this is #debianan IRC-Channel at freenode
(freenode IRC service closed
2021-06-01)
0[00:00:04] <CaptainDusty> jadax: What is your explicit
requirement of Ubuntu?
1[00:00:05] <jadax> could I run Ubuntu on Debian as a
container somehow?
2[00:00:06] <rant> coruja: you know what filename it'd
read as a config these days? I tried /etc/X11/Xorg.conf and
/etc/X11/xorg.conf and doesn't seem to be reading it
3[00:00:17] <CaptainDusty> I've not really run into any
particulars that could not be achieved with Debian, tbh.
4[00:00:17] <jadax> well, I work with few other people on that
project
5[00:00:28] <jadax> and Ubuntu is the blessed OS for all of
the configuration stuff
6[00:00:35] * alkisg even boots a windows installation both in bare
metal and as VM; and it's also possible to boot a VM in bare
metal, as long as it's in raw disk mode
7[00:01:10] *** Quits: darkstalker (~wolfie@replaced-ip) (Quit: No Ping reply in 180 seconds.)
8[00:01:43] <coruja> rant, xorg.conf - and have you read what
i wrote above? ;)
9[00:01:44] <CaptainDusty> jadax: If your development
lifecycle is dependent on an OS purely because it's been
configured that way - it's a bit of another problem. But
there's also a point in which going against the grain is not
logical.
16[00:02:30] <jadax> I don't want to hack it all together
17[00:02:32] <rant> coruja: yes, I dont want to do that. I
already ran X -configure and it generated an intel based config
18[00:02:38] <jadax> it's easier to just stick to the same
thing
19[00:02:50] <CaptainDusty> jadax: My point exactly. ;)
20[00:02:53] <rant> I hate my config files unnecessarily split
into pieces its annoying
21[00:02:54] <coruja> rant, ok
22[00:02:57] <CaptainDusty> jadax: Been there, too. >.<
Fun times.
23[00:03:48] <CaptainDusty> jadax: I guess I'd still stick
with a VM. Should it become inconvenient; you could always dual boot
(which you can also run as a VM, as per alkisg)
24[00:04:07] <CaptainDusty> Depending on the VM configuration;
you shouldn't really have any issues.
25[00:05:06] <rant> \o/ now its using the intel driver.. and
doesn't seem to be spewing errors anymore.. idk why anyone
would configure it that way to not use DDX in the first place
26[00:05:42] <jadax> CaptainDusty if I do bridge networking
then my VM should get IP address straight from router's DHCP
and my host will have another DHCP obtained IP address, is that
right?
27[00:06:37] <CaptainDusty> jadax: Until recently; I had
dedicated network interfaces for my VM's. But, yeah; depending
on how you configure the bridge, and the interface you're
bridging - that won't be an issue.
30[00:07:30] <CaptainDusty> jadax: Dedicated interfaces? It was
overkill - but I picked up some pretty good intel interfaces for
next to nothing.
31[00:07:36] <CaptainDusty> It worked well for a virtual
router, however.
32[00:08:06] <rant> I do know one thing.. thats the first time
I really used this machine in a virtual console and that damn
terminal bell is ANNOYING AF idk if you can still run a script like
for vc in `seq 1 6`; do setterm -blength 0 >
"/dev/tty$vc"; done in a boot script or not :P
33[00:08:11] <CaptainDusty> Finding a compatible configuration
of hardware, and then virt configuration /software to enable true
hardware passthrough was... tedious.
34[00:09:12] <rant> it used to work years ago pre-systemd linux
2.6 era
37[00:11:31] *** Quits: emOne (~emOne@replaced-ip) (Remote host closed the connection)
38[00:11:40] <rant> I'd imagine this intel driver has to
work better than the modesetting driver.. idk wth they'd
disable it by default, I haven't heard of any issues with it..
71[00:29:00] *** Quits: arka500 (~arka500@replaced-ip) (Quit: My MacBook Air has gone to sleep. ZZZzzz…)
72[00:29:01] *** debhelper sets mode: +l 1539
73[00:29:31] <coruja> ,i devscripts
74[00:29:32] <judd> Package devscripts (devel, optional) in
buster/amd64: scripts to make the life of a Debian Package
maintainer easier. Version: 2.19.5; Size: 1021.0k; Installed: 2488k;
RFH: #800413
89[00:42:07] <Dark> Hi can someone help me out with this issue
error while loading shared libraries: libQt5OpenGL.so.5: cannot open
shared object file: No such file or directory
170[01:31:06] <jmcnaught> HelloShitty: I can just add a USB host
device to a VM in virt-manager. It results in XML like this in the
VM domain:
replaced-url
176[01:32:48] <jmcnaught> I also have a controller for USB:
<controller type='usb' index='0'
model='qemu-xhci' ports='15'> <address
type='pci' domain='0x0000' bus='0x02'
slot='0x00' function='0x0'/>
</controller>
177[01:33:42] *** Quits: stefanos82 (~stephen@replaced-ip) (Quit: Quitting for now...)
250[02:18:51] <de-facto> i dont want interactive script, i want
to script it myself
251[02:19:14] <maxrazer> Ok, I almost figured it out. If I login
to i3 or anything use lightdm then hexchat won't freeze on
connect and dunst messages work. So lightdm is starting something or
it installed something. What could it be?
252[02:19:27] <whislock> de-facto: Try reading the manual page
that I referred to.
254[02:20:08] <de-facto> its the wrong tool i want useradd not
adduser
255[02:20:10] <maxrazer> If I uninstall lightdm then I have
problems with hexchat freezing and also dunst messages don't
work. What could lightdm be doing?
256[02:20:34] <maxrazer> If I login just using startx and i3 in
my .xinitrc is when I have problems. Lightdm no problems.
257[02:21:01] <SerajewelKS> is there a quick way to report any
differences between metadata in two ext4 directory trees? including
which files are hard linked, acls, xattrs, etc.
258[02:21:27] <SerajewelKS> basically everything except inode
numbers (except insofar as they tell what's hard linked) and
the contents of the files themselves
259[02:21:30] <themill> de-facto: there is nothing wrong with
adduser.
267[02:23:12] <whislock> --system changes a lot of behavior. In
the future, don't assume that someone is wrong just because
they're not using the command you asked about initially.
268[02:23:19] <whislock> They may be giving you a MUCH better
way to do what you want to do.
284[02:33:13] <de-facto> hmm indeed useradd does it wrong:
"The default is to use the smallest ID value greater than or
equal to UID_MIN and greater than every other user." <--
this is a lie from its manpage
489[05:11:58] <dpkg> #debian-next is the channel for
testing/unstable support on the OFTC network (irc.oftc.net), *not*
on freenode. If you get "Cannot join #debian-next (Channel is
invite only)." it means you did not read it's on
irc.oftc.net.
490[05:12:14] *** Quits: astro1 (~quassel@replaced-ip) (Remote host closed the connection)
491[05:12:50] *** Quits: electro33 (uid613@replaced-ip) (Quit: Connection closed for inactivity)
668[07:46:04] <alkisg> When running "reboot" in
buster, I see "Resuming from hibernation"; it's
displayed by initramfs-tools' local-premount/resume; but I
think it's confusing to have it shown when hibernation
isn't used... isn't it?
669[07:46:55] *** Quits: hybrid (~hybrid@replaced-ip) (Remote host closed the connection)
690[08:01:17] <Tordek> hi; I have 2 screens but whenever I boot
xorg puts them in the wrong location: the one physically on the
right behaves as if it was on the left; do I need to change
xorg.conf or is there a tool I could use to make the screens be
where they should?
691[08:01:30] <Tordek> I'm using an AMD card if that's
relevant
693[08:02:27] <alkisg> Desktop environments have tools to
rearrange; it's usually a bit harder to apply those changes to
the login screen; so if you're using xorg, yeah a xorg.conf
with leftof/rightof is probably the best way
694[08:02:59] <Tordek> yeah, I can rearrange with systemsettings
but it doesn't apply on next boot
722[08:36:34] <pragomer> hi. how can I add a script folder (e.g.
/home/user/scripts) to my "path" (?) so that I can execute
these script from everywhere in the terminal?
