bgstack15 | yes, my Latitude D630 does have Ethernet. I was just going to try the wireless operations with connman. And it wouldn't surprise me if the hardware is actually broadcom. I certainly remember having to use b43-installer something something in the past, but I can't remember for which hardware. | 00:24 |
---|---|---|
gnarface | if ejr comes back, someone make sure they've tried something more recent than the stock ascii kernel on it | 00:34 |
gnarface | (it would probably also help to clarify for them that the problem is almost certainly solely linked to a kernel major version difference) | 00:35 |
* systemdlete2 reads the beowulf release notes and thinks, hmmmm... | 02:11 | |
systemdlete2 | eudev and elogin: "e" prefix is for... what, emancipated? | 02:12 |
systemdlete2 | but | 02:12 |
* systemdlete2 has not installed official beowulf anywhere yet... | 02:12 | |
gnarface | i thought it was "e" for European but i don't know | 02:15 |
golinux | Ask gentoo. We got it from the, | 02:15 |
golinux | them | 02:15 |
* systemdlete2 likes "emancipated" | 02:16 | |
golinux | E also comes after D | 02:17 |
golinux | so alphabetically trumps D. OT here though | 02:17 |
systemdlete2 | as in, trumps Donald Trump? | 02:20 |
yeti | extracted? | 02:21 |
yeti | they call elogind extreacted from systemd | 02:22 |
yeti | maybe eudev has the same e as elogind | 02:22 |
phogg | eudev came first, so if anything it's the other way around | 02:27 |
yeti | havent touched gentoo since its 1.08 (or so days) | 02:32 |
yeti | 2003? | 02:32 |
systemdlete2 | My attempt at gentoo from 2003 is still compiling | 02:36 |
systemdlete2 | (I chose "everything") | 02:37 |
syco- | u run it on a x386 ? :p | 02:37 |
systemdlete2 | yes, but backwards, just for extra intrigue | 02:37 |
yeti | 2003 was more like K6-II+@500 | 02:38 |
systemdlete2 | does beowulf have any better support for wireless than ascii? I am finding that my asus laptop (cherry trail) loses wireless after a while... sometimes. What I mean is that it is very intermittent. | 02:41 |
systemdlete2 | laptop is on ascii atm. Thinking of upgrading, but won't bother for now if beowulf doesn't bring more relief. | 02:41 |
golinux | systemdlete2: Thank you for getting back OT | 02:48 |
systemdlete2 | so, inquiring about wireless support in Beowulf is not a support question? | 02:49 |
systemdlete2 | Sorry if so. | 02:49 |
man_in_shack | nah you're fine systemdlete2 | 02:52 |
golinux | Re-read what I wrote | 02:52 |
man_in_shack | or rewrite what he read | 02:52 |
golinux | OT can also mean On Topic | 02:52 |
systemdlete2 | Oh. | 02:52 |
golinux | Duh | 02:52 |
systemdlete2 | So now OT can be used to confuse EVERYONE. | 02:52 |
golinux | :) | 02:52 |
systemdlete2 | Good to know. | 02:53 |
man_in_shack | there's the problem with using the same acronym for two opposites | 02:53 |
man_in_shack | like "literally" being used as "figuratively" | 02:53 |
golinux | And this discussion is Off Topic here | 02:53 |
systemdlete2 | man_in_shack: YES! | 02:53 |
man_in_shack | in other grammatical news ... | 02:54 |
man_in_shack | (: | 02:54 |
bgstack15 | i am booting into that latest named iso file. | 03:38 |
golinux | BOOM! | 03:44 |
bgstack15 | so I had to enable the non-free repo and install firmware-iwlwifi. | 03:47 |
bgstack15 | I can see the iwl4965 module is loaded | 03:47 |
bgstack15 | but still no "iwconfig" useful output | 03:47 |
bgstack15 | is there something i need to do, without rebooting, because after all this is the live boot? | 03:47 |
gnarface | hmm. did you add the firmware package after the iwl4965 module was already loaded? if so, i'd try unloading then reloading it | 03:48 |
golinux | Isn't there a command to start it? | 03:48 |
syco- | in /proc/net/dev ? | 03:48 |
gnarface | in a fair world all you'd have to do is "/etc/init.d/networking restart" but in practice you have to have carefully setup the network first | 03:49 |
gnarface | bgstack15: i recall needing both ifconfig and iwconfig | 03:50 |
bgstack15 | i think the firmware package provided my iwl4965 module. but sure, i'll reload it | 03:50 |
gnarface | bgstack15: iwconfig doesn't completely reimplement all the features of ifconfig for wireless devices, it just suppliments a few missing ones | 03:50 |
gnarface | bgstack15: but you should be able to just ifup the wireless device if your network config is correct | 03:51 |
bgstack15 | thank you! the modprobe -r iwl4965 and modprobe iwl4965 did it! | 03:51 |
gnarface | whew | 03:51 |
gnarface | i'm relieved | 03:51 |
bgstack15 | yeah, i'm doing all this in a live setup, to test fsmithred's iso with connman. I'm one of those die-hard wicd guys, but I hear it's gone now... | 03:51 |
bgstack15 | RIP | 03:51 |
golinux | It could be resurrected | 03:53 |
bgstack15 | So yeah, my quick examination of connman is that it looks just like Network Manager. Which means it's familiar, a little ...too... familiar! | 03:53 |
bgstack15 | But it will do! | 03:53 |
bgstack15 | But then I have extremely simple use cases, so my opinion has a weight of 0.01 units. | 03:55 |
bgstack15 | I'd do /etc/network/interfaces but my users would object... | 03:55 |
bgstack15 | but if wicd gets resurrected, I'd probably just stick with that, because who wants to re-tool workflows, amirite? | 03:56 |
bgstack15 | Puddletag is another python2(+qt4) project getting resurrected currently. https://github.com/sandrotosi/puddletag | 03:57 |
fsmithred | bgstack15, I had to remove and reload the module to get it to work | 03:59 |
fsmithred | and yeah, wicd is in experimental, so it might migrate down in time for chimaera | 04:00 |
fsmithred | connman really doesn't look much like network-manager | 04:00 |
fsmithred | connman is a simpler interface | 04:01 |
nemo | fsmithred: hmmm | 04:15 |
nemo | fsmithred: I decided to look into that forum thread | 04:15 |
nemo | since it's a friday night and the kids are in bed | 04:15 |
nemo | fsmithred: I have to say, rolfie's response falls far short of a clear guide for someone like me | 04:16 |
nemo | updating to 5.5 kernel is easy | 04:16 |
nemo | it's just what else I have to do that's less than clear | 04:16 |
nemo | oooh wait. some commands in http://forums.debian.net/viewtopic.php?p=721727#p721727 | 04:16 |
nemo | ok. I'm going to try to uninstall the amd stuff and try that thread. | 04:17 |
Digit | anyone using min browser? i tried install it manually, but hit errors i'm not competent enough to resolve. wondering, maybe it's in a nother repo? maybe experimental? maybe packagers just havent heard of it yet? | 04:20 |
nemo | wassat? | 04:20 |
Digit | https://minbrowser.org/ | 04:20 |
Digit | s/a nother/another/ | 04:21 |
nemo | fsmithred: so, I'm on 5.5.0-0.bpo.2-amd64 from backports and I copied over the firmware Rolfie recommended... | 04:40 |
nemo | fsmithred: aaaaaand no graphics acceleration | 04:41 |
fsmithred | I don't know what to say. I didn't read what you wrote or what rolfie wrote. | 04:42 |
nemo | ah | 04:42 |
nemo | fsmithred: I was reading response to thing you proxied | 04:42 |
