same as prior prints
2023-06-29 03:51:26 <Saijin_Naib[m]> Hibernate again to verify?
2023-06-29 03:51:36 <LACampbell> yeah, in unchartered waters for me now lol
2023-06-29 03:51:54 <Saijin_Naib[m]> Here we go
2023-06-29 03:53:08 <Saijin_Naib[m]> I really do not like the flicker that happens upon hibernate... All goes black, then it shows the display for a few second and then poof off
2023-06-29 03:53:17 <LACampbell> yeah I have that too
2023-06-29 03:53:18 <Saijin_Naib[m]> Second hibernate worked fine 
2023-06-29 03:53:33 <Saijin_Naib[m]> Maybe LVM avoids whatever issue with the swap partition? 
2023-06-29 03:53:47 <Saijin_Naib[m]> Lsblk unchanged still
2023-06-29 03:53:52 <LACampbell> might be the issue. my rig is very basic. it's one of those mini PCs I put together
2023-06-29 03:54:03 <Saijin_Naib[m]> Sweet! 
2023-06-29 03:54:24 <LACampbell> (if anyone is thinking about a mini pc, I recommend getting the next size up, they get hot lol)
2023-06-29 03:54:25 <Saijin_Naib[m]> 4GB RAM 64GB eMMC Celeron N3450 craptop here 🤣
2023-06-29 03:54:55 <LACampbell> nice! what are you doing on it?
2023-06-29 03:55:12 <LACampbell> I was on similar stats until a few years ago
2023-06-29 03:55:31 <Saijin_Naib[m]> Work, haha. I use it with Alpine Edge XFCE, GNOME Builder, Firefox, and Evolution
2023-06-29 03:55:52 <Saijin_Naib[m]> Documentation, Community management, and tech support 
2023-06-29 03:56:05 <Saijin_Naib[m]> Real computer is a desktop but I don't use it often anymore 
2023-06-29 03:56:54 <Saijin_Naib[m]> Laptop is more convenient and I can be present with my family on it vs going to another room to work on the desktop 
2023-06-29 03:57:37 <Saijin_Naib[m]> Thanks for that guide! First time I have hibernate in Alpine 😎
2023-06-29 03:57:48 <Saijin_Naib[m]> Gotta do it to the desktop now
2023-06-29 03:59:44 <LACampbell> no problem, I've been through this like twice before
2023-06-29 03:59:50 <LACampbell> and I was llike "This time, write it down"
2023-06-29 04:18:20 <Saijin_Naib[m]> psykose: you were right, looks like some webkitgtk update fixed broken Evolution and Epiphany
2023-06-29 04:18:23 <Saijin_Naib[m]> Thank you! 
2023-06-29 04:18:28 <psykose> hmm
2023-06-29 04:18:37 <psykose> are you sure it was that and not all of the pam stuff above you did
2023-06-29 04:18:51 <Saijin_Naib[m]> Hmm 🤔 
2023-06-29 04:19:10 <psykose> having the pam-rundir with elogind would have probably done it because it breaks something with the rundir iirc
2023-06-29 04:19:12 <Saijin_Naib[m]> Still broken on laptop where I did all the Pam and elogind stuff but didn't update 
2023-06-29 04:19:18 <psykose> okie
2023-06-29 04:19:20 <psykose> well, update and check
2023-06-29 04:19:26 <Saijin_Naib[m]> Will do
2023-06-29 04:19:33 <Saijin_Naib[m]> Any logs to grab or no? 
