this is #debianan IRC-Channel at freenode
(freenode IRC service closed
2021-06-01)
0[00:01:00] <greycat> It's reasonable to expect that
systemd would reject the message, and the /sbin/shutdown client
program would report an error. So, file a bug report.
1[00:01:27] <greycat> They may never change it, but it's
reasonable to ask.
2[00:01:41] <plujon> greycat: Thanks; where to file..?
3[00:01:46] <greycat> *sigh*
4[00:01:50] <greycat> !bts
5[00:01:50] <dpkg> Bug Tracking System for Debian packages,
replaced-url
6[00:02:08] <jhutchins> Actually, it would be great if it
returned "you must be root", but I don't remember
which program actually does that.
7[00:02:34] <joepublic> "This failed, that failed, and
you are not root."
8[00:02:37] <greycat> I don't even think that's
strictly accurate any more. There are probably facilities for
non-root users to be allowed to request a system shutdown/reboot.
9[00:03:31] <plujon> FTR: /sbin/shutdown is from systemd-sysv
10[00:03:31] <jhutchins> greycat: There are, most GUIs allow a
user to shutdown/reboot.
11[00:03:47] <greycat> yeah, that's one of the cases I was
thinking of
12[00:03:49] <jhutchins> plujon: Possibly even earlier.
15[00:08:07] <gpunk> jhutchins: those GUIs go thru the DE (kde,
gnome etc) , and the DEs talk to "XDM" which runs as root,
that's why regular user can use the shutdown command in
general, unless the system is configured to allow a regular user to
reboot the machine with his/her real own uid
131[01:46:47] <mtlsw> "non-free" is part of debian
right boys? hence that's a reason why unsupported then becomes
somewhat "supported". Confusing to the new user. :)
134[01:48:41] <genr8_> theres a very good reason for it not
being default. convincing people to default auto-install everything
including the kitchen sink without knowing why is how you end up
with ubuntu
135[01:48:41] <mtlsw> leoni, fwiw, the unofficial installer is
available from the debian download locations --- they're hosted
on debian official servers -- this also adds some level of confusion
to new users....
136[01:48:51] <dka> I have then manually upgraded
linux-image-amd64 to use 5.8 :
replaced-url
137[01:49:03] <dka> I am now still rebooting on 5.8, to see if
that help
152[01:53:44] <dka> how can a non working kernel with non
upgraded dkms module can help to fix a previously stable kernel.....
153[01:54:03] <genr8_> thats not what happened.
154[01:54:06] <dka> Of course, but to me it was, otherwise I
would have fixed it like this. this is not expected that a kernel
upgrade fail, it shouldn't be on backport
157[01:54:30] <genr8_> your last paste shows you finally did the
proper course of action. it just did not work for some other reason
158[01:54:45] <genr8_> and now im out of ideas
159[01:54:49] <n4dir> recommending non-free repos or installers
to people without them complaining about related problems sure is a
bit odd
160[01:55:05] <dka> Should I remind you the history? my debian
was working super fine on 5.7, since this error, downgrading to 5.7
or keeping the upgrade to 5.8 keeps an issue with google chrome not
being able to full screen or maximize without it's window
freezing, but it's not crashed, I onloy need to minimize or to
disable full screen
162[01:55:43] <dka> since 5.8 can't be instlled without
failure of acpi-call-dkms from backport, I wanted to revert to 5.7,
now I am stuck with this bad video issue and 5.8 can't help
177[01:58:42] <genr8_> the bottom gives instructions
178[01:58:59] <dka> Run 'make oldconfig && make
prepare' on kernel src to fix it. <= yeah but what an
instruciton, it's like it believes i am building from ssource
179[01:59:06] <dka> and I am not
180[01:59:14] <sponix> dkms is a source build though
181[01:59:14] <genr8_> DKMS requires the kernel source, so yes,
you are.
182[01:59:44] <genr8_> cd
/usr/src/linux-headers-5.8.0-0.bpo.2-amd64; make oldconfig
&& make prepare
183[02:00:02] *** Quits: magenbrot (~magenbrot@replaced-ip) (Remote host closed the connection)
184[02:00:05] <genr8_> then redo the install for backport's
acpi-call-dkms
219[02:07:15] <maxrazer> How can I get no results from "apt
search g++-multilib" I know that packages.debian.org shows
"g++-multilib_10.2.0-1_amd64.deb" package
231[02:11:07] <dka> well, that's still not fixing anything
=/
232[02:11:35] <genr8_> wait
233[02:11:39] <genr8_> no that made progress
234[02:11:52] <dka> how? I dont see one
235[02:11:57] <dka> sudo apt autoremove
236[02:12:00] <dka> oups
237[02:12:10] <genr8_> consult the log of now
/var/lib/dkms/evdi/5.2.14/build/make.log
238[02:12:36] <n4dir> maxrazer: it might be the double ++ is
seen as a bash glob, and that is the problem. If you escape them, or
one of them, you get a result \+\+ like. I can't explain the
details
239[02:12:43] <mtlsw> you also need the headers-all package as
well for that version iirc
243[02:14:05] <maxrazer> n4dir, I thought maybe, but I tried
with double quotes too.
244[02:14:13] *** Quits: Gerowen (~Gerowen@replaced-ip) (Remote host closed the connection)
245[02:15:12] <n4dir> maxrazer: same here. double qoutes
didn't work :-) As said: i don't understand the details,
just a vague guess. You could ask in #bash, if no one knows for sure
here. They should know "what happens when" in and out
246[02:15:25] <genr8_> dka, i still think you're making
forward progress. even if nothing is fixed youve eliminated
variables of whats broken. if you wanna reinstall thats up to you. I
dont know how to fix it.
247[02:15:32] <dka> What package are you refering to mtlsw ?
248[02:15:53] <genr8_> I dont see an "all" for 5.8
249[02:16:01] <dka> genr8_, if I can't fix it, can I safely
downgrade to 5.7 ? And wait till a new kernel is coming out?
