74[01:24:51] <dbristow> What is the package name of llvm/clang
so that it will install everything I need to set up LLVM, I'd
like the packaged version, not a custom repo
75[01:25:34] <dbristow> I'd like to install all the distro
packages for whichever languages LLVM currently supports, presumably
at least C and C++
76[01:25:43] <dbristow> And their associated libraries and such
77[01:26:27] *** Quits: uvolmer (~uvolmer@replaced-ip) (Remote host closed the connection)
78[01:26:33] <sney> hmm, looks like it's probably
clang-tools
79[01:27:12] * dvs is glad it's not crang-tools
80[01:27:23] <sney> ,i clang-tools
81[01:27:38] <sney> ah, no bot.
82[01:28:11] <H-var> I have an issue with debian - it keeps on
permanently locking up at random times. How can I track down the
source of the problem?
83[01:29:19] <dbristow> Hmmm, I had to install clang as well.
90[01:33:57] <sney> H-var: try to make it less random. test
things and see if you can reproduce crashes or similar symptoms. and
maybe stick something like rsyslog on a rpi so you can collect your
logs on a separate machine.
107[01:43:14] <superstar64> can someone explain to me why `.deb`
files are `.ar` files that contain `.tar` files? why have archive
files inside archive files?
115[01:44:32] <H-var> but when I leave it with the apps running
constantly, and taking up to 50% cpu at all times, that's when
it locks up
116[01:44:39] <H-var> probably an overheating issue?
117[01:44:42] <Tenkawa> superstar64: there ya go
118[01:44:48] <dvs> H-var: but memtest is doing something so it
isn't idling
119[01:45:07] <Tenkawa> H-var: what kind of machine?
120[01:45:20] <superstar64> that doesn't explain why
121[01:45:38] <superstar64> why not just a single tar file for a
package?
122[01:45:44] <H-var> mini pc. It's beelink with an atom
cpu inside
123[01:46:00] <Tenkawa> superstar64: because they felt like it?
124[01:46:16] <zykotick9> superstar64: <channeling crystal
ball> note a tar file isn't "compressed" its just
a collection of files. ar is a compression "I believe".
125[01:46:46] <Tenkawa> zykotick9: I just pasted a article with
the entire structure
126[01:47:13] <Tenkawa> H-var: out of curiosity... whats your
sysctl -a | grep swappiness ?
127[01:48:02] <Tenkawa> if its pushing the vm off to swap a lot
that could push the machine fairly harrd
128[01:48:06] <Tenkawa> er hard
129[01:48:09] <H-var> integrated sd card as a storage device.
I've actually already created a return request on amazon
because of this issue. From my experience, it usually overheating,
and bad cpu management that causes lockups
130[01:48:17] <Tenkawa> even if its not doing too much
131[01:48:54] <superstar64> so it because they wanted to allow
for an uncompressed control archive and a compressed data archive?
132[01:49:00] <superstar64> or other combinations?
133[01:49:56] *** CommunistWolf is now known as CapitalistWolf
134[01:49:57] <Tenkawa> superstar64: the explaination on the
manpage here gives more context:
136[01:50:08] *** CapitalistWolf is now known as lupine
137[01:50:24] *** lupine is now known as CommunistWolf
138[01:50:25] <H-var> tenkawa: command not found
139[01:50:48] <Tenkawa> H-var: you got to use sudo
140[01:50:55] <superstar64> "File sizes are limited to 10
ASCII decimal digits, allowing for up to approximately 9536.74 MiB
member files"
141[01:50:55] <superstar64> this seems future proof
142[01:51:13] <H-var> tenkawa: 60
143[01:51:59] <Tenkawa> H-var: try editing /etc/sysctl.conf and
setting that down to like 10 and run sudo sysctl -p
144[01:52:14] <Tenkawa> lte it run for a little while and see if
the tepm changes
145[01:52:22] <Tenkawa> er temp
146[01:53:01] <H-var> I think I cannot recommend intel to anyone
anymore, especially atom cpu. I will change it with ryzen 5 3550H.
147[01:53:05] <H-var> mini pc
148[01:53:21] <Tenkawa> we have to tweak stuff like this a lot
for non-x86 soc's
149[01:53:51] <Tenkawa> to tune every ounce we can get out of
them and reduce heat
150[01:54:44] <H-var> you can order a ryzen 5 3550H mini pc for
240$ from China, + I'll take 16gb 2400mhz ddr4 sodimm ram for
40$, and 1tb nvme drive for 87$.
151[01:54:52] <Tenkawa> my last mini-itx was probably 10+ years
ago now
152[01:55:31] <Tenkawa> my current non-mac x86 is a 12 core hp
laptop i7
153[01:55:34] <Tenkawa> :)
154[01:56:14] <Tenkawa> er 12 thread
155[01:56:20] <H-var> tenkawa I think it's the nano-itx
platform
156[01:56:33] <Tenkawa> oh nano.. those were fun
157[01:56:39] <H-var> the mini pc format is usually within 15cm
box limits
158[01:56:59] <Tenkawa> pain in the back to work on though
161[01:58:47] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
162[01:59:11] <H-var> don't stay with your back bent for
too long
163[02:02:16] <Tenkawa> lol
164[02:02:21] <H-var> ryzen 5 3550H cpu is actually insane -
it's using so little power, it can be placed into a mini pc
computer (35w), and it gives out almost 91% of power of an intel
core i7 8700k desktop cpu (95w), and if you want to save on power
even more, you can take ryzen 5 3500U which is only 15w, and reaches
up to 80% of the mentioned intel cpu
172[02:04:23] <Tenkawa> (yes its a gaming laptop originally..
although I bought it for development)
173[02:04:50] <growly> hi everyone, i'm having a problem
with the installer. I get "no kernel modules found" when
trying to install and subsequently my network card driver is not
loaded, so I cannot continue installation. the module i need is
e1000e (common). i've tried a testing (daily release) netinst
image, a testing (bullseye alpha 3) netinst image, a testing (daily
release) DVD image, and the 10.7 netinst image. all do the
174[02:04:56] <growly> same thing
175[02:06:50] <sney> growly: how did you create the usb? many
common tools (rufus, etcher) will mangle the installer and can
create the problem you describe
176[02:07:37] <Tenkawa> sney: this sounds like proprietary
firmware blob missing
179[02:07:53] <sney> Tenkawa: "no kernel modules
found" is not the firmware error, no.