2023-06-29 04:20:16 <psykose> nope
2023-06-29 04:23:41 <Saijin_Naib[m]> Hibernate on desktop successful, which is nice. And fast! Thought 32GB RAM would take a bit onto SATA3 SSD
2023-06-29 05:26:48 <Saijin_Naib[m]> Upgrading webkit2gtk (2.40.2-r2 -> 2.40.3-r0)
2023-06-29 05:27:03 <Saijin_Naib[m]> This seems to have fixed Epiphany and Evolution after the past two weeks or so of breakage
2023-06-29 05:29:13 <psykose> well
2023-06-29 05:29:20 <psykose> good i guess, though it makes little sense :D
2023-06-29 05:29:32 <psykose> the diff of commits isn't anything relevant
2023-06-29 05:39:48 <LACampbell> Saijin_Naib[m]:your desktop is quite a bit beefier than the laptop haha
2023-06-29 05:39:58 <LACampbell> 
ACTION always forgets if he has 16 or 32 gb of ram)

2023-06-29 05:41:47 <Saijin_Naib[m]> <psykose> "good i guess, though it makes..." <- Hey I'm happy. Clean installed both machines apparently for nothing, but at least now I'm using XFS, I guess 🤷
2023-06-29 05:41:54 <LACampbell> https://twitter.com/LewisCTech/status/1674290661284261888/photo/1
2023-06-29 05:42:12 <Saijin_Naib[m]> LACampbell: Yeah, and that's not even close to enough for what I need
2023-06-29 05:42:21 <Saijin_Naib[m]> 128GB would be more comfy 
2023-06-29 05:44:02 <Saijin_Naib[m]> I bought my craptop new for $110... It's totally excellent for that price, especially when you factor in it came with a year of Office 365, a license of Minecraft, and Windows 10 S (which I unlocked to Pro) 
2023-06-29 05:44:46 <quinq> So craptop with pre-installed crapsoft
2023-06-29 05:44:49 <Saijin_Naib[m]> Desktop I built for far cheaper than retail, thanks to a perk from a past job. i7-6700k year of launch for $115 was very nice to build on
2023-06-29 05:45:07 <Saijin_Naib[m]> Ain't happening again 🤣 I'm holding onto that thing until it won't boot
2023-06-29 05:45:49 <Saijin_Naib[m]> 🤷 It was useful for my job at the time. Linux isn't suitable for running the software required for the teaching I was doing
2023-06-29 05:46:42 <Saijin_Naib[m]> Minecraft was a nice bonus 🤣 finally got to try it out, and still use it via Flatpak now
2023-06-29 05:50:20 <LACampbell> at this point computers are pretty much powerful enough to do everything for me
2023-06-29 05:50:25 <LACampbell> and the internet is fast enough to
2023-06-29 05:53:12 <psykose> always just need a few more electron apps
2023-06-29 05:53:17 <LACampbell> lol
2023-06-29 05:53:41 <LACampbell> I have more PWAs than electron apps installed
2023-06-29 05:53:53 <LACampbell> which are not ideal, but they at least use your machines resource more efficiently
2023-06-29 05:54:44 <quinq> What's a PWA?
2023-06-29 05:54:50 <quinq> Pirate With Attitude?
2023-06-29 05:55:09 <LACampbell> progressive web app. basically instead of getting an electron version, you just click the 'install' button on a web page
2023-06-29 05:56:03 <quinq> For installing what?
2023-06-29 05:56:20 <LACampbell> web apps
2023-06-29 05:56:20 <quinq> Files in cache? Or is it browser-specific?
2023-06-29 05:56:33 <LACampbell> it's broadly supported across browsers. 
2023-06-29 05:56:51 <psykose> famously firefox doesn't
2023-06-29 05:57:13 <psykose> chromium/webkit does tho
2023-06-29 05:57:15 <LACampbell> like you can 'install' twitter on your phone, there'll be a little icon on the homescreen, it will open in its own branded widnwo,.. but its just your main browser
2023-06-29 05:57:15 <psykose> convenient
2023-06-29 05:57:45 <LACampbell> psykose: well PWAs are kind of a grab bag of features. I don't test with FF anymore, what doesn't it support?
2023-06-29 05:57:59 <psykose> pwas in general at all
2023-06-29 05:58:04 <psykose> you can't "install" a page with firefox
2023-06-29 05:58:15 <quinq> Yeah, firefox is a web browser
2023-06-29 05:58:20 <LACampbell> really? I installed my home page on my android phone with firefox
2023-06-29 05:58:26 <psykose> android is different
2023-06-29 05:59:08 <quinq> That's too much smelly topics for one morning, bbl :)
2023-06-29 05:59:20 <psykose> https://developer.mozilla.org/en-US/docs/Web/Progressive_web_apps/Guides/Making_PWAs_installable
2023-06-29 05:59:47 <psykose> tl;dr everyone on a phone and on desktop 'Firefox and Safari do not support installing PWAs on any desktop operating systems.'