251[02:16:42] <n4dir> apt-cache search --names-only g++ multilib
did work too. But way too many results. And as "gobject"
was in the results, i thought of that globbing thing (assuming it is
called like that)
252[02:16:44] <mtlsw> dka, I think it may be the
"common" named package -- I think that particular name may
change to saying "all" instead, but I think the
dependencies check would also tell you.
253[02:16:44] <genr8_> yes, your grub config said you still had
the 5.7 installed
254[02:17:05] <mtlsw> dka, so nothing to worry, if your repos
are set correctly.
255[02:17:31] <dka> well, there's stil this acpi-call-dkms
that have been upgraded (with failure), should I also downgrade this
one ?
256[02:18:42] <mtlsw> dka, actually afaict when "apt-cache
search linux-headers|grep -i all" , it's all about kernel
4.x with things like linux-headers-4.19.0-10-all -- but they are
described as "meta packages" -- so they're virtual
packages likely pointing to common I suppose --- I don't see
all here on Debian -- I may have been drunk and recall something
with -all on Shoe-Buntu :)
257[02:19:08] <genr8_> i dont think thats necessary, but you
would need to have the 5.7 headers then. which you'd need to
get seperately from snapshots now that you forced 5.8 in
258[02:19:21] <genr8_> mtlsw, thats correct
259[02:19:43] <genr8_> 4.19 has all, 5.8 doesnt. idk why
260[02:19:58] <genr8_> but it is a meta package that points to
your arch + common
261[02:20:07] <mtlsw> its ok, it's not needed.. it's a
meta-virtual package that instead queries to install those other
-common header packages.
262[02:20:21] <dka> so how do I revert?
263[02:20:38] <mtlsw> dka, just ignore everything I said.
267[02:21:26] <dka> I am not ignoring, but you are discussing
with genr8_ about meta packages not being here for 5.x kernel, I
also agree that they are 4.x package that are not server for 5.x,
can't tell why
268[02:22:26] <mtlsw> if you have things for dkms-compiling for
external/proprietary drivers, then package "dkms" afaik
needs to be installed
284[02:25:42] <sponix> dka: it might come down to that -- I had
a LMDE user earlier today that had the backport 5.8 blow up his rig
also
285[02:26:06] *** Quits: cdown (~cdown@replaced-ip) (Remote host closed the connection)
286[02:26:23] <genr8_> dka, do you remember exactly what 5.7 you
had before ?
287[02:26:26] <sponix> dka: his kept saying something about
couldn't create a link error. Luckily he had a split off /boot
-- and unmounting that in installing the kernel to the root boot
directory worked out
318[02:37:26] <mtlsw> if linux-image-5.8.0-0.bpo.2-amd64 is a
problem you can try the -unsigned package -- which is also available
from bp
319[02:37:28] <mtlsw> ./bpo
320[02:37:48] <dka> i'll try it mtlsw
321[02:37:55] <genr8_> dka, thats why i have linked to the
webpage, said you have to download the .deb's manually and
install them manually with dpkg -i
374[02:48:58] <dka> why would I keep one that couldn't
succeed it's install?
375[02:48:58] <genr8_> 5.7 still should boot
376[02:49:10] <genr8_> dka, it did install. the error failed on
evdi now
377[02:49:10] <mtlsw> dka, the postinst scripts run again when
you just type "apt-get -f install"
378[02:49:16] <dka> Are you sure => rc linux-image-amd64
5.8.10-1~bpo10+1 amd64 Linux for 64-bit PCs (meta-package) ?
379[02:49:21] <mtlsw> dka, the 5.8 headers are optional if
you're not compiling anything
380[02:49:31] <genr8_> dka, as long as /boot/ has the 3 5.7
files then yes
381[02:49:47] <genr8_> and grub can find them
382[02:49:47] <mtlsw> dka, for a 5.8 kernel, you only need the
kernel package
383[02:50:10] <genr8_> and grub did find it, so
384[02:50:25] <mtlsw> dka, but for anything that you want built
in /var/lib/dkms, you'll need to have the headers installed
385[02:51:12] <genr8_> he's getting those build errors from
the headers, and i dont get why.
386[02:51:20] <mtlsw> dka, the headers package was failing to
compile things under that /var/lib/dkms path -- on the evdi driver
--- if you don't know what the evdi driver is for, you can
delete its package (dpkg -S /var/lib/dkms/evdi/)
387[02:51:39] <dka> Am I safe to reboot :
replaced-url
389[02:52:03] <mtlsw> dka, well you said you want 5.8
390[02:52:07] <dka> dpkg-query: no path found matching pattern
/var/lib/dkms/evdi/
391[02:52:10] <jim> this is really weird... it won't let me
-build- the 5.4 kernel, without all the -images- for that kernel
installed... how nuts is that? so now I'm looking for
linux-image-5.4.0-4-amd64-unsigned but it's not on snapshot
that I can find
392[02:52:15] <mtlsw> dka, dpkg -l|grep evdi ?
393[02:52:17] <dka> mtlsw, I can't upgrade to 5.8
:'(((( it fail
394[02:52:27] <mtlsw> dka, actually you should be able to
403[02:53:42] <mtlsw> dka, or just move that /var/lib/dkms/evdi
to /root/backup
404[02:53:43] <dka> I want the SOF audio fix for my soundcard so
it detect the microphone, that's a human use case, I don't
know what kernel-source-headers helps a real human use case
405[02:54:24] <dka> ok it's in backup, now what? reinstall
5.8?
463[03:03:29] <mtlsw> dka, you might have to use -t , if you
can't access backports --- its because your apt preferences
sees bpo as not a high-priority -- that can be fixed much later as
desired.
480[03:05:18] <mtlsw> acpid is a separate package -- the
acpi-call-dkms is something external and named differently
481[03:05:41] <mtlsw> if acpid should work, then
"acpid" shoudl be installed -- it tends not to be
installed sometimes -- that can be installed if whoever needs it.