180[02:07:54] <slop> hi all <3
181[02:08:05] <Tenkawa> good point
182[02:08:19] <Tenkawa> man I'm getting sleepy sloppy early
183[02:08:21] <sney> 'modinfo e1000e' shows no
Firmware headers eitehr
184[02:08:50] <Tenkawa> i can tell a storm is coming in
185[02:08:56] <Tenkawa> its hitting me hard
186[02:09:56] <Tenkawa> surprised he got ths far with a broke
write though..
187[02:10:39] <sney> it's not like an unsynced write, the
rufus tool unpacks the iso and recreates it. so all the pieces are
there but not in the way d-i expects.
188[02:11:10] <Tenkawa> yeah and if it wasnt synced possibly
before eject too
189[02:11:39] <Tenkawa> thats why I always just dd and be done
191[02:12:43] <sney> install guide says to use cp, but that and
dd and cat all work fine, win32diskimager is the rec for those
coming from windows. or in short, don't use a tool that
modifies the image.
195[02:13:26] *** Quits: sparky4 (~sparky4@replaced-ip) (Remote host closed the connection)
196[02:13:28] <sney> I think the installer team or the debian-cd
team should probably consider distributing the images as *.img
instead of *.iso, so this wouldn't be an issue
197[02:13:42] <Tenkawa> indeed
198[02:13:45] <sney> why? it just copies the bits to the flash
drive. no fuss. cp debian-foo.iso /dev/sdX;sync
199[02:14:10] <Tenkawa> because cp "itself" actually
has different io backend than dd
200[02:14:20] <Tenkawa> to the underlying os
201[02:14:29] <Tenkawa> it "can" do different things
202[02:14:48] <Tenkawa> especially once you get into filesystems
203[02:15:00] <Tenkawa> thats where cp can burn you
204[02:15:08] <sney> sure, and this isn't a filesystem
operation. :)
205[02:15:22] <Tenkawa> it still has to follow the rules
206[02:15:31] <Tenkawa> its still block io
207[02:16:52] <slop> Hi all, I have a new debian install thats
hanging on AMD drivers/firmware - how do you recommend getting
those?
208[02:17:01] <sney> !firmware image
209[02:17:01] <dpkg> There are <live> system and
<netinst> and DVD images containing non-free Debian
<firmware> packages available from
replaced-url
210[02:17:13] <slop> <3
211[02:17:36] <Tenkawa> cheers for now.. time to go get some
rest and grumble at this snowstorm
214[02:17:52] <sney> if the hardware is newer than what's
reasonably supported by the 4.19 kernel, you may need to take a
different approach, but try the firmware image first.
222[02:21:37] <slop> the guides are GREAT if you can get network
access
223[02:22:12] <slop> like, it showed what was needed step by
step but I couldnt get a terminal with access :D
224[02:22:21] <growly> @sney - thanks for your reply; i did `cp
/path/to/image /dev/sda` where /dev/sda is my flash drive block
device
225[02:22:27] <sney> yeah, the installer environment
doesn't give you much.
226[02:22:31] <slop> was trying to add non-free and contrib and
go that route
227[02:22:33] <sney> growly: did you 'sync'
afterwards?
228[02:22:57] <growly> i did a reboot; i expected it to block on
sync anyway?
229[02:23:29] <sney> reboot should sync, but I always do it
manually for any removable devices
230[02:24:18] <growly> hmmm
231[02:24:39] <growly> well i just did it again and followed up
with sync, which returned instantly, so i'm not sure if
it's worth retrying
232[02:24:59] <growly> but it seems weird to me that all four of
these are broken in the same way
233[02:25:27] <sney> e1000e is definitely in the buster
installer and doesn't require extra firmware, so the problem is
with *something* at your end
234[02:25:37] <growly> interesting
235[02:25:58] <sney> how many times have you reused that
particular flash drive? got another one?
236[02:26:15] <growly> i used it a couple of days ago no
problem. maybe if i blow away the partition table or use dd
237[02:26:41] <growly> oh interesting
238[02:26:43] *** Quits: BenNZ (~Ben__@replaced-ip) (Quit: Everytime I think IQ's must have dropped recently, I
remember that this is the internet)
239[02:28:27] <sney> there's a reason those things are 2
for $5 in the checkout at staples, they're not expected to
last. dd or wipefs isn't going to do anything that cp to the
device didn't already do.
240[02:29:59] <slop> :D tgats great
241[02:30:02] <slop> thats*
242[02:30:34] <slop> 32 gib devices for 5.99 and I am like I
MUST BUY AT LEAST 3
243[02:31:10] <growly> it strikes me as weird that the flash
drive is 16GB, but parted gives me "the driver descriptor says
the physical block size is 2048 bytes, but Linux says it is 512
bytes", and reports 64 GB
244[02:31:21] <growly> very weird
245[02:31:43] <sney> that sounds like one of the bootleg ones
that used electrical tricks to lie about the size
246[02:31:44] <cjoke> howto disable sddm on boot ?