nemo | he'd linked to some threads | 04:42 |
fsmithred | yeah | 04:43 |
nemo | after clicking on it, it seemed to involve 5.5 kernel from backports, cloning a repo for linux firmware and copying over the amd ones, adding a "4" to grub commandline (dunno what that does) and updating everything | 04:43 |
nemo | dmesg reports drm stuff | 04:43 |
fsmithred | 4 boots to runlevel 4 | 04:43 |
fsmithred | which is the same as 2 unless you do something to change it | 04:43 |
nemo | fsmithred: well that's the confusing part for me since it seemed debian made no distinction between runlevels | 04:43 |
fsmithred | was I on that thread? | 04:44 |
nemo | um... the one rolfie linked to? probably not. was just mentioning results since you'd reposted that thing in my name | 04:44 |
nemo | anyway trying to see if anything obvious I missed jumps out. otherwise back to amdgpu-pro I guess | 04:45 |
nemo | [ 132.677] (II) UnloadModule: "amdgpu" | 04:45 |
nemo | huh... what is Xorg doing... | 04:45 |
nemo | oh. oups. still on 1.19? thought I updated that 😃 | 04:45 |
nemo | lemme try that | 04:45 |
nemo | that could certainly do it | 04:45 |
gnarface | you might need to specify amdgpu in your xorg.conf to keep auto detect from forcing a generic driver on you | 04:46 |
gnarface | it seems to gravitate heavily to the modesetting driver even when it isn't the best one | 04:47 |
gnarface | but then once modesetting takes over nothing else can unseat it | 04:47 |
nemo | ok. lemme check the conf | 04:48 |
nemo | (1.20 didn't help btw) | 04:48 |
gnarface | in the conf you may have to specify the PCI ID for the card | 04:49 |
gnarface | though that's rare i've seen it before | 04:49 |
nemo | gnarface: stuff like that is kinda fun. us attributing mysterious impulses to software. it's FOSS, but there's a limit to how much RTFS we can do 😃 | 04:49 |
fsmithred | sleep time here. good luck. | 04:49 |
nemo | g'nite | 04:50 |
nemo | gnarface: /usr/share/X11/xorg.conf.d/10-amdgpu.conf seems to specify an amdgpu | 04:50 |
nemo | gnarface: not too familiar with how debian organises config | 04:50 |
gnarface | nemo: well if that's present then i would think it should be enough | 04:51 |
nemo | /usr/share/X11/xorg.conf.d# cat 10-amdgpu.conf | paste -sd '' | 04:51 |
nemo | Section "OutputClass"Identifier "AMDgpu"MatchDriver "amdgpu"Driver "amdgpu"EndSection | 04:51 |
gnarface | nemo: unfortunately i don't have any recent amd video cards here to check it with, my comparative experience is mostly with nouveau on nvidia hardware | 04:51 |
nemo | gnarface: well... my fallback is simply 1.19 Xorg, ascii xorg-server pin | 04:51 |
nemo | and the amdgpu-pro package | 04:51 |
nemo | gnarface: I was just trying to give the more foss-y variant an honest effort | 04:52 |
devuan | hello all | 04:52 |
nemo | gnarface: but the other one was definitely working fine yesterday with Witcher 2 on high everything graphics | 04:52 |
gnarface | nemo: i was under the impression that either way you need amdgpu-pro unless you want only 2d graphics and no compositing or video accel | 04:52 |
devuan | I am trying to setup alsa | 04:52 |
nemo | OH REALLY | 04:52 |
nemo | gnarface: grrrr darn debian forum folks | 04:52 |
nemo | gnarface: so context here | 04:52 |
gnarface | nemo: again, it's hearsay to me already | 04:53 |
nemo | gnarface: I post https://m8y.org/tmp/amdgpu.html | 04:53 |
gnarface | nemo: i just heard that from someone else | 04:53 |
nemo | gnarface: fsmithred here reposts it to forum in case anyone else is ddging for keywords | 04:53 |
devuan | what packages do I need to setup alsa sound | 04:53 |
nemo | gnarface: aaaand rolfie says I totally don't need that stuff | 04:53 |
devuan | on the console I tested with: | 04:53 |
nemo | gnarface: but you know. I should have been suspicious - I've had a lot of trouble with games and the nouveau driver. witcher for example did not work with it at all | 04:53 |
devuan | speaker-test -c2 -twav | 04:54 |
nemo | gnarface: bah. time to restore everything again | 04:54 |
devuan | and I hear, front left...front right | 04:54 |
devuan | but on desktop, no sound :( | 04:54 |
nemo | devuan: are you using pulseaudio? | 04:54 |
gnarface | devuan: alsamixer is what you want | 04:54 |
gnarface | devuan: also, pavucontrol if you have pulseaudio (which is likely) | 04:54 |
nemo | (that's an honest question since people have it installed by default and alsa can use it) | 04:55 |
nemo | gnarface: indeedy. had to play with profiles in pavucontrol to get pulseaudio to play on my speakers and TV hdmi. basically try them one by one until one worked ☺ | 04:55 |
nemo | gnarface: also there's the devuan config issue that was still unfixed in the beowulf rc2 I installed | 04:55 |
devuan | my card is "sun4i-hdmi" | 04:55 |
devuan | when I run alsamixer | 04:56 |
devuan | It selects that card but sayd that.. | 04:56 |
nemo | do you have pulseaudio installed | 04:56 |
devuan | "This sound device does not have any controls" | 04:56 |
devuan | :( | 04:56 |
nemo | gnarface: you know though, even if I'd managed to get the FOSS one to kinda work, I *was* looking forward to trying out opencl and I think I really would've needed amdgpu-pro to play with that | 04:57 |
nemo | gnarface: ah.. here's the answer. | 04:58 |
nemo | [ 601.906] (--) AMDGPU(0): Chipset: "Unknown AMD Radeon GPU" (ChipID = 0x731f) | 04:58 |
nemo | hmmmm | 04:58 |
nemo | maybe I can get a newer xorg driver | 04:58 |
devuan | Advanced Linux Sound Architecture Driver Version k5.7.0. | 04:59 |
nemo | devuan: do you have pulseaudio installed? | 05:00 |
devuan | nemo, I believe not | 05:00 |
nemo | please check | 05:00 |
nemo | xserver-xorg-video-amdgpu is dated 201902 - so definitely would not know about this chip... | 05:01 |
devuan | I only have this: | 05:01 |
nemo | aaaand nothing newer in backports | 05:01 |
devuan | ii libpulse0:armhf 12.2-4+deb10u1 armhf PulseAudio client libraries | 05:01 |
devuan | I am in OpenBox | 05:02 |
devuan | I don't even know what is the default way to control sound in OpenBox | 05:02 |
devuan | its a experiment that I am running :) | 05:02 |
nemo | devuan: aight. I'm regrettably out of practice with alsa personally. sorry. if gnarface has no suggestions, maybe try their channel. | 05:02 |
nemo | devuan: I *will* say be realllly careful on arm playing around with linux sound settings in alsamixer as I learned once | 05:03 |
nemo | I accidentally set my chromebook on fire | 05:03 |
nemo | literally | 05:03 |
nemo | sorry. I should say my SO's chromebook. | 05:03 |
nemo | devuan: I just wanted to make sure you didn't have 2 sound systems fighting for lock on the card | 05:04 |
devuan | cards I only have one, thanks god.. | 05:04 |
devuan | but the controls are not set.. | 05:04 |
devuan | I am on Olinuxino Lime2 | 05:05 |
devuan | with the experimental Audio kernel driver | 05:05 |
devuan | but I listen the woman, with: | 05:06 |
devuan | speaker-test -c2 -twav | 05:06 |
devuan | loud and clear | 05:06 |
devuan | well 16 bits :D | 05:06 |
devuan | its not yet 24 bits but.. | 05:07 |
