8[00:03:11] <pfred1> I never used more than 2.2 GB and that
was compiling something
9[00:03:52] <dob1> 5gigs?
10[00:04:04] <dob1> with free how much cache memory you have=?
11[00:05:21] <pfred1> now I am down to 285 MB I closed a bunch
of stuff
12[00:05:40] <pfred1> still have 4 xterms open though
13[00:06:01] <sney> this discussion made me realize that I
didn't have my chromium window with my inbox open, so
here's my real average memory use:
replaced-url
14[00:06:11] <sney> I'm not anywhere near the total so nbd
16[00:07:02] <pfred1> I have empty RAM slots on my motherboard
but they'll never get filled i just don't use RAM
17[00:07:28] <pfred1> 8 GB is more than enough for me
18[00:07:30] <dob1> I tried before: I closed/stopped all the
programs/services, no X session, and the memory was like 300-400 but
on htop every process was in 0-2% no more and on this vm I have 2gb
19[00:08:08] <pfred1> yeah with systemd my system uses more RAM
idle than it did with sysv
53[00:32:11] <hramrach> I have 8GB ram on a notebook and I only
get by because I am usingit as a terminal. I run web browser there
and everything else is offloaded elsewehre. Runinng a mail client as
well reliably kills it
54[00:32:48] *** Quits: alle_37 (~alle_37@replaced-ip) (Remote host closed the connection)
55[00:33:00] *** Quits: filohuhum (~filohuhum@replaced-ip) (Remote host closed the connection)
69[00:35:42] *** Quits: filohuhum (~filohuhum@replaced-ip) (Remote host closed the connection)
70[00:36:00] <hramrach> the top memory consumer is the browser
but when I run evince it takes a lot as well. Ironically for most
PDF files opening in browser gives way better experience than evince
94[00:46:02] <quadrathoch2> hramrach so how would I do that?
copy all pdfs I create per day to a usb device and then go over to
my printer (which doesn't have a usb plug?)
99[00:48:02] <hramrach> I don't know your printer. I deal
with a printer that accepts USB and SD cards (which is a PITA) and a
printer that can print PDFs sent over network witl lpr
102[00:49:31] *** Quits: tyzef (~tyzef@replaced-ip) (Remote host closed the connection)
103[00:49:37] <hramrach> but cups can adapt most printers to the
latter
104[00:49:39] <pfred1> I just got a new card reader
105[00:50:08] <quadrathoch2> hramrach so, again. I have to
"create" 50-100 pdfs a day, which when created all have
the same name (so I still need to rename them). then go over to a
printer which doesn't have a usb plug (let's assume it has
one). and then print them out. rather than just hit print in firefox
(or any browser) and print it. (I just need to file a bugreport to
106[00:50:08] <quadrathoch2> firefox at some point :/)
111[00:52:44] <hramrach> also when the PDF is created by a web
application and opens in the browser directly it makes sense to send
it to the printer from there but that's not how I get PDFs
113[00:55:59] <unixbsd> in bullseye there is a big problem with
libfltk1.3-dev package.
114[00:56:27] <unixbsd> in the libfltk1.3-dev bullseye, the libs
arent included where they should !! please fix as soon as possible.
115[00:56:52] <quadrathoch2> !debian-next
116[00:56:52] <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. See also
replaced-url
117[00:56:55] <quadrathoch2> unixbsd ^
118[00:58:49] <tyzef> hi! sometimes, someone may delete
importants .deb from from 'apt remove ... but someone
don't know he deleted important .deb for others softwares... is
there a way to check on our system if something is missing? regards
122[01:00:04] *** Ischwitch is now known as Ingvix
123[01:00:07] <smurfke> Hello, sometimes a network mount (NFS)
does not seem to mount after a reboot. It is a hit and miss. How can
I debug / find logs of the failed mount (specified via /etc/fstab)?
130[01:05:05] *** Quits: Jerrynicki (~niklas@replaced-ip) (Remote host closed the connection)
131[01:08:41] <smurfke> quadrathoch2: I remember someone telling
me to you 'journal' or something like that
132[01:08:54] <smurfke> a tool to read out logs I believe
133[01:08:55] <quadrathoch2> smurfke yup there is journalctl
134[01:08:57] <smurfke> thanks
135[01:10:24] <unixbsd> print server, 6-1 debian on raapberry pi
rp3ib: ok, it works. I have now a raspberry pi print server, using
the printer-driver-brlaser. It is a pure debian, if you need the
files, cmdline, fstab, ... and so on to get your print server.
please let me know.
136[01:12:38] <abrotman> smurfke: there's an option to
require networking before mount in the fstab I believe
138[01:16:43] <smurfke> abrotman: I believe I am using that
option: x-systemd.after=network-online.target
139[01:17:18] <smurfke> maybe there is something wrong with
vmware in combination with the latest debian. After upgrading debian
I started having this issue.
173[01:58:20] <smurfke> Is anyone familiar with this message
(it's a system unit): debian systemd[1]: local-fs.target: Found
ordering cycle on home-tom-storage.mount/start
174[01:58:38] <smurfke> I have no clue where to start digging
175[01:58:58] <smurfke> also no clue on what it actually means
176[01:59:21] <alexrelis[m]> I have a Realtek WiFi/Bluetooth
card and with firmware-realtek the WiFi works fine but for some
reason the Bluetooth will not activate. Here's what `dmesg`
gives me: `[154202.890014] bluetooth hci0: firmware: failed to load
rtl_bt/rtl8822cu_config.bin (-2)[154202.890018]`
177[01:59:21] <alexrelis[m]> `bluetooth hci0: Direct firmware
load for rtl_bt/rtl8822cu_config.bin failed with error -2`
200[02:38:29] <smurfke> What is the recommended way to reboot
debian? I always issued the command "reboot". Now I'm
reading that in debian 10 it should be "systemctl reboot"?