482[03:06:35] <genr8_> ACPI in the kernel is different than
acpid userspace daemon
483[03:06:58] <mtlsw> yeah I have no idea what that is
acpi-call-dkms, it looks like it is for developers.
484[03:07:12] <genr8_> could be useful for his laptop
494[03:08:02] <genr8_> well maybe another program relies on
those calls
495[03:08:07] <mtlsw> (apt show acpi-call-dkms)
496[03:08:41] <genr8_> i have no idea why he installed it to
begin with but its probably from some tutorial about fixing his
laptop to do some specific hardware command
497[03:08:42] <mtlsw> it's an uncompiled kernel module
500[03:08:59] <mtlsw> looks like its for some development
purpose.
501[03:09:18] <mtlsw> acpi is actually hardware
502[03:09:21] <mtlsw> :)
503[03:09:29] <genr8_> i hear you, but you're not
processing what im trying to convey
504[03:09:33] <mtlsw> dmesg |grep -i acpi
505[03:10:14] <dka> So I am on 5.8.0-0.bpo.2-amd64 and I still
have issue with chrome not being able to maximize or full screen
without freezing it's window, (i can recover by minimize or
exiting full screen still) ... AND ... the microphone works but
it's super low
506[03:10:16] <genr8_> ACPI on laptops controls essential
hardware functions, like fans, backlight, buttons, switches, any
number of proprietary things for his X1 Carbon laptop. its not clear
why he has that module, but some program could be relying on those
ACPI calls
507[03:10:26] <mtlsw> dka, that's good
508[03:10:35] <mtlsw> dka, at least getting somewhere woohoo
513[03:11:20] <dka> genr8_, thanks a lot for your time and
support
514[03:11:36] <genr8_> yw
515[03:12:45] *** Quits: Grldfrdom (uid391113@replaced-ip) (Quit: Connection closed for inactivity)
516[03:12:51] <mtlsw> genr8_, he was claiming no mic support on
5.7
517[03:13:08] <mtlsw> genr8_, I wonder why you think it is
necessary to remove 5.8 if it was booting?
518[03:13:22] <genr8_> this was his idea
519[03:13:23] <mtlsw> genr8_, the kernel priority can be changed
by editing /etc/default/grub if that needs to be the caes
520[03:13:32] <mtlsw> he's obvious too new to know?
521[03:13:42] <genr8_> he knows when somethings not working
522[03:13:59] <genr8_> did it boot? yes. did it work? no
523[03:14:14] <mtlsw> sure, but no need to remove kernel
packages. You can specifiy the order in /etc/default/grub with
GRUB_DEFAULT=
524[03:14:22] <dka> mtlsw, I know, and I never used that way, I
rather rename the kernel file so it gets weaker priority, because
this work fine
525[03:14:23] <genr8_> ok
526[03:15:02] <mtlsw> dka, another way is to just do -> mkdir
/boot/_dis and move the vmlinuz-*number-version into /boot/_dis and
then re-applay update-grub -- the highest-number kernel is the
default.
527[03:15:10] <mtlsw> but no need to apt-get remove things...
530[03:16:05] <mtlsw> you can always checck what is the default
kernel with "less /etc/grub/grub.cfg"
531[03:17:11] <genr8_> <dka> Should I remind you the
history? my debian was working super fine on 5.7, since this error,
downgrading to 5.7 or keeping the upgrade to 5.8 keeps an issue with
google chrome not being able to full screen or maximize without
it's window freezing, but it's not crashed, I onloy need
to minimize or to disable full screen
532[03:17:11] <genr8_> <dka> since 5.8 can't be
instlled without failure of acpi-call-dkms from backport, I wanted
to revert to 5.7, now I am stuck with this bad video issue and 5.8
can't help
536[03:18:15] <mtlsw> the dkms is irrelevant because -- the
errors around kernel-headers is just for things you want built in
/var/lib/dkms
537[03:18:26] <mtlsw> but it looks like he doesn't seem to
care or need it...
538[03:18:43] <genr8_> ok, he just brought up the mic, the acpi
may be irrelevant, but the chrome video freezing was the first issue
he had.
539[03:18:49] <dka> I don't know what's DKMS to be
honest
540[03:19:03] <dka> yes genr8_ is right, I just reboot and
purged google-chrome-stable
541[03:19:13] <dka> I am a web developer, I can't have such
a big issue with my primary browser
542[03:19:27] <dka> I'll try to reinstall it, without any
extension, see if it goes better, etc...
543[03:19:47] <dka> this bug started with acpi-call-dkms failing
to install when upgrading from 5.7 to 5.8
544[03:19:52] <genr8_> do you know if maybe chrome auto-updating
to 86 caused the bug, and the kernel was just a coincidence ?
545[03:19:57] <dka> downgrading didn't helped, upgrading to
5.8 didn't help
546[03:20:05] <dka> perhaps.
547[03:20:52] <genr8_> i would look into going back to older
chrome then. maybe you still have the .deb in your
/var/cache/apt/archives
548[03:21:27] <mtlsw> dka, what do you want to get fixed?
549[03:21:30] <mtlsw> dka, chrome?
550[03:21:38] <mtlsw> dka, "chrome full-screen" ?
551[03:21:53] <mtlsw> dka, that sounds like your video-driver is
not being utilized with your X11/Wayland instance.
552[03:22:02] <mtlsw> dka, you'd have to check the X logs
for that.