723[08:37:32] <alkisg> pragomer: mkdir ~/bin, then relogging for
it to be automatically inserted to PATH
724[08:37:39] <alkisg> *re-login
725[08:37:48] <alkisg> (or open another shell etc)
864[09:33:42] *** Parts: COOurb (COOurb@replaced-ip) ("Once you know what it is you want to be true, instinct is
a very useful device for enabling you to know that it is")
995[10:34:40] <aZz7eCh> hi guys. LO6.3 is out ... I installed
6.2.* in addition to the default 5.* that comes with debian, by
downloading the 6.2.* deb package and dkpg'ing it ... all
installed fine when I did that. As I now want to do the same with
6.3, i'm a little concerned i will have conflicting system
integration if I dont first get rid of the 6.2 deb files I installed
the other day ... I aksed in #libreoffice but they told me to follow
thte wiki to uninstall it
996[10:34:40] <aZz7eCh> ... of course - i'm NOT goign to
follow the LO wiki to uninstall it as it will remove the OS default
LO install, along with my Gnome and every other default application
... i ONLY want to uninstall all the 6.2.* deb files from week or
two ago, so I can repeat the process with the new deb pacakage for
6.3 ...
997[10:35:10] *** Quits: hybrid (~hybrid@replaced-ip) (Remote host closed the connection)
1067[11:03:52] <CleymSpaze> i had only 1 single problem. i needed
to switch to a text console during boot and install
linux-firmware-nonfree so that X could be loaded
1111[11:15:46] *** Quits: sedrosken (~sedrosken@replaced-ip) (Quit: You shouldn't be seeing this!)
1112[11:15:54] <BCMM> i think this might actually be a bug, since
gpus from the major vendors tend to work well enough to show a UI,
just no acceleration, power management, etc.
1169[11:51:10] <brutser> for some unknown reason when i put
autologin with lightdm on a mate desktop, it takes ages for the
desktop to load, sometimes i have to move my mouse for a minute to
finally get to the desktop < it's debian 10 with
lightdm/mate
1170[11:51:13] *** Quits: Jo-Anna (uid57472@replaced-ip) (Quit: Connection closed for inactivity)
1177[11:54:37] <BCMM> centrix: hang on, was it wayland or x that
couldn't load?
1178[11:54:40] <brutser> alkisg: i only got 2 laptops here and
they are the same, i install debian 10 minimal, just basics
mate-core etc and lightdm, then i enter my user name in lightdm.conf
and reboot
1180[11:55:19] <alkisg> brutser: I mean, the long delay happens
ALL the times when you try it with autologin, and NOT at all when
you try with manual login?
1181[11:55:20] <centrix> BCMM, ??
1182[11:55:40] <BCMM> centrix: ah, sorry. meant to say CleymSpaze
(tab completion failure)
1183[11:55:41] <brutser> alkisg: yes, it's always with
autologin
1188[11:56:07] <BCMM> CleymSpaze: if this is Gnome, iirc the
wayland compositor has a hard requirement for accelerated opengl,
but i thought it was supposed to automatically fall back to Xorg if
that doesn't work
1189[11:56:43] <alkisg> brutser: when it delays, is your
background loaded already?
1191[11:56:52] <brutser> but i should not say always, i had one
install where it for some reason auto-logged-in normally to desktop,
but i am testing things, so a many new installs happened after,
after that i only managed to get that delayed
1192[11:57:21] <brutser> alkisg: no, it's a black/gray-ish
(not fully black, it's lit somewhat)
1195[11:58:44] <alkisg> I would edit
/usr/share/xsessions/mate.desktop, change Exec=mate-session with
Exec=xterm, then login, get an xterm at top left, then run
mate-session from that xterm to see when exactly it delays wrt the
stdout shown in the xterm
1196[11:59:09] <alkisg> Maybe it's related to the keyring
and entropy... maybe not...
1197[11:59:16] <person3473847> hey all -- are snaps safe to use?
what precautions should I take before installing one?
1229[12:11:24] <CleymSpaze> BCMM: good that you told me. i
checked and believe xorg was loaded as a failback and that failed
due to the missing firmware. is that possible?
1243[12:15:04] *** Quits: person3473847 (aa009cd4@replaced-ip) (Remote host closed the connection)
1244[12:15:11] <brutser> to be 100% clear, i only added [Seat:*]
autologin-user=bob and auto-user-timeout=0
1245[12:15:19] <brutser> nothing else
1246[12:16:10] *** Quits: vadPerevad (~VadPereva@replaced-ip) (Remote host closed the connection)
1247[12:16:20] <brutser> alkisg: should i change back >>
change Exec=mate-session with Exec=xterm << ?
1248[12:16:39] <alkisg> brutser: yes, put it back as it was,
although it's strange that it didn't work,
1249[12:16:53] <alkisg> brutser: if you have a password protected
keyring, it can't be automatically unlocked by PAM if you
autologin. So the delay might be a bug related to that.
1273[12:22:55] <jadax> but then on 1st or second screen I'm
getting stuck as installer complains about lack of some driver...
1274[12:23:18] <jadax> I don't want to discuss any win 10
related topics, just curious if I would ever want anything other
than dd to push that win10.iso image to USB stick
1275[12:23:47] <alkisg> "yes dd won't work with
win10.iso"; although this isn't at all debian related :)
1276[12:24:58] <jadax> unetbootin then?
1277[12:25:01] <brutser> alkisg: i have no seahorse installed and
never set any keyring password.. sorry for my lack of skill, but i
only added bob to the lightdm.conf, nothing more
1297[12:38:11] <alkisg> brutser: try to rename the
~/.local/share/keyrings directory to something else, then restart
and see if it logs in when it doesn't have a keyring
1298[12:38:20] <alkisg> *logs in quickly
1299[12:38:27] <alkisg> Remember to put it back after the test
1382[13:27:10] <jim> what could be happening (since python2 -and-
python3 are preinstalled, is that the maintainers are being asked to
port their code over
1386[13:29:00] *** Quits: hays (~quassel@replaced-ip) (Quit: No Ping reply in 180 seconds.)
1387[13:29:23] <mcmx> well, they effectively aren't being
asked. their inaction is being accomodated
1388[13:29:23] <BCMM> mcmx: /usr/bin/python is symlinked to
python2 for backwards-compatibility. it will probably stay that way
until python2 is actually removed from debian.
1389[13:29:45] <mcmx> but I'm looking for a a link, for the
wikipedia article
1390[13:30:06] <mcmx> I guess I will keep looking, thanks
1391[13:30:10] <BCMM> (stuff that requires python3 usually
specifically has python3 in the #!, stuff that requires python2
often just says "python")
1394[13:31:16] <kryl> hi, I'm looking for a squid
alternative? I don't need caching feature but I'm
interrested with acl / many user auth (sql or whatever) / logs
1395[13:31:29] <jim> BCMM, so is my statement accurate, some
maints still have python2 code in their maint scripts?
1396[13:32:00] <BCMM> jim: i don't know about actual debian
packages/scripts; my understanding was that it was for user scripts
with defective #!s
1398[13:34:00] <BCMM> jim: honestly not entirely sure what a
maint script is
1399[13:34:16] <jim> so then the question becomes... is python2
still an acceptable language/interpreter for those packages that use
python to set up the package?
1400[13:35:18] <jim> BCMM, ok, I mean the preinst, postinst,
prerm, postrm scripts
1405[13:37:44] *** Quits: gyroscope_excent (~gyro@replaced-ip) (Remote host closed the connection)
1406[13:37:55] <jim> (which means that before any part of a
package is installed (either unpack or configure phase), the
predepends must first be installed
1428[13:50:49] *** Quits: czesmir (~stefan@replaced-ip) (Quit: Lost terminal)
1429[13:51:21] <centrix> BCMM, I am aware of it. I am just
extending an already set solution that still depends on NTLM ... No
time, space or will to push Kerberos ....
1431[13:52:14] <BCMM> centrix: well, that bug is the explanation
for its removal from debian, anyway. you could probably still build
it yourself from the old Debian source package, though
1432[13:53:09] <centrix> BCMM, Yes. That is what I
"wanted" to hear :) I respect your warnings though. Thank
you.
1444[14:00:40] <jadax> and I can boot off that stick and
installation starts, but then during 'load installer components
from CD' it fails saying 'there was a problem reading data
from the CD-ROM...'
1445[14:02:44] <diogenes_> Hello guys, how do i remove all lines
that start with # and leave only the lines without #?
1477[14:22:56] <rizoongarbage> jadax: no, avoid swap it's
not 1990
1478[14:23:04] <bad_cat> wat
1479[14:23:11] <RoyK> jadax: you should always have swap
1480[14:23:19] <bad_cat> jadax: if you want to use hibernation,
enough to contain your runtime memory
1481[14:23:28] *** Quits: vadPerevad (~VadPereva@replaced-ip) (Remote host closed the connection)
1482[14:23:43] <jadax> I don't use hibernation
1483[14:24:02] <RoyK> rizoongarbage: you're wrong - linux
uses swap all the time to throw things not in use out in the swap to
allow for memory to be used actively for useful things like
caching/buffering
1516[14:29:14] <jadax> I don't need separate mounts for
/home etc.
