deb2dev | doing the buster to beowulf upgrade and have a problem. Have sysvinit installed but eudev won't install without removing systemd first | 06:08 |
---|---|---|
deb2dev | says nothing about this problem in the installation guide | 06:09 |
deb2dev | wondering is I can use --force-yes as I'll have to reboot anyway | 06:10 |
deb2dev | as long as eudev gets installed won't matter I'm thinking. Anyone have the same problem, should be like a cookie cutter, same result every time | 06:11 |
deb2dev | maybe aptitude can get it done, need to put a hold on systemd, something fancy like that | 06:13 |
deb2dev | damn son, I ain't no debian virgin no more | 07:57 |
gnarface | deb2dev: you were already gone by the time i saw your message. did aptitude figure it out for you? | 07:58 |
deb2dev | I boot back into the system and was able to get eudev and libeudev1 installed | 07:58 |
deb2dev | ok, here's what I did | 07:58 |
deb2dev | aptitude download eudev libeudev | 08:00 |
deb2dev | dpkg --force-conflicts --auto-deconfigure -i | 08:01 |
deb2dev | then the 2 .deb packages | 08:01 |
deb2dev | I'd recommend doing --dry-run first | 08:01 |
deb2dev | I need to get the website updated now | 08:02 |
gnarface | hmm, i think it would have worked to just let it remove systemd if you made sure you reinstalled everything you needed from the devuan repos instead | 08:02 |
gnarface | you're right that a lot of people have been through this, and the main issues should be mentioned in the release notes... | 08:03 |
gnarface | but if you got it working that's what is important | 08:03 |
deb2dev | this info would be a nice update to the migrate document | 08:03 |
deb2dev | they say in the guide to remove sysyemd after reboot, but you can't do it that way until eudev is installed | 08:04 |
deb2dev | if I was a gnulinux newbie I would have reboot not realizing eudev was not installed | 08:06 |
deb2dev | you know who I can contact about this, I'd like to help anyone else out with the same problem | 08:07 |
deb2dev | I went from a Buster clone to Beowulf clone, everyone should have the same problem | 08:08 |
gnarface | it's a slow channel, be patient. people will see this though. i can't be 100% sure myself if the docs are out of date or if there's some other cause. | 08:13 |
gnarface | i assume that actually you would have been able to install eudev after the reboot too... the regular udev used to still work with sysvinit anyway | 08:13 |
gnarface | but you'd be surprised how many variables there are | 08:14 |
deb2dev | document just needs an update to help out if anyone else has the same problem | 08:14 |
gnarface | even if you didn't overlook something, it is possible for you to run into something nobody else did | 08:14 |
gnarface | you might be right, it might just be an old document too | 08:14 |
gnarface | which one did you use? | 08:15 |
gnarface | was it the README.txt in the directory with the installer, or some forum post? | 08:15 |
deb2dev | ok, that was a question I couldn't find an answer for, whether sysvinit would work with udev after a reboot | 08:15 |
gnarface | eudev is supposed to provide equivalent functionality to udev | 08:15 |
gnarface | the primary reason it exists was that this wasn't expected to be a stable state of affairs after they renamed udev to "systemd-udevd" | 08:16 |
deb2dev | buster-to-beowulf document on devuan.org website | 08:16 |
gnarface | also the word "systemd" just showing up as a substring in anyone's process list output generated an abnormally high amount of support traffic | 08:17 |
deb2dev | after the reboot I ran apt-get --fix-broken to remove systemd | 08:18 |
deb2dev | then I was able to run apt-get dist-upgrade | 08:18 |
leizaola | Hello I have a question about this project.... what is the reason for doing a full distro instead of releaseing a set of scripts that would convert debian into debuan and could be applied to a normal debian installation? | 08:19 |
deb2dev | debian has lost it's way my brother | 08:20 |