247[02:32:18] <sney> cjoke: systemctl disable sddm. or if you
want to switch to a different display manager, I think it's
dpkg-reconfigure display-manager
249[02:33:00] <growly> ok this weirdness only occurs _after_ i
cp the image on
250[02:33:06] <cjoke> sney: I want to boot in text, as I
configured in grub, but mate and debian is new to me, I did look in
systemctl and I thought I had to run --system.
251[02:33:25] <sney> cjoke: when you run systemctl as root,
--system is the default.
252[02:34:16] <sney> for more about disabling things in systemd,
this blog post is nice:
replaced-url
253[02:34:49] <robobox> heh wikipedia is 20 years old
254[02:34:51] <growly> does everyon still hate systemd or have
we moved on now?
255[02:34:56] <cjoke> sney: it dont recognize sddm as a
operation :/
265[02:36:45] <sney> growly: most peoples' reasons for
hating systemd were stupid, while there are some legitimate
criticisms, the stupid reasons never really stop coming around.
it's not as bad as it was ~6 years ago though.
296[02:53:01] <gholinbrown> thanks i'm just going to purge
mono.
297[02:53:15] <sney> monopd is not related to mono, despite the
name. see 'apt show monopd'.
298[02:53:40] <sney> but maybe you did something like 'apt
install mono*' accidentally.
299[02:53:59] <gholinbrown> i think i did.
300[02:54:14] <sney> hungry globs, heh
301[02:54:30] <gholinbrown> i wanted the sec certs . the actual
software is just icing
302[02:56:59] <gholinbrown> shux. appears i must install mono
for those anyway
303[02:57:15] *** semeion is now known as mnemonic
304[02:59:10] <gholinbrown> btw it would be nice if password
managers didn't require deps. for example if you acquire the
debian dvd, and install offline. then you find out you can't
access your passwords without internet, that's not fun
305[03:00:00] *** Quits: kawaiipunk (~from@replaced-ip) (Quit: Leaving this Club)
306[03:00:35] <sney> just one of those things you have to plan
ahead for, I suppose. dvd1 has a huge number of packages but it
can't contain everything.
308[03:01:17] *** Quits: Tobbi (~Tobbi@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
309[03:01:45] <gholinbrown> sure but password manager stores all
your passwords. including those for vpns/proxies
310[03:02:40] <sney> only you can prevent catch-22s in your own
environment
311[03:02:43] <gholinbrown> so it's actually asking users
to connect potentially through a bad channel, just to gets the deps
for none other than the password manager
312[03:02:57] <gholinbrown> yeah pencil paper sney but a
password manager come on now
313[03:03:04] *** Quits: T-zef (~tyzef@replaced-ip) (Remote host closed the connection)
314[03:03:10] <sney> if you have a specific bug report against a
package, file it
315[03:03:44] <gholinbrown> does debian have a native password
manager?
326[03:11:20] <gholinbrown> last question. are the packages we
download for example through apt-get , are these checked locally to
be what was expected? and if it does, surely those hashes are
pre-secured in the initial installation or ar they literally fetched
across http at the time you give an apt command
331[03:13:59] <sney> checksums are in the package. if
you're concerned about http, https has been built in to apt for
some time, and most mirrors support it too, so feel free to tweak
sources.list accordingly
332[03:14:03] <sney> !debsums
333[03:14:03] <dpkg> debsums is a utility that will check a
package's files against their checksums. The "-a"
argument will instruct it to also check configuration files:
"apt install debsums; debsums -a -s". Almost all packages
come with md5sums included in the package or apt will have generated
them for you; generate missing ones with "apt-get install
--reinstall `debsums -l`". Ask me about <md5sums>.
353[03:26:31] <themill> ghormoon: you're not subject to
MITM attacks with apt downloading packages, no matter whether you
use http, https or some other protocol
354[03:26:36] <themill> !secure-apt
355[03:26:37] <dpkg> [apt-secure] Starting with version 0.6,
<APT> performs signature checking of the Release file for all
archives. See
replaced-url
369[03:32:29] <ryouma> what physical sector size should one
expect from most consumer drives? all of mine are NOT showing 4096
when i run fdisk -l. logical and physical are showing as the same.
just wnat to know if this is normal.
371[03:33:34] <ryouma> (also want to know if debian does or
requires anything special about this, such as in mount options or
partitioning tools or something)
376[03:36:26] <sney> ryouma: to the second point, IME with a
wide range of hardware, I've never had to do anything special
to get normal/expected performance from any disk I've owned.
there are certainly tweaks people can do to make things faster/more
efficient, but for general "works" I've never seen a
disk that conflicted with the defaults.
382[03:38:14] <ryouma> that also answers my next question about
relatime vs. lazytime vs. noatime on extn vs. btrfs and various
types of drives
383[03:40:08] <sney> and looking at fdisk -l on my hypervisor:
mechanical disks report 4096B sectors, SSDs report 512B, and the
logical zvols it created for kvm guests are 8192B.
384[03:40:21] <sney> I remember hearing something about SSD
firmware possibly lying about this as well
385[03:41:13] <ryouma> i didn't think my spinning rust
drives were /that/ old, but maybe they are
386[03:41:26] <ryouma> but if it doesn't matter much, great
483[05:06:59] <b00b_bot> hi. im doing several server tasks on my
vps atm. i did a portscan with nmap for example but it only outputs,
the ports of the webserver im maintaining, ssh is missing, ehy is
that?