201[02:38:39] <ddsys> check first in /lib/firmware/rtl_bt to see
if you have it
202[02:38:53] *** Quits: tagomago (~tagomago@replaced-ip) (Remote host closed the connection)
209[02:47:43] *** Quits: queip (~queip@replaced-ip) (Remote host closed the connection)
210[02:48:08] <nevivurn> smurfke: they do essentially the same
thing. In fact on my debian10 system reboot is a symlink to
systemctl, which behaves differently depending on how it was called.
220[03:07:10] <smurfke> Question: How do I determine if i am
using NetworkManager or systemd-networkd? (In my /etc/fstab I see
the following options when mounting a NFS share:
,x-systemd.automount,x-systemd.after=network-online.target)
221[03:07:28] <smurfke> would that mean that I am using
systemd-networkd for the nfs mount?
222[03:09:51] <ndegruchy> smurfke, it means you're using
systemd to automount it
223[03:10:22] <smurfke> thank you
224[03:10:48] <ndegruchy> if you've not set up either, you
could just be using the interface scripts.
240[03:24:36] <shifty> found it: checkinstall. Allows you to
fake the "make install" component and leaves you with a
.deb that can be easily removed later.
301[04:25:27] <mehwork> i know i can `ls` to get filenames, and
`pwd` to get the current dir. But how can i combine the two, such
that ls will show me the full path to the files listed?
302[04:25:51] <mehwork> do i have to write a script to do it
303[04:26:39] <mehwork> i'm kind of surprised this
isn't a flag for ls, since it's really annoying to do it
in 2 steps, a lot of copy and pasting
304[04:26:57] <Poster> try "find ."
305[04:27:01] <Poster> see if that gives you the desired outptu
306[04:27:16] <dvs> Poster: I just tried that but that gets
everything below that directory
307[04:27:24] <mehwork> doesn't, that just says ./foo.txt
308[04:27:33] <mehwork> there might be a flag for find though
that does
309[04:27:41] <Poster> ls -R will list recursively
310[04:27:59] <Poster> but it doesn't show the full path
per se
311[04:28:05] <Poster> well, per line
312[04:28:13] <mehwork> i don't need recusive, just the
current dir's files absolute paths + name
313[04:28:51] <Poster> oh ok so for the directory of interest
"/foo", try "find /foo"
526[07:30:02] <sklav> true that the redirect is not working
527[07:30:20] <sklav> just noticed i had a bunch of =1 1 2 3
etc...
528[07:30:42] <sklav> seems the redirect might need escaping
529[07:31:26] <sponix> It is 00:30 here right now, at about
16:00 my time you might try to catch sney and greycat. They both
tend to know there shit really well, and could likely help better
530[07:31:57] <sklav> yeah i figured different timezone
531[07:32:05] <sklav> 1:30 am here
532[07:32:23] <sklav> was not expecting much but figured morning
in Europe :)
533[07:32:31] <sponix> > just doesn't seem like the char
that should be used to define a variable of this nature.
534[07:32:58] <sklav> i agree just it is listed in the wiki and
i figured ok i will try it
843[12:20:48] <jolt> I'm not sure what the psql usecase is,
but the organisation creates nice packages as well if you need psql
12 in buster for example.
848[12:25:26] <GNU\colossus> packaged by the same people that
package the "official" Debian packages, btw. pgdg is one
of the very few 3rd party repos I can recommend using.
850[12:27:45] <n4dir> i can add librazik repos to that list.
851[12:28:01] *** debhelper sets mode: +l 1186
852[12:28:26] <ratrace> can anyone confirm firefox-esr is
opening a new tab to gigster dot com, on FIRST start of a NEW
profile (~/.mozilla not existing at all, so it's that kind of
first start)?
853[12:28:34] <ratrace> this is MALWARE if so.....
854[12:29:11] <n4dir> would that be the very first start too? If
so i can install it now, which i planned anyway
855[12:30:00] <ratrace> I have created a /home/firefox user that
I plan to use with a separate, much stricter apparmor profile, for
anonymous random site browsing, where I use my main user for
browsing trusted sites where I have accounts.
856[12:30:21] <ratrace> so yes, this is a fresh new, never
before started in that user account, start. no ~/.mozilla at startup
857[12:30:28] <GNU\colossus> ratrace, if that is true, I think
you are overreacting. because the only alternative is
replaced-url
858[12:30:42] <ratrace> I know FF used to have sponsored LINKS
on the default homepage ... but THIS.... opening a new tab to some
random gigster dot com?
859[12:31:01] <ratrace> I am absolutely NOT overreacting. I want
a CHOICE to click or not. I'm fine with ADS, I'm fine with
sponsored LINKS that I have a CHOICE to click.
860[12:31:10] <ratrace> I am NOT FINE with it openig a tab to a
rando site without asking.
872[12:33:46] <n4dir> yeah, well, i wouldn't know. I
installed it and opened it, and it opens that usual tab and a second
one, but from mozilla itself. Default behaviour, iirc
873[12:34:07] <ratrace> the second one is mozilla's privacy
blah blah, yes. but the first one is gigster dot com
874[12:34:09] <n4dir> sorry, i would have loved to confirm it.
875[12:34:16] <ratrace> I'm in EU ... you?
876[12:34:19] <quadrathoch2> ratrace just tried it with a new
account. nope, but I am sitting here in germany
877[12:34:24] <n4dir> ratrace: yes, Germany
878[12:34:27] <ratrace> ... wth then
879[12:34:34] <iflema> suite?