leizaola | i know... | 08:20 |
leizaola | but all that we need to do is remove systemd, purge in install sysvinit and do a few things to clean it up... | 08:21 |
deb2dev | I like debians package manager, but I don't like some of the stupid shit they adopt, it's probably more then I need | 08:24 |
deb2dev | I like the old gnulinux philosophy of keeping things simple | 08:25 |
deb2dev | I understand I need the latest libraries, but some things I know are going on that suck | 08:26 |
leizaola | me too... now it is chaos, nmcli, network manater, netplan, systemdctl resolvctl i mean time to pull your hair | 08:26 |
leizaola | we waste so much time.... | 08:26 |
deb2dev | so much shit | 08:26 |
leizaola | why change things that worked so for for 20-30 years..... | 08:26 |
gnarface | leizaola: most the packages aren't even altered, they're served by http redirect through devuan servers from debian servers, actually. mostly. | 08:27 |
deb2dev | to complicated for me, I'm a flat earther | 08:27 |
gnarface | leizaola: they didn't fork the whole distro, just the broken packages they could afford to fix. | 08:27 |
leizaola | my question is whether this project warrants a full distro.... instead of a script that removed all the crap out of debian and turns it into a usable Linux again... | 08:28 |
gnarface | leizaola: trust me, devuan is what you want, after the idea you've started with permutes through the necessary iterations to face reality, be future-proofed, and work as a drop-in replacement that can't be moved out from underneath the users on the fly by a hostile upstream | 08:28 |
leizaola | I personally run debian on proxmox lxc containers, I have a master template where I removed systemd then I clone it and spawn all my other containers from it.... | 08:29 |
gnarface | leizaola: yea, and that works only so far. there's a few hundred packages that need to actually be rebuilt to remove systemd dependencies. | 08:29 |
gnarface | leizaola: luckily it's not all of them. not even most of them. | 08:29 |
gnarface | leizaola: (yet - but they baked in a plan in case that changes) | 08:30 |
leizaola | i've seen a couple but generally I have survived.... | 08:30 |
gnarface | some people just want to use those | 08:30 |
gnarface | devuan has made an attempt to preserve as much as they could | 08:30 |
leizaola | is there a debuan proxmox template? | 08:30 |
gnarface | i don't know | 08:30 |
gnarface | there's build scripts on git.devuan.org if you want to see how changes are merged with debian | 08:31 |
gnarface | as for proxmox, maybe check the forum (url in /topic) | 08:31 |
gnarface | this is a slow channel, but people generally respond if you stay connected long enough | 08:32 |
leizaola | no probs.... | 08:33 |
deb2dev | I have another system to upgrade from stretch, going to experiment and go from stretch to devuan, see if I have the same problem | 08:45 |
deb2dev | I'll contact devuan about what needs to be updated on the migration document, especially the need to mention sysvinit will work with udev, but even if you can reboot, you still have the same problem, eudev conflicting with systemd | 08:47 |
deb2dev | off to bed for me | 08:47 |
ham5urg | Where should I put my little Python application in the Linux/Debian filesystem? If it only consists of a single my_app.py it would be easy to copy it to /usr/local/bin. But it consists of a bunch of other.py files too. Where should I put these? Should I use pip, via the Makefile, to install these? | 09:40 |
Helle | ham5urg: turn it into a python package, the my_app.py goes into bin (without the py extension) (you could even make it just a simple load and call the library at an entry point), the library goes into the Python library path. Once you have the Python package you can turn that into a Debian package without too much hassle if need be | 09:42 |
Chain-Q | I was just about to write the same | 09:42 |
ham5urg | Helle, thanks. I took a look with sys.path http://paste.debian.net/1186392/ and will copy it under /usr/local/... | 11:03 |