553[03:22:35] <dka> chrome full-screen and maximisez, yeah it
look like the video driver failed, but it used to work and I
didn't change anything
554[03:22:39] <mtlsw> dka, but even if you're trolling that
doesn't bother me, at least a few new other users are learning
something ;-)
555[03:22:48] <dka> except that "coincidance"
556[03:22:57] <mtlsw> yeah yeah "coincidence" :)
557[03:23:05] <mtlsw> try spelling that correctly kiddo ;-)
558[03:23:09] <mtlsw> good luck
559[03:23:14] * mtlsw yeeps
560[03:23:15] <genr8_> hes not trolling lol. it started with
chrome
561[03:23:16] <dka> Sorry, it's late, I am not trolling
562[03:23:50] <genr8_> he left cause fixing video driver issues
are too complex :P way beyond "check the logs"
563[03:23:50] <dka> mtlsw, and I appreciated your help, see I
told genr8_ I was grateful for he's support and time, why would
I do that If I am not concerned with my own problem =/
565[03:24:31] <dka> anyhow, I think you are on the right track,
it sounds like a xvideo issue, but I don't know how it came up,
the one thing i did that started this instantly was upgrading a
working 5.7 to a 5.8 and hit that acpi-call-DKMS error , which I
don't know anything about
566[03:24:33] <mtlsw> I don't know what the problem is. I
disgressed long ago. I was just trying to understand why kernel
headers was needed at all.
567[03:24:45] <dka> I left, because I reboot , that's all I
try to fix things
568[03:25:15] <dka> mtlsw, I am also grateful for your help, I
feel less lonely and more confident by reading other people opinion
575[03:26:06] <dka> mtlsw, well, in that case, may I ask what
did you disagree on?
576[03:26:17] <mtlsw> dka, are you sure you know the difference
between kernel/+_+-userspace___?
577[03:26:17] <genr8_> revert to old version would achieve more
578[03:26:39] <dka> mtlsw, no I am not
579[03:26:52] <dka> and now I do understand this was a mistake
580[03:26:52] * mtlsw thinks dka is a b*t
581[03:26:59] <mtlsw> lol
582[03:27:12] <dka> And since I know that, I would say that
kernel and user space are seperated concerns, and I am safe to keep
N kernels on my laptop
583[03:27:17] <mtlsw> they're getting better
584[03:27:22] <mtlsw> ;-)
585[03:27:52] <dka> it was a mistake to downgrade, but since I
saw that ERROR on DKMS module and I am not used to see that error on
my kernel upgrade step, I freaked out
592[03:28:24] <dka> I am not getting your point sorry mtlsw
593[03:28:27] <mtlsw> you feed it text, and it splindes things
on different strays.
594[03:28:44] <dka> perhaps, time to sleep for both of us ^^
595[03:29:12] <dka> I won't bother your any longer, thanks
for your support and your time mtlsw
596[03:29:38] <mtlsw> you probably shouldn't have wasted
time with genr8_ who doesn't know what he's doing...
597[03:29:40] <dka> Can xvideo issue cause error only for chrome
?
598[03:29:45] <genr8_> wow
599[03:30:34] <mtlsw> !dkms
600[03:30:34] <dpkg> Dynamic Kernel Module Support (DKMS) is a
framework for generating out-of-tree Linux kernel modules. Packages
using DKMS build modules during installation and when updated Linux
versions are installed, provided Linux kernel headers are present
(these are not installed by dkms). DKMS-managed modules are
installed to /lib/modules/$(uname -r)/updates/dkms/ . The module
build log file is at /var/lib/dkms/$module/$version/build/make.log
601[03:30:39] <mtlsw> !acpi
602[03:30:39] <dpkg> Advanced Configuration and Power Interface
(ACPI) is a power management interface that gives more control to
the BIOS, or horribly horribly broken, or "run this arbitrary
code in ring 0", or generally better than <APM>. For a
list of supported systems/BIOSes, see
replaced-url
603[03:30:43] <mtlsw> !headers
604[03:31:13] <genr8_> we still have no idea why he was using
that acpi-call-dkms. I still think its for his Lenovo Thinkpad X1
Carbon Gen 7
614[03:33:13] <genr8_> dka, ahhh this is where the
"evdi-dkms" came from
615[03:33:37] <dka> to be more specific, I installed this, It
used to work and I was able to send SMS or use 4G with it, but I
don't use it often and I lost it's benefits I dont know
when, when I upgraded the kernel probably, long ago.
616[03:33:54] <mtlsw> dka, do you have like a medical condition
of some sort? there's something wrong over here.
617[03:34:10] <mtlsw> dka, it's ok to say it, because I can
tell you have real life issues.
618[03:34:17] <mtlsw> dka, you're very repetitive.
619[03:34:25] <dka> mtlsw, I do, I am hyperactive and stressed,
probably because of my childhood condition, should we fix that first
?
620[03:34:27] <dka> x]
621[03:34:47] <mtlsw> dka, try to read up about those things
that "!" the dpkg factoid pointed out, if you're
interested on what dkms is about that you mention a million times.
622[03:35:01] <dka> I did read every dot
623[03:35:02] <mtlsw> dka, ok. you do have issues. That'
why I was patient.
624[03:35:11] <mtlsw> dka, and I did provide an answer to your
questions.
625[03:35:29] <mtlsw> dka, DKMS -- as I said at least couple
times, pertains to "driver source code"
626[03:35:37] <mtlsw> dka, if you don't know what it is,
you don't need it.
627[03:35:47] <mtlsw> dka, that doesn't mean you need to
remove packages. Just leave it as is.
628[03:35:52] <mtlsw> dka, and it'll be fine.
629[03:35:57] <dka> I got it 100%
630[03:36:37] <dka> what about the google-chrome-stable issue?
Won't you advise on that one too ?
631[03:36:39] <mtlsw> you can /msg dpkg on terms and that
factoid can also point out things to help out while learning debian
636[03:37:04] <mtlsw> and I answered the issue around the
google-chrome issue -- I did say about checking your X logs, but you
refused to do that as well.
637[03:37:20] <mtlsw> so that's why I disgressed and
signaled good-luck to your endeavors.
638[03:37:22] <dka> genr8_, it used to work, on jessie with a x1
carbon gen 1, now I gave up, I don't use that screen since I
only use linux and it does not work with debian, only with ubuntu,
apparently, and it's not so good
661[03:40:43] <mtlsw> abrotman, I suppose that individual is
very noisy-cluttery -- despite how much help me and genr8_ provide
to him on solutions he just ignores all answers we provide to him.