2023-06-29 06:00:21 <LACampbell> well, I wonder how I did it with firefox on an android and safari on an iphone :)
2023-06-29 06:01:03 <psykose> because firefox on android and safari on iphone supports it
2023-06-29 06:01:13 <LACampbell> ah right
2023-06-29 06:03:10 <LACampbell> if chrome uses manifest v3, it might revitalise FF
2023-06-29 06:07:22 <psykose> i don't think the reason ff doesn't support it is a very technical one, but idk
2023-06-29 06:07:34 <psykose> they used to support them then removed it
2023-06-29 06:07:55 <psykose> ah, or were working on it
2023-06-29 06:07:59 <psykose> was years ago
2023-06-29 06:08:09 <LACampbell> I mean to be honest, in 2023 firefox isn't really a factor anymore for this stuff
2023-06-29 06:55:44 <idkrn[m]> <psykose> "android is different" <- Based*
2023-06-29 06:56:31 <abby> based on what
2023-06-29 07:00:03 <psykose> u
2023-06-29 07:00:42 <abby> :O
2023-06-29 07:01:53 <psykose> not gonna like
2023-06-29 07:02:02 <psykose> this is Definitely me when i'm Being based
2023-06-29 07:29:32 <LACampbell> abby: the number of people using it
2023-06-29 07:29:52 <LACampbell> oh nvm
2023-06-29 07:30:36 <psykose> fell for zoomer meme
2023-06-29 07:32:27 <abby> gottem
2023-06-29 08:57:54 <Gnd> Hello everybody
2023-06-29 08:58:08 <Gnd> please help me configure hd-idle for 2 hdd
2023-06-29 09:00:32 <Gnd> in the file /etc/conf.d/hd-idle
2023-06-29 09:00:46 <Gnd> i only have DISK="/dev/sdb"
2023-06-29 09:02:24 <Gnd> how add harddisk 2 /dev/sdc ?
2023-06-29 09:04:45 <aron> looking its init.d and conf.d, I see only an utterly ugly solution
2023-06-29 09:05:12 <aron> according to the manpage, this is legal:   hd-idle -i 0 -a sda -i 300 -a sdb -i 1200
2023-06-29 09:05:48 <aron> thus, writing more stuff into a variable in conf.d "solves" it, but it is ugly AF
2023-06-29 09:06:13 <aron> (see, no quotation marks around the variable expansions in conf.d)
2023-06-29 09:08:22 <itd090fso> since busybox-initscripts package was removed, what is the prefered way to run p.a. crond?
2023-06-29 09:11:30 <ikke> What is p.a. grond?
2023-06-29 09:11:38 <ikke> crond*
2023-06-29 09:12:30 <psykose> busybox-openrc that you already have installed
2023-06-29 09:12:53 <psykose> nothing changed
2023-06-29 09:15:51 <ikke> 8b30ff10ab691ce8380423c551dd9c327a5a1f9a
2023-06-29 09:17:24 <itd090fso> nice, thanks a lot
2023-06-29 09:19:19 <bananicorn> Hello :) I noticed that pppd now crashes when starting the service with default settings, with the error message "no device specified and stdin is not a tty"
2023-06-29 09:20:34 <bananicorn> I'd be thankful for any pointers about what I have to do to get it working again, unfortunately couldn't figure it out yet on my own :)
2023-06-29 09:23:28 <ikke> Sounds like some configuration that needs to be set, don't know pppd though 
2023-06-29 09:35:31 <psykose> i imagine you need to make a 
2023-06-29 09:35:37 <psykose> /etc/conf.d/pppd
2023-06-29 09:35:42 <psykose> with an arg for some device
2023-06-29 09:36:14 <psykose> like `PPPD_OPTS="call name-of-interface"`
2023-06-29 10:24:12 <Someone[m]1> Hi, why does no shell on Alpine do source /etc/profile when run? Is there a solution?
2023-06-29 10:25:53 <Someone[m]1> On login, yes, but on execute, no
2023-06-29 10:26:13 <ikke> That's how it is defined 
2023-06-29 10:26:33 <psykose> because profile is for login
2023-06-29 10:26:36 <ikke> /etc/profile will only be sourced for login shells 
2023-06-29 10:27:43 <psykose> you can replicate the same effect with shell -l or whatever
2023-06-29 10:29:13 <Someone[m]1> is there any file (global or non-global) that is executed by ash, mksh or dash?
2023-06-29 10:29:31 <Someone[m]1> * or dash on non-login shell?