1517[14:29:17] <jadax> I will keep everything in /
1518[14:29:19] <RoyK> ok
1519[14:29:53] <RoyK> but one PV, that is, the partition, then
one VG on that PV, and then smaller LVs for the rest, like the root
and swap and whatever you want to isolate
1520[14:30:16] <jadax> what's PV and VG?
1521[14:30:16] <RoyK> putting home on a separate lv/fs is a good
idea, btw, in case you want to reinstall
1555[14:37:26] <RoyK> I've described most of this on
replaced-url
1556[14:37:33] <RoyK> the manual way, though
1557[14:37:48] <RoyK> but hell - just start off with 4GB swap and
50GB root and take it from there
1558[14:38:42] <jadax> is it good idea to put swap on SSD?
1559[14:38:57] <RoyK> greycat: I know this nice, little cafe in
Reykjavík called Grái kötturinn, meaning
"[the] grey cat" :)
1560[14:39:00] <RoyK> it is
1561[14:39:51] <RoyK> jadax: you can adjust how much linux is
using swap later, by chainging vm.swappiness (default 60) to
something lower if you see it's swapping more than anticipated
1562[14:40:05] <RoyK> but generally, yes, it's a very good
idea to put swap on ssd
1563[14:40:18] <RoyK> with 32 gigs of RAM, it probably won't
swap much anyway
1564[14:40:43] <RoyK> so you could perhaps start off with even
1GB swap
1565[14:40:55] <RoyK> but you want *some* swap - linux is
designed for it
1566[14:41:49] <rizoongarbage> why would you want to reduce the
life of your ssd though
1570[14:43:19] <RoyK> rizoongarbage: today's SSDs can handle
writes quite well. Logging and caching will probably be *far* worse
for the SSD than swapping
1578[14:47:43] <humpled> is there any point to lvm for ordinary
users? just seems like an extra layer of complexity
1579[14:47:44] <greycat> The ideal is to have enough RAM to run
all your applications (web browsers are *huge* RAM hogs). The unused
RAM gets used to cache file system contents. Adding a small amount
of swap lets idle daemons get swapped out, which frees up some more
RAM for file caching.
1580[14:48:04] <RoyK> rant: that depends on the load - on some
fileservers, I've seen swap climbing skyhigh, just because of
the pressure of buffering and perhaps samba processes not being in
used being swapped out
1581[14:48:31] <RoyK> humpled: the performance is the same, but
it adds flexibility
1589[14:50:18] <RoyK> humpled: I know, but lvm is *far* more
flexible. Why do you think redhat/centos has been using it by
default for years? for fun?
1590[14:50:41] <RoyK> humpled: I use it for all my stuff, except
perhaps the pi computers
1591[14:50:41] <colo-work> out of spite!
1592[14:50:43] <colo-work> (I kid)
1593[14:50:45] <jadax> my freshly installed debian doesn't
boot; it hangs at 'sev command 0x4 timed out, disabling PSP;
SEV: failed to get status. Error 0x0'
1597[14:50:54] *** Quits: BlueByte (~walther@replaced-ip) (Quit: This computer has gone to sleep)
1598[14:51:00] <humpled> they add and take away drives on servers
i guess
1599[14:51:14] <davis> hello, I have a new pc. It has two video
ports on back. One is for nvidia and the other is for intel. I have
configured bios to use primary as pci express, the nvidia, I have
installed the nvidia drivers. glxinfo shows that in X windows, the
nvidia drivers are enabled. However, I can not get the built in
video to work simulatenously. Monitor settings app only shows one
monitor. Can I have two monitors
1600[14:51:20] <davis> working on this desktop?
1601[14:51:21] <RoyK> humpled: just learn it and you won't
go back
1633[15:03:50] <rant> I mean making sure kernel drivers are
loaded, examining the pci bus, writing an xorg config file, working
in the console with X turned off.. etc
1635[15:04:56] <rant> xrandr might work for manipulating outputs
of devices already configured, but as far as I know it can't
reconfigure the X server or kernel
1638[15:06:49] <davis> i'm looking at two pages on web. one
is a stackoverflow post and one is a blog. Both are out of date but
they mention using the Nividia X server SEttings app, the X server
display configuratin tab. On that tab there is a button for Detect
Displays but it always shows just the one display.
1639[15:07:07] <rant> davis: I'd start by going to tty1 and
stopping whatever display manager you may be using and running X
-configure as root to get yourself a config file to work with
1640[15:07:34] <davis> i wonder if the bios settings for primary
display "pcie express" is really disable chipset video.
1641[15:07:36] <rant> davis: you are not trying to use two
displays you are trying to use TWO DIFFERENT CARDS by TWO DIFFERENT
MANUFACTURERS
1642[15:08:08] <rant> its a whole different ballgame when you are
trying to use two cards with totally different drivers
1643[15:08:22] <rant> you can't use nvidia anything to make
an intel graphics card work
1644[15:08:29] <davis> indeed. i'm currently using the
nivida card. I'm trying to get x to recognize the chipset video
card.
1645[15:08:45] <rant> to do that it has to be told its there
1662[15:13:24] <rant> the primary card should have absolutely
nothing to do with any of this
1663[15:13:39] <davis> btw, greycat, i recognize3 your nick. you
have helped me a million times. I am happy to see you are still
helping out folks. tip of the hat to you.
1664[15:13:51] <rant> its whether or not the onboard video is
enabled, whats primary makes absolutely no difference
1665[15:14:05] <davis> rant, so you don't think that setting
is the reason lspci is not showing the chipset video?
1666[15:14:25] <davis> i would imagine for two displays, lscpi
should show two video cards.
1667[15:14:35] <rant> not unless you dont actually have two video
cards.. and you are actually using something like nvidia optimus
1668[15:14:38] <greycat> If you disabled it in the BIOS and
it's not showing up in lspci at all, then Linux won't be
able to use it.
1669[15:15:50] <rant> the difference between two video cards and
hybrid video is, hybrid video only has one set of outputs that can
be changed/remapped to different hardware.. rather than outputs hard
wired to specific devices
1670[15:16:11] <davis> greycat: agreed. the bios setting has a
config titled "primary vga" and it has two settings,
chipset or pci express. its currently set for pci express. but it
appears to disable the builtin
1671[15:16:53] <rant> which if you are only changing the primary
output device, and not disabling anything, then you probably dont
have anything to disable.. you only have one video device for all
practical purposes.. its a hybrid
1673[15:17:29] <greycat> or it *is* in lspci's output but he
didn't see it, either because he didn't read all the
lines, or because he grepped incorrectly
1679[15:18:55] <rant> davis: first you need to figure out if you
are dealing with hybrid graphics or not.. if its nvidia its called
optimus, and it'd be mentioned somewhere.. probably a sticker
on the machine among other places
1680[15:18:57] <davis> i asked my boss for a new pc. He gave me
some options and I simply purchased a bundle from new egg.
1681[15:20:00] <davis> im lookint at the box. it says intel b365
chipset
1691[15:24:01] <davis> i noticed it has two video ports on the
motherboard, one dvi and one hdmi. I have both connected and a third
monitor connected to dvi on nividia card.
1732[15:45:32] *** Joins: davis (~davis@replaced-ip)
1733[15:45:39] <davis> hello
1734[15:45:46] *** Quits: TomyWork (~TomyLobo@replaced-ip) (Remote host closed the connection)
1735[15:45:48] <abrotman> new? 8?
1736[15:46:51] <davis> ok, so I dis some tests. When the bios
setting is configured for chipset as primary display, lspci shows
the intel and nvidia on the buss. When the biso setting is pci
express, lspci only shows nvidia.
1739[15:47:40] <davis> currently I am in primary display is
chipset, so lscpi shows two video cards, but now X server fails.
I'm currently using the virtual console.
1740[15:48:31] <davis> in /etc/X11 there is not a xorg.conf file
or Xerver.conf
1775[15:57:06] <davis> nvidia-detect said it was compatible with
nvidia-driver, when I installed that driver via apt-get it enabled x
windows to work.
1776[15:57:24] <alkisg> davis: and the output of "kernel
drivers" (not in use) for intel? does it mention something like
i915 etc?