880[12:34:39] <GNU\colossus> can you reproduce that behaviour at
will, ratrace?
881[12:34:44] <n4dir> i wouldn't know such internal
questions. just a casual user. Good luck
883[12:35:51] <ratrace> GNU\colossus: yes, I delete ~/.mozilla
and start it gain, and there it is alright
884[12:36:45] <quadrathoch2> i just assume, it's mozilla at
its best, as they always try new stuff out in the usa
885[12:38:12] *** Quits: alle_37 (~alle_37@replaced-ip) (Remote host closed the connection)
886[12:38:24] <mrkramps> hm, also germany and cannot confirm on
buster
887[12:38:26] <ratrace> okay, I banned gigster dot com via local
DNS resolver, removed ~/.mozilla, started, now it of course
can't resolve the name, but it also doesn't open the
second, mozilla's privacy blah blah page either. this is
definitely something in the code
913[12:51:17] <antto> hm, actually, i should've first asked
how to find out which device this thing is, i think it's a
composite USB device that has some HID thing and also a
"virtual serial port"
935[13:20:32] <ratrace> okay, false alarm. found the issue.
it's not firefox' doing, or mozilla, or cloudflare, or
anyone, but PEBKAC. tl;dr due to the way I'm calling firefox in
private mode from my main user into firefox specific user, using
passwordless sudo and passing params for URL ... it interpreted
--private-window "sudo" as URL and returned first site
that matched via "feeling lucky" search.
936[13:21:01] <ratrace> it's PEBKAC, false alarm,
y'all can go back to bed now ...... but I blame FF for doing
the random site feeling lucky search result when you give it a
malformed URL.
941[13:23:32] <ratrace> security is hard. all this for security
reasons. separate user, apparmor profile, isolation, sudoers,
passwordless script to allow only firefox to be run and nothing
else, plus a bug in param passing. :)
942[13:24:03] <ratrace> but holy hell, internet being https
mostly now makes debugging HARD. can't see a damn thing in
wireshark except alotta encrypted stuff going up and down
948[13:26:47] <grady> hi, i have very odd problem with md raid
array with fstab. if i try to add gid or uid settings to it, it
dosnt allow me to mount the mount point
949[13:27:07] <grady> ideas?
950[13:27:33] <ratrace> grady: start with pastebin'ing your
fstab line, please
974[13:44:43] <ratrace> a filesystem that supports DAC security
paradigms, multiple users, xattrs, ACLS, etc.... I doubt you can
mount it forced under a single UID or GID. it's one thing to do
that with FAT which has no concept of DAC, users, permissions.... so
it's always one user everywhere. and quite another for xfs or
ext4.
975[13:45:57] <ratrace> meanwhile, you can always chown the
mount point, or even use a shared user or group as the filesystem
root owner
976[13:47:48] *** Quits: magenbrot (~magenbrot@replaced-ip) (Remote host closed the connection)
980[13:52:08] <ratrace> a specific UID or GID that all your
systems would have ACLs to read and write to
981[13:52:53] <ratrace> if you mount that on multiple machines,
the simplest thing is to have tha owned by single UID eg 1000 which
is default first human user on most distros
983[13:53:38] <ratrace> the permission checks are actually
against numeric IDs, not names. so on your computer it could be bob,
on another alice, but both being UID=1000, if you mount a filesystem
that's chown'ed to UID=1000, both will see it as their own
984[13:54:00] <grady> where or how i do that :)
985[13:54:20] <ratrace> which part, I mentioned several options
986[13:55:47] <grady> chown dosnt work because every mounting
/booting it revers it back to root
987[13:56:14] <ratrace> you must chown _after_ mounting. chown
the filesystem, not the local directory you mount _to_
988[13:57:06] <ratrace> but you have to understand what
you're doing. you're chowing the root of a filesystem.
with -R you'll chown everything on it. make sure that's
what you want.
990[13:57:39] <grady> im not chown the /, just /mnt/storage :)
991[13:57:45] <ratrace> like I said, if you have a disk you
migrate around between several machines, if all the users are the
same UID, and they would be if they're first human user on most
distros, that gets UID=1000, then they'll all have normal
access to the filesystem
998[13:59:12] <ratrace> mount /dev/md/my_cute_array
/mnt/somewhere ; chown bob:bob /mnt/somewhere will chown the
"root of the filesystem" under /mnt/somewhere, without
descending into dirs. you need -R to descent. AGAIN: make sure
that's what you want.
999[14:02:05] *** Quits: hanasaki (~hanasaki@replaced-ip) (Remote host closed the connection)
1004[14:03:46] <ratrace> grady: then you need to make the files
and directories accessible to everyone? that's usually default,
644 for files, 755 for dirs.
1005[14:03:59] <ratrace> owners can modify, group and others only
read (and descend into directories)
1006[14:04:15] <grady> so xfs support those at least :)
1010[14:05:02] <ratrace> you can further limit this with shared
groups, like I said. eg you have a group called
"shared_files", then you chown all the files on
"storage" to that group, and make all your users members
of that group. might require a setgid flag on dirs so any file
created by anyone automatically inherits the group. you'll also
need g+w on them if you want all those users to write
1012[14:06:13] <ratrace> and also you can use ACLs with
setfacl(1) and friends to specifically allow specific users and
groups, specific kind of access to specific files, regardless of
main DAC properties.
1013[14:07:43] *** Quits: pvdp (~pvdp@replaced-ip) (Remote host closed the connection)
1014[14:07:50] <ratrace> you can also have wide open 777
directory that allows read and write by anyone. with a sane umask
that renders files 644, and dirs 755, you can have individual dirs
for individual users, where they can write into, but everyone can
read. the possibilities are multiple.
