libera/#devuan/ Tuesday, 2021-12-07

nero_hey yo00:37
nero_so... do you know where to find some rad Devuan backgrounds?00:37
gnarfacei don't actually sorry, i always liked digitalblasphemy though00:38
nero_I mean, I just installed it and I was welcomed with just a dark background00:39
gnarfacei only know of onedevuan back ground00:39
gnarfaceone devuan background00:39
gnarfaceyea, the red one00:39
gnarfacewell there's a blue one too i think now00:39
gnarfacedid you check the fourm?00:39
nero_I will00:40
gnarfacei remember e16 used to be bundled with some neat ones but it's not in the repo anymore00:40
nero_kinda weird they didn't name the forum dev1devuan or something00:42
nero_there must be a story behind it00:42
gnarfacedid yoyu check out "apt-cache search backgrounds" ?00:43
nero_there's only plasma's00:47
nero_and something called UKUI00:47
nero_somebody in the forum made this one
nero_guess I'm gonna use that one00:48
nero_see ya00:51
guest127hello, can i install netinstall without wired connection?01:14
guest127or should I use server ISO instead?01:15
fsmithreduse server iso.01:16
guest127I want to do minimal xorg install01:16
guest127and add stuff as needed01:16
fsmithredthen desktop iso01:16
fsmithreduncheck the desktop stuff at the tasksel window01:16
fsmithredcheck or uncheck standard system utilities as you prefer01:17
fsmithredthen add what you want later01:17
fsmithredif you will have network after the initial install, then either server or desktop will do.01:17
fsmithredif you want xorg without network, check the contents of both isos to see which one has what you need.01:18
guest127how do i check contents of isoZ01:18
fsmithredoh, good question01:19
fsmithredlet me see if I can find it01:19
adhocan iso file?01:20
adhocyou could mount it via loopback01:20
gnarfacewait, the netinstall will "work" without a network connection to make a bootable install, won't it?  it used to01:21
gnarfaceif you'll have a network connection later, then a netinstall will still work afaik for that purpose, you just won't have much of use on there until you do get a network connection to download more01:22
gnarfaceotherwise yea, get the big cd or dvd set...01:23
gnarfaceand maybe ALSO the server and desktop iso01:23
gnarfaceand a live image too for good measure01:23
gnarfaceif you're not gonna have a connection i'd download it all...01:23
gnarfacemaybe that's overboard and just a copy of the server iso will do01:24
fsmithredyeah, you'll have a bare system with a no-network netinstall01:24
gnarfaceyea but it'll still boot01:25
guest127will wireless work?01:25
fsmithredthe iso contains enough to connect to repo and get packages01:25
guest127i could not get ceni t work with mine01:25
fsmithredyes, all the installer isos have firmware01:25
guest127with antix core01:26
fsmithredif it's certain broadcom hardware, you must have a network connection to get what you need to have a network connection. (e.g. wire first.)01:26
gnarfacei think in those cases it might still work with a static unencrypted setup, it's just a really bad idea for security01:27
fsmithredok, I just checked for a list, and the person who makes the installer isos said mount the iso and use 'find'01:27
fsmithredcertain broadcoms require a download to get the actual firmware. The deb package is just an installer.01:28
gnarfaceoh that's annoying01:28
guest127how big of a difference between netinstall and server?01:28
fsmithredthey're the ones I don't include in the live isos because of the restrictive licensew01:28
gnarfaceguest127: there's no effective difference after you've installed all the same packages01:29
fsmithrednot a lot, guest. Everything is more bloated with every release.01:29
fsmithredoh, I was thinking of the iso size.01:29
fsmithredwhich is easy to read01:29
guest127contents wise01:29
guest127is netinstall for super experts?01:30
fsmithredthe desktop iso has desktop environments on the disk01:30
gnarfacewell what's already on the netinstall is just the bare minimum to boot and download more01:30
fsmithredthe server has server software, xorg and some desktop or window manager, I think01:30
gnarfacethe other stuff is intended for possible offline use and is an attempt to contain everything you'd need01:30
fsmithredtime for dinner. back in a little while.01:31
guest127I guess I'll just get server01:31
guest127thanks for your time01:31