devuan | does any one knows what is the alsa irc chanel? | 05:07 |
devuan | I have a very strange pcm device.. | 05:09 |
devuan | 00-00: SUN4I-HDMI PCM 1c16000.hdmi-0 : SUN4I-HDMI PCM 1c16000.hdmi-0 : playbac| | 05:09 |
devuan | +âââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââ | 05:09 |
nemo | heh | 05:12 |
man_in_shack | gotta agree with that last line | 05:12 |
man_in_shack | aaaaaaaaaaaaaAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA | 05:12 |
nemo | gnarface , fsmithred - https://m8y.org/tmp/amdgpu.html I updated 😃 | 05:12 |
nemo | (yes, I know he's sleeping) | 05:12 |
devuan | It should mean the controls are messedup in some way | 05:12 |
man_in_shack | devuan: can you unplug it, dust the port, and plug it back in? | 05:13 |
man_in_shack | cos that looks like the string isn't properly terminated, which indicates a signal error | 05:14 |
devuan | https://paste2.org/bE5MnzNh | 05:16 |
devuan | yeah. I unpluged it, and plugged it back | 05:16 |
devuan | now I am listem noise | 05:16 |
devuan | without touching anything else :D | 05:16 |
devuan | heheh | 05:16 |
nemo | moar cleaning | 05:16 |
man_in_shack | devuan: so the name is more "normal" now? | 05:18 |
devuan | I played again with 'speaker-test -c2 -twav' | 05:19 |
devuan | and the noise dissapeared and the woman spoken :D | 05:19 |
devuan | a normal behaviour | 05:19 |
man_in_shack | good news | 05:19 |
man_in_shack | next time, don't break it | 05:19 |
devuan | but the chanels are at levels 0:0 | 05:19 |
devuan | with would mean 'mute' or so | 05:19 |
man_in_shack | in theory :P | 05:20 |
devuan | 00-00: SUN4I-HDMI PCM 1c16000.hdmi-0 : SUN4I-HDMI PCM 1c16000.hdmi-0 : playback 1 | 05:20 |
man_in_shack | looks better, yes | 05:21 |
devuan | I don't know how to set the sound in OpenBox WM | 05:22 |
devuan | :S | 05:22 |
devuan | https://paste2.org/aLA8ybev | 05:27 |
devuan | I believe that sound is muted.. | 05:27 |
man_in_shack | have you tried muting it? | 05:28 |
devuan | how can I do that? | 06:03 |
devuan | in the console? | 06:03 |
frabbit | xdm doesnt show up after boot | 07:14 |
frabbit | nothing special in the logs | 07:14 |
frabbit | when i sotp it and start it again it does show up | 07:15 |
frabbit | *stop | 07:15 |
frabbit | but thats no solve on taht machine | 07:15 |
frabbit | *not a solution | 07:15 |
gnarface | frabbit: nothing in the Xorg log either? | 07:28 |
gnarface | frabbit: you don't have 2 displays connected, do you? | 07:28 |
frabbit | gnarface: i have | 08:02 |
frabbit | but the vga gives no signal on tty7 and the laptop screen stays black with frozen undertline cursor in the upper left corner | 08:03 |
frabbit | *underline = _ | 08:03 |
gnarface | frabbit: hmmm, not sure about that. i did once have a issue with xdm showing up on the display i didn't expect when it was off | 08:03 |
gnarface | frabbit: but in that instance, i wouldn't expect a cursor on the other screen | 08:04 |
gnarface | just a plain black screen | 08:04 |
frabbit | will check the xorg log again but i couldnt find anyzhing in first place. i cleared them and rebooted | 08:04 |
frabbit | brb | 08:04 |
frabbit | xorg log is empty and in xdm log nothing special (version etc.) | 08:12 |
frabbit | when i restart xdm the same will be posted to xdm.log | 08:13 |
aitor | good morning | 08:13 |
frabbit | i know that on boot it shows that xdm is started | 08:13 |
frabbit | aitor: good morning =) | 08:14 |
Guest35977 | frabbit: i replaced xdm in gnuinos by wdm | 08:14 |
* frabbit checks out wdm | 08:14 | |
Guest35977 | anybody knows how to install internet explorer under beowulf? I tried it yesterday, but it was a fiasco | 08:14 |
frabbit | Guest35977: wdmis inactive sinde 2005... o_0 | 08:15 |
frabbit | *wdm is | 08:15 |
frabbit | *since | 08:15 |
frabbit | ie... | 08:15 |
Guest35977 | the most used web browser in the world | 08:15 |
frabbit | interent explorer... ok i do better if i dont follow your software suggesstions xD | 08:16 |
Guest35977 | the most used... for downloading other web browsers | 08:16 |
gnarface | Guest35977: i'd guess you would probably need wine-staging from winehq.org and maybe some add-on winetricks (check the winehq appdb for hints) but if you have an actual copy of windows it might just be easier to install in a qemu guest | 08:16 |
frabbit | yea and google is the most used search engine, facebook the most used "social" medie platform etc. and they all suck! =P | 08:16 |
Guest35977 | frabbit: wasn't wdm supposed to be a replacement for xdm? At least, the code seems to be more updated | 08:18 |
frabbit | with apt u get taht wdm is oldstable now | 08:19 |
frabbit | *that | 08:19 |
frabbit | and xdm too xD | 08:19 |
frabbit | any small dm that is still maintained? (except slim) | 08:20 |
Guest35977 | synaptic says: an xdm replament with a WindowMaker look (this is its short description) | 08:20 |
frabbit | dont use synaptic it can cause problems... | 08:21 |
frabbit | just use apt | 08:21 |
gnarface | frabbit: make sure you don't have consolekit and elogind both installed together | 08:21 |
frabbit | gnarface: for getting xdm to wotk otb? | 08:21 |
gnarface | frabbit: just check to be sure, or nothing will work right you know | 08:22 |
gnarface | frabbit: could be a driver issue for all i know but you don't have anything in the logs... so it suggests a configuration issue of some sort | 08:22 |
gnarface | frabbit: something could have gone wrong in the updates | 08:22 |
Guest35977 | i wonder why apt-cache overlooks long descriptions... is it an amprolla issue? | 08:22 |
frabbit | ok wait a sec... (need always to check such things on that beginer system, cause all the programs brings tons of dependencies with them...) | 08:23 |
gnarface | frabbit: it's really easy to just run dpkg -l|grep consolekit | 08:23 |
frabbit | gnarface: no package that contains the phrase consolekit installed there | 08:24 |
frabbit | what? | 08:24 |
gnarface | frabbit: dpkg -l |grep login | 08:25 |
frabbit | gnarface: i know how to look for packages... | 08:25 |
frabbit | yeah i know | 08:25 |
frabbit | o_0 | 08:25 |
gnarface | frabbit: the issue is that i'm trying to help you debug it, and you'd rather argue with me about whether you should have to or not | 08:25 |
gnarface | frabbit: it would take less typing to just answer my questions instead | 08:25 |
frabbit | no consolekit but elogind as i said | 08:25 |
gnarface | ok, pretty sure from memory that xdm needs the other one | 08:26 |
gnarface | but i'll have to boot a machine to check | 08:26 |
frabbit | consolekit? | 08:26 |
frabbit | but now elogind lets u run startx without set that bit thing | 08:26 |
frabbit | man... | 08:26 |
Guest35977 | neither xdm nor wdm allows to root running X from the command line (i'm not talking about gksu or pkexec, but from the command line), and root needs to copy the .Xauthority from the current X's ownership | 08:27 |
gnarface | frabbit: wait startx.... i thought you were trying to get xdm working? | 08:27 |
gnarface | frabbit: using xdm would preclude using startx... | 08:27 |