484[05:08:29] <sney> vps providers often implement a firewall
that makes it harder for botnets to find your listening ssh daemon
486[05:09:03] <sney> 'ss -lt' from the vps itself will
show what services are actually listening
487[05:10:11] <b00b_bot> sney: oh i didnt know that. yeah localy
isnt a problem
488[05:10:31] <rk4> presumably ssh $ip works? what of nmap -Pn
-p 22 $ip?
489[05:11:04] <b00b_bot> rk4: i dont use the default ssh port
490[05:11:22] <rk4> well then sub in whatever port you're
using.
491[05:11:33] <b00b_bot> but i will try with the custom one
492[05:12:43] *** Quits: kini (~kini@replaced-ip) (Remote host closed the connection)
493[05:13:01] <b00b_bot> yes that shows me the port is open, but
the problem is, that someone told me my custom ssh port before... i
dont know how he found out what it is...
498[05:15:18] <sney> there's only 65535 tcp ports, if
someone wants to find out what port you're using for ssh all
they need is your ip and some time. but also, it doesn't matter
if someone knows your ssh port, if your sshd is secured in a sane
way
499[05:16:17] <sney> it can get a little annoying when botnets
discover it, but even then it only eats cpu cycles
500[05:17:18] <b00b_bot> rk4: i used the default one
501[05:17:22] <sney> they called it "secure shell" for
a reason.
502[05:17:33] <rk4> b00b_bot: which was...
503[05:17:47] <b00b_bot> sney: yeah i know.. but i was curious
how he found out (he wont tell me...)
508[05:19:11] <sney> were I to do it, I'd use check_ssh
from monitoring-plugins, and script bash to run it through the
22,222,2222s and then just count from 1 to 65535. and then go have a
sandwich, and come back to the result. and I'm not even a h4x0r
509[05:19:17] <rk4> b00b_bot: well then, i don't think your
ssh port is on the default list of ports scanned then
510[05:20:06] <rk4> 'By default, Nmap scans the most common
1,000 ports for each protocol.
512[05:20:31] <b00b_bot> ok thx so far ;) next question: i would
like to force all users to wait a specific amount of seconds before
they can run any commands after the logged in onto the server, no
matter if physical or over ssh or anything else.
518[05:21:26] <dpkg> Slow down for a bit! Are you sure that you
need to jump through that particular hoop to achieve your goal? We
suspect you don't, so why don't you back up a bit and tell
us about the overall objective... We know that people often falsely
diagnose problems because they are too close to them -- it's
easy to miss that there is a better way to proceed. See
replaced-url
519[05:22:13] <slop> what will that gain? any attacker worth
their salt is not going to get in and make noise.
520[05:22:46] <b00b_bot> when i do with a simple sleep 10 in
/etc/profile it can be canceled with a simple crtl+c
521[05:22:56] <foxide> b00b_bot: That doesn't answer the
"why."
522[05:23:30] <b00b_bot> foxide: doesnt matter
523[05:23:35] <sney> trying to implement stupid stuff is a good
way to improve one's linux skills
524[05:23:36] <foxide> Yes, it actually does.
525[05:23:48] <foxide> It really, really does.
526[05:23:51] <sney> ofc if that's the goal then it also
precludes IRC randos giving one the answer :V
527[05:23:56] <b00b_bot> foxide: for you?
528[05:24:16] <foxide> When you have multiple users going
"what, why?" it might be a good idea to listen.
529[05:24:33] <foxide> So, yes, the why matters.
530[05:24:46] <b00b_bot> foxide: answering your questing doesnt
help me getting my question answered
531[05:25:05] <themill> it does
532[05:25:07] <foxide> It certainly does.
533[05:25:16] <foxide> b00b_bot: What won't help you get
your question answered is your attitude.
534[05:25:33] <slop> it sounds like your goal is to harden your
debian install(s) - have you checked this out:
replaced-url
535[05:25:37] <b00b_bot> foxide: ok how is this.. the users
should relax 20 seconds after log in to get a calm mind
538[05:26:00] <themill> b00b_bot: sleep 20; is perfect for that
usage.
539[05:26:13] <b00b_bot> themill: no that can be canceld be
everyone
540[05:26:19] <b00b_bot> they should be forced
541[05:26:26] <themill> b00b_bot: that's not an issue for
the use case you describe however
542[05:26:32] <b00b_bot> forcing relaxation
543[05:26:54] <themill> b00b_bot: I'm bored now. Just
thought I'd put that out there
544[05:27:17] <slop> How many users are we talking about logging
in?
545[05:27:35] <b00b_bot> 20 lines of replies... how many solving
answers would fit there...?
546[05:27:45] <foxide> b00b_bot: None, because what you're
trying to do is idiotic.
547[05:27:50] <b00b_bot> slop: 12
548[05:28:03] <b00b_bot> foxide: ignore. bye ;)
549[05:28:07] <rk4> so basically the room thinks you're
trying to do some weird shit, any of us could achieve what you want,
but we're not inclined to help since you're not inclined
to help us understand the why
555[05:31:31] <slop> For hardening, I recommend following these
steps first.
replaced-url
556[05:33:06] <slop> ah, he is gone. still odd wanting to do
that annoyance for a very small group of users. Id think hardening
and restricting user permissions would be a better route than
delays.
557[05:33:10] <slop> anyway.
558[05:33:31] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
559[05:35:05] <foxide> I can
560[05:35:16] <foxide> 't fathom a single use case where
that would be valuable.
561[05:36:05] <slop> agreed. other than wanting someone to
replace you quickly.
609[06:33:31] <sney> and on linux your print "server"
is used for local printing operations too. so even if it was just
your pc and 1 direct attached printer you would still need cups.