1032[14:15:19] <ratrace> you need executability on ALL files on
the storage?
1033[14:15:36] <grady> yep thats the point
1034[14:15:53] <ratrace> so they're all executables?
1035[14:16:05] <ratrace> (shell) scripts, C programs, pythons,
... ?
1036[14:16:46] <grady> no but it not make much a difference
1037[14:16:58] <phogg> why would you +x a file that you
can't execute?
1038[14:17:13] <phogg> that's somewhere between pointless
and dangerous
1039[14:17:14] <ratrace> grady: until one day it does and you
accidentally :)
1040[14:17:46] <ratrace> Principle of Least Privilege, should be
everyone's prime directive. but oh well... whatever floats your
boat, right.
1041[14:18:18] <grady> why it matter?
1042[14:18:45] <ratrace> grady: because $PATH issues and
executability don't need malware, viruses or haxx to break your
system
1043[14:19:06] <ratrace> case in point:
replaced-url
1044[14:19:19] <grady> well that storage arent on my $path :)
1045[14:19:34] <phogg> grady: Windows users suffer from "I
double clicked the file to open it, but it was actually malware and
it hurt me" but on Linux this doesn't happen because +x is
required to execute and you don't +x your JPEGs, or whatever.
1046[14:19:35] <ratrace> until one day it becomes on your path,
or something symlinks .... just .... don't +x stuff that
needn't be +x.
1051[14:20:23] <grady> ok this is again that nonsense talking,
like "blaablaa because of raid0" :D it really dosnt
matter.
1052[14:20:35] <dvs> heh
1053[14:20:42] <ratrace> phogg: the most fun part was when GNOME
got the vulnerability to start Windows Malware through the
thumbnailer in file manager, if WINE was also persent. fun fun fun.
1054[14:21:11] <phogg> ratrace: the most fun part is the way
.desktop files re-break this
1055[14:21:12] <ratrace> grady: whatever floats your boat, as
long as you're aware of what you're doing, I'm fine
with t hat.
1056[14:21:44] <ratrace> phogg: that too
1057[14:21:47] <phogg> grady: you are doing something unwise. You
have been warned. On your head be it.
1058[14:22:03] *** Quits: tyzef (~tyzef@replaced-ip) (Remote host closed the connection)
1061[14:23:03] *** Quits: ndegruchy (~nathan@replaced-ip) (Remote host closed the connection)
1062[14:24:34] <grady> i literally got this odd thing with on
fstab on xfs, something what i never got before. i got solution to
it, thank you for waking my brains again and i figured out how i can
manually fix the issue. but for this +x is bad :D is this new raid0
thing? :D
1063[14:25:25] <grady> i have been on linux over 20 years btw :D
1064[14:25:45] <phogg> appeal to authority?
1065[14:25:56] <ratrace> grady: I doubt that, judging from your
questions so far.
1066[14:26:41] <ratrace> unless you spent 20 years learning
nothing about the permission model (you did come here asking about
it and asking "how to do that" when I mentioned chowning),
in which case those 20 years have no value of mention.
1067[14:26:59] <ratrace> coulda been 20 minutes linux user, same
thing.
1068[14:27:15] <grady> everyone has these "low" moments
time to time :)
1070[14:27:25] <ratrace> also, you came here asking for FREE,
unpaid, volunteer help. you got it and you got sane advice and
warnings. if you don't respect that, if you're gonna omg
and roll your eyes and call it "nonsense".... well then.
:)
1071[14:27:41] <grady> haha
1072[14:27:45] <ratrace> tells a lot about who you really are ;)
1073[14:27:47] *** Quits: gonzo (~gonzo@replaced-ip) (Remote host closed the connection)
1074[14:28:00] <pizzaiolo> ratrace did that wine exploit have a
name? i want to read up on it
1075[14:28:06] <ratrace> pizzaiolo: yes, lemme look it up
1076[14:28:30] <phogg> I can't think of a reason why
.desktop files aren
1136[15:19:19] <dpkg> To set the time on Debian systems, use date
--set. The truly enlightened user, however, will install <ntp>
to have their clock set automatically on reboot and slewed to match
time servers. Also verify the system's UTC setting and set the
correct time zone; ask me about <utc>, <timezone>. see
systemd-timesyncd instead of installing ntp
replaced-url
1137[15:20:34] <cluelessperson> Hi there. I've found an area
with debian I believe I can assist with some tooling to make
mirroring apt repos easier.
1162[15:43:01] <cjoke> Im about to intall debian on my laptop. if
I got this correct , it is the same kernel in unstable as in
testing? Im going to use qtile wm, and have some gitrepos I need for
some audiowork. stripped down question is what would I gain using
unstable vs testing ?
1165[15:48:21] <cjoke> im so tired of ubuntu and snap dont suit
me :o) my thought where a minimal installation on laptop, where I
only intstall the gtk libraries needed for audioproduction. I think
it was 15 years since I installed through netinstall and base only
and from there gather what I need. I try to avoid bloated things
that sucks up energy.