2023-06-29 10:30:15 <Someone[m]1> psykose: I'm using sway and launching foot will start without the arg
2023-06-29 10:30:25 <psykose> yeah that's normal
2023-06-29 10:30:39 <psykose> shells are not meant to source /etc/profile every time
2023-06-29 10:31:56 <psykose> for ash you can use ENV= to have something sourced every start
2023-06-29 10:32:05 <psykose> e.g. https://unix.stackexchange.com/a/418715
2023-06-29 10:32:45 <psykose> that way you can create an ashrc that does what you want
2023-06-29 14:04:39 <minimal> Gnd: Alpine hd-idle maintainer here. Actually the conf.d file is misleading. In general, by default, hd-idle will spindown *all* disks. However as this may be undesired the current Alpine package specifies "-i 0" for all disks (so they do NOT spindown by default") and then also uses the DISK variable to indicate a single disk that should spindown (overriding the default non-spindown)
2023-06-29 14:05:07 <minimal> I'll have to rewrite the conf.d/init.d to handle specifying multiple disks.
2023-06-29 14:14:42 <minimal> actually from re-reading the upstream README "-a" *does* appear to accept multiple device names
2023-06-29 18:45:05 <elagost> is there a status page for package build systems? The repos on edge have not seen updates on arm64 for a little bit.
2023-06-29 18:46:26 <ikke> elagost: There is https://build.alpinelinux.org/
2023-06-29 18:46:46 <ikke> elagost: but: https://gitlab.alpinelinux.org/alpine/infra/infra/-/issues/10800/
2023-06-29 18:48:13 <elagost> thank you ikke - bookmarked!
2023-06-29 18:48:44 <ikke> The builders have been unreachable and thus were not able to upload built packages
2023-06-29 18:58:53 <psykose> 3.18 should be soon, would do edge after
2023-06-29 18:58:56 <psykose> but that will take a long time
2023-06-29 18:58:59 <psykose> lots of queue
2023-06-29 18:59:10 <ikke> I'll have to do edge tomorrow
2023-06-29 19:00:30 <elagost> cool, thank you for the hard work. infrastructure can be finicky!
2023-06-29 22:09:43 <jcatv> Any tips on setting up a vmware hosted alpine linux instance to run a mail server?
2023-06-29 22:15:56 <Habbie> i have zero tips, but for what part are you looking for tips?
2023-06-29 22:19:47 <minimal> Nike tip: "Just do it!" ;-)
2023-06-29 22:20:13 <Habbie> yes
2023-06-29 22:20:15 <Habbie> agreed
2023-06-29 22:20:24 <Habbie> especially on a VM, can always just redo it
2023-06-29 22:22:45 <jcatv> nothing in particular. but it occurred to me if there were any gotchas it would be nice for an Alpine linux newbie to know about them.
2023-06-29 22:41:09 <raspbeguy> A useful tip : if you power off your VM, then it won't be able to serve emails
2023-06-29 22:41:32 <Habbie> but the cloud provider will still charge you the exact same amount of money
2023-06-29 22:41:38 <Habbie> (or a few cents less with some)
2023-06-29 22:42:02 <Habbie> as much as raspbeguy might be half joking, my addition is a real thing that has surprised people around me
2023-06-29 22:42:13 <Habbie> and a thing that you can easily "do wrong" in some clouds
2023-06-29 22:54:51 <raspbeguy> Reminds me this https://turnoff.us/geek/security-expert/
2023-06-29 22:55:27 <Habbie> lol yes
2023-06-29 22:55:51 <Habbie> wonder why the RSS says .sign()
2023-06-29 22:55:56 <Habbie> did they mean .signup() ?
2023-06-29 23:22:30 <jcatv> raspbeguy You reminded me that back before the cloud was the thing we had a networking security tool that was almost foolproof. 
2023-06-29 23:22:35 <jcatv> http://purplebark.net/maffew/scissors.pdf
2023-06-29 23:24:43 <jcatv> I was looking at this wiki page as a starting point but that big "work-in-progress" notice has me thinking. 
2023-06-29 23:24:46 <jcatv> https://wiki.alpinelinux.org/wiki/ISP_Mail_Server_3.x_HowTo
2023-06-29 23:26:47 <jcatv> It is probably an opportunity to update the wiki however.
2023-06-29 23:28:04 <jcatv> One question I had is how much RAM would that use and the RAM estimate wiki page seemed inadequate. 