1777[15:57:25] <davis> in current bios setting with both cards on
bus, x windows does not work.
1778[15:57:40] <alkisg> You need to see 2 lines of "kernel
driver in use" for it to work
1779[15:58:13] <BCMM> davis: if you have an internet connection
on the machine in question, you can use a pastebin even if you
can't copy/paste
1780[15:58:15] <BCMM> !pastebinit
1781[15:58:16] <dpkg> A command-line tool to send data to a
<pastebin>. To paste e.g. your sources.list do "aptitude
install pastebinit; pastebinit /etc/apt/sources.list"; to paste
the output of a program do e.g. "dmesg 2>&1 |
pastebinit". See also <pastebinit config>,
<nopaste>.
1782[15:58:17] <davis> agreed. I wonder if I should apt-get
install something for the intel driver
1783[15:58:43] <davis> i did not know I c ould use a console
based pastebin
1784[15:59:06] <alkisg> (04:50:00 PM) alkisg: davis: what's
the output of: lspci -nn -k | grep -A 3 VGA | nc termbin.com 9999
==> this was console pastebin :)
1786[15:59:54] <davis> my problem is I don't have copy and
paste in console. so if I could pastebin the output, I would have to
type all the text flipping between virt terms
1787[16:00:07] <rant> davis: as root: (dmidecode;lspci
-nn;dmesg)|nc termbin.com 9999
1788[16:00:46] <BCMM> davis: you don't have to type all the
output!
1789[16:00:55] *** Quits: MenschZwoNull (~MenschZwo@replaced-ip) (Remote host closed the connection)
1790[16:01:08] <BCMM> davis: you've been given two solutions
where you can just pipe the output of a command to a website
automatically
1804[16:03:59] <alkisg> It's possible that you just need a
newer kernel...
1805[16:04:01] <davis> if i configure bios to have primary video
to be chipset, then lspci shows only nvidia and xwindows works.
1806[16:04:26] <alkisg> In both times you said "only
nvidia"
1807[16:04:41] <davis> i was hoping with three phyical video
connections on the computer, two on botherboard and one on nvidia, i
could get a dual monitor setup working
1808[16:04:58] <alkisg> You can get dual monitor with 1 nic
1867[16:16:46] <rant> [ 0.000000] Linux version 4.19.0-5-amd64
(debian-kernel@replaced-ip) (gcc version 8.3.0 (Debian 8.3.0-6)) #1 SMP Debian
4.19.37-5+deb10u1 (2019-07-19)
1868[16:16:55] <rant> we have your dmesg output, we don't
need your uname :D
1869[16:17:43] <rant> idk about anyone else, but knowing there
are few other debian options, I would perhaps recommend trying the
experimental kernel for argments sake
1870[16:17:46] *** Quits: superduperuser (~superdupe@replaced-ip) (Remote host closed the connection)
1871[16:18:07] <rant> thats a linux 5.0 which should be happy to
support your newer hardware I'd imagine
1872[16:18:14] <greycat> might be safer to build your own .deb
from kernel.org sources (5.2.x) than to use an experimental 5.0.x
package
1873[16:18:51] <rant> greycat: I see two problems with that.. 1)
after what you've seen you think this user can do that? and 2)
have you tried building 5.x? It requires a LOT of time and space
1874[16:19:18] <rant> back in the day with 2.6 and kernel-package
or such, yeah, I'd have also recommended that :D
1875[16:19:25] <davis> hmm. well on upside, I switched
minotors/cables and now I have two monitors mirrored on the dvi and
hdmi connector on motherboard working.
1876[16:19:44] <greycat> No, I haven't built a kernel since
2.6.
1878[16:20:13] <rant> me either except for trying 5.x when it
first build using a deb target.. and I couldnt free up enough space
on my 120GB ssd to complete the operation..
1879[16:20:36] <rant> I just gave up after freeing well over 20GB
and it still failing due to out of disk space
1880[16:20:46] <davis> so the new kernel would support xvideo
drivers for the motherboard chipset?
1881[16:21:04] <davis> as opposed to a console output which is
working now?
1882[16:21:05] <greycat> maybe ubuntu has a 5.x kernel
that's safer than experimental?
1883[16:21:22] <rant> davis: the xvideo and motherboard chipset
are both irrelevant here.. the graphics on on the CPU itself
1884[16:21:26] <karlpinc> davis: In the past I've solved
video issues with a newer kernel.
1885[16:21:50] <rant> and the CPU is newer than the kernel
you're running by far..
1886[16:22:03] <rant> so it'd be nothing short of a miracle
if it could somehow make that work
1890[16:22:22] <alkisg> I've successfully booted debian with
ubuntu kernels and the opposite; ubuntu has a "mainline
ppa" that has kernel 5.0 that does support this card, as show
in modules.alias,
1891[16:22:23] <davis> im wondering if xlog output is relevant
1892[16:22:38] <rant> I can already imagine what xorg log says
1893[16:22:39] <alkisg> but I don't think the user is
interested in multiseat; he just wants 2 monitors; so using jus
nvidia should be fine
1901[16:24:23] <davis> rant, perhaps this is not possible but I
wanted a dual display setting. one monitor via nvidia and one from
either of the two connectors on the motherboard
1902[16:24:55] <davis> right now, i have two monitors working but
they are connected via the two ports on the motherboard. one is dvi
and the other is hdmi.
1903[16:25:05] <karlpinc> davis: And we can assume that you want
both monitors to be part of the same login session, right?
1905[16:25:22] <davis> the third monitor connected via nvidia
card is not working.
1906[16:25:26] <davis> karlpinc: yes
1907[16:25:35] <rant> the dual outputs are on the motherboard for
the intel, not the nvidia pcie
1908[16:26:06] <davis> if I change the bios setting, then only
the nividia connected monitor work, but it also works with x
windows.
1909[16:26:16] <uio> Hi, I have an external monitor set
'above' my laptop. Both are on. Usually they have
different wallpapers. I changed the laptop one recently and after
logging out and back in, they both appear to have disappeared. Also,
I can't right-click the desktop for a menu like I could before.
What is wrong? Thanks.
1913[16:26:50] <alkisg> Ah, davis, you mean you got dual output
but only text mode?
1914[16:28:02] <davis> alkisg: yes, with bios setting primary
video, chipset, then i have a console mode output on the two
monitors connected via motherboard in mirrored mode, no xwindows and
the third monitor connected to nividia card is not working.
1916[16:28:20] <rant> alkisg: right, thats what is happening fb0:
EFI VGA frame buffer device
1917[16:28:55] <rant> I can't gather from the info I have
what the nvidia card has.. some of them of that model have only one
output, some have 4 outputs
1918[16:28:57] <alkisg> OK sounds like he needs a newer kernel in
any case then
1919[16:29:19] <alkisg> let's ask him; davis, does nvidia
have 2 outputs, e.g. dvi and hdmi?
1920[16:29:24] <rant> my guess is the nvidia card has multiple
outputs and can support many displays easily, but they probably lack
proper cables/adapters for them
1955[16:36:22] <han_> do anyone know how can i speed up the
booting debian?
1956[16:36:28] <greycat> That's why one person suggested the
5.0 kernel in *experimental* which is pretty old and pretty unsafe,
and another person mentioned building your own, or trying for one
from Ubuntu.
1957[16:36:29] <rant> davis: you only have two options, install
the 5.0 kernel from experimental or build your own kernel
1958[16:36:44] <Bushmills> han_: a faster computer may do
1959[16:37:01] <davis> rant: how would i try the experimental
path?
1960[16:37:10] <greycat> han_: remove stuff you don't need
1961[16:37:24] <uio> I kind found a solution, but not an
explanation: Ubuntu had some help on the issue:
replaced-url
1962[16:37:24] <Bushmills> boot from flash/SSD
1963[16:37:27] <davis> is that simply adding experimental in
place of stable as an additional source?
1964[16:37:28] <greycat> han_: or, is there a specific *problem*
you are having, that we could help you with?
1965[16:37:31] <rant> davis: you could just grab it from the
mirror and install it just for testing places
1968[16:37:59] *** Quits: Andocromn (~quassel@replaced-ip) (Remote host closed the connection)
1969[16:38:01] <greycat> !experimental
1970[16:38:01] <dpkg> experimental is the bleeding edge of Debian
Development. Packages here have been deemed
unfit/DANGEROUS/untrustworthy/etc for release by the maintainer
responsible for them. DO NOT INSTALL PACKAGES FROM EXPERIMENTAL
WITHOUT KNOWING EXACTLY WHY AND WHAT YOU ARE DOING. #debian does
_not_ support experimental. For an actual description, see section
4.6.4.3 of the Developer's Reference.
replaced-url
2008[16:44:19] <rant> thats even more surprising to me.. cause
last I checked all we had beyond 4.19 was a 5.0.0 in experimental..
to now have 5.2 in sid is quite a surprise
2009[16:44:31] <rant> I may even consider installing that on my
machine :D
2010[16:45:02] <uio> Should I file a bug report?