1166[15:48:37] <Poster> in general you will be using newer
revisions of software, this can mean new features in the software,
but the drawback is that it may also contain bugs or other
undesirable behavior
1167[15:50:14] <Poster> the netinstall is a great way to start to
keep things minimal, that is the only way I install myself
1177[15:57:04] <nkuttler> cjoke: ^^ looks like currently they are
the same
1178[15:57:34] <nkuttler> !debian-next
1179[15:57:35] <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. See also
replaced-url
1180[15:58:02] <nkuttler> cjoke: also, support for testing/sid is
not here :) that being said, personally, i think testing is much
worse than sid because when it breaks it can stay broken for a long
time
1185[16:03:18] <Poster> I generally will use virtualization to
keep things tidy, if I need say a lot of development packages to
build or test something, I will spin up a VM to do so, when the
testing is done the VM can be turned off or deleted without adding
any more "bloat" to the base OS
1187[16:04:53] <cjoke> nkuttler: interessting point
there(testing/sid). I above average user so I guess sid would work
just fine for me then. kernel on stable is to old for my audiocard,
support for my card came with kernel5.4 I would have to compile
kernel anyways.
1188[16:06:14] *** Quits: gonzo (~gonzo@replaced-ip) (Remote host closed the connection)
1189[16:06:37] <towo`> no, you haven't to compile a kernel,
since you can have 5.9 over backports in stable
1198[16:13:15] <smurfke> Is it normal for debian to report:
"ifconfig not found"
1199[16:13:23] <smurfke> Is it normal for debian to report:
"ifconfig command not found"*
1200[16:13:32] <cjoke> towo`: yes, I should have that in mind
aswell.
1201[16:14:21] <cjoke> smurfke: not normal to me.
1202[16:16:09] <rk4> do you have /sbin/ip at least
1203[16:16:24] <cjoke> smurfke: try which ifconfig &&
echo $PATH
1204[16:17:25] <cjoke> smurfke: which ifconfig should show a path
to ifconfig and that path should also be in the echo $PATH
1205[16:18:18] <rk4> ifconfig comes from the apparently optional
net-tools package
1206[16:18:19] <Poster> if installed, it's likely
/sbin/ifconfig, though ifconfig itself has been deprecated in favor
of "ip". It is still available by installing the
"net-tools" pacakge
1218[16:33:14] <Ede|Popede> netcrash: with a big
"BUT!": depending if you want to keep configs or not,
apt-get remove gnome OR apt-get purge gnome. and now the BUT:
1220[16:34:25] <Ede|Popede> package "gnome" is
<Section: metapackages> - so it will take a lot with it.
depending if you want this or not you may want to install some
smaller packages it depends on before removing it.
1221[16:34:56] *** Quits: jim (~jim@replaced-ip) (Read error: Connection reset by peer)
1222[16:37:35] <Ede|Popede> heh, removing orage or xfce here
wouldn't make a lot of difference. a screenfull of packages
would disappeared with each of them
1259[17:08:10] <converge> I used snap running PopOS, and I had
some performance issues. will I experience the same running Snap on
Debian?
1260[17:08:37] <converge> *I tried
1261[17:09:17] <greycat> There isn't *nearly* enough
information in that question to be able to answer it.
1262[17:11:26] <converge> let me re-phase it, I need to install
an app that needs libindicator3-1, that inst available on unstable,
then I thought using Snap would solve my issue. but Im afraid it
will break something, or create other issues
1263[17:11:44] <converge> question is, should I try to build
libindicator3-1 instead of using Snap?
1269[17:14:11] <greycat> When you say "install an app",
you mean a black box precompiled binary from some disreputable
third-party commercial vendor?
1270[17:14:21] <Ede|Popede> converge: do you have enough ram to
install half a dozen packages on a live system? snapd seems less
invasive than i thought.
1272[17:14:50] <greycat> I have no freaking clue what a
"libindicator" is, but I did an apt-cache search
libindicator locally, and there is a libindicator3-7 available. So
I'm guessing libindicator3-1 is really *old*. Which leads to my
guess that you're installing a binary.
1282[17:19:51] <greycat> OK, let's see... according to
packages.d.o the source package for libindicator3-7 is called
simpmly "libindicator". According to snapshot.d.o the
versions of libindicator (source pkg) go from 0.3.8-1 to 0.5.0-4.
1283[17:19:52] *** Quits: lesless (~lessless@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
1284[17:20:05] <greycat> The binary packages from 0.3.8-1 include
"libindicator0" so that's too old.
1298[17:24:08] <converge> Im really afraid to messy my system
with snapd, but sounds like there is no other way
1299[17:24:34] <rby813> I just bought a laptop with a Ryzen 7 Pro
4750U, and I'm concerned about how to get Debian running X on
it. From my experience with Buster on a 3200G, I would need to use
backports to get a newer kernel and update the AMD drivers. If I was
to instead install Bullseye, would it work out of the box? Does
anyone know what kernel version Bullseye is using right now, and
what version of amdgpu is currently in the ports?
1300[17:25:26] <greycat> converge: that's why Ede|Popede
suggested testing it from a live instead of a real system, but even
if you tested on a real system, if the worst happens, you just
delete everything you installed for the test.
1304[17:26:38] <karlpinc> rby813: You could try a kernel from
backports
1305[17:26:46] <Ede|Popede> alternatives depending on the
machine: VM of some kind, extra mini partition with a test system.
plain dd for backup to reset it to a working state.
1306[17:26:48] <greycat> The URL you gave me for
"insomnia" doesn't work very well, but it links to
replaced-url
1307[17:26:57] <greycat> Note that Windows 7 and Ubuntu 14 are
both rather old.
1308[17:27:09] <greycat> I have no idea how old macOS 10.12+ is
1310[17:28:09] <cjoke> experimental = make a deb from
git(Master/currentversion), make it work and upload it to repos?