Helle | ham5urg: Python has a whole toolset to avoid just copying | 11:04 |
ham5urg | Helle, yes, I plan to use pip for installing/copying the files. | 11:05 |
ham5urg | But I dislike the Pip/Python way to bring in it's own repository. I will stick to Python packages from Devuan-repo. | 11:07 |
ham5urg | I plan to use Pip only to copy my stuff to the right direction. | 11:08 |
Beer | Anyone else faced a "Logical volume *** contains a filesystem in use." error while trying to deactivate/remove an unmounted LV? | 11:24 |
eyalroz | I'm having trouble with the v4l2loopback-dkms package | 11:33 |
eyalroz | it is supposedly failing to bild | 11:33 |
eyalroz | scratch that, actually failing to build... | 11:34 |
eyalroz | Build log: https://pastebin.com/raw/JdkdKk2J | 11:34 |
eyalroz | Looks like it's some sort of header mismatch | 11:35 |
eyalroz | This has been reported on other Debianesque distributions, e.g. Raspbian: https://github.com/RPi-Distro/repo/issues/188 | 11:36 |
eyalroz | hmm... looks like this is resolved by downloading this package from Debian backports: https://packages.debian.org/buster-backports/all/v4l2loopback-dkms/download | 11:38 |
eyalroz | ... which is weird, because I have Devuan backports enabled in my APT sources files: | 11:39 |
eyalroz | deb http://il.deb.devuan.org/merged beowulf-backports main non-free contrib | 11:40 |
rrq | eyalroz: is that the running kernel ... I have a memory of it getting confused with header files | 11:51 |
eyalroz | rrq: No, the running kernel is older, I just installed a new kernel version. But I was getting the error before I installed the new version, with 5.9.whatever, as well. | 11:53 |
eyalroz | context: I'm trying to setup v4l2sink, which is a plugin for using the obs-studio output as a virtual camera, that in turn lets you do fancy presentation-casting via Jitsi, Zoom, etc. | 11:55 |
rrq | sorry; can't remember.. I did that some weeks ago, but it's blank how I dealt with it ... | 12:01 |
rrq | actually it was v4l2loopback | 12:05 |
eyalroz | rrq: v4l2loopack is a prerequisite for v4l2sink | 12:13 |
eyalroz | and the latter is not a Debian package | 12:13 |
rrq | can I tempt you with a built .ko ? | 12:21 |
rrq | 5.10.0-0.bpo.3-amd64/extra/v4l2loopback.ko | 12:21 |
eyalroz | @rrq: I've already succeeded, using the backported package... | 12:38 |
* eyalroz needs to reboot now for the new kernel... | 12:38 | |
eyalroz | rrq: Well, the v4l2loopback module loads fine, it seems, but | 13:45 |
eyalroz | even before I do that - I've lost my audio :-( ... or rather, my on-board audio out is no longer accessible with ALSA, only other outputs I don't use | 13:46 |
eyalroz | I can still access it in some round-about way via ALSA (audacious can, anyway) | 13:46 |
eyalroz | as "sysdefault:CARD=PCH - HDA Intel PCH whatever" | 13:47 |
rrq | eyalroz: if that's card 1 you might need a ~/.asoundrc with defaults.ctl.card 1\ndefaults.pcm.card 1\ndefaults.timer.card 1\n in it | 14:20 |
eyalroz | rrq: apparently, when I installed my new kernel, this got installed too: | 14:21 |
eyalroz | debian-pulseaudio-config-override | 14:21 |
eyalroz | I suspect that had something to do with it. | 14:21 |
eyalroz | and its not card 1... card 0: PCH [HDA Intel PCH], device 0: ALC1150 Analog [ALC1150 Analog] | 14:22 |
rrq | possibly.. I think that makes pulsaudio start automatically | 14:23 |
eyalroz | well, it is running currently | 14:24 |
eyalroz | ... although `pactl list` gives me: "Connection failure: Connection refused" and "pa_context_connect() failed: Connection refused" | 14:26 |
rrq | perhaps kill and restart pulsaudio? it might be running as root? | 14:27 |
eyalroz | rrq: Will restarting it. But better use my name so I get a notification. | 14:52 |
eyalroz | no, it's not running as root. | 14:53 |
eyalroz | No change when restarting pulseaudio, nor when rebooting | 14:54 |
eyalroz | VLC manages to find the right output with no configuration | 14:57 |