frabbit | gnarface: yes | 08:27 |
frabbit | i want xdm to start after boot and thats not happening here | 08:27 |
frabbit | but i can run startx as non root user without setting that bit thing before | 08:28 |
frabbit | setuid or what it is called.. | 08:28 |
gnarface | frabbit: confirmed, xdm it is using consolekit here | 08:28 |
frabbit | gnarface: ok thxä | 08:28 |
frabbit | so good bye xdm xp | 08:28 |
gnarface | frabbit: also, add yourself to the "video" group | 08:28 |
frabbit | gnarface: ... | 08:28 |
Guest35977 | i'm building an imgage without polkit | dbus | consolekit | elogind | 08:28 |
frabbit | I.am. | 08:28 |
Guest35977 | *image | 08:28 |
Guest35977 | and replacing sysvinit by runit | 08:29 |
gnarface | frabbit: you can have consolekit without policykit, just fyi (just uninstall policykit afterwards) | 08:29 |
frabbit | xdm doesnt work without consolekit, slim cant work with wicd-gtk.... | 08:29 |
Guest35977 | it works | 08:29 |
frabbit | gnarface: fyi? | 08:30 |
Guest35977 | xdm without consolekit, i mean | 08:30 |
frabbit | Guest35977: ok.. | 08:30 |
frabbit | do i have to install ie first? | 08:30 |
gnarface | frabbit: "for your information" | 08:30 |
Guest35977 | and slim has nothing to do with wicd | 08:30 |
frabbit | xD ;) | 08:30 |
Guest35977 | but slim depends on dbus | 08:30 |
frabbit | Guest35977: slim breaks when wicd-gtk is installed here | 08:30 |
Guest35977 | both depend on dbus | 08:30 |
frabbit | often not always ive tested this dozen times... | 08:31 |
Guest35977 | wicd has a lot of bugs | 08:31 |
frabbit | Guest35977: maybe i dont use these guishes things for years | 08:32 |
frabbit | last time i had wicd-gtk installed on my own system was 5 or 6 years ago or something... | 08:32 |
frabbit | gnarface: install cnsolekit, uninstall policykit | 08:33 |
frabbit | wasnt that policykit needed for that mount thing in gui fm? | 08:33 |
Guest35977 | i'm going to take a coffee, see you later | 08:33 |
frabbit | cya | 08:33 |
Guest35977 | hope to find a wizard for ie :) | 08:34 |
gnarface | frabbit: heh, oh yea, probably, nevermind | 08:34 |
frabbit | gnarface: lol xD | 08:34 |
gnarface | frabbit: at a certain point, you can't get that type of stuff without making sacrifices | 08:34 |
gnarface | frabbit: i prefer the extra typing method, myself. there's less moving parts | 08:34 |
gnarface | frabbit: you could just login at a text prompt and type startx every time | 08:35 |
frabbit | gnarface: and thats only one reason why i try to avoid gui stuff... | 08:35 |
frabbit | gnarface: i know | 08:35 |
frabbit | i do this for years | 08:35 |
frabbit | but thats a beginner system | 08:35 |
gnarface | they don't need it either | 08:36 |
gnarface | just add the call to startx in their bash login scripts | 08:36 |
frabbit | and.. well that old lady... run some commands as root yesterday accidentaly.. xD | 08:36 |
gnarface | don't even give them the root password... | 08:36 |
frabbit | nah thats no problem she can update the system | 08:36 |
gnarface | i'm not seeing a good argument against consolekit in your use case here | 08:37 |
frabbit | she can open a terminal in xfce run su type root passphrase and then run a script i wrote for her to update the system | 08:37 |
gnarface | also, that's suicidal if she can't be trusted to remember "startx" | 08:37 |
frabbit | gnarface: hm? u said i need to deinstall policykit then... | 08:37 |
frabbit | gnarface: she did this first time yesterday | 08:38 |
gnarface | just forget about that, you don't need to deinstall anything | 08:38 |
frabbit | startx thing | 08:38 |
frabbit | ... | 08:38 |
frabbit | ok u confusing me | 08:38 |
gnarface | the only thing you need to deinstall is elogind | 08:38 |
gnarface | because it's probably what is breaking xdm for you right now | 08:38 |
frabbit | first i should make clear if both are installed, then i should install consolekit and remove policykit now i should install consolekit besides policykit... | 08:39 |
frabbit | oh | 08:39 |
frabbit | and that elogind is only needed for run startx as non root? | 08:39 |
gnarface | uh... no, startx doesn't require either | 08:39 |
frabbit | huh? | 08:40 |
frabbit | wait | 08:40 |
gnarface | consolekit and elogind compete | 08:40 |
gnarface | effectively | 08:40 |
gnarface | you need to make sure only one or the other is installed, and it needs to be the right one for the gui. xdm needs consolekit. maybe slim needs elogind, i'm not sure. | 08:40 |
frabbit | ok ill check | 08:41 |
gnarface | startx doesn't require any | 08:41 |
gnarface | some drivers might require the suid root wrapper | 08:41 |
frabbit | *errm* apt-get autoremove --purge elogind wants to temove alot of libs and blinken, kio and kollision.... | 08:43 |
frabbit | *remove | 08:43 |
gnarface | that's expected. your only choice is to let it | 08:44 |
frabbit | blinken and kollision are games | 08:44 |
frabbit | linux is broken xD | 08:44 |
* frabbit creates #linux-rant | 08:44 | |
gnarface | anything that is still working in the distro will allow you to reinstall it again after you install consolekit | 08:45 |
gnarface | if a game actually got left behind in the release it's probably because the maintainer abandoned it | 08:46 |
gnarface | or because you got it from another 3rd party repo to begin with | 08:46 |
frabbit | ganerr.. u mean i can reinstall the games after consolekit is installed? | 08:46 |
frabbit | no, no 3rd party stuff here except libdvdcss2 | 08:46 |
gnarface | if they were from the distro and they weren't obsoleted with this release, they should work still | 08:46 |
gnarface | i would assume | 08:46 |
gnarface | but i don't actually know anything about those games specifically | 08:47 |
frabbit | ok =( | 08:47 |
gnarface | if you just try to install consolekit instead of remove elogind, what does it say? | 08:47 |
gnarface | does it say it will install consolekit and uninstall elogind, or does it just blindly install consolekit? | 08:47 |
frabbit | kollision is one of her favorite games (good for older people to train concentration and reaction) i wont remove that one | 08:47 |
frabbit | reverse | 08:48 |
frabbit | or.. wait | 08:48 |
gnarface | does it keep the games ? | 08:48 |
frabbit | it will install consolekit when i purge elogind | 08:48 |
frabbit | no | 08:49 |
gnarface | hmmm | 08:49 |
frabbit | as i said | 08:49 |
frabbit | but i didnt execute that install | 08:49 |
gnarface | i see kollision in beowulf | 08:50 |
gnarface | you said this is beowulf, right? | 08:50 |
gnarface | oh | 08:51 |
frabbit | yes | 08:51 |
gnarface | kollision, sorry i just got it | 08:52 |
gnarface | these are kde games | 08:52 |
gnarface | kde requires elogind | 08:52 |
frabbit | yes | 08:52 |
frabbit | ah i see | 08:52 |
gnarface | you're stuck with elogind and you're stuck with kde | 08:52 |
gnarface | sorry | 08:52 |
frabbit | shitty de here again xD | 08:52 |
frabbit | gnarface: not ur fault | 08:52 |
frabbit | devs of the games fault | 08:52 |