2023-06-29 23:28:39 <jcatv> https://wiki.alpinelinux.org/wiki/Requirements
2023-06-29 23:30:31 <jcatv> Has anyone done any real world testing on a similar scenario? How about a live server that you adjusted the RAM in and created a performance boost?
2023-06-29 23:31:02 <jcatv> Or am I overthinking it because it is a new system that runs mostly in RAM and I am new.
2023-06-29 23:32:36 <Apachez> for baremetal I prefer 8GB minimum for 64bit OSes
2023-06-29 23:32:49 <Apachez> and 4GB for 32bit (sure there is PAE but meh)
2023-06-29 23:33:06 <Apachez> when it comes to VM's it depends what you run inside it
2023-06-29 23:33:40 <Apachez> except for the kernel itself you normally have stuff like (give or take) ntpserver, dnsserver, logging towards syslogserver
2023-06-30 05:51:26 <Saijin_Naib[m]> Yeah, FireFox not supporting PWAs when they pioneered OWAs that seem to have been the basis of current PWAs boggles the mind.
2023-06-30 05:51:29 <Saijin_Naib[m]> Damn, do I miss b2g/FireFoxOS
2023-06-30 07:09:19 <oleovit> hi. short question. i'm using chrony(d), nts, with rtcsync directive. can/should i set clock_systohc="NO" in /etc/conf.d/hwclock. i'm a little bit lost on this topic..
2023-06-30 07:57:31 <oleovit> + CONFIG_RTC_HCTOSYS=y CONFIG_RTC_HCTOSYS_DEVICE=rtc0
2023-06-30 07:58:18 <oleovit> is set, so i guess i have to set systohc to NO
2023-06-30 08:51:24 <raspbeguy> jcatv, this tool is not very reversible
2023-06-30 11:14:22 <minimal> oleovit: "is set, so i guess i have to set systohc to NO" - why?
2023-06-30 11:15:39 <minimal> oleovit: actually you don't need /etc/init.d/hwclock enabled at all, just enable /etc/init.d/osclock instead
2023-06-30 11:21:40 <minimal> oleovit: the kernel has CONFIG_RTC_SYSTOHC=y and CONFIG_RTC_HCTOSYS=y set and so during boot the *kernel* will set the system clock from the RTC (before OpenRC is even run) and, once chrony achieves NTP sync, then the *kernel* will copy the time from system clock to RTC every 11 mins, so there is no need for the hwclock script to be used to do either of those. /etc/init.d/osclock is a dummy script just to meet the "clock" service dependency
2023-06-30 11:23:33 <oleovit> @minimal thx for response. i don't want to change too much, so i want to stick with the hwclock service. your first answer: even if these parameters are enabled in the kernel, i guess, one has still to enable it "explizit": https://0x0.st/HQkY.txt
2023-06-30 11:24:47 <minimal> oleovit: the kernel RTC->SYS copy is done automatically on every boot, "dmesg" will show it happening
2023-06-30 11:25:21 <minimal> the kernel SYS->RTC every 11mins is controlled by the chrony.conf "rtcsync" setting which you are using
2023-06-30 11:26:15 <minimal> if you have hwclock enabled then *both* the kernel and the hwclock init.d script will copy RTC to SYS during boot
2023-06-30 11:26:46 <minimal> (this is assuming the drivers for your RTC are compiled into the kernel - for x86/x86_64 machines they are, for other machines they may or may not be)
2023-06-30 11:28:07 <minimal> as you have "rtcsync" in your chrony.conf then the kernel will copy time from SYS->RTC every 11mins *plus* during shutdown hwclock will copy from from SYS-RTC as well, the 2nd thing is unnecessary
2023-06-30 11:28:45 <minimal> so you have 2 things manipulating both RTC and SYS clocks - the kernel/chrony and also hwclock
2023-06-30 11:32:18 <minimal> oleovit: I think you should only have 1 thing manipulating the clocks, the kernel (together with chrony "rtcsync")
2023-06-30 11:32:20 <oleovit> i understand, the chrony wiki is clear on this topic with this directive set. so, if i decide to stick with hwclock, and to avoid additional not necessary stuff from this service, i guess, i have to set clock_systohc="YES" , or am i wrong  
2023-06-30 11:32:50 <minimal> oleovit: ?
2023-06-30 11:32:58 <oleovit> yes?