662[03:40:43] <dka> I installed in on a given server, a big Dell
things, it was horrible, I didn t get anything, was reading linux
for dummies, all I do was destroying it not on purpose, long after I
went into a company and fixed what I was doing wrong, at least a
little
663[03:40:46] <dka> but that's off topics
664[03:41:00] <dka> ok
665[03:41:04] <abrotman> then use #debian-offtopic ?
666[03:41:06] <dka> Sorry for this, I am out
667[03:41:09] <genr8_> i have no issue with dka
668[03:41:33] * dka things mtlsw should rest
669[03:41:48] <abrotman> move on, eat tacos, drink a beer ..
678[03:46:33] <mtlsw> for whatever record I recall last mandark
release was around 2011 -- but it wasn't even called that
anymore -- it was name change years prior as well.
681[03:47:27] <mtlsw> resurrected projects around its base are
connected to mageia, pclinuxos, and of course the respun/revitalized
Mandriva-->to "OpenMandriva" iirc around 2013-2015...
684[03:47:52] <dpkg> #debian is primarily a support channel for
Debian users. Please keep the discussions in #debian on-topic and
take longer discussions and non-support questions to
#debian-offtopic. Imagine the chaos if each of the hundreds of
people in the channel felt the need to wander off topic for a few
minutes every day.
703[04:02:08] <leoni> mtlsw: Hey, I'm sorry for not
replying earlier, your message at 20:44 I mean. I already added the
contrib and non-free repositories because my computer needed those
anyways. But thanks.
704[04:05:22] <mtlsw> leoni, good -- but you'd be delighted
to know, that the reason why I mention the "unofficial"
(but it is hosted on the debian official servers -- and is safe)
installer, is because some network cards require the
"non-free" firmware in order to work -- so you're
stuck in this catch-22 situation...
705[04:06:13] <mtlsw> leoni, so it might be a good idea to keep
this in mind, when you install debian, and it "fails" the
network, but Ubuntu/Mint etc work because those distributions come
with the non-free firmware on their installers..
706[04:06:52] <mtlsw> leoni, (and ^ that I mean the
"official" Debian installer, which does not come with the
non-free firmware)
707[04:07:35] <mtlsw> leoni, (-- which you then added the
non-free repositories with apt --- but then of course your
"network" should already be happy enough with what is
installed) ... glad it worked. :)
793[06:26:51] <leoni> I'm still really bad with regex so I
don't use things like grep and awk too often, I mostly remember
basic regex and that's it. But I don't really have a need
for it either, for now.
836[07:31:07] <aliceussr> Hello! I detect security issies with
packages: exim4, exim4-light - its packages have suid in debian and
devuan repositories - it`s is viruses or troyans!
841[07:33:32] <jelly> aliceussr: setuid on executables serve a
specific purpose, if an application is designed to need them, it
probably needs them to continue working correctly
843[07:34:04] <aliceussr> jelly: The mta agent does not need
suid. It works great without it. And with it, he independently
changes the logging security settings.
844[07:34:50] <aliceussr> Ok! Deal with this security issue as
you wish. I warned you about the potential danger and compromise of
Debian-based systems with exim4 * packages installed
851[07:37:16] <jelly> aliceussr: this channel is an end-user
support channel; if you know of a way to reduce attack surface and
keep the functionality intact, please file a bug report
892[07:58:44] *** Quits: dez (uid92154@replaced-ip) (Quit: Connection closed for inactivity)
893[07:59:18] <themill> Except that you'd then have to have
to teach every mail client to talk to the network and there are lots
that don't, they just talk to /usr/sbin/sendmail
980[09:13:59] <jelly> > The standard service for printing
under Unix, Cups, is used in many Linux distributions and also under
Apple's MacOS. Its public development has stagnated since the
then main developer Michael Sweet left his employer Apple at the end
of last year . With the support of the OpenPrinting project, Sweet
has now started a "temporary fork" of Cups, which is
supposed to collect bug fixes and patches from distributions .
1116[10:55:56] <deadrom> hi. previously: banged together low
power pc from leftovers, at the core embeded Celeron J1900 board
with integrated graphics and boatloads of random crashes trying to
pull backups from an NFSv4 server. for shits and giggles tried via
CIFS. machine did not crash overnight, but cifs module had a world
class fit:
replaced-url
1117[10:56:39] <deadrom> couple minutes before that X reset the
i915 module. not sure if there is a correlation.
1122[10:59:54] <deadrom> I installed the intel microcode package
and created an xorg.conf explicitly stating intel video drivers as I
was told random system freezes in debian 10 might go away then. did
not: NFS transfer still crashed the machine. now unsure: cifs issue
or is that hardware simply shot? intel has a history of neglecting
old iGPU hardwware after a while and
1123[10:59:54] <deadrom> the J1900 iGPU has a couple years on its
back, so possibly a vector.
1182[12:02:58] <jelly> try a differeny kernel version for fun,
then
1183[12:05:42] <ws2k3> i have a question. im researching on a
project where i would like to setup a ssh server where users can
upload files. and im thinking of i should go with proftpd or i just
should just sshd. cause the setup with sshd requires chroot and
everything to be setup. my feeling is that proftpd handles this a
bit better. but i was curious what the opinion here would be on this
matter.
1187[12:06:32] <ksk> ws2k3: just use sshd with sftp-only (or
internal-sftp? kind of mixing them up all the time..) and chroot it.
Its really just 3 or 4 things you have to edit for that
1188[12:08:20] <jelly> ksk: proftpd can do sftp these days I
think
1212[12:17:01] <jelly> deadrom: whatever you got nearby. If
it's debian 10, there's a 5.something in buster-backports
1213[12:17:19] <jelly> remember to also grab firmware from there
1214[12:17:26] <jelly> !bdo kernel
1215[12:17:26] <dpkg> Newer kernels for Debian stable releases
are available from the <buster-backports> repository. After
modifying your sources.list, run «apt update». To
install the current backported kernel: «apt -t
buster-backports install linux-image-`uname -r|sed
's,[^-]*-[^-]*-,,'`». To list available backported
kernel image packages: «aptitude search
'?narrow(~nlinux-image,?origin(Debian Backports))'».