1311[17:28:26] <cjoke> in short terms :->
1312[17:28:53] <greycat> experimental is where a debian developer
puts a package that they don't believe is ready or appropriate
for unstable users to test en masse; it might have known issues, or
it might be a build intended only for a select audience to test
1313[17:29:21] <cjoke> greycat: ok, thanks. =)
1314[17:29:54] <rby813> greycat: Sierra (10.12) is 4 major
updates behind current MacOS
1320[17:32:13] <rby813> karlpinc: Wouldn't it be easier to
just install bullseye? The packages are freezing in less than a
month so I figured it was pretty close to stable at this point
1321[17:32:43] <rby813> greycat: initial release date was
9/20/16, so pretty close
1322[17:32:43] <greycat> Right before a freeze is the MOST
unstable point. It's when everyone throws their half-baked crap
in to get it in before the freeze.
1323[17:33:50] <rby813> greycat: I suppose that makes sense. Is
there any hand-holdy guide to updating kernel and amdgpu drivers
from backports? I'm still not super familiar with drivers on
the Linux kernel
1324[17:34:07] <greycat> Or... the second most unstable. The most
unstable would be right after a release, when everyone throws in the
crap they'd been holding back during the freeze.
1325[17:35:07] <greycat> rby813: installing a buster-backports
kernel is usually pretty safe. I don't know whether you'd
need a new X or whatever to go with it.
1332[17:38:25] *** Quits: gonzo (~gonzo@replaced-ip) (Remote host closed the connection)
1333[17:38:55] <rby813> greycat: Laptop came with Windows, so
I'm on that right now because I'm hesitant to reformat the
computer without being ready to deal with the drivers. Let me see if
I can find what the lspci would print
1335[17:40:14] <greycat> It's common to leave Windows in
place, just shrinking its partition(s), and dual-boot between that
and your Linux/BSD of choice.
1336[17:40:29] <greycat> You could also try a live thing.
1337[17:40:59] <greycat> I'd say "live CD" because
that's how they were originally branded, but people tell me CDs
are old-fashioned now.
1343[17:45:44] <greycat> hmm, kernel.org shows 5.9 as stable
[EOL], 5.10 as stable, but no longterm newer than 5.4 yet... wonder
what kernel bullseye is going to land on
1418[18:16:03] <Gerowen> If I wanted to resize an encrypted
partition, how would I do so; luksopen it and then use regular tools
like cfdisk to resize the ext4 partition that's inside it?
1419[18:18:40] <jelly> greycat, for us that don't have time
to read lkml, announcements for longterm are usually mentioned on
LWN
1427[18:24:51] <Gerowen> The partition is part of a RAID volume
that's currently re-shaping, so I can't do anything right
now, but when it finishes I'm gonna have about double the
capacity that I had when I first created the RAID and will need to
grow the partition to take advantage of it.
1430[18:26:00] <Gerowen> The reason for sdb1 instead of sdb is I
created partitions about 1GB less than the total capacity of the
drives, that way in the future if I have to replace or upgrade a
drive, and the capacity isn't exactly the same, it gives me
some wiggle room.
1431[18:27:16] <jelly> you could have avoided partitions
_completely_ and put md0 directly on whole disks
1432[18:27:45] <jelly> mdadm can be told not to use whole size
1433[18:28:02] *** Quits: lesless (~lessless@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
1436[18:28:35] *** Quits: Frenn_ (~Frenetick@replaced-ip) (Remote host closed the connection)
1437[18:28:54] <jelly> but that's irrelevant. You'd
resize all the members first, then the md array, then just close and
reoopen luks (a reboot would do that), then resize2fs
1438[18:29:36] <jelly> if I remember correctly, luks does not
keep track of how large the usable space is, at all
1443[18:30:30] <jelly> it will show junk data past the end of
your filesystem in the newly added space after reopening, but that
does not hurt
1444[18:30:52] *** Quits: Frenn_ (~Frenetick@replaced-ip) (Remote host closed the connection)
1445[18:31:21] <Gerowen> jelly: If I have to, I've got a
backup I can restore from and re-make the group, it's currently
converting from a two disk RAID 1 to RAID 5 with a third disk, hence
the reshape, but if possible, I'm gonna try to resize the
existing partition to take advantage of the new capacity.
1449[18:32:08] <Gerowen> And honestly, with RAID 5, the
encryption isn't strictly necessary since, from what I
understand, you can't get anything useful from a single disk of
a RAID 5 group. The encryption was for if I RMA'd a drive from
the RAID 1 group where each drive has everything.
1450[18:32:57] <jelly> you get slices of actual data. The slices
can be large enough to compromise yourself or your customers
1481[19:02:41] <alexrelis[m]> I closed the lid to put my laptop
in sleep mode and when I wake up to open the lid I'm greeted
with a black screen. I'm using KDE Plasma What gives?
1484[19:04:29] <zutat> alexrelis[m]: maybe there's some hope
left, then.
1485[19:04:56] <alexrelis[m]> I had a similar problem yesterday
when I woke the laptop up from sleep and the login bar wouldn't
show up no matter what I do.
1486[19:05:53] <alexrelis[m]> How do I make it so that I never
have the problem again?
1490[19:09:49] <kurufu> where can i find historical discussion on
why PATH does not include /sbin. It doesnt seem to in bash's
changelog but im not sure what package would generally own the users
path.
1492[19:11:41] <greycat> It's got nothing to do with bash
per se. It's the traditional behavior of login(1) and the
/etc/profile as installed by the system administrator and/or Unix
vendor.
1494[19:12:00] <greycat> Unless you're talking about su,
which is a whole different issue.
1495[19:12:01] <greycat> !buster su
1496[19:12:01] <dpkg> In buster, su no longer overrides PATH by
default, requiring that you use "su -" or "su
-l" for login shells (which is not really a new thing at
all...). See
replaced-url
1497[19:12:36] <kurufu> im not really interested in any
particular maintainers opinions on paths and FHS, im just interested
in when the decision was made.