612[06:37:30] <growly> so, i've now tried bootdisks from
three different debian installers from two different machines on two
different flash drives. the problem is either my pc or the way
debian is generating the images. i doubt the latter because then
someone else will have run into this problem by now
630[06:57:22] <growly> iflema: i've tried both cp x.iso
/dev/sdx and dd
631[06:57:24] <growly> with sync
632[06:57:38] <iflema> conv=sync
633[06:57:55] <growly> to dd?
634[06:57:57] <iflema> m
635[06:58:05] <growly> sure why not
636[06:58:11] <iflema> status=progress
637[06:58:14] <iflema> :D
638[06:58:50] <growly> the actual command i used was 'sudo
dd bs=4M
if=/home/arya/Downloads/debian-bullseye-DI-alpha3-amd64-netinst.iso
of=/dev/sdb status=progress oflag=sync'
639[06:59:11] <growly> well. the last one. i also tried bs=512,
no flags
641[06:59:57] *** Quits: growly (~arya@replaced-ip) (Quit: Lost terminal)
642[07:00:30] <zerocool> guys, somehow mariadb-common is cock
blocking me from upgrading. i said f it and just removed all mariadb
but it still isn't removing
665[07:19:33] <growly> the only weird thing i've noticed is
that the logical sector size of the filesystem when you dd the .iso
to the flash drive is 2048 instead of 512 b, which means the
partition table doesn't make much sense
853[10:56:14] *** Quits: EmleyMoor (42b789682f@replaced-ip) (Remote host closed the connection)
854[10:58:02] <muffindrake> I successfully set the program using
xdg-mime default pcmanfm.desktop inode/directory, but chromium
apparently does not use xdg-open for this purpose, as it continues
to use nautilus instead after restarting it
855[10:58:45] <muffindrake> Er, now I'm not sure whether
that's the correct program to use
857[11:00:24] <muffindrake> ... what's the correct way to
set a default program for a mime type on a DE-less system in debian
sid? All the information I'm finding on this appears to be
outdated or wrong.
869[11:09:33] <alex11> muffindrake, maybe this helps you
replaced-url
870[11:10:52] <alex11> muffindrake, however, you might also be
hitting a bug; maybe ask #debian-next on oftc
871[11:11:13] <growly> brigand: asus prime x299-deluxe ii, intel
i7 9940x, 128gb ram, samsung 970 pro nvme ssd, lsi sas raid card
which forces me to use legacy boot room along with uefi
873[11:11:26] <growly> tomorrow i will take out the lsi card and
try using no CSM at all
874[11:11:45] <growly> but this is clutching at straws
875[11:11:59] *** Quits: freebench (~freebench@replaced-ip) (Remote host closed the connection)
876[11:12:28] <growly> yeah unfortunately installing a uefi
system through the legacy installer mode seems to be quite a lot
more complicated than the push-button walkthrough i did last time
or, at least, was expecting
879[11:14:04] <muffindrake> Hitting a bug with setting a default
application? I created ~/.local/share/applications/mimeapps.list and
put in the content as should be correct according to that wiki
entry, but the change is not reflected when I use xxdg-open on any
directory.
949[12:08:58] <vanfanel> hi! I am trying to access an vnc server
(tightvnc) from the local network. I first launch it with
"vncsever", then create an ssh tunnel with: ssh -L
59000:localhost:5901 -C -N -l <username>
<local-network-ip>
950[12:09:50] <oxek> !tell haba713 bat
951[12:10:01] *** Quits: dselect (~dselect@replaced-ip) (Quit: ouch... that hurt)
952[12:10:01] <vanfanel> however I get, on vncviewer I get:
vncviewer: ConnectToTcpAddr: connect: Connection refused
953[12:10:09] <oxek> !bat
954[12:10:09] <dpkg> [Basic Apt* Troubleshooting]. To diagnose
your problem, we need ALL OF THE FOLLOWING information: 1. complete
output of your apt-get/apt/aptitude run (including the command used)
2. output from "apt-cache policy pkg1 pkg2..." for ALL
packages mentioned ANYWHERE in the problem, and 3. "apt-cache
policy". Use
replaced-url
977[12:38:28] *** Quits: Jad (~Nashmi@replaced-ip) (Quit: Benefits I derive from freedom are largely the result
of the uses of freedom by others, and mostly of those uses of
freedom that I could never avail myself of.)
983[12:43:42] <relipse> Hi, I'm on Debian Jessie, how
exactly do I backup my stuff, I am on linode so I clicked the take
snapshot, but is there anything else I need to do before I upgrade?
1001[12:49:16] <ratrace> kwami: suggestions for what precisely?
your question isn't clear
1002[12:49:44] <daysun> thanks oxek
1003[12:49:46] <oxek> daysun: assuming you're on debian 10
(buster) of course
1004[12:49:47] <kwami> Of what can be causing this issue and how
can I fix it
1005[12:50:28] <daysun> kwami: lost after each reboot or
randomly?
1006[12:50:34] <ratrace> kwami: have you checked the logs to see
if there's any message along with a dropped route?
1007[12:50:51] <daysun> never mind
1008[12:50:56] <relipse> Managed hosting is . +$100/month per
Linode , my own server is ~$20/month, that's a huge price
difference that's why, I know linux well enough to keep the
sucker running, but it's time for an upgrade
1009[12:51:10] <kwami> daysun: randomly
1010[12:51:26] <ratrace> kwami: also give us more details. what
route? ipv4? v6? how do you lose it? one moment it's there,
another it's gone from the list of routes, or it IS three but
no packets are going through?