1216[12:18:12] <deadrom> jelly, if that's ok with X and DRI
to go with nvidia... (not installed for debugging right now but will
be in the final build... if this crate ever runs smooth)
1217[12:18:44] <jelly> oh, you have an nvidia card in there? that
can introduce all sorts of instability
1218[12:19:17] <jelly> disregards if you just plan to put it and
it's not there right now
1221[12:34:33] <dka> I am on an updated debian buster system, on
linux 5.8. I have issue with google chrome, I can use it minimized,
but as soon as I maximize or full screen it's windows, the
content is frozen, I can recover by minimnizing or exiting full
screen. This is painful for web development, watching video in full
screen, I am stuck, it used to work, no clue what changed. How can I
fix chrome ?
1222[12:37:10] <ksk> dka: since "google chrome" is not
part of Debian, eh, ask Google? ;)
1224[12:38:12] <ksk> there is chromium available, but I have
heard it does not have the best support (meaning versions are
outdated, stuff might not work for some users)
1225[12:38:12] <dka> this sounds to be an issue with xvideo-org
not chrome
1226[12:39:01] <ksk> dka: If you can reproduce the issue with
firefox-esr from Buster it could be, why not.
1249[12:53:04] <ratrace> willow_: I mean snapd is designed for,
default and integrated on ubuntu. it is not on debian and some snaps
don't work without further fiddling with config
1250[12:53:36] <willow_> i only use snaps for a few things:
eclipse, intellij. android-studio, libreoffice, etc. - versions are
always current and they are automatically updated.
1251[12:53:51] <ratrace> despite their claims, snaps really
aren't cross-platform and frankly not even on ubuntu (some
snaps fail to run on 16.04 for example)
1257[12:55:06] <willow_> and in my experience it works great.
only a couple minor sandbox-related issues may occur with some apps,
need to use --no-sandbox switch
1258[12:55:27] *** Joins: garo (~garo-ircc@replaced-ip)
1259[12:55:35] <ratrace> installing snaps doesn't integrate
well on debian is what I mean
1260[12:55:42] <dka> What is better for working on debian ?
Chromium or chrome ? If chrome is a fork of chromium, is it just for
keeping things up to date (ksk stated chromium as a poor support)
1263[12:56:26] <willow_> i don't notice any problems. seems
to integrate will AFAICT. as I mentioned, a couple minor
sandbox-related issues. for example atom editor must be launched
with atom --no-sandbox
1264[12:56:27] <dka> Chromium can full screen
1265[12:56:59] <ratrace> willow_: that's okay, you're
only using a few snaps that, luckily, worked out of the box :)
1283[12:59:46] <joepublic> chromium is in no way de-googlized
1284[12:59:59] <ratrace> dka: start chromium follow the
"sign up to google account!" nagware popups
1285[13:00:26] <dka> I didn't have it
1286[13:00:37] <willow_> i just tried the new Microsoft
"dev" .deb release of Edge. it's actually pretty
decent but not what i would use. it's also based on chromium
1289[13:00:48] <ratrace> garo: personally I think the whole
concept of LTS debian is moot. if someone really has such needs,
then RHEL and maybe CentOS are way better fits.
1294[13:02:52] <ratrace> willow_: except L for debian stable is
really only 2 years. ubuntu is 5, or 10 for paid support
1295[13:03:36] <willow_> ratrace: i guess that would be good for
the army or something where they would not upgrade their system for
like 20 years ;-)
1296[13:03:50] <ratrace> and if one needs actual long term
support (you have enterprisey software under "don't
touch!" conditions) then it's only in their best to pay up
for such support
1297[13:04:29] <willow_> missle-silo OS LOL
1298[13:04:49] <garo> I'm replacing someone that was
managing some servers, so i run "lsb_release -r"
1312[13:07:00] <ratrace> garo: what services is that running?
1313[13:07:18] <garo> I'll do it in steps anyway otherwise
i'll miss configs, data, ...
1314[13:07:48] <ratrace> garo: maybe it'd be way better to
install 10 in a VM, install services, configure, test them, and if
all is fine, just repeat the procedure for production
1315[13:07:55] <ratrace> automating that with something like
ansible will help quite a lot
1316[13:08:59] <Matrox> what is this about? in apt-get update: E:
Release file for
replaced-url
1317[13:09:30] <garo> there are too much services running:
i'll do just have to do in steps so that i don't miss
anything
1318[13:09:48] <ratrace> garo: i'd really recommend the VM
way
1320[13:10:08] <ratrace> unless you're okay with services
failing and general downtime until you catch up with major config
changes and backward incompatibilities
1321[13:10:12] <Matrox> should i somehow notify them? change
server in apt/sources , or just wait?
1322[13:10:28] * willow_ remebers my last major upgrade and dealing
with postgres pg_upgradecluster and a bunch of dovecot conf file
changes. not fun.
1323[13:10:31] <ratrace> Matrox: yes
1324[13:10:39] <garo> ok, i'll follow your suggestion. I
have backups anyway
1325[13:11:58] <ratrace> postgres upgrades will be a mess,
because you need both bins for every upgrade step. better dump and
restore
1326[13:12:56] <willow_> IIRC i had to upgrade minor versions
incrementally before it allowed to upgrade to the next major version
1327[13:13:21] <ratrace> yes because in postgres the minor
versions WERE major vesions until 10.x
1328[13:13:32] <willow_> basically it *wasn't smooth* :-/
1329[13:13:46] <jelly> Matrox: if it's a mirror in the
official lists, there are automated tests in place which will detect
their mirror's gone too old (that happens after 14 days of no
updates I think)
1330[13:14:06] <ratrace> now 10.x, 11.x, 12.x are whwat 9.7.x,
9.8.x and 9.9.x would've been
1331[13:14:10] <jelly> pick a different one in the meantime
1365[13:28:14] <jelly> garo: it's 2-4 hours per release
upgrade if it's not your first time doing that particular step.