1498[19:12:54] <kurufu> consider it an archeological expedition
1499[19:12:55] <greycat> Uh... some time around 1985?
1500[19:13:11] <kurufu> it seems more like issues cropped up
around 2000.
1508[19:14:51] <greycat> Are you talking about "I used su
and then things still didn't work because my PATH wasn't
right"? That issue would definitely have hit other Linuxes long
before it hit Debian.
1509[19:14:58] <kurufu> No im not talking about su
1510[19:15:38] <kurufu> and unfortunately it comes up much more
often in google searches than ancient mail list archives about
people bickering about FHS.
1511[19:15:59] <greycat> what issue?
1512[19:16:09] <kurufu> about people having issues with paths in
su
1562[19:42:31] <greycat> You can apparently retrieve historic
package version numbers from UDD. I just asked it: select package,
version from public.packages_summary where package like
'dpkg%' and release = 'squeeze';
1620[20:26:05] <Gerowen> jelly: One more question if you
don't mind. Since I'm "increasing" the size of
the partition, is that something that will take a considerable
amount of time, or since it's not necessary to move any data
around, will it be basically instant? I'll be doing it over SSH
and am wondering if I need to send the resize process to the
background or not in case I forget and close my laptop lid or
something.
1632[20:28:11] <Gerowen> greycat: That's a good point
actually. It's my personal/home server, and have never bothered
doing anything other than just using regular old gnome-terminal.
1707[21:31:33] <jolt> Casper26: great question! I guess it would
be possible to do something with a script in /etc/network/if-up.d
1708[21:32:04] <jolt> Casper26: the script there is called with
some env vars, so it should be possible to utilize that in a script
that flips the interface.
1710[21:32:36] <jolt> There is probably some more fancy software
that can handle it thought. I'm more of a server person myself,
so I use the script way in my daily stuff.
1766[22:22:18] <rby813> greycat: I think I might have
disconnected there for a few hours. Did I miss anything, there? I
see something about 5.9 and 5.10, which both support my GPU (I
think)
1768[22:23:03] <rby813> On a related note, is there a decent
guide somewhere on how to deal with backports? Never done it before,
I ended up finding a custom debian image specifically for the other
Ryzen APU I had
1770[22:24:04] <greycat> I don't know much about video
support in post-stable kernels.
1771[22:24:30] <greycat> This was the line I spotted and pasted
here earlier from your URL: 11:42 greycat> 06:00.0 VGA compatible
controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Renoir
[1002:1636] (rev d1)
1772[22:24:43] <greycat> If anyone can help rby813 with this
card, I'm sure they'd be happy.
1773[22:25:19] <rby813> Alright I'll stop hassling you now.
Thanks for the help so far, though!
1774[22:25:23] <Mister00X> rby813: Are you sure you need 5.10
1775[22:25:44] <greycat> Oh, also I should point out that they
didn't get this from their OWN system. They found it on the
internet from someone else's system that they're assuming
is similar to theirs.
1779[22:26:19] <rby813> and the lspci is from the same laptop
model, just one that ships with more RAM and SSD space
1780[22:26:32] <rby813> Lenovo Thinkpad L15 (AMD)
1781[22:26:35] *** Quits: gonzo (~gonzo@replaced-ip) (Remote host closed the connection)
1782[22:27:04] <greycat> A thing you'll learn as you work
with computers is that model numbers don't mean much.
Manufacturers swap out components without changing the labels, *all*
the time.
1783[22:27:12] <Mister00X> rby813: are you able to install buster
? Or does the installer work
1788[22:28:22] <rby813> I haven't attempted an install on
the laptop yet (running Pro 4750U), but I did a successful install
with a non-functional X install on a 3200G
1790[22:28:44] <Mister00X> rby813: but the command line interface
works?
1791[22:28:54] <hramrach> you can always do something like take
out disk, connect disk to some other machine over usb, run installer
in qemu to install on the disk, add any kernels you like, put back
disk into laptop
1796[22:30:22] <another> rby813: buster-backports has 5.9
1797[22:30:37] <hramrach> and experimental 5.10
1798[22:30:41] <rby813> Mister00X: Yes, it installs successfully.
It's the display driver that has issues. If I recall correctly,
it also has trouble booting, I may have needed to use nomodeset?
This was a while ago. Unfortunately I do not have any alternate
drives at this point in time, so I'm unable to do a hard drive
swap. This is my first m.2 drive
1809[22:33:03] <rby813> The unsuccessful installs were with the
official image, and the nonfree image. When I finally got it
working, I had found a custom image that had the newer kernel and
non-free drivers baked in
1815[22:34:09] <rby813> Ah, rolling release would have no issue
with newer hardware, but I have issues with rolling release haha
1816[22:34:27] <another> did you try a backports kernel?
1817[22:34:56] <Mister00X> rby813: if debian fails to boot after
the install I can't help you. However If the tty works I can
explain to you how to get a kernel from backports
1818[22:36:47] <oxek> I get this error in wget
replaced-url
1819[22:37:46] <rby813> Mister00X: I found a user who had the
same issue I did, and it's refreshing my memory. Debian
successfully installs and boots to a blinking underscore that does
not respond to keyboard input
1820[22:38:31] <rby813> Is my best bet to roll to Ubuntu until
packages in bullseye stabilize?
1821[22:38:33] <Mister00X> rby813: Have you tried alt+F2 or
ctrl+alt+F2 to get to a tty
1829[22:40:32] <rby813> Would it be possible to just recompile a
debian iso with a newer kernel version for the install, since I have
a working rig right now?