eyalroz | ... though it fails to find the right mixer device | 14:57 |
eyalroz | (like it was failing before) | 14:57 |
deb2dev | anyone else have problem with startx and a user logging in to the X server after doing a migration from debian? | 20:10 |
deb2dev | I can startx as root | 20:11 |
deb2dev | modeset(0): drmSetMaster failed: permissiin denied | 20:13 |
deb2dev | that's in the xorg.9.log | 20:14 |
nemo | are you a member of the video group? | 20:14 |
deb2dev | but I get this error to the terminal | 20:15 |
deb2dev | yeah,the first forum I searched had a guy that said he checked that, I'll do it as well | 20:16 |
deb2dev | xf86EnableIOPorts: failed to set IOPL for I/O (Operation not permitted) | 20:17 |
deb2dev | in the terminal after startx fails, will search that next | 20:18 |
nemo | please check your groups and /dev/dri permissions first | 20:18 |
fsmithred | deb2dev, is elogind installed? | 20:25 |
fsmithred | if not, is xserver-xorg-legacy installed and did you edit Xwrapper.config? | 20:25 |
fsmithred | see https://files.devuan.org/devuan_beowulf/Release_notes.txt for starting X from a terminal | 20:26 |
deb2dev | fsmithred, that worked, used the Xwrapper.config with the needs_root_rights=yes | 20:51 |
deb2dev | now it's good ol lenny again | 20:52 |
deb2dev | ratpoison and gimp, till death do we part | 20:53 |
deb2dev | who should I contact that handles the Migration document on Devuan? | 20:55 |
deb2dev | best to update it with what I encountered to help anyone else streamline the process | 20:56 |
fsmithred | lenny was the best | 21:05 |
fsmithred | deb2dev, what happened? | 21:05 |
deb2dev | seems to be up and running now | 21:13 |
gnarface | fsmithred: he felt the removal of systemd wasn't clearly enough described in the setup document he hread | 21:14 |
gnarface | *read | 21:14 |
deb2dev | going to do fresh Beowulf installs with more security in mind | 21:14 |
fsmithred | thanks | 21:15 |
fsmithred | fresh install is easier | 21:15 |
fsmithred | migration is more variable | 21:15 |
deb2dev | because they assume you will install a window manager that doesn't use startx anymore there is no info in the migration doc for xinit and startx launches from a terminal | 21:16 |
deb2dev | I could help with the write up | 21:16 |
deb2dev | plus there was no way to install eudev without some tang team work of aptitude, dpkg | 21:18 |
mason | My cut-down cheat sheet: https://bpa.st/W377W | 21:18 |
mason | deb2dev: You can get eudev to install if you do it in two steps. | 21:18 |
deb2dev | the default options are to not allow eudev to be installed with systemd in place | 21:19 |
mason | Something that saddened me the other day was seeing how much systemd is in our sysvinit packages. Unit file handling, anyway. | 21:20 |
mason | But to be fair, there was also a ton of openrc stuff. | 21:21 |
deb2dev | great, that's another way to do it, only problem is the non-free and contrib repositories, this is where we become fatal enemies | 21:22 |
mason | Part of the notion of freedom is letting people use whatever they want. | 21:23 |
mason | Free will, as it were. | 21:23 |
deb2dev | and be free to choose your friends and foes | 21:24 |
mason | Whatever floats your boat I guess. | 21:24 |
deb2dev | we all have to choose our path, and you've lost your way | 21:27 |
deb2dev | mason my foe, I'll have to include your migration technique in the requested update to the documentation | 21:29 |
eyalroz | Could someone help me with an audio issue? | 21:29 |
deb2dev | foss 4 life | 21:29 |
mason | deb2dev: I just distilled stuff from the forums. Nothing original there. | 21:30 |
gnarface | eyalroz: don't ask for permission, just ask the actual question and be patient | 21:30 |
mason | fsmithred pointed me at the original | 21:30 |
eyalroz | My audio config is busted somehow after installing debian-pulseaudio-config-override. | 21:31 |
eyalroz | Details: https://unix.stackexchange.com/q/635799/34868 | 21:31 |
eyalroz | Basically - most apps stopped seeing my on-board audio output device | 21:31 |