gnarfacegood luck01:32
systemdletewhat could trigger an unsolicited mount request?  I was restarting a couple of VMs and suddenly there was a popup for authorization to mount a file system.  The file system it was asking for is not even one I use.  It has a hyperbola linux install on it, but I haven't accessed it in over a year, maybe years.05:10
systemdleteIt's on an md raid partition and it IS listed in my mdadm.conf file.  But why would it all of a sudden request a password to mount it?05:10
systemdlete(the VMs are on other file systems, so it is not related to virtualization)05:11
systemdleteI wasn't even working at the command line, nor was I browsing files or directories.  I was simply launching a couple of VMs from desktop icons05:13
systemdleteI launch these VMs regularly and never noticed this previously05:13
blockheadsounds very windows-like.  unfortunaley, i have no ideas (save maybe to click cancil to the popup as a matter of principle?)05:14
onefangThen see what breaks, might provide a clue.05:15
rrqis there a monthly mdadm resync involved?05:15
systemdleterrq:  Ah... maybe that's it.   I don't know.  But if there is, I have never been prompted like this before.05:16
systemdleteThis machine has been up and running beowulf now for months.05:17
systemdleteI did a "grep -ir /dev/md/mdX /etc" looking for such a reference. Nothing05:17
systemdleteThere are no desktop icons for that file system (or others which I do not usually mount)05:18
systemdleteso it can't be an accidental click of the mouse (which I am known to do sometimes)05:18
rrqsomething in some /etc/cron*/mdadm05:20
rrqask for password seems strange05:21
systemdleteI checked.  Not seeing anything like that...05:21
rrqit suggests something doing "sudo" (?)05:22
systemdleteok, now there is a daily (not monthly) job that runs mdadm05:23
systemdleteI was grepping through files, not looking for files named mdadm, that's why I missed it05:23
systemdletethe file is clearly named "mdadm"05:24
systemdletethat daily cronjob is running a scan on the device(s)05:24
systemdleteI've commented out the mdadm devices I don't mount in the /etc/mdadm/mdadm.conf file05:26
systemdleteHopefully that will avoid recurrences.  But  it is still strange.05:27
systemdleterrq: Afaik, I am the only one here.   That doesn't mean there can't be some trouble.05:27
rrqI suppose if the timing of that mdadm resync is right then that's a likely cause05:29
rrq(or "scan" it was)05:30
systemdleteI wonder if using thunar is connected to this.  I have my VM icons in a desktop folder.05:31
systemdleteperhaps, somehow, even though that mdadm device is not IN that folder, thunar was getting a click on it anyway05:32
systemdleteMaybe a bit farfetched but I think within possibility.  I have been having some problems with my mouse here.  I've tried a couple different ones and they are clicking when I don't want them to, and not when I do.05:32
systemdlete(sorry I did not think of this sooner, actually)  So some sort of phantom click was generated.05:33
systemdleteI have another mouse...05:34
rrqsounds more realistic though :)05:34
systemdleterrq:  I'll use this (3rd) mouse for a while and see how that goes.05:35
onefangLots of mice running around clicking on things, you need a cat.  Luckily, Devuan comes with one.05:35
* systemdlete grumbles deeply about M-I-C-E05:35
systemdleteonefang:  Tell me more05:35
systemdleteI know it comes with a guard dog, called "firewall"05:35
onefangman cat05:36
systemdleteoh, that cat.  yes.  of course05:36
systemdleteI call my cat all the time, too.05:36
systemdleteSo really linux is just a game of cat-and-mouse then?05:37
onefangIf we keep this doggerel up, one of the admins will scream #devuan-offtopic at us.  Though sometimes that's my job.05:38
systemdletegood thing you only have 1 fang05:41
systemdleteI'll be back with a follow up in a few hours.  Thanks for all the input and feedback.05:42
onefangI got a dentist appointment first thing tomorrow morning.  That might change.05:42
systemdlete(browsing the system logs I happened to notice a few messages mentioning Thunar, which made me think about the fact that my VM icons were in a folder)05:43
systemdlete(the messages themselves were not warnings or errors, but they did make me think back to the mouse issue I've been having)05:46
