libera/#devuan/ Friday, 2021-11-12

rwpSince "upgrade" can't add additional nor remove packages.  It's trivial for apt to do it.00:00
rwp"upgrade --with-new-pkgs" can upgrade and can add but cannot remove packages.00:00
rwp"dist-upgrade" is the one I need to stare at to ensure that the solution does not want to remove half the system.  If it does then I wait and debug why.  Unstable migrations mostly.00:00
brocashelmgood calls00:01
rwpI often see people in this channel and elsewhere that jump immediately to dist-upgrade, coupled with having a Frankenstein of 3rd party packages installed.00:02
rwpWhich most often is why they are trying to get help.  Because the solution was to remove half of their system!00:02
brocashelmi just do the apt update/upgrade daily and 99% of the time have no issues00:03
darkeyeHi,05:23
darkeyeHow's it going?05:23
critrit's still going. i don't know how.05:23
darkeyehow can I update lxqt to 1.0.005:23
darkeyeis it possible right now05:23
darkeyeor should I wait05:24
rwpdarkeye, The standard answers when people ask "can I get something newer" are to look in backports, to upgrade to testing, to upgrade to unstable.  In that order.06:10
rwphttps://pkginfo.devuan.org/cgi-bin/policy-query.html?c=package&q=lxqt&x=submit06:12
darkeyerwp thanks06:12
rwpSo it looks like lxqt 1.0.0 released on 5-November just less than a week ago.  That's bleeding edge stuff there!06:15
rwpFor something that new you would need to compile it from source.  Just like the maintainers packaging it for unstable.06:15
brocashelmi tested lxqt in ceres and it's fine06:36
brocashelmtoo bad it also pulls pulseaudio06:36
darkeyeif I upgrade to ceres can I revert to chimaera again?09:06
N3utr1nogood morning09:33
YonleMorning~09:38
djphdarkeye: I don't think "going backwards" is posible (or if it is, it's difficult)12:34
Yonlequit14:00
neutral(if [[ ! -e /tmp/.X11-unix/X0 && ${XDG_VTNR} == 1 ]]; then startx fi)14:07
neutralthis is what my .bash_profile looks like for login to tty114:07
neutralin online they also check for (-z $DISPLAY). is this necessary14:07
neutralhi14:10
LXGHTNXNGNot much clarity on the Website so I will ask here. Does the armhf architecture support armv6l?14:59
UsLLXGHTNXNG: check with the folks in #devuan-arm if no one responds here15:11
LXGHTNXNGI'll take it no though.15:11
UsLpeople in there build their own stuff, so no to your no until they answer.15:13
ShorTieno LXGHTNXNG, Debian does not do armv615:14
LXGHTNXNGGot it.15:14
neutral(error running install command 'modprobe -i nvidia-current' for module nvidia: retcode 1)15:31
neutrali have nvidia installed15:31
neutralX runs fine although that error pops on every boot15:31
brocashelmdarkeye: from what i understand, if you decide to switch back to chimaera or daedalus (testing), the updates will have to be "slowly rolled out" for weeks/months at a time until all the dependencies can be safely met. otherwise, it's a pain to have to individually downgrade packages and correct further dependencies. you can either upgrade to newer builds in chimaera/daedalus over time (without updating ceres ever again) or do a clean install of15:34
brocashelmchimaera for the safest routes (IMO)15:34
brocashelmthe good news is chimaera and ceres have been very close for the longest time in terms of package versions15:36
ascii``Hi there! I'm getting ssh login delays on a `chimaera` installation: pam_elogind(sshd:session): Failed to create session: Failed to activate service 'org.freedesktop.login1': timed out (service_start_timeout=25000ms). Do you know how to fix?16:07
nemoare you really still on ascii?16:07
ascii```busctl` says `org.freedesktop.login1       - -       -    (activatable) -    -       -          `16:07
nemooh. chimæra16:07
ascii``@nemo my nickname is ascii :D16:07
nemoheh. just thought nick had meaning ☺16:07
ascii``is has :D16:07
ascii``s/is/it/16:08
nemoascii``: I realise all of these answers are systemd oriented, but in reading https://unix.stackexchange.com/questions/239489/dbus-system-failed-to-activate-service-org-freedesktop-login1-timed-out  it mentioned excess session files and out of memory which might be generically applicable16:10
nemoesp since /run/systemd/ exists in devuan too16:10
nemohm. but not /system/ ...16:10
nemoascii``: (no idea, was just ducking for your error string)16:10
ascii``@nemo thanks let me investigate16:13
Afdalhmm how does catfish get set to my file manager16:45
Afdalthat's not a file manager...16:45
used____busybox question: Beowulf has busybox installed, but it is not local-installed ( -l ). Is it used only for ramdisk rebuilding?17:41
used____-l == --install17:41
used____man busybox has no info on options -s and --install...17:43
used____-s softlinks applications and --install installs obviously, -s is needed with non root installs.17:45
used____I propose a "policing" script to find all programs which do not have a manpage and also those which have a --help option which lists options, where said options are not mentioned in the existing manpage. The latter is hard(er).18:38
used____The question is, who does one pester with the resulting list of missing manpages?18:39
blockheadthat's a thing a like about debian, and this devuan: somtimes the debian devs make man pages for programs that don't have them18:49
used____You mean they run commands like `$prog --help >/usr/man/man1/fake-$prog.1` ? No, that is not "make man pages" alas.18:50
used____It's just reformatting --help output.18:50