gnarface | that's the case for pretty much most stuff following this kname convention | 08:52 |
gnarface | not just the games | 08:53 |
gnarface | just fyi ;) | 08:53 |
frabbit | probably... | 08:53 |
malade_mental | Glad to hear about the newsletter I received today! Happy third release to all the Devuan contributors! | 11:38 |
malade_mental | gonna install this quick quick quick! | 11:41 |
crhylove | Hey all! How do I get Compiz going in Devuan? | 12:21 |
crhylove | Everything else seems to be working, although this is a very old version of OBS. LOL | 12:21 |
crhylove | But it works pretty well, so.... | 12:21 |
crhylove | Man is Devuan fast compared to Mint on the same machine.... | 12:24 |
crhylove | HD crashed, so.... Here I am! | 12:24 |
gnu_srs | Hi, whit polkit etc packages do I need to get the shutdown/reboot back with beowulf/mate. It worked with ASCII :( | 12:29 |
frabbit | gnu_srs: it works with elogind i think | 12:35 |
frabbit | but im not sure i forgot ^_^ | 12:35 |
gnu_srs | I prefer to use the consolekit version. | 12:36 |
fsmithred | gnu_srs, usually policykit-1-gnome is the missing piece | 12:39 |
crhylove | hmmmmmm Got compiz MOSTLY working, but I can't switch workspaces. Or enable 3d cube. LOL | 12:54 |
crhylove | OK, Got Desktop Wall to work and am able to switch workspaces.... | 13:01 |
crhylove | Figured it out! I needed apt install compiz-plugins | 13:03 |
gnu_srs | fsmithred: Thanks but no improvement. I don't have any menu entries for shutdown/reboot?? | 13:13 |
crhylove | Got almost everything working, except transparent desktop on rotate and transparent cube caps. | 13:16 |
crhylove | Going for a reboot.... Maybe that will fix it? | 13:22 |
fsmithred | gnu_srs, I think maybe there is no fix for that. In Release Notes it says that XFCE and Cinnamon are the only desktops that work with consolekit. | 13:29 |
crhylove | Well, got it all working mostly! Any idea how to get a newer version of OBS? | 13:38 |
crhylove | Compile from source, or.... ? | 13:39 |
fsmithred | what is OBS? | 13:47 |
gnarface | screen recorder | 13:48 |
gnarface | it doesn't do anything you can't do with ffmpeg, but it's popular because it is easy to set up | 13:48 |
climbingturtle | version 25 of obs-studio is possible to get from chimaera-repos | 13:49 |
ShorTie | better to compile it then try to pull it from chimera, imho | 14:02 |
gnarface | did anyone check for it in beowulf-backports? | 14:06 |
gnarface | the work may have already been done | 14:06 |
gnu_srs | fsmithred: I have a Debian box running mate with policykit where the System menu has a Shut Down entry! | 14:08 |
fsmithred | gnu_srs, I'm guessing your debian box does not use consolekit | 14:10 |
gnu_srs | policykit versions: 0.105-26, mate-* versions 1.20.2-1 (same as devuan) | 14:10 |
fsmithred | not a devuan version of policykit-1? | 14:11 |
fsmithred | 0.105-25+devuan8 | 14:12 |
fsmithred | running beowulf here | 14:12 |
fsmithred | gnu_srs, while I have your attention, I have a question about eudev. | 14:13 |
gnu_srs | You are probably right: On that box I have elogind 241.3-1+debian1, and no polkit backends at all?? | 14:13 |
gnu_srs | shoot | 14:13 |
fsmithred | between ascii and jessie, the start command changed from 'udevd -d' to start-something | 14:13 |
fsmithred | start-stop-daemon? | 14:13 |
climbingturtle | gnarface: I think I checked beowulf-backports :) | 14:13 |
fsmithred | can you give me any information about that change? | 14:14 |
fsmithred | my problem is related to booting live-isos - when booting from optical media, I have to stop and start eudev an extra time to get it to recognize all the hardware and load right modules | 14:14 |
gnu_srs | do you mean: between ascii and jessie->ascii and beowulf? | 14:16 |
fsmithred | version in ascii vs. version in beowulf | 14:16 |
gnu_srs | I run beowulf on a laptop: ps aux|grep udevd; /sbin/udevd udevd; eudev 3.2.7-6 | 14:19 |
fsmithred | if start-stop-daemon --start $NAME --user root --quiet \ | 14:19 |
fsmithred | ^^^ that's in beowulf | 14:20 |
fsmithred | in ascii: if udevd --daemon; then | 14:20 |
gnu_srs | Remember: I was not involved in eudev until after 3.2.7-6 | 14:20 |
fsmithred | ok, I was missing that detail | 14:20 |
fsmithred | so I need to talk to Andreas? | 14:21 |
gnu_srs | see the changelog | 14:21 |
gnu_srs | Mainly centurion and katolaZ | 14:22 |
fsmithred | yeah, it looks like you were the first to get 3.2.9. I probably need to look at upstream changelogs | 14:24 |
fsmithred | do you know anything at all about start-stop-daemon? | 14:25 |
fsmithred | oh, a man page! | 14:25 |
ShorTie | or the commits on the eudev github ?? | 14:28 |
gnu_srs | There is a bug in the eudev script in two places: mkdir -p $OMITDIR/$NAME; ln -sf $PIDFILE $OMITDIR/$NAME | 14:28 |
gnu_srs | Should be: mkdir -p $OMITDIR/; ln -sf $PIDFILE $OMITDIR/$NAME | 14:28 |
gnu_srs | We should probably fix this bug and create a 3.2.7-7 version before enabling 3.2.9-7~beowulf1 (that does not have that bug) | 14:30 |
fsmithred | oh, that's in the older version? | 14:31 |
gnu_srs | Yes, that's why elogind runs as /sbin/udevd udevd | 14:31 |
fsmithred | any idea what that bug would do? | 14:31 |
gnu_srs | Runs udevd twice?? | 14:31 |
fsmithred | hm | 14:31 |
fsmithred | I'm running eudev a second time to fix a problem | 14:32 |
fsmithred | in beowulf | 14:32 |
gnu_srs | Maybe you can fix that bug in your live ISo to check? | 14:32 |
gnu_srs | ISO** | 14:32 |
fsmithred | might be interesting to see if fixing the bug will cause the problem in ascii isos | 14:32 |
gnu_srs | fsmithred: Looking ate the bug above reveals that the $PIDFILE variable gets lost. | 14:35 |
fsmithred | oh, that bug is in the beowulf version. | 14:35 |
fsmithred | yeah, there's a leftover pidfile | 14:36 |
fsmithred | ok, I need to fix that in the beowulf version, right? | 14:36 |
gnu_srs | And PIDFILE="/run/udev.pid" | 14:36 |
fsmithred | looking at it now | 14:36 |
teknopaul | re: the lack of ext4, in expert mode, trying again on different hardware and its repeatable, this server has no wifi issue. Just Ext2, FAT16, FAT32 or swap are available | 15:26 |
karborsan | hello folks | 15:36 |
karborsan | could somone help me?? I would like to wildcard my localhost for development purposes. on MX Linux I used dnsmasq & it was working fine, however it is not working on Devuan 3.0 using the same configurations | 15:38 |
teknopaul | what do you mean by "wildcard my localhost" | 15:40 |
karborsan | I mean to be able to forward abc.localhost:8075 to 127.0.0.1 | 15:41 |
karborsan | the used the following comands | 15:43 |
karborsan | sudo apt-get install dnsmasqecho "address=/localhost/127.0.0.1" > /etc/dnsmasq.d/localhost.confecho "prepend domain-name-servers 127.0.0.1;" >> /etc/dhcp/dhclient.conf/etc/init.d/dnsmasq restart | 15:43 |
karborsan | sudo apt-get install dnsmasq | 15:43 |
karborsan | echo "address=/localhost/127.0.0.1" > /etc/dnsmasq.d/localhost.conf | 15:43 |
karborsan | echo "prepend domain-name-servers 127.0.0.1;" >> /etc/dhcp/dhclient.conf | 15:43 |