rwpsystemdlete, That asking for a password thing is definitely the Desktop Environment.  No self respecting window manager would do ask for it.06:06
rwpNot related to mdadm in any way.  I am confident.06:06
rwpThunar though might have done it.06:06
systemdleterwp:  I am inclined to agree, esp.given the problems with the mouse here06:06
systemdleteI've not seen ANY problems of any kind with mdadm in the decade or so I've been using it06:07
rwpmdadm operates at a lower layer of the software.  It's already running as root to do its job.  So it would not need to ask for a password in order to gain root priviledges.06:08
systemdleteright.  In fact, beyond root.  It is kernel driver06:08
rwpThat mouse click problem...  It isn't unheard of for bad things to generate mouse clicks.  But I think it extremely unlikely it is a malicious agent.06:08
systemdleteonly a portion of it runs outside06:08
rwpI think your best option is to switch to a mouseless widow manager like i3 and then throw the mouse away! :-)06:09
systemdletesome of the articles I found online said that it has to do with a switch inside the mouse.  Sometimes the switch becomes faulty and one guy had even bridged a capacitor in to fix it!06:09
systemdleterwp:  I like that idea.  It IS appealing.06:10
systemdleteBut actually, I still prefer a mouse over a touchpad.  Like my smartphone which is a touch pad.06:10
rwpI do still use the mouse for cutting and pasting.06:10
systemdletecutting and pasting can be done with control keys06:10
rwpAt this moment I am on my desktop and it does not have a touchpad.  My keyboard here though is a Lenovo with the trackpoint.  But that is not so good for cutting and pasting.06:11
rwpSome cutting and pasting can.  Shift-Insert of course.  But copying out of a web browser is tricky without a mouse.06:11
systemdleteI didn't like the tiny trackball at first -- I've worked on a laptop that had one built in between the G and H keys.  But I got used to it, and I kind of liked it.06:11
systemdleteYou don't have to move your fingers off the home keys to use it.06:12
pingpongballif i want to ask question regarding sysVinit where can i ask13:22
GyrosGeierlikely here13:25
pingpongballthis is my question13:26
GyrosGeiersysvinit uses start/stop scripts13:27
pingpongballso what can i do it for it?13:28
GyrosGeierbecause there's a lot of common code, there is a library of common functions13:28
GyrosGeierI'd start with a template script and delete what I don't need13:28
pingpongballr u a devuan maintainer13:28
pingpongballi asked same question on other forum13:29
pingpongballthey told13:29
pingpongballregarding on sysVinit13:29
pingpongballsysvinit doesn't have user-level services. the common workaround was an @reboot job in crontab13:29
GyrosGeier is an acceptable starting point13:29
GyrosGeierah wait13:30
GyrosGeierthere are no user services indeed13:30
pingpongballis there any  library to do so, without making such scripts13:30
pingpongballi should put that template to bashrc13:31
GyrosGeieryou could define start/stop functions in bashrc, that would work13:33
GyrosGeierthe template assumes that it's a full script that is called with a "start" or "stop" parameter13:33
pingpongballThank you GyrosGeier13:35
pingpongballSir i can see readme tells13:37
pingpongballCopy template to /etc/init.d and rename it to something meaningful13:38
pingpongballinit.d is folder or should i rename template to init.d13:39
GyrosGeiersysvinit has a script collection in /etc/init.d, and symlinks in /etc/rc0.d, /etc/rc1.d, ... /etc/rc6.d that tell which script need to be run13:40
GyrosGeierthat's a folder13:40
pingpongballyes yes :)13:40
GyrosGeiera daemon can be started and stopped by calling the init script directly, or by init using the run-parts utility13:41
GyrosGeier"/etc/init.d/bluetooth stop" shuts down bluetooth for example13:41
GyrosGeierand /etc/rc0.d/K01bluetooth is a symlink to ../init.d/bluetooth13:42
pingpongballoh wow cool ,13:42
pingpongballThen edit the script and enter that name after Provides: (between ### BEGIN INIT INFO and ### END INIT INFO).13:43
pingpongballwhat should i do inside it13:43
GyrosGeierso when shutting down (aka "runlevel 0"), all the K.. scripts in /etc/rc0.d are called with a parameter "stop", and all the S.. scripts are called with a parameter "start"13:43
GyrosGeierthe comment at the bit provides extra information13:44