1830[22:40:40] <Mister00X> but I'm not a chroot expert
1838[22:43:32] <Mister00X> Just a question, would it be possible
to install a backports kernel in the installer. Like switch to a tty
install it from bpo and then complete the install?
1840[22:44:47] <rby813> That sure would be useful if it was
possible
1841[22:45:14] <Mister00X> It is possible on debian live to
install packages so maybe it's possible in the installer too.
I'm just unsure about the kernel part
1850[22:48:16] <hid3> Greetings everyone. Slightly offtopic
question but maybe anyone runs HP ProLiant / Microserver Gen8 and
have upgraded iLo firmware? Will it take entire machine for a reboot
or will iLo just restart itself without influencing the OS?
1851[22:48:17] <jelly> oxek, wget on debian 10, built with
gnutls, fails. testssl.sh says that web page does TLSv1.2 and strong
ciphersuites, so who knows why gnutls fails
1859[22:49:19] <ArsenArsen> is this news? tyrants have been
trying it for a while
1860[22:49:38] <spacedust> and khazahstan with its root certs
1861[22:49:47] <spacedust> no not news since its q3 news :)
1862[22:50:01] <jelly> oxek, well, wget could be built with
openssl as well, but isn't.
1863[22:50:03] <rby813> quadrathoch2: Any decent guides for that?
1864[22:50:25] <spacedust> isnt there an option for certs in wget
?
1865[22:50:32] <spacedust> like not check certs
1866[22:50:44] <Mister00X> quadrathoch2: nice. Is it also
possible to install the new kernel in the installer, like into the
ram and after that the installer installs the newer kernel? Do you
now any about that?
1867[22:50:48] <spacedust> sure its skipping "securety"
1868[22:51:04] <jelly> hid3, dunno about microservers, but on
proliants flashing iLO reboots just iLO and does not affect any
running OS at all
1869[22:51:05] <spacedust> Mister00X: you talking about kexec ?
1870[22:51:17] <spacedust> ah no
1871[22:51:38] <quadrathoch2> rby813 eh, i guess it's like
this (i know it s a screenshot by kali)
replaced-url
1872[22:51:42] <quadrathoch2> exit, exit :)
1873[22:51:49] *** Quits: wsky (uid453465@replaced-ip) (Quit: Connection closed for inactivity)
1874[22:51:51] <hid3> jelly, microserver is basically a proliant.
Thanks for sharing your experience!
1875[22:51:55] <oxek> jelly: so I guess the solution to my
problem is to use curl instead of wget, unless I want to recompile
wget to use openssl instead of gnutls?
1876[22:52:07] <spacedust> oxek: try gentoo :P
1877[22:52:20] <jelly> oxek, please onsider filing a bug report
against wget
1878[22:52:22] <spacedust> ohh im in #debian :D misstook it for
#linux sorry :D didnt ment to
1882[22:54:22] <jelly> spacedust, its nice to be enthusiastic
about technology, but if you're trying to help, please focus on
things you actually have hands-on experience with, and maybe not
jump to "use different distro" right away
1883[22:55:16] <spacedust> oxek: ahaaaa you also in #linux thats
why i got distracted :)
1884[22:55:18] <oxek> jelly: I checked the bugpage and looks like
it has not been reported yet
1928[23:23:02] <simpledat> I just installed debian, LVM with
encryption for the entire harddrive. Is it possible to shrink the
size of the drive and use it for something else?
1934[23:25:26] <bolt> simpledat: gparted may be able to help you,
though I haven't resized an encrypted LVM with it before. make
sure you have backups.
1948[23:31:27] *** Quits: seanrobert (~seanrober@replaced-ip) (Remote host closed the connection)
1949[23:31:31] <simpledat> oxek: Why?
1950[23:31:44] <simpledat> I used to su to root and run openvpn
as usally.
1951[23:32:00] <quadrathoch2> !buster su
1952[23:32:00] <dpkg> In buster, su no longer overrides PATH by
default, requiring that you use "su -" or "su
-l" for login shells (which is not really a new thing at
all...). See
replaced-url
1953[23:32:12] <oxek> simpledat: because it is installed in
/sbin/ which is not in a non-root users PATH by default
1954[23:32:57] *** Quits: lesless (~lessless@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
1955[23:33:45] *** Quits: notobvious (b029ae57@replaced-ip) (Remote host closed the connection)
1973[23:39:06] <pasiz> simpledat: is debian hard and long task to
install?
1974[23:39:14] <simpledat> I would rather not
1975[23:39:22] <Mister00X> simpledat: If you need to resize the
root partition use a live image
1976[23:39:38] <pasiz> then boot your debian live and resize from
there
1977[23:40:02] <pasiz> you may get it almost as fast as
installing new system in text mode
1978[23:40:17] <simpledat> Sorry, I dont mean to bothering you.
But what would it make for difference if I did it from GParted or
Windows Disk manager?
1979[23:40:43] <Mister00X> simpledat: Windows does not know how
to open LUKS container
1980[23:41:18] <Mister00X> and has troubles with LVM too
1981[23:41:20] <oxek> or LVM, or any underlying linux filesystem
1982[23:42:12] <simpledat> Hmh, I thought it wouln't matter
because its unused data anyway.
1983[23:42:47] <pasiz> with that knowledge, i would avoid
encrypting my file systems...
1986[23:45:40] <pasiz> simpledat: if you have encrypted volume,
you cannot know where your data are and how much there is it.
That's kind of the reason to encrypt whole volume...
1987[23:45:56] <simpledat> pasiz: I see, thank you.
2011[23:58:47] <cluelessperson> quadrathoch2, My intention is to
fully mirror all the main debian repos. I'm looking at and
playing with the archvsync docker implementation, but haven't
gotten it to work yet.