2011[16:45:18] <greycat> experimental doesn't mean
"new". it means "here's a thing that will
probably break, please tell me how it breaks exactly".
2014[16:45:57] <greycat> The 5.0 kernel was uploaded to
experimental for some specific reason, months and months ago.
2015[16:46:38] <greycat> There hasn't been a need to test a
suspected-broken kernel since then, so there isn't a newer one
in experimental, and that's normal.
2028[16:48:53] <greycat> as one would expect a month after a
stable release
2029[16:49:05] *** Quits: ZombyWoof (~ZombyWoof@replaced-ip) (Remote host closed the connection)
2030[16:50:19] <rant> yeah.. I installed guixsd in a VM last
night... it obliterates a lot of these debian type issues, but at
the same time makes me love them all the more.. this approach to
development is rather... crazy..
2031[16:50:53] <greycat> Debian doesn't even attempt to herd
the cats, most of the time.
2032[16:51:30] *** Quits: hays_ (a2faf302@replaced-ip) (Remote host closed the connection)
2033[16:52:01] <rant> the guix approach seems to be there
isn't anything a bag of parenthesis and a web of symlinks
can't fix..
2034[16:53:38] <tarzeau> alkisg: putting /usr/bin and /bin
together, they could've made /Applications right away ;)
2035[16:53:44] <rant> greycat: you ever need a laugh, I watched
this while installing this last night
replaced-url
2073[17:14:30] <Jellyg00se> Hi, when running apt-get update,
I'm given the " There is no public key available for the
following key IDs" warning, where can I find the specified ID
to remove it please?
2088[17:21:58] <Bushmills> by default, unless specified as
options, rsync doesn't remove any directories, regardless ether
empty or not. therefore uit's not refusing to, it just
won't do even attempt to
2091[17:22:48] *** Quits: mibo (~mibo@replaced-ip) (Remote host closed the connection)
2092[17:23:55] <greycat> One can only assume (because he
didn't bother to TELL US) that he used some kind of --delete
option, of which rsync has several. But we don't know which
one, or whether he's trying to remove stuff on the source or
dest, or why there are files remaining in the directory he wants to
remove, etc.
2093[17:23:58] *** Guest83712 is now known as argusbr
2100[17:26:51] *** Quits: zeSoup (~jsc@replaced-ip) (Remote host closed the connection)
2101[17:27:12] <greycat> !esp
2102[17:27:12] <dpkg> I don't have extra-sensory perception
(mind-reading powers). If you're asked a question or to run a
command and provide output, it's because information is needed
to help _you_ solve your problem. If you don't cooperate, or
don't clearly state your problem, we can't help, and will
move on to someone else. Understand?
2106[17:28:20] <j_wrok> Well, sorry. There are a lot of things
that may or may noy be involved. So I am actually using --delete and
--delete-excluded, but also --execlude and --filter. It is quite a
complicated rsync command, really, so it would be really, really
hard to provide context.
2107[17:28:40] <j_wrok> The reason this has to be so damn
complicated is that rsync does not parallelize.
2108[17:29:18] <greycat> there's an #rsync channel
2146[17:50:04] *** Joins: davis (~davis@replaced-ip)
2147[17:50:08] <davis> hello
2148[17:50:22] <Tom-_> !ask
2149[17:50:23] <dpkg> If you have a question, just ask! For
example: "I have a problem with ___; I'm running Debian
version ___. When I try to do ___ I get the following output ___. I
expected it to do ___." Don't ask if you can ask, if
anyone uses it, or pick one person to ask. We're all
volunteers; make it easy for us to help you. If you don't get
an answer try a few hours later or on debian-user@lists.debian.org.
See <smart questions><errors>.
2150[17:50:26] <Tom-_> :-)
2151[17:50:52] <davis> I installed the experimental 5.2 kernel,
and viola I now have X windows working with the bios setting for
primary video=chipset
2165[17:54:25] <greycat> The sid page on packages.debian.org has
"signed" not "unsigned". I don't know
exactly what that unsigned kernel is. rant apparently found it by
manually looking through the pool and picking one.
2171[17:55:57] *** Quits: Haudegen (~quassel@replaced-ip) (Quit: Bin weg.)
2172[17:56:17] <greycat> There are many linux-headers-* packages
in
replaced-url
2173[17:56:19] <rant> davis: if this works well for you, for the
time being you may want to setup pinning with a sid source.. you
need the pinning to avoid breaking your system
2174[17:56:43] <rant> I would imagine within a month or so
however it'll be available via backports
2175[17:56:43] <greycat> there shouldn't be any dependency
issues with an isolated kernel package (and headers?)
2176[17:56:57] <davis> it has a dependency problem.
2177[17:56:57] <rant> greycat: I was thinking more to be able to
maintain updates
2196[17:58:36] <rant> davis: use this one instead
replaced-url
2197[17:58:41] <greycat> He might want to go to the
packages.debian.org site and download the image/headers from there,
instead of manually from the pool as you did
2198[17:58:53] <rant> yeah, you could do that too :D
2199[17:58:55] <greycat> ... or, you could go fishing for him.
2200[17:59:10] * rant goes to have a look at the tracker
2205[18:00:43] <rant> I'd guesstimate by next month
it'll be able to be migrated to testing and probably available
for backports
2206[18:01:24] <rant> my only concern with regard to pinning sid
was to be able to have a source to obtain updates to mitigate any
security concerns with use of this
2207[18:01:37] <rant> the manual fetching was more of just seeing
if a newer kernel would resolve the issue
2208[18:02:18] <rant> its not really a long-term solution as it
could compromise the security of the system using a manually fetched
kernel from its first day in debian :P
2209[18:02:45] <greycat> If it were me, I think I would *not* add
a sid source, and just continue using the kernel, security bugs and
all, until a buster-backport is made.
2210[18:03:00] <rant> yes, that is another option
2211[18:03:13] <rant> either way I just would like to make the
point this is not a permanent solution
2214[18:04:13] <uio> Hi, I have an external monitor set
'above' my laptop. Both are on. Usually they have
different wallpapers. I changed the laptop one recently and after
logging out and back in, they both appear to have disappeared. Also,
I can't right-click the desktop for a menu like I could before.
What is wrong? Thanks.
2215[18:04:34] <rant> it was all just bad timing.. buying
hardware that is less than 6mo old within weeks of a Debian release
2218[18:05:17] <geek1011> uio: what is the output of xrandr?
2219[18:05:24] <rant> uio: sounds like whatever app that draws
your desktop is not running. We'd need a bit more information,
the release and the DE you're using
2220[18:05:51] <uio> Sorry, yes Stretch XFCE.
2221[18:05:58] <uio> I kind found a solution, but not an
explanation: Ubuntu had some help on the issue:
replaced-url
2222[18:06:11] <uio> It's in settings manager ->Settings
and Startup->Session tab... as 'on'
2223[18:06:15] <uio> Hi, I have an external monitor set
'above' my laptop. Both are on. Usually they have
different wallpapers. I changed the laptop one recently and after
logging out and back in, they both appear to have disappeared. Also,
I can't right-click the desktop for a menu like I could before.
What is wrong? Thanks.
2224[18:06:20] <davis> hmm. install of kernel headers, requires
common, requires, kbuild. did all of that then during kernel header
install it fails on nvidia.
2225[18:06:38] <uio> *couldn't, now it's
'working', but I don't think that should happen and
don't get it.
2226[18:06:39] <rant> uio: usually I'd run the app from a
terminal to answer that question.. it may produce some kind of error
running it manually from a terminal
2228[18:07:21] <uio> rant, No error output for
'xdesktop'.
2229[18:07:28] <davis> looking at log it shows that during build
of headers for nvidia it failed because of need to make config on
the kernel.
2230[18:07:32] <uio> But it just didn't want to run at
login.
2231[18:07:34] <rant> uio: personally I consider XFCE a bit
half-baked.. and more so being that you're on stretch.. its a
nice DE, and its mostly functional, but it is not without issues
2232[18:07:45] <uio> rant, What is better?