GyrosGeier*at the start13:44
GyrosGeierthat isn't actually used by init itself, but by more efficient replacements for run-parts13:44
GyrosGeiernormally, the number behind K.. and S.. gives the order in which scripts are run13:45
GyrosGeierbut for some people, that is too slow13:45
GyrosGeierso there are "dependency-based" boot systems where you can declare dependencies like in systemd13:45
pingpongballGyrosGeier how did you learn this stuff, would you provide this tutorial13:46
GyrosGeierI learned that a bit over twenty years ago13:46
pingpongballoh   :)13:46
GyrosGeiermostly by reading the existing scripts13:46
GyrosGeierthe sysvinit system is very simple13:46
pingpongballshould i change required-start or which part should i change to add "emacs --daemon"13:47
pingpongballI see cmd, is it cmd to put13:47
GyrosGeiercan you read shell scripts?13:48
GyrosGeierin this template, cmd is the command13:48
GyrosGeierdir is the working directory13:48
GyrosGeierand user is the username13:48
GyrosGeieryou can see where these variables are used later13:49
pingpongballif i write something bad, would it break or is it be fine?13:49
GyrosGeieras long as you don't create the symlinks, nothing can break, because system startup ignores your script, and you have to call it manually13:49
GyrosGeieryou then test it manually, then use update-rc.d(8) to create the symlinks13:50
pingpongballyes i've written13:50
pingpongballcmd: "emacs --daemon"13:50
pingpongballuser: "ping"13:50
pingpongballhow can i test it? without creating sysmlinks13:51
GyrosGeierjust call it13:52
pingpongball### BEGIN INIT INFO and ### END INIT INFO13:52
pingpongballI've just uncommented , nothing changed13:52
GyrosGeierfrom the shell's point of view, that's just comments13:53
GyrosGeierthese should remain comments13:53
pingpongballoh i commented it13:53
GyrosGeierthe shell will ignore them, but if you use startpar(8) instead of run-parts(8) during system startup, that info will be used to start things in the right order13:54
av6i think the header is used by update-rc.d that builds /etc/init.d/.depend.foo13:56
pingpongballi should run this afterwards13:57
pingpongballalgorithms is my name of template13:57
pingpongball /etc/init.d/algorithms start13:57
av6i guess then the is used by startpar13:57
av6(IOW update-rc.d is a required step)13:57
pingpongballwhat is startpar and run-parts, and update-rc.d13:58
pingpongballbash permission denied13:58
GyrosGeiertools to manage symlinks13:58
pingpongballit denied13:58
GyrosGeieryou can look in the manual to understand them13:58
pingpongballusing sudo also denied13:59
GyrosGeierthe script needs to be executable13:59
* av6 experiences ltt flashbacks13:59
pingpongballchmod +x /etc/init.d/algorithms13:59
pingpongballis it13:59
pingpongballah i get error14:00
pingpongballemacs command not found14:00
pingpongballyes it worked14:02
pingpongballyeah it worked14:02
pingpongballsir now does it just run on startup14:03
pingpongballi'll update my answer later14:03
GyrosGeierto run it on startup, you also need to use update-rc.d to create the symlinks14:09
GyrosGeierthe best starting point for that would be the manual page14:09
fsmithredWhy would all of /var/log/*.1 files be empty?14:12
fsmithredThis is a laptop, and it looks like all those logs got turned over on Nov. 414:12
pingpongball51how can i use update-rc.d to create the symlinks?14:16
pingpongball51again thanks:)14:16
jonas1Hi everyone, I'm trying to install devuan on my laptop. I already have Windows 10 and I want to keep it installed. I have a gpt and already created a partition for devuan.So now I've booted from an usb-stick with devuan live. I runned the installer, verified my partitions and now I'm asked where to install GRUB. If I understand it correctly I don't have a MBP on gpt(?). What do I have to do to18:16
jonas1install grub? Thank you18:16
ham5urgI'm missing the xscreensaver-settings-icon in xfce4. Is there a package I missed?20:24
dlcusaham5urg, I doubt it.  I've never seen that icon in CRUX.  If you can't look into it, I'll attempt to make that a priority.20:29
dlcusaI just open the xscreensaver app and tweak the basic and advanced panes.20:30
dlcusas/app/settings -> xscreensaver20:31
dlcusaham5urg, ^20:32
ham5urgdlcusa, there is no xscreensaver entry in settings, using daedalus20:36
dlcusaham5urg: Sorry, it's Screensaver, not xscreensaver.20:37