teknopaul | potentially you need an extra period echo 'address=/.localhost/127.0.0.1' | 15:45 |
karborsan | still not working | 15:48 |
teknopaul | whats in /etc/resolv.conf | 15:56 |
teknopaul | devuan does not install resolvconf | 15:56 |
teknopaul | nameserver 127.0.0.1 | 15:56 |
teknopaul | nameserver [real server ip] | 15:56 |
teknopaul | seemed to work for me | 15:56 |
karborsan | nameserver 1.1.1.1nameserver 1.0.0.1 | 15:59 |
karborsan | so should I change something | 15:59 |
teknopaul | Top trick: I added that to my network's dnsmasq server, and it seems to work across the network | 16:00 |
teknopaul | in /etc/resolv.conf what do you have now | 16:01 |
karborsan | nameserver 1.1.1.1 | 16:01 |
karborsan | nameserver 1.0.0.1 | 16:01 |
teknopaul | put nameserver 127.0.0.1 infromt | 16:01 |
crashoverride | try nameserver 0.0.0.0 | 16:01 |
crashoverride | I heard it's really fast. | 16:02 |
teknopaul | I suspect you have setup dnsmasq correctly but not told Linux to use it. | 16:02 |
teknopaul | Something other than 127.0.0.1 probably does not have your *.localhost trick setup | 16:03 |
karborsan | well, thank you guyz. it works fine now | 16:03 |
teknopaul | np, you might want to make sure DNSmasq is a caching DNS too: Linux is always going to hit 127.0.0.1 first | 16:04 |
teknopaul | thanks for the trick, this will save a bunch of junk in my /etc/hosts | 16:05 |
karborsan | thank you all | 16:06 |
teknopaul | deb cdrom: is left uncommented in /apt/etc/sources.list after fresh install please insert the disc labeled 'Devuan GNU/Linux 3.0 (beowulf) amd64 - server 20200526' | 16:51 |
teknopaul | its there commented and uncommented | 16:51 |
gnu_srs | Hi again. Seems like installing elogind (replacing consolekit) in Beowulf also removes slim and installs lightdm. slim+mate+elogind not possible? | 17:37 |
gnu_srs | On the other hand: installing libpam-elogind (end elogind) does _not_ remove slim?? | 17:39 |
specing | Will "fakeroot debian/rules binary" start rebuilding things if the source package changes? | 18:24 |
specing | I forgot to exclude .git from the source package, and now it's huge and taking up a lot of space | 18:24 |
specing | and the thing takes 3h to build | 18:25 |
gnu_srs | seems like installing libpam-elogind works with slim. Some dependency makes installation of elogind removing slim and installing lightdm?? | 18:48 |
fsmithred | gnu_srs, I've seen some weird things like that around the whole login/policykit ecosystem. | 18:55 |
gnu_srs | fsmithred: Any luck with eudev? | 19:05 |
fsmithred | no | 19:05 |
fsmithred | I changed it, made an iso, booted. No change. But then I noticed that I only changed it in the start case but not restart. | 19:05 |
fsmithred | playing with it again now | 19:06 |
fsmithred | Sat Jun 6 17:10:30 2020: [....] Setting up LVM Volume Groups...no db file to read /run/udev/data/b8:0: No such file or directory | 19:20 |
fsmithred | There are a couple of places where /run/udev (or its absence) bites me in the ass. | 19:21 |
fsmithred | That log line is from /var/log/boot | 19:22 |
GLAT-agent1 | Hello. I am selling GNU/Linux licenses. Contact me if you want one ($99 for single computer, $49 for each computer if purchasing for 25 or more machines). | 20:33 |
krypton | Hi, has someone here migrated to beowulf from ascii? | 20:46 |
humpelstilzchen[ | some time ago, yes | 20:47 |
krypton | I've lost the automatic module loading by udev.... | 20:47 |
krypton | After booting, not all the necessary modules are loaded, but when I stop/start ude, they get loaded | 20:48 |
humpelstilzchen[ | are the links in rc?.d intact? | 20:49 |
krypton | Yes, looks like it | 20:51 |
krypton | Hmm, but there is no udev-start link in rc5.d | 20:51 |
krypton | There is a start link in rcS.d, however. | 20:52 |
humpelstilzchen[ | seems to be normal | 20:53 |
humpelstilzchen[ | I also have only rcS.d/S02eudev -> ../init.d/eudev | 20:53 |
fsmithred | krypton, I have the same problem but ONLY when booting live-DVD | 20:55 |
fsmithred | and not when booting toram | 20:55 |
krypton | Well, I'm booting from disk | 20:55 |
fsmithred | first I've heard of it happening there | 20:55 |
krypton | From my previous ascii install... | 20:55 |
fsmithred | rc5.d shouldn't really matter unless you're using runlevel 5 | 20:56 |
fsmithred | default is 2 | 20:56 |
fsmithred | my workaround for the live isos was to add a line to /etc/rc.local | 20:57 |
fsmithred | /etc/init.d/eudev stop && /etc/init.d/eudev start | 20:57 |
krypton | Well, in rc2.d there is no udev link, but humpelstilzchen[ does not need one there... | 20:57 |
fsmithred | don't have that here, either | 20:58 |
fsmithred | yeah, it starts in rcS and that's all | 20:59 |
humpelstilzchen[ | makes sense to start in rcS | 20:59 |
fsmithred | one bug was found today, you could try this: | 21:00 |
fsmithred | <gnu_srs> There is a bug in the eudev script in two places: mkdir -p $OMITDIR/$NAME; ln -sf $PIDFILE $OMITDIR/$NAME | 21:00 |
fsmithred | <gnu_srs> Should be: mkdir -p $OMITDIR/; ln -sf $PIDFILE $OMITDIR/$NAME | 21:00 |
fsmithred | in /etc/init.d/eudev | 21:00 |
fsmithred | it's in the start and restart cases | 21:00 |
krypton | OK, will try that out, I've added a link in rc2 and tried a reboot | 21:01 |
fsmithred | oh, ok. Let me know if that works. | 21:01 |
krypton | nope, didn't | 21:02 |
krypton | I'll apply the fix now | 21:02 |
krypton | I've gonna add /sbin to root's path real soon now... | 21:04 |
fsmithred | krypton, echo 'ALWAYS_SET_PATH yes' >> /etc/default/su | 21:05 |
fsmithred | that will make it act like it used to | 21:05 |
krypton | I don't use su, plain old root login | 21:06 |
fsmithred | then you should have the sbins in your path | 21:06 |
fsmithred | already | 21:06 |
_abc_ | Hello. Is there an update command for going from ascii straight to beowulf in one command? | 21:07 |
krypton | Right you are, fsmithred, my fault | 21:07 |
_abc_ | Keeping $HOME mostly untouched I assume. | 21:07 |
fsmithred | _abc_, yeah, gimme a sec | 21:07 |
_abc_ | Hi fsmithred . | 21:08 |
fsmithred | hi | 21:08 |
fsmithred | it's actually more than one command | 21:08 |
_abc_ | Up to 10 is ok :) | 21:08 |
fsmithred | https://devuan.org/os/documentation/dev1fanboy/en/upgrade-to-beowulf | 21:08 |
_abc_ | thanks | 21:09 |
krypton | but still only a small number of modules loaded, even after the init script change... Perhaps something in the initramfs is missing? The list of the files in there is a lot shorter than the in the ascii initramfses. | 21:09 |
fsmithred | I'm looking at that now | 21:10 |
fsmithred | about to compare to ascii | 21:10 |
fsmithred | beowulf has all the same udev rules files as ascii plus one more. | 21:12 |
krypton | Hmm.. udevd is now in /usr/sbin in the initram. It was in /sbin before.. | 21:12 |
fsmithred | another thing that changes is in ascii the start command was 'udevd --daemon' | 21:13 |
fsmithred | and in beowulf it uses start-stop-daemon | 21:13 |
fsmithred | we tried changing it and it did fix the problem, but I don't know if it's a good idea to do that. i.e. comment out 20 lines or so and replace with one line. | 21:14 |