2233[18:07:56] <uio> I'm open to other DEs too, if they are
light.
2234[18:08:00] <rant> uio: I consider MATE and Cinnamon to be a
bit more mature and stable
2235[18:08:06] <uio> Hmm.
2236[18:08:11] <uio> But much heavier, no?
2237[18:08:17] <rant> ,de usage
2238[18:08:22] <rant> !de usage
2239[18:08:22] <dpkg> methinks de usage is The HDD/RAM usages of
the 7 Stretch DE on amd64 VirtualBoxes with 1GB RAM / 32GB HDD are
as follows as reported with only their terminals running df -Th and
free -h: GNOME 4.2G 726M, KDE 4.1G 604M, Cinnamon 3.7G 482M, MATE
3.1G 215M, LXQt 3.1G 184M, LXDE 3.0G 180M, XFCE 2.9G 226M
2245[18:10:07] <uio> Interesting. Lighter on ram than XFCE.
2246[18:10:09] <uio> Surprised.
2247[18:10:18] <geek1011> I actually prefer XFCE
2248[18:10:25] <uio> to MATE?
2249[18:10:31] <geek1011> Certainly.
2250[18:10:35] <uio> Why?
2251[18:10:35] <rant> uio: I typically use MATE if I use a full
DE, and I recently had switched to XFCE for a month or so.. and a
couple issues I had that annoyed me back to MATE was due to the fact
I was using various setups with my laptop, including a TV on the
display port and an extra monitor on the VGA and in MATE it
automatically restores the last known display AND sound output
settings where XFCE I had to keep doing it
2264[18:11:50] *** Quits: sz0 (uid110435@replaced-ip) (Quit: Connection closed for inactivity)
2265[18:11:50] <geek1011> rant: I use ARandr for that.
2266[18:12:00] <geek1011> ,i arandr
2267[18:12:01] <judd> Package arandr (x11, optional) in
buster/amd64: Simple visual front end for XRandR. Version: 0.1.9-2;
Size: 62.1k; Installed: 320k; Homepage:
replaced-url
2268[18:12:09] <greycat> If "low RAM use" is the main
goal, then consider not using any desktop environment at all. Just
use a regular, traditional window manager.
2269[18:12:46] <geek1011> ^ i3 - I boot in 3s (SSD) and use 180
mb RAM on boot
2270[18:12:51] <rant> geek1011: yes but did you manage to make it
so that it automatically restores the configuration on plugging in
displays? Configuring the layout the way you had it, changing the
audio output to HDMI..etc
2271[18:12:56] <uio> greycat, I'm too noobèsque for
that!
2272[18:13:21] <geek1011> rant: the layout, yes, but doesn't
pulseaudio restore the audio properly?
2273[18:13:41] <rant> I should make some updated benchmarks.. but
I been contemplating a better system to give a more detailed info of
the differences between the DEs
2276[18:14:15] <rant> geek1011: not in xfce it doesnt.. for
whatever reason
2277[18:14:50] <geek1011> rant: I guess. The funny thing is I
have more automated in i3 than I did in XFCE :D
2278[18:14:54] <rant> I also got annoyed by inconsistencies in
their panel applets and debated hacking them.. but then considered
it too much like work cause they're written in C which I think
is silly
2279[18:15:18] <rant> some applets can easily work in rows,
others cant for example..
2280[18:15:42] <geek1011> rant: the code is actually pretty easy
to work with once you get started, but I agree the panel
inconsistencies are annoying (especially the useless calendar)
2281[18:16:07] <rant> idk why everyone writes panel applets in
C.. I think applets should be written in something more flexible
like python or perl, or something interpreted where you can just
hack them at will without rebuilding anything
2282[18:16:13] *** Joins: Abdullah (~ak@replaced-ip)
2288[18:18:00] <rant> I was thinking though for updating my de
usage thing I may perform a more typical use benchmark and produce
something stupified like a Windows(tm) experience index :P
2290[18:18:19] <davis> that fails because of nvidia
2291[18:18:30] <davis> i have apt-get remove nvidia-driver
2292[18:18:37] <geek1011> I had a plank dock on the bottom, and
on the top panel: (whisker menu, vala-panel-appmenu,
expanded-spacer, window-buttons-iconified, folders, applets, time,
workspaces-4-rows)
2293[18:18:48] <davis> but it seems the soures are still around
so it tries to build them or something
2294[18:18:56] <rant> cause that factoid I made, only installs
them from tasksel, then I just boot a fresh install and run du -Th
and free -h in their terminals.. its not really a good benchmark of
typical usage like web browsing and such
2300[18:20:07] <rant> geek1011: yes well the point of making a de
usage factoid was to satisfy new users who ask silly questions and
want some sort of silly answer :P
2301[18:20:33] <geek1011> rant: I was like that 8 years ago when
I was still a newbie :P
2302[18:20:40] <rant> so far it has been very useful it seems to
new users wanting some sort of information about the differences
between the choices
2308[18:22:32] <rant> geek1011: they have a new i7-9700 cpu, and
installed a sid kernel to get support for the 9th gen video on the
CPU, but now their fancy nvidia pcie card is not working
2309[18:22:40] <geek1011> oh, and a quick rant about the non-free
office suites for linux: the deb packages are all messed up! the
pre/post-install scripts modify icons, override file associations,
remove config files, depend on nonexistent packages, and mess
everything up!
2319[18:23:49] <judd> [10de:1f07] is 'TU106 [GeForce RTX
2070 Rev. A]' from 'NVIDIA Corporation' with kernel
modules 'nouveau', 'snd-hda-intel' in stretch.
See also
replaced-url
2320[18:23:57] *** Quits: mase-tech (~mase-tech@replaced-ip) (Remote host closed the connection)
2322[18:24:11] <davis> hmm. so I apt-get remove nvidia-driver,
apt-get autoremove, I am booted with a 5.2 kernel, and xwindows
works using i910 chipset. But I can not install the 5.2 kernel
headers.
2323[18:24:43] <davis> it keeps saying consult nvidia log.
2329[18:26:51] <geek1011> davis: what does sudo modinfo
nvidia-drm say?
2330[18:27:35] <rant> I highly doubt he has any nvidia modules
for the running kernel as we manually fetched the sid kernel and
didnt have any headers installed for it
2331[18:28:07] <rant> the more important things would be what the
output of trying to install the headers says.. EXACTLY
2333[18:28:20] <rant> rather than running around uninstalling
nvidia packages and such
2334[18:28:37] <geek1011> rant: he said he needed to remove
nvidia-driver, so I'm wondering if anything got left behind
built for an old kernel
2335[18:28:56] *** Quits: spiros (~spiros@replaced-ip) (Remote host closed the connection)
2336[18:29:05] <davis> so this apt-get --purge autoremove
nvidia-driver followed by dpkg -i
linux-headers-5.2.0-2-amd64_5.2.7-1_amd64.deb says Consult
/var/lib/dkms/nvidia-current/418.74/build/make.log for more
information.
2337[18:29:07] <rant> he didn't /need/ to.. he thought he
needed to.. and modinfo isn't going to tell you that
2339[18:29:21] <rant> he's running the 5.2 kernel, its only
going to show you info about that kernel
2340[18:29:42] <geek1011> rant: unless something managed to load
that module
2341[18:29:53] <rant> @.@
2342[18:29:54] <geek1011> rant: but I admit I'm not too
familiar with the NVIDIA drivers
2343[18:29:54] <davis> and if yhou look at that log, it says the
kernel config is wrong for building of the nvidia. which makes
sense, since when I originally installed nvidia it was for kernel
4.x
2344[18:30:04] <davis> the module is not loaded
2345[18:30:14] <rant> there is no driver, can't be, it had
no headers to make one, and it can't load a 4.19 module on a
5.2 kernel
2346[18:30:15] <davis> i've booted from the 5.2 kernel
2347[18:30:51] <davis> rant: exactly. not sure why its trying to
build the headers for a module which is not installed nor loaded.