dlcusaham5urg, I'm using whiskermenu.20:38
barrett9hEverything was working fine with ASCII, but now that I upgraded to Chimaera audio is not working well.22:05
barrett9hI get distorted audio with Firefox, and sometimes audio stop working at all until I reboot.22:07
barrett9hAfter updating from ASCII, I tried reformating / and installing Chimaera from scratch, with the same results. :-(22:07
barrett9hI use the MATE desktop environment, if that matters.22:08
barrett9hCan anyone help me solve this? I'm about to give up Devuan (for now) and try another distro, or even *BSD.22:09
barrett9hCome to think of it, pulseaudio came from the same guy that invented systemd, and it has been and endless source of problems.  Shouldn't we get rid of it too?22:10
gnarfacebarrett9h: trying it without pulseaudio is a very good and strongly advised troubleshooting step, but if it used to work with ascii that strongly suggests the possibility of a kernel regression too.  were you using pulseaudio on ascii?22:11
golinuxDid you go directly from ASCII to Chimaera or did you go through Beowulf to Chimaera?22:13
gnarfacebarrett9h: fyi though, kernel 5.14 is available in chimaera-backports and may behave better for you.22:13
av6you can call me names, but i personally think there's a way bigger chance of audio breaking because of pulseaudio than because of kernel22:13
gnarfaceav6: eh, it's about 60/4022:13
gnarfaceav6: maybe 50/50 if you're using snd_hda_intel22:14
av6i'm on multiple machines with snd_hda_intel and never had any regressions, since forever22:14
gnarfacebarrett9h: it's worth mentioning that there's almost no chance this is devuan-specific; the same versions of kernel and userspace tools in debian or any other debian derivative that share them are going to behave exactly the same22:14
barrett9hgnarface: yes, I was using pulseaudio on ascii22:15
barrett9hgolinux: sorry, I meant beowulf to chimaera22:15
barrett9hwell, I'm using devuan since alpha days22:15
barrett9hgnarface: I'll try 5.14 before giving up22:16
av6is it firefox only?22:16
gnarfaceyea, run some tests with speaker-test and aplay/arecord22:16
gnarfacemake sure it's not just firefox's fault22:16
barrett9hsnd_hda_intel here22:17
gnarfacein general, firefox and wine both make bad choices for debugging anything other than themselves22:17
av6maybe you'll see something in about:support22:17
av6there's a media section22:18
barrett9hI use mpv a lot to play things22:18
gnarfacehmm. yea i can think of one thing specifically; disable the Cisco OpenH264 plugin22:18
gnarfacein firefox about:plugins22:18
barrett9hthe Firefox thing is just distorted audio.  it stops as soon as I close Firefox (or the offending tab).  the issue of the audio stopping working at all has (as far as I know) nothing to do with Firefox.22:19
gnarfaceit's provided for cross-platform compatibility but it's inferior to the native one that ships with every modern linux distro22:19
av6so audio in firefox affects other audio streams? what about audio output using two mpv at the same time?22:20
av6(thinking maybe it's the mixing)22:21
av6preferably with different sample rates, e.g. 44.1 and 4822:21
gnarfaceav6: with the snd_hda_intel thing, you've just been very lucky or only used it for stereo half-duplex.22:27
gnarfaceav6: it is a catch-all driver kinda like the usb-audio one but primarily for supporting products mostly ironically made by Intel's competitors22:28
gnarfaceav6: i've used it across several motherboards over almost 2 decades now; support for them is not uniform in quality22:28
gnarfaceav6: i agree that pulseaudio is probably higher up on the list of usual suspects but in the case of the snd_hda_intel driver i wouldn't exactly turn your back on it, either22:29
gnarfaceav6: it *almost* never fails at outputting stereo, but if you try testing using 5.1 or 7.1 surround sound, or doing any output while also recording at the same time to mic input, that's where the seams start to appear frayed22:30
barrett9hstereo-only here22:31
gnarfacebarrett9h: i think you can test without pulseaudio even if it's installed by doing something like this as root:  pasuspender -- speaker-test -c 222:32
barrett9hinstalling linux-image-5.14.0 now22:32
av6fair enough, i didn't try to squeeze every bit of functionality out of my audio, always going stereo and occasionally full-duplex, but nevertheless it worked fine for me22:33