So 6-12 hours, and then a week fixing things at the worst
1366[13:28:46] <jelly> doing a fresh install is always nice, but
one of the nice things about Debian is you don't _have to_
1367[13:29:01] *** debhelper sets mode: +l 1170
1368[13:29:44] <shtrb> jelly, assuming you CAN fix it
1369[13:30:15] <jelly> shtrb: I can always fix it.
1412[14:03:51] <deadrom> jelly the nv was not in until a moment
ago, which indeed introduces a new thing, it recovers every couple
seconds telling from Xorg.0.log. but otherwise I have been a hppy
nv/linux users for *years*
1413[14:05:51] <shtrb> jelly, do you have a clue about my
question about duplicting audio source ?
1428[14:11:59] <jelly> shtrb: saying "rtfm" seems
useless, so if that's the best I can say, I keep quiet. That
also means you don't have to ask me if I know something about
foo separately. If I knew, and I had the time, I'd have
commented.
1435[14:17:28] <shtrb> yes i know , every day I promise to myself
I would move back to it as soon as I could , and I fail (as it is
also not in debian as far as I remember )
1491[15:05:12] *** Quits: yonder (~yonder@replaced-ip) (Remote host closed the connection)
1492[15:05:52] *** Quits: nifker (~Thunderbi@replaced-ip) (Remote host closed the connection)
1493[15:05:58] <deadrom> aptitude -t buster-backports search
nvidia <- won't list the nvidia 450 driver from
buster-backports. how do I properly search the bp repos?
1509[15:25:00] <codedmart> I tried running `sudo btrfs fi du /
-s` and I get a few "failed to walk dir/file: ... permission
denied" messages and then `ERROR: cannot check space of
'/': Permission denied`?
1510[15:25:33] <LtL> deadrom: apt search nvidia wil show all
matches from backports, and all sources
1541[15:52:21] *** Quits: pvdp665564 (~pvdp@replaced-ip) (Remote host closed the connection)
1542[15:52:55] <LtL> deadrom: aptitude search outputs a wide
format, on a small screen i imagine multiple lines would be normal.
when i do aptitude search i get only one line entries
1545[15:54:38] *** Quits: hanasaki (~hanasaki@replaced-ip) (Remote host closed the connection)
1546[15:54:53] <deadrom> Brigo, when I use apt-cache, say
apt-cache search intel-microcode it shows me the package name, but
not which repos it was found in.
1587[16:29:36] <deadrom> only other board I have right now is
legacy, non uefi - does anything speak against using that?
1588[16:30:27] <n4dir> when using "apt-cache search"
pattern with special characters, man apt-get says it uses . ? and *
as regular expressions, while man apt-cache speaks of Posix regular
expressions. How does it look at a "+" in the search
pattern ?
1589[16:31:07] <n4dir> to put it different: how to search for a
pattern with a + character in it (and, if possible, why) ?
1606[16:40:41] <pathompong> Hello everybody, I have a little
problem about autoScorlling. I've searched and found that my
'Page Down' key is auto-pressing. The problem is like the
one which is demostrated in this video:
replaced-url
1608[16:41:49] <delial> @n4dir, you need 2 backslashes to get a
backslash to escape the + sign: apt-cache search \\+
1609[16:42:12] <pathompong> The video's solution is by
downloade Keytweak, which I guess not work in Debian. But if it
works then it is not by free software way. Does anybody have any
idea how to fix it?
1620[16:46:38] <pathompong> My laptop is acer E5-572G-78HY. And
this problem has been arising since this week. It's maybe
affected by some update/upgrade?
1623[16:48:07] <pathompong> I've tried to use backward
kernel but the problem didn't solved. Now my 'uname
-a' command show me my machine system is Linux aceraspiree15
4.19.0-10-amd64 #1 SMP Debian 4.19.132-1 (2020-07-24) x86_64
GNU/Linux
1625[16:49:11] <delial> n4dir: `apt-cache search` expects a
regex, and backslashes have meaning in bash so they need to be
escaped. Interestingly, `apt-get install` has special handling for
'+' and '-', so it first tries to see if you
want to add/remove the package before falling back to a posix regex
if the package name has a '.', '?', or
'*'
1626[16:49:20] *** Quits: ctrlbreak_MAD (~ctrlbreak@replaced-ip) (Remote host closed the connection)
1632[16:51:43] <pathompong> Or I need to do something with my
hardware?
1633[16:51:46] <delial> Hello, newb contributor here. I have a
manpage to contribute to resolve a bug, but I'm not sure the
process for contributing it. Should I just email it to the package
maintainer?
1634[16:53:42] <greycat> Attach it in your email to the bug.
1690[17:39:38] <zerocool> hey ya'll. im experiencing
something weird. i have a dell pe t440, stock nics (broadcom
netxtreme bcm5720), used the non-free firmware media for
installation (debian 10), fully updated.
1691[17:39:54] <zerocool> problem is, i cannot seem to vlan.
1698[17:42:11] <zerocool> in /etc/modules i have... 8021q,
vhost_net, loop, lp, rtc, bonding.