gl4t-ag3nt | I am selling GNU/Linux licenses. $99 for each computer if less than 25 computers, else $49. | 21:15 |
fsmithred | krypton, here's a discussion with some things we tried: https://dev1galaxy.org/viewtopic.php?id=3520&p=2 | 21:17 |
krypton | I think here's something fishy in my filesystem: ls -al /sbin/udev* /usr/sbin/ude* : ls: cannot access '/usr/sbin/ude*': No such file or directory lrwxrwxrwx 1 root root 12 Feb 9 2019 /sbin/udevadm -> /bin/udevadm -rwxr-xr-x 1 root root 405896 Feb 9 2019 /sbin/udevd | 21:18 |
krypton | Shouldn't udev be in /usr/sbin in the final install, too? | 21:19 |
fsmithred | not here | 21:19 |
fsmithred | I have /sbin/udevd | 21:20 |
fsmithred | and /sbin/udevadm is a symlink to /bin/udevadm | 21:20 |
krypton | OK, same as in my install, same timestamps? | 21:20 |
fsmithred | Feb 9 2019 | 21:21 |
fsmithred | yeah | 21:21 |
krypton | OK, and binary works, so that should be OK. Must be something with the scripts or the initramfs, I suppose. | 21:22 |
fsmithred | what is different about your installation? | 21:22 |
fsmithred | seems like a timing problem | 21:22 |
fsmithred | like it stops detecting hardware too soon | 21:22 |
krypton | Well, it's an upgrade from ascii. | 21:22 |
fsmithred | or it gets stupid and forgets what it needs to look for | 21:23 |
fsmithred | I think I've only booted upgrades in a VM | 21:23 |
fsmithred | not on hardware | 21:23 |
krypton | And the box is a quick 24 core Xenon | 21:23 |
krypton | So, timing will be different :-) | 21:23 |
krypton | (2 CPUs, HP Z600) | 21:24 |
_abc_ | nice box | 21:24 |
_abc_ | Is the kernel built with 24 cpu support? ... | 21:25 |
krypton | Yeah, was a cheap (~300 Euros) buy on ebay, used of course | 21:25 |
fsmithred | pretty sure it is | 21:25 |
krypton | All CPUs are there | 21:25 |
krypton | Well, perhaps the wait during udev start does not work? | 21:27 |
krypton | Hm, uses udevadm settle... | 21:28 |
ShorTie | that don't work all the time, imho | 21:29 |
ShorTie | an extra sleep is needed sometimes | 21:30 |
krypton | Do you have a suggestion for the time of the sleep, ShorTie? | 21:31 |
ShorTie | i use 4 | 21:31 |
krypton | OK, I'll use 10 to be sure for the next reboot :-) | 21:32 |
ShorTie | i'd stick it at the end of start | 21:33 |
krypton | I tried it in the if-section of the settle - that did not work | 21:35 |
krypton | Trying your suggestion now | 21:36 |
ShorTie | just above ;; | 21:36 |
ShorTie | so it's the last thing | 21:36 |
krypton | Yeah, maybe it runs into the timeout | 21:37 |
krypton | Still no go :-( | 21:37 |
krypton | lsmod | wc 35 125 1339 | 21:38 |
krypton | /etc/init.d/eudev stop && /etc/init.d/eudev start && lsmod | wc: 78 274 3293 | 21:39 |
krypton | But I could see from the bootup sequence that the script executed the sleep. | 21:40 |
fsmithred | we tried adding sleep in a couple places | 21:42 |
krypton | Hmmm... There's a warning in dmesg now: random: udevd: uninitialized urandom read (16 bytes read) | 21:42 |
krypton | Just after udevd[224]: starting version 3.2.7 | 21:43 |
krypton | urandom gets inititalised way later than udev starts (udev: 02, urandom: 11) | 21:48 |
krypton | I'll try pushing it before udev | 21:48 |
krypton | Eh, that looks strange: | 21:51 |
krypton | Run /init as init process | 21:51 |
krypton | udevd[224]: starting version 3.2.7 | 21:51 |
krypton | random: udevd: uninitialized urandom read (16 bytes read) | 21:51 |
krypton | random: udevd: uninitialized urandom read (16 bytes read) | 21:51 |
krypton | udevd[225]: starting eudev-3.2.7 | 21:51 |
krypton | in that order in the dmesg | 21:52 |
fsmithred | that's how mine starts | 21:52 |
fsmithred | krypton, did you initially install ascii with a desktop from the installer iso? | 21:56 |
fsmithred | or did you start minimal and build up? | 21:57 |
krypton | fsmithred: from a minimal | 21:57 |
fsmithred | I guess I should upgrade my laptop. So far, to test each change I have to make a new live-iso and then burn it to dvd. | 21:58 |
krypton | As long as you use rewritable DVDs ;-) | 21:59 |
fsmithred | nope | 21:59 |
krypton | upps | 21:59 |
fsmithred | don't have any rw | 21:59 |
krypton | Wait, shouldn't be /dev mounted before udev starts? | 22:00 |
fsmithred | maybe | 22:00 |
fsmithred | or does udev create /dev? There's no /dev in the initramfs | 22:00 |
fsmithred | early in the init script there are some tests for what udev needs | 22:04 |
fsmithred | and there's a comment about the time between mounting new /dev and running udevadm trigger. | 22:05 |
fsmithred | which I don't fully understand | 22:05 |
fsmithred | krypton, did you try this fix? https://dev1galaxy.org/viewtopic.php?pid=22172#p22172 | 22:09 |
krypton | no, but I'll try | 22:09 |
krypton | Hey, fsmithred, that did the trick! | 22:14 |
fsmithred | yeah, it would be nice to know if cutting out all that other code will cause trouble | 22:14 |
krypton | lsmod | wc: 78 274 3293 | 22:14 |
fsmithred | lsmod |grep snd | 22:15 |
krypton | lsmod | grep snd | 22:15 |
krypton | snd_hda_codec_realtek 122880 1 | 22:15 |
krypton | snd_hda_codec_generic 86016 1 snd_hda_codec_realtek | 22:15 |
krypton | snd_hda_codec_hdmi 57344 1 | 22:15 |
fsmithred | I get around 140 when everything loads correctly, around 80-90 when it screwed up | 22:15 |
fsmithred | ok | 22:15 |
krypton | snd_hda_intel 49152 0 | 22:15 |
krypton | snd_hda_codec 151552 4 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec_realtek | 22:15 |
fsmithred | stop | 22:15 |
krypton | snd_hda_core 94208 5 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek | 22:15 |
fsmithred | shit | 22:15 |
krypton | snd_hwdep 16384 1 snd_hda_codec | 22:15 |
krypton | snd_pcm 114688 4 snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hda_core | 22:15 |
krypton | snd_timer 36864 1 snd_pcm | 22:15 |
krypton | snd 94208 8 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hwdep,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek,snd_timer,snd_pcm | 22:16 |
krypton | soundcore 16384 1 snd | 22:16 |
fsmithred | you're still here? I thought for sure you would get bounced. | 22:16 |
krypton | No I'm here | 22:16 |
fsmithred | snd were among the missing in my tests. Also wireless. | 22:17 |
krypton | I was missing my radeon DRM | 22:17 |
krypton | But it's all there now, even coretemp, kvm, i7core_edac pcc_cpufreq, etc. | 22:18 |
krypton | Was all missing before | 22:18 |
fsmithred | yup | 22:18 |
fsmithred | I've been fighting with this for a couple weeks | 22:18 |
fsmithred | still don't really know what the problem is | 22:18 |
krypton | Booting is different now as well, as the fbdev is now used as expected. | 22:19 |
krypton | Well, I still have the urandom warnings in my dmesg as well. | 22:22 |
krypton | But later on udev gets started twice again: | 22:23 |
krypton | random: crng init done | 22:23 |
krypton | [ 7.373001] random: 3 urandom warning(s) missed due to ratelimiting | 22:23 |
krypton | [ 7.861093] udevd[617]: starting version 3.2.7 | 22:23 |