2023-06-30 11:33:41 <minimal> clock_systohc="YES" is *already*  the default config in /etc/conf.d/hwclock. Which means that during shutdown hwclock will set RTC time to that of SYS
2023-06-30 11:34:08 <minimal> however this is unnecessary as kernel will be doing the same sync every 11mins in general so why ALSO do it during shutdown?
2023-06-30 11:35:03 <oleovit> it's the chat, i cant concentrate when i chat, its overkill for me, sry. no, i mean set it to NO : lock_hctosys="NO"
2023-06-30 11:35:34 <minimal> and you haven't mentioned the default setting clock_hctosys="YES" - the comment just above that in /etc/conf.d/hwclock even mentions this is unnecessary with a modern kernel (as the kernel does the sync during boot)
2023-06-30 11:36:14 <oleovit> wait a second pls
2023-06-30 11:40:29 <oleovit> 1st: i have to set: clock_hctosys="NO", because YES is the default. thats for bootup. clock_systohc="YES" is the default, during shutdown. this, i have to leave it like it is.
2023-06-30 11:41:23 <oleovit> of i do not set clock_hctosys="NO" , we have these multiple stuff going on
2023-06-30 11:41:40 <minimal> oleovit: why do you "have" to leave it as clock_systohc="YES"? There's no need to write the SYS to RTC at shutdown as the kernel will already be copying from from SYS to RTC every 11mins
2023-06-30 11:42:00 <minimal> s/from from/time from/
2023-06-30 11:46:07 <minimal> looking at dmesg output from a x86_64 VM here, I see "[    0.747086] rtc_cmos rtc_cmos: setting system clock to 2023-06-30T11:38:11 UTC (1688125091)"
2023-06-30 11:47:29 <oleovit> pls wait, again, cant concentrate while chat ..
2023-06-30 11:48:04 <minimal> that is the *kernel* setting the SYS from RTC 0.74 seconds after kernel started initialising, rootfs is not even mounted until after 2.7 seconds and so OpenRC starting and running hwclock service will take longer still
2023-06-30 11:49:43 <oleovit> yes, but let's concentrate on the file. everything else will turn out anyway, when i take a look after the chat is finished. so, step for step. one moment
2023-06-30 11:52:09 <minimal> oleovit: I don't understand, I have already explained how things work and why what you are doing is unnecessary
2023-06-30 11:57:27 <oleovit> 1st bootup: change to clock_hctosys="NO". YES, the default is not needed. 2nd: shutdown: leave clock_systohc="YES" , in order to set hw-clock to system time. ahh, i dont know whats wrong with me, chat, silent migraine, i'm fucked today 
2023-06-30 11:58:27 <minimal> oleovit: WHY do you want hwclock to set RTC from SYS clock at shutdown when you have already configured chrony so that kernel will set RTC from SYS *every 11mins* ?
2023-06-30 11:59:13 <minimal> you can set RTC at shutdown but there is no need to do so with the 11min auto-sync in place
2023-06-30 12:01:32 <oleovit> yes, thats what the gentoo source in my uploaded file is telling me/us. so, clock_systohc="NO" too? booth to f.. NO ? i was confused because of these 2 configs (alpine/gentoo)
2023-06-30 12:02:30 <oleovit> the kernel params and the chrony directive are enough, so set NO to all of them 
2023-06-30 12:02:46 <minimal> yes I have been repeatedly saying both set to "NO" - in which case the hwclock init.d script then does not actually do anything and therefore why enable it? Disable it and enable osclock instead
2023-06-30 12:05:10 <oleovit> thats my chat-stress-cant-concentrate+silent+migraine thing. sry. so, yes, i begin to understand. so, NO NO because of this and that.. or disable the service and enable osclock instead. right? ;)
2023-06-30 12:05:46 <dwfreed> do the latter, disable hwclock completely and enable osclock
2023-06-30 12:06:13 <minimal> you're just repeating back to me what I said
2023-06-30 12:07:50 <oleovit> repeating back is a good thing in my case. i said: silent migraine, cant concentrate while chat. so thx for your patience. 