2381[18:36:29] <rant> if you WANTED for arguments sake to remove
them now, you should've just used nvidia*
2382[18:36:32] <davis> i am removing them so that the headers
will install
2383[18:36:51] <rant> the headers do no good without the drivers
to build :D
2384[18:36:52] <davis> now I assume I can reinstall them and
since the headers are there it can rebuild
2385[18:37:07] <geek1011> this is why I use aptitude for
dependency resolution.....and actually read the output to make sure
it is doing what I intend to do
2386[18:37:20] <davis> lol
2387[18:37:25] * rant assumes you're going to see the same problem
from a different angle
2388[18:37:35] <greycat> geek1011: in theory, maybe, but davis is
attempting to use sid packages on buster
2389[18:37:55] <rant> manually fetched packages no less
2390[18:38:07] <rant> that there are no policy/source for
2391[18:38:13] * geek1011 thinks it would be easier just to use sid
from the beginning
2392[18:38:22] <davis> i am doing an apt-get install
nvidia-driver now
2393[18:38:32] <rant> we don't recommend that here,
especially to new users
2394[18:38:35] <davis> geek1011: i was going to try unstable, but
this was suggested
2395[18:38:42] * geek1011 hasn't had any stability issues except
the recent initramfs bug with e2fsprogs and logread stuff
2396[18:38:59] <rant> we only suggested the newer kernel just to
see if a newer kernel would properly support their new hardware
2397[18:39:03] <davis> lol, now it says errors during
nvidia-driver and dkms
2401[18:39:47] <rant> you mean to tell me you still have to
resolve the ACTUAL problem? *gasps* the horror..
2402[18:39:50] <davis> i'm guessing the one via apt is not
setup to build for a 5.2 kernel
2403[18:40:03] <rant> man, you're just full of surprises
2404[18:40:11] <geek1011> davis: can I see nvidia-installer.log?
2405[18:40:16] <davis> i wonder if i am in the archlinux channel
2406[18:40:35] <rant> now you're telling me you think the
nvidia drivers for a branch that doesnt have a 5.2 kernel might not
work with a 5.2 kernel.. my mind is blowing here
2407[18:40:39] <davis> geek1011: yah, i don't know how to
get that though
2408[18:40:49] <geek1011> davis: try /var/log
2409[18:40:54] <rant> geek1011: you might be on to something :D
2410[18:41:04] <geek1011> davis: I think the issue might be at
least partly due to the drm symbol changes in linux 5.*
2439[18:45:23] <davis> cat
/var/lib/dkms/nvidia-current/418.74/build/make.log | nc termbin.com
9999 pasted. But, fwiw, this log file was noted from a different
command and not the apt-get install nvidia-driver which simply
failed and did not give a log file to examine.
2469[18:48:08] <geek1011> note these lines:
/var/lib/dkms/nvidia-current/418.74/build/nvidia-uvm/uvm8_tools.c:209:13:
error: conflicting types for ‘put_user_pages’ /
/usr/src/linux-headers-5.2.0-2-common/include/linux/mm.h:1080:6:
note: previous declaration of ‘put_user_pages’ was here
2497[18:54:18] <geek1011> I don't have time to go through
the source and tell you what to change (I need to go in a moment),
but maybe rant can help with that.
2504[18:57:00] <greycat> "this nvidia driver doesn't
compile with linux 5.2 because data types have changed" is way
beyond the scope of #debian. Maybe #nvidia can advise.
2505[18:57:05] <Nothing4You> is it possible to create a raid1
btrfs for /boot and raid10 btrfs for / and boot the system with that
setup?
2511[18:59:57] <jelly> greycat, typically solved by "wait
until upsteam fixes and debian maintainers put the updated one into
experimental or unstable or stable-backports"
2518[19:01:36] *** Quits: Insanity_ (uid179350@replaced-ip) (Quit: Connection closed for inactivity)
2519[19:04:04] <davis> hmm. its an either or situation. i either
switch back to 4.x kernel which supports the nvidia card. use that
and then I can at least have the nvidia working. Or, I can limp
along with the chipset video and hope in future 5.2 kernel becomes
available in stable.
2520[19:05:03] <greycat> It is pretty likely that a 5.2 kernel
will be in buster-backports long before bullseye becomes stable. The
issue for YOU right now is whether the nvidia driver will be ported
to the 5.2 (or later) kernel API.
2521[19:05:04] <davis> its probably a better bet, to revert back
to 4.x kernel. I wonder how to do that. apt-get install a 4.x
kernel?
2524[19:05:28] <greycat> You already HAVE the 4.x kernel unless
you did something moronic.
2525[19:05:32] <greycat> ls /boot
2526[19:05:50] <geek1011> Unless you already removed it, it
should still be there and you can just remove the 5.x one AFTER you
boot to the old one.
2527[19:05:50] <davis> geycat, moronic is my middle name! lol
2528[19:06:31] <greycat> After confirming that you didn't
remove it (by doing ls /boot), at the GRUB menu when you reboot,
choose "advanced options for Debian" and pick the kernel
you want.
2552[19:14:08] *** Joins: davis (~davis@replaced-ip)
2553[19:14:13] <geek1011> !ask
2554[19:14:13] <dpkg> If you have a question, just ask! For
example: "I have a problem with ___; I'm running Debian
version ___. When I try to do ___ I get the following output ___. I
expected it to do ___." Don't ask if you can ask, if
anyone uses it, or pick one person to ask. We're all
volunteers; make it easy for us to help you. If you don't get
an answer try a few hours later or on debian-user@lists.debian.org.
See <smart questions><errors>.
2555[19:14:31] <davis> hmm. ok, so changed bios setting to be pci
express for primary video.
2556[19:15:10] <geek1011> davis: are you booted to 4.19?
2557[19:15:49] <davis> it booted the kernel selection menu, i
selected 4.x. it fails to load some modules. I figure apt-get
install nvidia-driver would fix it. it says removing files, starts
building and then fails because its trying to build 5.2.
2558[19:15:57] <davis> yes, i am booted as 4.19
2559[19:16:09] <davis> uname -a says 4.19
2560[19:16:33] <davis> i assume it tried to build modules for
both kernel versions and it failed on 5.2 as before
2561[19:16:43] <geek1011> First, remove the 5.x kernel the way
you installed it.
2562[19:16:45] <davis> i assume i need to remove the 5.2 packages
2563[19:16:55] <geek1011> Yes
2564[19:17:27] <davis> i'm guessing there is a dpkg command
to list what is installed and then grep for 5.2 and note the ones
which i installed and remove them
2565[19:17:49] <geek1011> If you pastebin your dpkg -l, I can
give you the commands.
2651[20:11:00] <EdePopede> how good do the spam reports work on
bdo? especially what about reaction time? not really relevant for
spam, but malware is meh.
2683[20:24:51] <johnfg> Owner of /var/replaced-url
2684[20:24:51] <geek1011> !ask
2685[20:24:51] <dpkg> If you have a question, just ask! For
example: "I have a problem with ___; I'm running Debian
version ___. When I try to do ___ I get the following output ___. I
expected it to do ___." Don't ask if you can ask, if
anyone uses it, or pick one person to ask. We're all
volunteers; make it easy for us to help you. If you don't get
an answer try a few hours later or on debian-user@lists.debian.org.
See <smart questions><errors>.
2692[20:26:34] <johnfg> greycat: See, that's the good simple
answer I needed, as I was overthinking it.
2693[20:26:37] <johnfg> greycat: Thanks!
2694[20:26:40] <greycat> Also, *why* do you want it to be owned
by
replaced-url
2695[20:26:56] *** Quits: Ekchuan (~RandyMars@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
2696[20:27:28] <johnfg> greycat: Perhaps best practices?
Don't remember exactly when I did it, but something complained
about root.
2697[20:27:52] <johnfg> I'm running the typo3 cms, and not
sure if that's why. I know I can do whatever though.
2698[20:27:53] <greycat> If some web-application needs to write
to it, then that's a somewhat-valid reason. If it's
"because I want it to match the directory", that's
not.
2699[20:28:35] <greycat> if "typo3 cms" needs to write
to certain subdirectories under /var/replaced-url
2700[20:29:02] *** debhelper sets mode: +l 1541
2701[20:29:23] <stefandxm> Yes, keeping as few items writable
/owned by the serving application makes sense
2702[20:29:49] <stefandxm> Consider adding a
replaced-url
2704[20:30:19] <geek1011> there's also ACLs too for more
complex setups
2705[20:30:34] <pasiz> have anyone idea why my wifi is lagging on
lenovo X201 (debian testing). There is approx second or two pauses.
This happens only on 2.4 ghz band... If i switch on 5ghz, no hiccups
occurs.
2706[20:31:11] <Habbie> pasiz, in my house, all devices do that
on 2.4.. it may not be a debian problem
2707[20:31:19] <johnfg> greycat: So, is it required to
delete/remove a symbolic link before you recreate it to a different
source?
2724[20:36:23] <greycat> so you're *in* whatever directory
this symlink is in
2725[20:36:41] <johnfg> When I ran: sudo ln -fs
../typo3_src-9.5.8 typo3_src, it didn't change that existing
link, but created a link in typo3_src.