1011[12:52:09] <ratrace> *there
1012[12:53:50] <kwami> ratrace: I set up a route in the interface
file wich works perfectly, but after several minutes the route
disapear from the route table anda the server turn itself unable to
send packets to hosts on another network
1019[12:58:51] <kwami> ratrace: realy, I checked /var/log/syslog
1020[12:59:03] <ratrace> relipse: then you should know exactly
wht to back up. image the server has blown up and you have to set it
up anew. you back up each "Gee, I wish I had this file from
before" you think would happen :)
1027[13:01:08] <ratrace> the main/default debian repos _should_
allow for smooth upgarde. actually, now that I think about it, PHP
was a bit problematic because it was installed in parallel, so
you'd have to explicitely remove old packages and configs
1028[13:01:20] <ratrace> (the configs are in version specific
directories, so keep that in mind)
1039[13:05:39] <ratrace> kwami: and when does the route
disappear? after some set time? randomly? do you have before and
after exaples of the `ip route` output ?
1040[13:06:19] <ratrace> daysun: apt no, apt-get yes, I
don't know about aptitude
1051[13:08:41] <ratrace> relipse: maybe but it's not
supported or tested. just to through stretch.
1052[13:09:37] <OzFalcon> Found an interesting race condition
with xorg
1053[13:09:47] <ratrace> kwami: do you have networkmanager, or
some other network configuration tool that might conflict with
ifupdown? a leftover cron debugging or something script that would
remove the route?
1054[13:10:27] <ratrace> kwami: there's also
systemd-networkd installed on Debian. are you 100% sure it is not
enabled or started somehow?
1055[13:10:59] <ratrace> kwami: it'd be super awesome if you
could know _precise_ time when the route drops, then pastebin the
logs surrounding that time, say a minute before and after
1063[13:14:28] <ratrace> ideally you'll need the logs for
the time surrounding the route loss. the logs cna show if some other
framework piped up and reset the network
1064[13:16:09] <ratrace> also check if your (r)syslog
configuration is default. by default, it logs *.* into
/var/log/syslog. if you changed that, then there might be other
files with clues. ideally you use the journal.
1077[13:31:42] <relipse> Can someone help me with my sources
file, the upgrade guide tells me to rename everything to stretch,
but i see a lot of them commented out? What do I do?
replaced-url
1078[13:33:31] <ratrace> uncomment them. also comment out third
party repos.
1079[13:33:34] <ratrace> !stretch sources.list
1080[13:33:34] <dpkg> A suitable /etc/apt/sources.list for
"Stretch" has three lines: "deb
replaced-url
1081[13:35:27] <n4dir> relipse: deb-src lines you only need if
you plan to compile from source. Most of the time you want to
backport a higher version, and for that you would need a different
release than the usual one (say testing or sid if you run stable)
1082[13:35:42] <n4dir> as in: it being commented is not unusual
1083[13:36:18] <n4dir> deb cdrom you only need if you install
packages from, well, a CD or DVD. If you don't, you can remove
it altogether
1084[13:36:37] <relipse> ok is this good?
1085[13:36:41] <n4dir> notesalexp : i got no clue.
1087[13:37:25] <n4dir> relipse: looks good to me, besides the
notesalexp entry, no clue about it. The link ratrace gave will offer
you something for copy and paste, if you want to be sure
1088[13:38:18] <relipse> n4dir: I don't understand what you
mean? What is notesalexp ?
1089[13:38:33] <n4dir> you got such a line in your sources list
1092[13:39:25] <relipse> i removed it can you look at the second
link I posted
1093[13:40:02] <ratrace> relipse: is "main" the only
component you used?
1094[13:40:20] <n4dir> relipse: i sure can, but really, your best
bet is to comment everything in sources.list, copy and paste the
entry from the link of ratrace at the top, then go for it
1095[13:40:43] <relipse> ratrace probably I mean that's the
only thing in that file
1098[13:41:07] <n4dir> i tend to remove everything i don't
need, instead of commenting it out, so i can understand the
sources.list at a quick glance. Your choice though
1099[13:41:28] <n4dir> if you have everything commented at the
bottom, it is just as clear
1110[13:43:50] <unixbsd> anyone capable to run stratagus with
strategic games? where to place the data files?
1111[13:44:16] <ratrace> unixbsd: wrong channel?
1112[13:44:25] <n4dir> i sure would, but, you know, there is
always something someone could oversee. So i hesitate to shout a
"GO", and then be the bad guy. relipse To me it looks good
1113[13:44:32] <ratrace> relipse: I would enable all components,
unless you're 100% sure "main" is the ONLY one
you've possibly ever used
1114[13:44:52] <n4dir> i would enable contrib and non-free only
if i knew i need it
1115[13:44:58] <ratrace> enabling the components won't
automagically install unwanted software, so it's safe for
_upgrade_
1116[13:45:32] <n4dir> if jessie worked fine being free i see no
reason why stable shouldn't ...
1117[13:46:23] <ratrace> non-free probably isn't needed, but
contrib could be. is there harm in enabling them for the upgrade?
1125[13:49:19] <n4dir> ratrace: to be done with this i agree with
you. To me it doesn't matter that much, and i hardly am able to
use main only anyway.
1126[13:49:24] <ratrace> most guides suck, and this one
ain't an exception
1127[13:49:56] <relipse> ok i'm going to go upgrade
1128[13:50:00] <relipse> thanks for your help
1129[13:50:01] <ratrace> n4dir: unless I've gone crazy, I
think there was at some point a situation where zfs components where
spread between main and contrib, which I thought was rather unusual.
1130[13:50:07] <n4dir> yee-haw ! good luck relipse
1131[13:50:40] <n4dir> ratrace: ah, yeah, hence i don't
wanna discuss. I live in my world, with my problems. And i know only
that much (or: little :-) )
1132[13:50:40] <ratrace> relipse: before you do....