2023-06-30 12:10:18 <psykose> `rc-update -a del hwclock; rc-update add oslock boot`
2023-06-30 12:12:27 <oleovit> yes, thx to you too psykose. again, the 2 configs distracted me. i need time to process information, because nothing can be "known"by being told
2023-06-30 12:14:24 <oleovit> and dwfreed.. see, i'm blind of too much is going on in a chat. thx to all, and your patience
2023-06-30 13:08:40 <oleovit> did it. sorry @minimal for stressing him too. don't be mad at me. i've read the gentoo/alpine comments + chrony faq and some lorthing - with chat closed - and time - now it becomes clear, that this is a logical step. maybe, some thing for the wiki too. thx
2023-06-30 13:09:46 <oleovit> "lore thing", not lorething 
2023-06-30 19:08:56 <msiism> Splitting off documentation info -doc sub-packages is done to kee the main packages small, right?
2023-06-30 19:09:02 <msiism> s/kee/keep/
2023-06-30 19:09:31 <psykose> packaging granularity
2023-06-30 19:10:18 <msiism> And does size play a role, too?
2023-06-30 19:10:28 <ikke> yes
2023-06-30 19:51:00 <jcatv> reading https://wiki.alpinelinux.org/wiki/Installation#Data_Disk_Mode 
2023-06-30 19:51:54 <jcatv> yesterday I asked about RAM. It seems a system requirement in Data Disk Mode. Does any one use this? How? What For?
2023-06-30 20:34:12 <ikke> jcatv: Your question is not clear to me
2023-06-30 20:35:13 <msiism> So, how do man-pages and man-pages-posix relate? Why would I install the latter as well?
2023-06-30 20:35:57 <ikke> msiism: man-pages are the linux kernel man pages, mostly related to the libc interface
2023-06-30 20:37:05 <msiism> Yeah, I've read a bit about that.
2023-06-30 20:37:22 <msiism> Then I discovered man-pages-posix int he Git tree.
2023-06-30 20:37:26 <msiism> s/he/the/
2023-06-30 20:37:38 <msiism> s/int/in/
2023-06-30 20:37:51 <ikke> posix man pages document the POSIX standards
2023-06-30 20:37:59 <ikke> which is more general then just linux
2023-06-30 20:38:26 <msiism> Bu there will be no conflict when installing both, right?
2023-06-30 20:39:04 <ikke> posix provides man0
2023-06-30 20:39:30 <msiism> I see.
2023-06-30 20:40:08 <ikke> They both provide several sections, but they don't overlap
2023-06-30 20:40:20 <msiism> Good to know.
2023-06-30 23:05:13 <malmi70[m]> I need to dual boot and encrypted Alpine will be the first installation. Should I shrink the existing partition or manually do the partition during installation (how?).
2023-06-30 23:05:13 <malmi70[m]> Asking because the wiki article "lvm on luks" says: "if you want to dual boot, stop here and ask an expert". Funny statement.
2023-06-30 23:05:53 <Habbie> hehe
2023-06-30 23:06:06 <Habbie> probably means "this is a hard problem and i cannot begin to write down what could go wrong"
2023-06-30 23:06:26 <Habbie> i see how it's not useful documentation, other than stopping people from going in and ruining their setup because they didn't know
2023-06-30 23:06:51 <Habbie> that said, i happen to dual boot with lvm on luks, so maybe i can help
2023-06-30 23:06:57 <Habbie> i'm not running alpine but that doesn't seem too relevant
2023-06-30 23:08:07 <Habbie> debian wasn't the first OS i installed, it was the second
2023-06-30 23:08:28 <Habbie> i shrunk windows first
2023-06-30 23:08:37 <Habbie> then i did indeed partition manually
2023-06-30 23:08:48 <Habbie> and assigned 100GB to a LUKS partition, which i then put LVM in
2023-06-30 23:08:52 <malmi70[m]> Setup-alpine provides very easy install but shrinking the existing install might be pain.
2023-06-30 23:08:52 <malmi70[m]> The manual partitioning without using the whole disk isn't well documented.
2023-06-30 23:09:00 <Habbie> right
2023-06-30 23:09:55 <Habbie> i've always been comfortable just using the command line tools for partitioning when necessary, but this does not help you much
2023-06-30 23:11:42 <malmi70[m]> Me too (coming from Arch, btw), but only if setup-alpine had an option to not use the whole disk for root would solve all my problems right now lol
2023-06-30 23:11:58 <Habbie> right :)
2023-06-30 23:20:45 <minimal> malmi70[m]: dual boot complexity also depends on whether it is MBR or UEFI booting, whether it is dual boot Windows/Alpine or Alpine/another Linux distro etc