2726[20:36:48] <johnfg> greycat: Yes.
2727[20:37:39] <greycat> Well, if it doesn't work, then just
remove the symlink first. Symlinks to directories might have
different behavior than symlinks to files.
2756[20:57:33] <Nothing4You> does it work to change the rootfs
(/) which doesn't include /boot to a new partition if i do it
like this? 1. stop all non-essential services, 2. rsync all files
from the old rootfs over, 3. chroot to new root, 4. update-grub to
change rootfs for boot, 5. reboot?
2757[20:58:21] *** Quits: maggotbrain (~maggotbra@replaced-ip) (Remote host closed the connection)
2759[21:01:35] <Bushmills> there'll be file updated while
doing so. log files, for example, some others in /var. those
won't reflect the very latest state after rsync, and
you'll lose some at the end from those
2760[21:01:52] <Nothing4You> i don't care about lost logs
2827[21:21:54] <jhutchins_wk> Nothing4You: Sounds like a new
build is in order then.
2828[21:22:08] <trek00> Nothing4You: may be it's not
relevant to you, but check the contents of your initramfs, may be
you can spot something needed by your system: lsinitramfs
/initrd.img
2829[21:22:10] <Nothing4You> btw what's better performance
wise - mdadm raid0 for swap or multiple separate swapdevices?
2841[21:25:37] * geek1011 hates that memory-hogging giant
2842[21:25:42] <jhutchins_wk> Nothing4You: We did a group
research project in our LUG a few years ago and concluded that
anything less than $350 for the controller was a waste when it came
to performance, and anything less was questionable on reliability.
2843[21:26:34] <jhutchins_wk> geek1011: We use oomkiller on our
java boxes (not reeally on purpose).
2857[21:28:34] <jhutchins_wk> geek1011: I just configured one
with 30G yesterday.
2858[21:28:52] <rant> I noticed when I use juicessh with mosh it
connects no matter what, but when I mosh from Debian to my Debian
server it will not connect if there is an existing mosh-server like
earlier I drove off with it still connected to mosh, it lost wifi
and then I powered the machine down.. so the mosh-server was still
active but the client was dead
2859[21:28:55] <trek00> swap is the last line before an out of
memory event, that is rarely handled by applications
2860[21:29:02] *** debhelper sets mode: +l 1530
2861[21:29:24] <geek1011> jhutchkins_wk: this is my personal one,
as I don't have endless amounts of money :D
2863[21:29:42] <alkisg> Nothing4You: in ltsp, we have a method of
cloning a server / as a template for the clients; we're
basically using `mount --bind / /tmp/dir; cp -a /tmp/dir target`;
that bind mount ensures that no submounts are inherited like dev,
proc, gvfs or whatever, so it makes cp/rsync/mksquashfs very easy
2880[21:35:12] *** Quits: VadPerevad (~vadpereva@replaced-ip) (Remote host closed the connection)
2881[21:35:52] <CleymSpaze> 1000 schools can afford only 1
supervisor?
2882[21:36:53] <alkisg> CleymSpaze: i'm developing, writing
guides, solving the difficult issues; then each school has an IT
teacher with no linux expertise that is able to read the
documentation and maintain the easy parts himself
2883[21:37:15] <Nothing4You> looks like everything worked, thanks
2884[21:37:33] <alkisg> It works fine for everyone that way; no
stress at all
2888[21:39:49] <alkisg> jhutchins_wk: I'm recommending dual
windows + ltsp (linux) installations; ltsp needs only one
installation per school, so it's very easy to maintain; some
teachers go for diskless linux only though, without windows
2889[21:40:08] <alkisg> (some of the teachers do need windows, so
it's best to offer both and not make a fuss of it)
2902[21:49:01] <CleymSpaze> then you open windows and it changes
again?
2903[21:49:13] *** dingen is now known as dreamer
2904[21:50:09] <alkisg> We packaged around 20 GB of software in
.deb format using wine, java, flash etc, so most of the teachers
were fine without windows too (except for the air :D); but in some
cases they need some software, local (greek) or not... it's
usually best to give them all options; and then they usually end up
using linux most of the time anyway :)
2905[21:50:36] <alkisg> Some preferred ltsp because they had
ancient computer labs; but a couple of years later, when they got
new computers, they asked for linux again...
2906[21:50:43] <jhutchins_wk> I think it's cool that they
have access to linux; I think there's nothing wrong with
windows, it's a necessary work skill for most people.
2907[21:50:53] <Tordek> yeah, like that, commonly used in
「いいえ！待って、お兄ちゃん」
2960[22:29:44] <trek00> han_: it depends by the panel you are
using, they all have different configurations
2961[22:30:38] <han_> i didn't make any changes. ,
2962[22:30:49] <stefandxm> is there a channel for arm debian?
2963[22:31:06] <greycat> #debian-arm on OFTC
2964[22:31:17] <trek00> han_: probably you are using gnome
desktop, but i have no experience with that
2965[22:31:19] <stefandxm> oftc?
2966[22:31:24] <greycat> !oftc
2967[22:31:24] <dpkg> OFTC is the Open and Free Technology
Community, a support/collaboration service. They have an IRC
network: irc.oftc.net. You may (or may not) be connected to
OFTC's network.
replaced-url
2968[22:31:46] <stefandxm> cool! thank you :)
2969[22:32:08] *** Quits: han_ (~han@replaced-ip) (Remote host closed the connection)
3006[22:58:46] <rant> do they make a tool for building random
sources that wraps ./configure or something? so I dont sit here
running it over and over finding and satisfying deps one at a time?
3037[23:02:54] <trek00> rant: i remember configure creates cache
files for things it found, may be you can reuse these cache files
with multiple builds
3065[23:13:13] <FIUBOT> Whenever savelog runs, log.txt is
converted to log.txt.0, log.txt.0 to log.txt.1, and so on. My
problem is that the program's output, once savelog runs,
doesn't keep going to log.txt, it goes to log.txt.0
3066[23:13:32] <FIUBOT> If savelog is run again, the
program's output is sent to log.txt.1
3067[23:13:45] <pasiz> FIUBOT: it's like log rotation
3068[23:13:54] <pasiz> and what is the problem?
3069[23:14:06] <trek00> FIUBOT: the program should close and
reopen log.txt after receiving a SIGHUP, if not you should restart
the program after logrotation
3070[23:14:08] <pasiz> seems your savelog is working as expected
3099[23:17:46] <FIUBOT> Still the program output goes to the
log.txt.0 file
3100[23:18:09] <jhutchins> FIUBOT: What is the program?
3101[23:18:18] <pasiz> savelog
3102[23:18:51] <trek00> FIUBOT: when a log is begin rotated, it
is simply renamed to another name, but the program still having a
file handler opened to that file, whichever the filename is
3123[23:30:07] <jhutchins> FIUBOT: /user/sbin/logrotate - sbin is
not in regular users, you need to call it as root. This makes sense,
as it is operating on daemons and system files.
3124[23:31:06] <jhutchins> not in reagular users' $PATH
3125[23:31:40] <FIUBOT> $ ls -ltra | grep -i log
3126[23:31:45] *** Quits: xin_ (~textual@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
3156[23:40:46] <FIUBOT> Regardless, the only way that I can get
the output to the same log.txt is by restarting the program each
time I run savelog?
3157[23:41:25] <FIUBOT> jhutchins: $ man logrotate // No manual
entry for logrotate
3158[23:42:20] <LtL> FIUBOT: /usr/sbin/logrotate -f
/etc/logrotate.conf #force rotation of log files# if this is what
you want
3159[23:42:26] <jhutchins> FIUBOT: your pipe does not write to
the file by filename, it already has the actual file open and it
doesn't even know you've renamed it. In order to switch
files, the pipe has to close/end and a new one started.
3160[23:42:30] <jordanm> FIUBOT: for applications that don't
use SIGHUP or similar to tell it to reopen it's log files, you
can use the "copytruncate" option in logrotate
3161[23:42:52] <Habbie> but copytruncate is terrible
3171[23:44:38] <trek00> FIUBOT: yes, restarting the program each
time will fix it
3172[23:46:07] <jordanm> hmm packages.d.o doesn't list
package package priority
3173[23:46:35] <FIUBOT> Yes, my shell server uses Gentoo
3174[23:47:11] <jordanm> yeah, it's priority: important.
available on every debian install, but I guess it could be stripped
from debian docker containers