1133[13:51:15] <relipse> Does this look good?
replaced-url
1134[13:51:18] <ratrace> relipse: linode used to have situations
in the past where "snapshots" were supposedly done but
when you needed them, they were broken. I'd _never_ trust my
provider to create backups. if you can, make sure you have your own,
independent backups
1135[13:51:54] <relipse> i already backed up my databases
1161[13:59:30] <jim> wheezy is 7... is this asdfasd?
1162[13:59:42] <daysun> relipse: and uname -a and see if
it's says jessie there
1163[13:59:54] <daysun> jim: asdfqwerty
1164[13:59:55] <n4dir> daysun: i wouldn't bet, but i seem to
recall upgrade doesn't install or remove packages not yet on
the system, dist-upgrade does that if needed. In practice i just do
upgrade first, then dist-upgrade.
1170[14:01:07] *** Quits: raekuul (pokota@replaced-ip) (Remote host closed the connection)
1171[14:01:16] <jim> I was talking to a guy yesterday who was
thinking about upgrading from wheezy to buster, to make sure he
could mount a windows share
1172[14:01:22] <n4dir> if it was me and i was not sure if i would
need it in the future, i would comment it for now, and remove it
when being sure later
1173[14:01:34] <relipse> ok i'll comment it out
1174[14:01:44] *** Quits: JohnML (~john1@replaced-ip) (Remote host closed the connection)
1175[14:01:54] <jim> last I knew, he had backed up his hd
1185[14:04:46] <ratrace> what jim said pretty much, first upgrade
current distro to latest ...... assuming the repos are still
available for jessie
1186[14:04:50] <raekuul> daysun: upgrade by itself only changes
the packages already installed. If A.1 depends on B.1 but A.2
depends on both B.2 and C.1, regular upgrade won't upgrade A.1
because it needs a new package, but dist-upgrade will install C.1 in
order to upgrade A.1 to A.2
1187[14:05:12] <jim> isn't he on wheezy right now?
1188[14:05:51] <jim> aka debian 7
1189[14:06:17] <n4dir> jessie
1190[14:06:18] <ratrace> I thought it was jessie
1191[14:06:24] <ratrace> they said 8.1
1192[14:06:31] *** Quits: Grldfrdom (uid391113@replaced-ip) (Quit: Connection closed for inactivity)
1194[14:06:44] <n4dir> that back and forth with versions and
nicknames of version is killing me softly
1195[14:06:48] <ratrace> which BY THE WAY.... just reinforces my
argument earlier that they should NOT be running a public server
without getting some basic knowledge first.
1196[14:07:14] <relipse> no i'm on jessie
1197[14:07:28] <relipse> !codenames is wrong
1198[14:07:30] <dpkg> ...but codenames is already something
else...
1199[14:07:48] <relipse> !codenames
1200[14:07:48] <dpkg> Debian release codenames are buzz (1.1),
rex (1.2), bo (1.3), hamm (2.0), slink (2.1), potato (2.2), woody
(3.0), sarge (3.1), etch (4.0), lenny (5.0), squeeze (6.0), wheezy
(7) jessie (8) stretch (9) buster (10) and sid=Unstable.
"lsb_release -sc" (lsb-release package) will display a
Debian system's codename.
replaced-url
1201[14:07:50] <daysun> raekuul: so which one should you use? :P
1202[14:07:56] <daysun> upgrade or dist-upgrade
1203[14:08:03] <ratrace> that server apparently saw zero upgrades
and is vulnerable AF..... if not already pwnt. people should really
get licensed before taking out a 20t server truck on the internet
highway.... you'll kill someone.
1204[14:08:15] *** Quits: jpe (~jp@replaced-ip) (Remote host closed the connection)
1205[14:08:22] <ratrace> _literally_ (botnets attacking medical
hardware)
1206[14:08:22] <n4dir> daysun: i was always told first upgrade,
then dist-upgrade. I never thought about the why.
1207[14:08:24] <raekuul> Which one *should* I use, or which one
*do* I use?
1208[14:08:37] <ratrace> raekuul: apt upgrade does it all.
1209[14:08:38] <relipse> are you guys talking to me or someone
else?
1210[14:08:46] <ratrace> relipse: I was referring to you :)
1211[14:08:49] <n4dir> on stable both ain't that exciting
anyway
1212[14:08:51] <daysun> n4dir: dist-upgrade does upgrade for you.
it's pointless to do both
1213[14:09:09] <ratrace> upgrade jessie to max first, then switch
to stretch and apt update, apt upgrade
1214[14:09:09] <n4dir> daysun: might be. I was told, over and
over, and hence that is what i do
1215[14:09:12] <relipse> so what should I do, i'm about to
upgrade to stretch
1223[14:11:09] <daysun> dvs: does it depend on whether
you're using apt or apt-get?
1224[14:11:33] <dvs> daysun: the release notes usualy does
apt-get
1225[14:11:40] <relipse> i already ran apt-get update; apt-get
upgrade
1226[14:11:48] <relipse> under jessie
1227[14:11:54] <relipse> so i should be good to go
1228[14:11:59] <hejux> debian is so fucked up. why apt and
apt-get
1229[14:11:59] <raekuul> daysun: depends on if you're
scripting the update process or not. apt for interactive use,
apt-get for scripted use
1230[14:12:01] <daysun> I read that apt full-upgrade upgrades
first before doing a "dist-upgrade". It differs from
apt-get if iirc where you got to apt-get update first
1231[14:12:07] <dvs> relipse: no, you need dist-upgrade as well
1232[14:12:28] <relipse> dvs: so you want me to change my sources
back to jessie and run apt-get dist-upgrade ?