blockheadthat's not what i mean. but if you want to change the meaning of all i say, have yourself your own little conversation18:51
used____I apologize if I misunderstood. In my experience, all Debian provided manpages for other programs which come without, are rather worthless stubs, just --help output reformatted.18:51
used____Typical example: man xfce4 -> Xfce4 - NAME Sensors application for xfce4 ?!19:54
used____`xfce4-sensors -h` shows the same thing, and the "manpage" references this fact.19:55
used____man xfce4-screenshooter -> terse, send off to "full manual" under `info xfce4-screenshooter` which is ... a copy of the terse manual.20:29
used____Because the info pages are probably not installed20:29
used____A package called xfce.*doc-info does not exist.20:33
used____Arch has much more "info" -- I find I use Arch's wiki extensively to set up and get rid of annoyances in Beowulf at least. https://wiki.archlinux.org/title/Xfce#Troubleshooting20:36
used____(some aspects of which are systemd contaminated -- need separate steps, f.ex. laptop lid switch)20:39
[NoClan]GoAwaygreetings. I got some problem running Devuan 4.0 in a HVM enviroment. it's  using up 99% CPU. when looking with "top", it's showing "xen-balloon" using up the CPU.20:50
used____:)20:51
[NoClan]GoAwayis this a known issue? since it's going away when reverting back to 3.120:51
[NoClan]GoAwaysomeone over on OFTC/#xen suggested it might be a kernel issue.20:51
used____The kernel is identical to Debian's, you could ask there?20:54
ascii``Feedback on my issue: commenting out `#session        optional                        pam_elogind.so` in `/etc/pam.d/common-session` fixes the `pam_elogind(sshd:session): Failed to create session: Failed to activate service 'org.freedesktop.login1': timed out (service_start_timeout=25000ms)` issue (delays connecting to SSH). Cheers :)20:55
[NoClan]GoAwayI thought of that, yet I didn't try the Debian one or any flavour of Debian since 2011 because of systemd ... that's why I ask here.20:56
used____https://forum.qubes-os.org/t/high-dom0-cpu-usage-by-xen-balloon/6732 last message >> [NoClan]GoAway -- this is not straight linux but similar20:56
[NoClan]GoAwaythanks, I'll have a look20:56
used____[NoClan]GoAway: you can usually use a Debian kernel as is on Devuan. Check version numbers and config identical.20:57
[NoClan]GoAwayandyhhp over on #xen linked this: https://lore.kernel.org/xen-devel/20211029142049.25198-1-jgross@suse.com/20:57
used____Certain kernel versions seem to generate the high load.20:57
used____[NoClan]GoAway: what kernel do you have? The problem one?20:58
[NoClan]GoAwaynope, currently I'm running Beowulf 3.1 sinc the problem went away with this one. I'll check the 4.0 one...just a second...20:59
[NoClan]GoAwayI would've tried a PV version, but I haven't installed Devuan as a host currently, just wanted to check out if it would run as HVm for testing...as it does, yet with a slight problem ;)21:01
[NoClan]GoAway4.0 is running on kernel 5.10.0-9-amd6421:03
[NoClan]GoAwayI started my tests two weeks ago, so I got the 4.0 and thought it was ok...well, it did install fine and it runs...yet it uses 99% of the cpu21:04
used____5.4 is mentioned as having the problem in my link21:04
used____They suggest updating the kernel to 5.4.151 (out of tree...)21:04
[NoClan]GoAwayyeah, I stumbled on that about a minute ago...thanks again :)21:04
[NoClan]GoAwaywell, upgrade it is then...21:22
neutralfatal error: X11/Xlib.h: No such file or directory21:50
neutral21:50
neutralhelp21:50
neutralwhat is the package in devuan21:50
neutral?21:50
[NoClan]GoAwaymaybe "libx11-dev"?21:59
jk hello, I am having some system-freeze on a relatively fresh devuan (~2 months old) how can I debug this? newbie alarm :)22:14
sadoon_albader[mTry seeing if there's any obvious errors is dmesg22:18
sadoon_albader[mAlso can you ssh into the machine when it freezes? Could be only a GPU freeze22:18
jki can't ssh to it..22:21
jkehmm.. any hints on how to navigate the output of dmesg? like getting the time of the day each entry?22:23
jk... cheching man dmesg now ^^22:23
ltsdmesg -T22:23
gnarfacejk: which video card?  just because you can't ssh in afterwards doesn't mean it might not still be the GPU's fault23:17
_ds_jk, https://wiki.ubuntu.com/Kernel/Netconsole may help23:18
gnarfacejk: *leading* cause of lock-ups is Nvidia's shitty driver support.23:18
gnarfacejk: (but sometimes just getting a different version of them fixes it)23:18
_ds_(it may provide useful log text)23:18
gnarfacehmm, bad timing though with there being no updates for it in chimaera-backports yet23:25
gnarfaceyou might get some extra log output in some cases (i forget exactly which, but this tip was directly Nvidia related) by adding no_console_suspend=1 to the kernel commandline23:27
gnarfacei forget if that only helped when the lockup was related to suspend/sleep states or nto23:28
gnarfaceor not*23:28
gnarface("help" in this case though merely confirmed Nvidia's fault and the total lack of recourse)23:28
gnarfacenote that there's also a time factor often23:29
gnarfaceas in, when the GUI freezes up completely, in my experience with Nvidia hardware you have maybe 2-3 seconds to react and kill it before it locks up the whole system23:29
gnarfaceif you have a ssh terminal attached already *before* that happens, often it can still retain enough control to figure out what caused it (in lieu of a power management problem it's often related to opengl stuff, particularly through Wine or Firefox)23:30

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!