gnarface | eyalroz: you should have used apulse with firefox, or firefox-esr | 21:32 |
mason | eyalroz: FWIW, you don't actually need PulseAudio for much. ALSA still works fine. | 21:33 |
eyalroz | @gnarface: audio in firefox wasn't working before and isn't working now; | 21:33 |
gnarface | eyalroz: i'm not aware of other programs having problems aside from steam... might be a bad diagnosis | 21:33 |
eyalroz | but Chromium and most other apps, and the configuration app, no longer see my output device | 21:33 |
mason | gnarface: Steam also working fine here, as a data point. | 21:33 |
gnarface | eyalroz: perhaps you added another soundcard without realizing it? | 21:34 |
eyalroz | I should mention I didn't voluntarily install this, it got pulled by some other packages | 21:34 |
gnarface | mason: not for remote-play (formerly known as "in-home streaming") | 21:34 |
eyalroz | gnarface: Well, I did install v4l2loopback | 21:34 |
gnarface | eyalroz: oh, loopback devices do count... | 21:35 |
mason | gnarface: Oh, I haven't tried that yet. | 21:35 |
gnarface | eyalroz: default alsa configuration plays to device #0 regardless of whatever it is | 21:35 |
eyalroz | But then I should just see another output device, not one less... | 21:35 |
gnarface | eyalroz: check "aplay -l" | 21:35 |
eyalroz | gnarface: follow the link, it has the first few lines of aplay -l ... | 21:36 |
mason | gnarface: Remote play worked last time I tried it, at least in its Proton guise. | 21:36 |
mason | but I'll try actual remote play | 21:36 |
eyalroz | gnarface: Or here's the full output: https://pastebin.com/raw/8nRNbjcb | 21:36 |
deb2dev | just remember this, everything you touch in life is a vote for what you support in this world | 21:39 |
gnarface | mason: if you don't have pulseaudio on the host end, you don't get audio on the client end | 21:40 |
deb2dev | you buy gmo shit food and some guy will have a patent on fruit that don't produce seed and you pay royalties on | 21:40 |
mason | deb2dev: There's #devuan-offtopic where we can debate that stuff. | 21:41 |
deb2dev | you support organic and we stay free and healthy | 21:41 |
gnarface | eyalroz: my man i see THREE hardware audio devices here | 21:41 |
deb2dev | fuck off | 21:41 |
eyalroz | Yes, there are lots. | 21:41 |
eyalroz | gnarface: But they were here before, too. | 21:41 |
mason | deb2dev: Clean up your language. | 21:41 |
deb2dev | blood in blood our sell out fag | 21:41 |
gnarface | eyalroz: what does "speaker-test -c 2" do? | 21:41 |
deb2dev | go fuck your shit face to hell | 21:42 |
gnarface | eyalroz: (should generate audio *and* console output) | 21:42 |
deb2dev | blood in blood out bitcb | 21:42 |
eyalroz | gnarface: Console output, no audio | 21:42 |
gnarface | eyalroz: any errors in the console output, or does it claim it is working right? | 21:42 |
eyalroz | no errors so far. | 21:42 |
mason | Third time's a charm. | 21:43 |
gnarface | eyalroz: if you run alsamixer, does it give you volume controls for "card #0: PCH [HDA Intel PCH]..." or just a single pulseaudio slider? | 21:44 |
gnarface | eyalroz: you can ctrl+c that speaker-test instance if you haven't, btw | 21:44 |
eyalroz | gnarface: alsamixer gives me: Card: PulseAudio, Chip: PulseAudio . | 21:45 |
gnarface | eyalroz: alright, that could be half the problem | 21:45 |
eyalroz | gnarface: I always knew pulseaudio is half of my problems... :-P | 21:46 |
gnarface | eyalroz: check pavucontrol, see if it's pointing to the wrong device too | 21:46 |
eyalroz | gnarface: pavucontrol only recognizes one output device: "HDMI/Displayport 2 (plugged in)" | 21:46 |
gnarface | eyalroz: check the furthest tab on the right (i forget what it's called... profiles or something?) | 21:47 |
eyalroz | Ah. | 21:47 |
eyalroz | Let me make a screenshot. | 21:48 |
gnarface | i'll look at it if you put it on imgur.com | 21:48 |
eyalroz | NM. The "built-in audio" option is : "Digital Stereo HDMI 2 output + Analog stereo in" | 21:49 |
eyalroz | So maybe that's the problem. | 21:49 |