1699[17:42:37] <sney> even with the firmware I'd trust bnx2
about as far as I could throw the whole server. maybe check if
something is disabled in the bios? otherwise better gigabit nics are
not super expensive
1780[18:52:18] <genr8_> i'm getting a lot of segfaults in
libglib2.0-0
1781[18:53:03] <genr8_> xfce4-panel[1808]: segfault at 1e ip
00007d0eb94629ea sp 00007fff4d4d4580 error 4 in
libglib-2.0.so.0.5800.3[7d0eb9412000+7e000] xfdesktop[1812]:
segfault at 1e ip 00007c048154e9ea sp 00007ffec5179ba0 error 4 in
libglib-2.0.so.0.5800.3[7c04814fe000+7e000] gmain[2088]: segfault at
1e ip 00007bd3ac6339ea sp 00007bd3a9b94470 error 4 in
libglib-2.0.so.0.5800.3[7bd3ac5e3000+7e000] traps:
mate-system-log[1715] trap int3 ip:7ecd6d712c75 sp:7ffde935b270
1782[18:53:03] <genr8_> error:0 in
libglib-2.0.so.0.5800.3[7ecd6d6da000+7e000]
1784[18:53:40] <genr8_> the mate-system-log one gave me some
clues : ../../../../gtk/gtktextbtree.c:4041: byte index off the end
of the line Byte index 6176 is off the end of the line
1785[18:53:56] <taman> What do the errors at lines 10 and 11 mean
in <replaced-url
1786[18:54:08] <taman> s/does/do/
1787[18:54:26] <jelly> karlpinc: gcc docs are not free by DFSG
standards; you have to enable contrib and non-free sources to be
able to install gcc-doc
1788[18:54:45] <jelly> !non-free sources
1789[18:54:45] <dpkg> Edit /etc/apt/sources.list, ensure that the
two main Debian mirror lines end with "main contrib
non-free" rather than just "main", then
«apt-get update». But bear in mind that you'll be
installing <non-free> software. These may have onerous terms;
check the licenses. See also <sources.list>.
1790[18:54:49] <karlpinc> jelly: Should have thought of that.
Thanks.
1791[18:54:49] <sney> taman: looks like dhclient tried to send a
bigger packet than your link supported, but was able to negotiate
the right size afterwards
1796[18:55:56] <taman> sney, ok... I don't think I've
seen it before. This is is since I've configured wireguard. It
isn't stopping me using irc for example.
1797[18:56:03] <genr8_> yea that was my next question
1798[18:56:14] <genr8_> I should be able to replicate the crash
1799[18:56:15] <taman> Is that a Known Thing?
1800[18:57:47] <sney> if it's a wireguard interaction it
might be known somewhere, but wg isn't widespread enough yet
for stuff like that to be common knowledge
1817[19:13:26] <astronavt> e.g. in firefox, if i do `apt install
webext-ublock-origin-firefox`, will it be added to FF automatically,
as if i had added it through the firefox addons interface? or will i
need to add it manually somehow?
1818[19:14:19] <sney> afaik it will be added automatically. you
may need to restart the browser.
1841[19:37:12] <karlpinc> jelly: AFAICT there are no gcc-7 doc
packages.
replaced-url
1842[19:37:18] <dead> well this is a very odd problem...all of
the sudden when I try to take a screenshot with `import` command.
Awesome-window manager stops responding, until the screenshot is
taken
1843[19:37:45] <dead> I can't switch screens or anything, it
just freezes the display until I take screenshot...never had this
issue in my life wtf?
1946[21:11:49] <genr8_> sney, I think its actually a bug. I found
out our libgtk-3-0 version 3.24.5-1 is behind, and a fix that I
believe to be related was committed in 3.24.19
1947[21:12:39] <genr8_> and I got a backtrace thats at least 80%
complete. still missing some symbols but its close
1948[21:13:35] *** Quits: MrTrick (uid181961@replaced-ip) (Quit: Connection closed for inactivity)
1949[21:14:03] * karlpinc wonders if the gcc maintainers are paid 0 in
dollars, euro, or bitcoin -- you know there are tax implications
1950[21:14:54] <greycat> zorkmids
1951[21:14:57] <genr8_> the problem is theres just so many dang
packages to debug when dealing with this. I thought it was confined
to one libglib-2.0 but its not, its linked to gtk-3 and gdk-3 as
well :/
2089[23:35:32] <Mat63> When you start in advanced options, you
need the user to enter username and password. Since it is installed
in the magnetic disk it does this and I can not enter until now
2090[23:35:46] <hop> that's not active by default, though
2091[23:36:30] *** Quits: thiras_ (~thiras@replaced-ip) (Remote host closed the connection)
2092[23:36:52] *** Quits: chele (~chele@replaced-ip) (Remote host closed the connection)
2093[23:37:07] <Mat63> I apologize, I will check if I have a
password
2094[23:37:43] <c-c> I wonder if name regex omits ., - and _
2095[23:38:18] *** Quits: Mat63 (a8b5d136@replaced-ip) (Remote host closed the connection)
2110[23:46:11] <phogg> huh, it allows a username to start with a
dot
2111[23:46:28] <hop> phogg: is that a question?
2112[23:46:37] <phogg> that's an exclamation
2113[23:46:42] <hop> ok :)
2114[23:47:17] <greycat> I wouldn't do that, though. Some
BSD-based utilities use . as separator between username and
groupname (SystemV/POSIX uses : instead).
2115[23:47:21] <hop> it seems to allow names ending with a $ as
well
2116[23:47:27] <phogg> super biased of it to permit . and _ but
not - as the first character. In practice I would expect bugs if you
start a username with anything outside [a-zA-Z]
2120[23:47:59] <delial> hop: that's what I found weirdest,
why optionally one $ at the end?
2121[23:48:00] <phogg> greycat: I thought the BSDs followed along
with the switch from . to :
2122[23:48:01] <greycat> Yeah, you definitely want to start your
usernames with a letter if at all possible.
2123[23:48:03] <hop> greycat: chown, etc will recognize . as well
as :, so yes, bad idea
2124[23:48:23] <phogg> delial: maybe it's a windows compat
thing? Some of their special usernames end in $
2125[23:48:26] <greycat> phogg: I'm referring to historical
practice which may linger in some tools, who the hell knows, there
are how many thousands of pkgs...
2126[23:48:59] <phogg> greycat: fair enough; just because the
standard *officially* changed 20 years ago doesn't mean a
little maintained tool won't bite you.
2127[23:49:56] <phogg> I know I've seen code naively doing
'if starts with digit then user ID else user name', and
I'
2128[23:50:14] <phogg> I'm sure there must be some out there
that does "if starts with alphabetic then username else user
ID"