barrett9hI tried this pasuspender.  As root it gives a connection failure.  As my user it "works", but I hear no sound.22:34
av6well, isn't it a catch-all driver because it implements hda intel, which is simply a protocol that audio chip talks?22:35
barrett9hgonna reboot now to test 5.14   brb22:35
gnarfacebarrett9h: maybe it's easier to just stop pulseaudio; i don't think it will relaunch it like with systemd22:35
gnarfacebarrett9h: oh, try that first yea22:36
av6barrett9h: did you try multiple audio streams with different sample rates yet?22:36
barrett9hhow do I stop pulseaudio?  If I kill it it just comes back.22:36
barrett9hav6: nope, atm the audio is not working, so I can't test22:37
gnarfaceav6: it's a protocol licensed by intel to 3rd party chip vendors to support a range of form factors and hardware capabilities; the fundamental support overhead is identical to the usb-audio scenario even if the licensing/ownership parts aren't22:37
gnarfaceav6: (most the devices in the wild using this driver aren't even made by Intel)22:38
av6yeah, i'm saying it makes sense that the protocol is handled by just one kernel module22:39
av6and then other modules can depend on it22:39
gnarfacei was just saying it doesn't make sense to design hardware this way, no matter who you foist the blame off to for support issues afterwards22:39
av6why doesn't it make sense if it's more or less standardized?22:40
gnarfacebecause it proverbially "bites off more than it can chew" so it's rather less standardized than more22:41
barrett9hrebooted, sound is working now, but it proves nothing, as it was usually temporarely "fixed" with a reboot before22:41
barrett9hI'll have to use it for some time to see if the problem went away.22:41
gnarfaceav6: and their cost-cutting results in support issues that trickled down to here22:41
av6barrett9h: well, you can try and wait for the problem to reappear and kill pulseaudio, that would restart it22:42
av6if that fixes it you'll know where's the problem22:43
gnarfacebarrett9h: if you can reliably trigger the failure maybe someone can come up with a workaround in the form of a module parameter or custom alsa pulseaudio config change22:43
gnarfacebarrett9h: if it never comes back just assume it was a regression that was already caught and count yourself lucky22:43
gnarfacebarrett9h: there is one known issue with that driver i think probably is not fixed still and probably never will be, but i've only managed to find one program that causes the issue, it was the Civ5 linux port on Steam and it doesn't happen until you've had it playing the in-game music constantly for a good several hours.  any chance this issue has only appeared to you after a very long constant audio stream source of22:45
gnarfacesome sort?22:45
gnarfacethe symptom is that after a very long time the audio just starts stuttering/chopping in a very annoying fashion, similar but not quite identical to the type of stuttering you get when you get a buffer underrun22:47
gnarfaceit didn't actually require a reboot or anything, just a restart of Civ522:47
gnarfacethe difference seemed to be that while most games would open a separate connection to the sound card for every sound, Civ5 would actually just open one connection, once at the opening menu music, then never put it down, presumably software-mixing all additional sound effects into this one stream internally22:52
gnarfaceand even additional music tracks22:52
gnarfaceso for some reason after several straight hours of this, the driver occasionally will drop a beat and never fully recover22:53
gnarfaceunclear if this is a hardware or software issue, but it is a common issue that is either the driver or the hardware's fault, that most people just blame on Civ522:55
gnarfaceand it might be related to your issue, barrett9h22:55
ham5urgIs there a tool to generate a .dot file from a package and all dependencies? (must-have, recommends, optional)23:22
ham5urgAnd dependecies of dependencies of ... to the very end.23:23
rrqham5urg: like debtree?23:25
ham5urgrrg, yes. Is there a way to recursively include all packages which will be involved?23:29
adhocham5urg: you mean dot files, as in graph viz ?23:36
ham5urgadhoc, yes23:37
ham5urgdebtree prints it recursively out. Nice. Thanks guys.23:41

Generated by 2.17.0 by Marius Gedminas - find it at!