gnarface | eyalroz: yea, it's sending the audio to your monitor it looks like. | 21:49 |
eyalroz | Unfortunately, there are about 30 options to choose from | 21:49 |
eyalroz | Ok, trying one of them. | 21:49 |
gnarface | eyalroz: yea, in my experience, the one that's supposed to work (the one that matches the name put next to card0 in the "aplay -l" output) won't work anyway. just try a few that look plausible. maybe you'll get lucky and not lose your surround speakers | 21:50 |
gnarface | eyalroz: plausible candidates would be the ones with "analog" in the output, i presume | 21:51 |
eyalroz | gnarface: I only have 2 speakers and a subwoofer | 21:51 |
eyalroz | Ok, got it | 21:52 |
eyalroz | When I choose 5.1 surround, it worked, magically. | 21:52 |
gnarface | nice, lucky you! | 21:52 |
mason | cool | 21:52 |
gnarface | that shit never works right for me | 21:52 |
eyalroz | gnarface: Wait, maybe audio in FF will work now, too... | 21:53 |
gnarface | eyalroz: yea, it should actually | 21:53 |
eyalroz | gnarface: Oh, wow, maybe this has also fixed my marriage and my cat! | 21:53 |
gnarface | eyalroz: you have to learn how to make pavucontrol and alsamixer play nice together to be a success at this over the long term. luckily in this case you didn't have to actually disable pulseaudio to fix alsamixer first. things could have been worse. | 21:54 |
gnarface | eyalroz: (note that i failed at such an endeavor and ended up pitching pulseaudio overboard) | 21:54 |
eyalroz | gnarface: This makes me wonder if the pulseaudio-let-me-rewrite-stuff-on-your-system package shouldn't be a bit more careful with its logic | 21:54 |
gnarface | hahaha | 21:55 |
gnarface | i can only assume adding and removing virtual loopback devices jumbled everything uyp | 21:55 |
gnarface | up | 21:55 |
gnarface | the firefox-esr build in the repos shouldn't actually need pulseaudio | 21:56 |
gnarface | and apulse won't help steam in-home streaming | 21:56 |
gnarface | but apulse should help anything else | 21:57 |
eyalroz | gnarface: This announcement: https://9to5linux.com/systemd-free-devuan-gnu-linux-3-1-distro-released-for-freedom-lovers | 21:57 |
gnarface | but if you can make alsamixer and pavucontrol play nicely and they both actually recognize all your channels in the right order, then congradulations, you're the only problem with your system (despite how that sounds, this is actually a huge victory in the grand scheme of things) | 21:57 |
eyalroz | says: "This release also ships with a new package (debian-pulseaudio-config-override) that promises to address issues with the PulseAudio sound system being off by default" | 21:57 |
gnarface | hmmm | 21:58 |
gnarface | news to me | 21:58 |
gnarface | but which part of the problem? the part where it's off or the part where it doesn't work while it's off? | 21:58 |
gnarface | and will that conflict with it actually being ON? | 21:58 |
gnarface | i don't actually know | 21:58 |
eyalroz | gnarface: Are you asking me? 8-\ | 21:58 |
gnarface | i thought you were asking me | 21:59 |
eyalroz | I was just emphasizing that the "we'll solve your audio problems" aspect of the release can introduce new ones sometimes | 21:59 |
gnarface | i threw pulseaudio overboard, remember? it may have seemed like i knew what i was doing when i helped you a moment ago but it was a lucky guess really... | 21:59 |
eyalroz | Let me try to figure out who maintains that package. | 22:00 |
eyalroz | Or maybe I should file a bug against PulseAudio somehow? | 22:00 |
eyalroz | I now have another challenge, which is getting the obs virtual camera device working properly. | 22:00 |
eyalroz | Now with my audio back, I'll give it another shot. | 22:00 |
fsmithred | the PA override package takes the place of editing /etc/pulse/client.conf.d/00_autospawn (or whatever it's called) | 22:01 |
miner49er_ | hello | 23:35 |
gnarface | hello miner49er_, if you have questions just ask | 23:42 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!