krypton | [ 8.052203] udevd[617]: specified group 'kvm' unknown | 22:23 |
krypton | [ 8.278318] udevd[618]: starting eudev-3.2.7 | 22:23 |
krypton | And after that all modules are inserted | 22:24 |
fsmithred | I have urandom warnings in ascii, too | 22:24 |
krypton | Yes, me too | 22:25 |
krypton | So, that's probably a red herring | 22:25 |
krypton | start-stop-daemon has been upgraded as well | 22:28 |
krypton | -rwxr-xr-x 1 root root 44200 Jun 3 2019 /sbin/start-stop-daemon | 22:29 |
krypton | on ascii it's smaller | 22:29 |
krypton | I think that start-stop-daemon is just wrong in that eudev script. udev is started before already and will not be started again by start-stop-daemon --start. | 22:35 |
krypton | I don't understand why udev is already started, though ;-) | 22:36 |
fsmithred | how do you know it's started? | 22:36 |
krypton | Because I have that line in dmesg: | 22:37 |
krypton | [ 2.195982] udevd[224]: starting version 3.2.7 | 22:37 |
krypton | And later: | 22:37 |
krypton | [ 7.861093] udevd[617]: starting version 3.2.7 | 22:37 |
krypton | [ 8.278318] udevd[618]: starting eudev-3.2.7 | 22:37 |
krypton | So it gets started 3 times. | 22:38 |
fsmithred | same here | 22:38 |
fsmithred | twice at around 3 seconds and once at 17 | 22:38 |
krypton | Only after the last of the starts all modules load up | 22:38 |
fsmithred | same here except in the live-dvd not all the modules load | 22:41 |
krypton | I just go back to the start-stop script to see if the timing is different | 22:42 |
fsmithred | [ 38.892669] udevd[765]: starting eudev-3.2.7 | 22:42 |
fsmithred | [ 44.132388] udevd[765]: cleanup idle workers | 22:42 |
fsmithred | [ 44.132452] udevd[765]: Validate module index | 22:42 |
fsmithred | [ 44.132554] udevd[1395]: Unload module index | 22:42 |
fsmithred | [ 44.132823] udevd[765]: worker [1395] exited | 22:43 |
fsmithred | [ 44.153090] udevd[765]: seq 2521 queued, 'add' 'module' | 22:43 |
fsmithred | [ 44.153317] udevd[765]: seq 2521 forked new worker [1569] | 22:43 |
fsmithred | [ 44.153478] udevd[1569]: seq 2521 running | 22:43 |
krypton | The first line here is: | 22:43 |
krypton | [ 2.192779] udevd[224]: starting version 3.2.7 | 22:43 |
fsmithred | [ 44.153529] udevd[1569]: no db file to read /run/udev/data/+module:sunrpc: No such file or directory | 22:43 |
krypton | Then the warnings about urandom and then: | 22:43 |
fsmithred | I'm running in debug mode | 22:44 |
krypton | [ 7.932107] udevd[619]: starting version 3.2.7 | 22:44 |
krypton | [ 8.092937] udevd[619]: specified group 'kvm' unknown | 22:44 |
fsmithred | set it in /etc/udev/udev.conf | 22:44 |
krypton | [ 8.308626] udevd[619]: starting eudev-3.2.7 | 22:44 |
krypton | Looks the same to me as in the non-working case. | 22:44 |
krypton | OK, rebooting with debug now, non-working case | 22:46 |
fsmithred | need coffee - back in 10-15 min | 22:47 |
aitor | hi | 22:48 |
krypton | I'll just check the debug output, then I'll go to bed... | 22:48 |
Guest902 | there is a mistake in the packaging of runit in debian, i guess... | 22:49 |
Guest902 | the location of the services is arbitrary: it can be /service, /etc/service, /etc/sv... | 22:51 |
Guest902 | Artix (base on arch), for example, uses /etc/runit/sv | 22:52 |
Guest902 | The folder "/etc/runit/runsvdir" should contain symlinks to those services you want to use | 22:53 |
krypton | During the last startup of udevd, the lines after startup are the same, including: | 22:54 |
krypton | [ 7.857477] udevd[616]: Failed to symlink /proc/kcore to /dev/core: File exists | 22:54 |
Guest902 | you can define a lot of services in, say /etc/sv/, and use only a few of them creating only a few of symlinks at /etc/runit/runsvdir | 22:54 |
krypton | until: [ 7.884542] udevd[616]: header size 80 bytes | 22:54 |
krypton | [ 8.274786] udevd[617]: starting eudev-3.2.7 | 22:55 |
krypton | After that, udev goes different ways. In the OK case the next line from udev is: | 22:55 |
krypton | [ 12.858428] udevd[819]: starting '/etc/console-setup/cached_setup_font.sh' | 22:55 |
krypton | And output is switched to the frame buffer | 22:56 |
krypton | Where in the fail case, the next line is: | 22:57 |
krypton | [ 14.078079] udevd[616]: cleanup idle workers | 22:57 |
Guest902 | Sorry krypton: i'll finish with the issue concerning to runit... | 22:59 |
Guest902 | But first i rectify: the symolinks are located in /etc/runit/runsvdir/default | 23:00 |
Guest902 | in the packaging of debian, the *whole* /etc/runit/runsvdir/default point to /etc/service, as you can see in debian/runit-init.links, but the folder used for the services is /etc/sv | 23:01 |
krypton | Guest902: fsmithred and me were debugging the udevd failures in beowulf, but I'm tired now anyway | 23:01 |
Guest902 | krypton: i know, i read it before | 23:02 |
krypton | OK, bye | 23:03 |
Guest902 | i'll build my own packages of runit, and i'll share an image of gnuinos using it as init system tomorrow | 23:03 |
Guest902 | krypton: bye :) | 23:04 |
fsmithred | aitor? | 23:06 |
fsmithred | Guest902, how much do you need to do for runit to work? | 23:07 |
Guest902 | hi fsmithred | 23:07 |
fsmithred | does it just work out of the box if you install it and remove sysvinit? | 23:07 |
Guest902 | this is my goal | 23:08 |
Guest902 | aside of the required changes in the grub meny entry | 23:11 |
fsmithred | what changes in grub entry? | 23:17 |
Guest902 | init=/sbin/runit-init | 23:17 |
Guest902 | sorry: /sbin/runit-init has been renamed to /sbin/init | 23:18 |
Guest902 | and also changes in the list of the allowed init systems: runit is missing | 23:18 |
fsmithred | where is that list? | 23:19 |
Guest902 | fsmithred: download the sources for grub and have a look at the following quilt patch: "mkconfig_other_inits.patch" | 23:20 |
fsmithred | you do know that beowulf point-release will have runit as a choice in the installer? | 23:21 |
Guest902 | line nº28: | 23:21 |
Guest902 | +SUPPORTED_INITS="sysvinit:/lib/sysvinit/init systemd:/lib/systemd/systemd upstart:/sbin/upstart" | 23:21 |
Guest902 | i didn't know | 23:21 |
Guest902 | then, you'll need to retouch the patch above | 23:22 |
fsmithred | thanks | 23:23 |
fsmithred | stuff for the point release is/will be in beowulf-proposed-updates | 23:24 |
Guest902 | ok | 23:24 |
Guest902 | i'm also enabling services for the virtual tty's, so that they can work with runit, leaving the tty7 (and subsequents) for the Xsession | 23:27 |
Guest902 | time to dinner, see you tomorrow | 23:28 |
devuan | ShorTie, til now I have this list of packages:https://paste2.org/dyL8ybh5 | 23:30 |
crhylove | How do I get an updated version of OBS-studio in Devuan? | 23:34 |
crhylove | Everything is working great, but .1 is pretty old and seems a tad buggy. | 23:34 |
crhylove | Also is there a way to add an android device on the local network as a secondary webcam? | 23:38 |
sixwheeledbeast | They have instructions on how to build the latest version for debian via there git on the website. | 23:58 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!