1251[14:16:30] <hejux> n4dir:that's why i stick with PACMAN
1252[14:16:37] *** Quits: binofaata_ (~binofaata@replaced-ip) (Remote host closed the connection)
1253[14:16:39] <n4dir> ha ha. yeah, good idea
1254[14:16:48] <raekuul> daysun: in practice 'apt
upgrade' and 'apt full-upgrade' differ on how they
handle conflicting upgrades. regular upgrade preserves what's
already installed, full-upgrade will remove older versions that
prohibit newer versions
1349[14:37:33] <jim> if so, copy the file into your home dir
1350[14:38:04] <relipse> is this bad? E: Sub-process
/usr/bin/dpkg returned an error code (1)
1351[14:38:12] <jelly> relipse, in general the release upgrade
will fare better if you use the vendor (distro) version of config
file, or if you immediately put your custom changes into the new one
1352[14:38:39] <relipse> the only error was I couldn't
upgrade phpmyadmin
1353[14:38:41] <jelly> relipse, the lines before that would be
useful.
1367[14:40:27] <jim> then let it replace with the package
mgr's version, then when the upgrade is finished, look at both
versions, and you should also see the changes as reported from just
when the dist-upgrade started (I dunno where else that's
stored, except it might be in an email to and from root
1380[14:42:16] <jelly> relipse, make sure there are no dpkg and
apt inconsistencies, reboot, then see if things still work, then
think about stretch->buster
1390[14:43:54] <relipse> ok it looks like fail2ban is not working
1391[14:44:04] <ratrace> it has its own logs with probably more
clues
1392[14:44:22] <jim> relipse, unfortunately the php language devs
did a huge disservice, when they repeatedly released incompatible
(features missing, new and incompatble features added) versions of
php
1393[14:44:48] *** Quits: eliotome3000 (~eliotime@replaced-ip) (Remote host closed the connection)
1394[14:45:13] <jelly> relipse, first fix dpkg and apt
inconsistencies. Then reboot and fix services that need fixing.
1395[14:45:38] <relipse> jelly: can you please tell me how to do
that
1396[14:46:12] <jim> systemd has a wonderful way of not being
informative (failed with result 'exit code')
1397[14:46:45] <jim> someone should file a bug report for that
1398[14:47:04] <jelly> relipse, I already have. "dpkg
--configure -a" and "apt-get -f install"
1400[14:47:30] <ratrace> I don't think that's a bug
really. it means the process exited unexpectedly with a nonzero exit
code. the process in question, however, is responsible for proper
logging
1401[14:47:49] <ratrace> and oftentimes, services are badly
configured.
1402[14:47:58] <relipse> fail2ban is not working
replaced-url
1403[14:48:03] <jim> ratrace, well, it might be helpful to know
what that exit code is
1404[14:48:05] <jelly> relipse, after reboot?
1405[14:48:18] <ratrace> relipse: are you reading the advice
you're given here?
1426[14:54:49] <ratrace> and even if you used daemntools. the
capabilities of service/process configuration and containment under
systemd, currently has NO alternative
1427[14:54:52] <jim> jelly, their last paste doesn't seem to
be the output of (say) dpkg --configure -a
1456[15:01:11] <jelly> Jan 17 13:54:37 cworklog.com
fail2ban-client[14965]: ERROR Failed during configuration: While
reading from '/etc/fail2ban/jail.local' [line 155]: option
'port' in section 'pam-generic' already exists
1457[15:01:25] <relipse> wow good eye again jelly and jim
1493[15:07:02] <relipse> i suppose you are going to tell me
rtfrigginQ
1494[15:07:05] <relipse> i suppose you are going to tell me
rtfrigginm
1495[15:07:11] <jelly> relipse, if you've chosen to keep own
versions of config files, you might have a mix of config files from
the old and the new version
1496[15:07:37] <ratrace> well that woudl be a good start, to know
the software you're using. back to may argument about you
shouldn't be running a public server and you claiming you know
linux well enough to keep it running. seems like you have zero clue
really.
1523[15:13:43] <jelly> also, it seems you have a copy of the
whole jail.conf in jail.local, that's needless; the point of
jail.local is to write down minimal changes to defaults that are
needed for this machine
1537[15:16:46] <pasiz> brute force attack or fail2ban?
1538[15:16:57] <ratrace> not completely no. it's useless
against distributed attacks, and I mostly see those. one IP will try
a few times, much less than the tresholds we keep
1539[15:17:31] <jelly> yep, it can't protect against one
user being probed from 100,000 different ip addresses
1540[15:17:54] <jelly> and it doesn't do blocking of whole
ranges
1541[15:18:23] <jelly> but some probes are not that smart and
still get blocked
1543[15:18:26] <ratrace> and from what I've seen in the
logs, for example smtp and imap bruteforcers, they try for exactly 4
times per IP as if 5 is some magickal round number treshold lol :)
1544[15:18:43] <jelly> 5 is probably f2b default to block things
;-)
1545[15:19:14] <ratrace> it is, yes. maxretry = 5 in default
jail.conf
1546[15:19:23] <pasiz> best practice would be disable password
auth
1547[15:19:26] <relipse> ok i've reboote and everything
seems fine.
1548[15:19:29] <relipse> d
1549[15:19:37] <ratrace> pasiz: yes, but you can't for all
services
1551[15:19:53] <jelly> relipse, good, now you can say you have a
stretch installation!
1552[15:20:05] <ratrace> relipse: now do it all over again for
stretch -> buster. again, note, you might have issues with php,
as the config dir changes
1553[15:20:09] <relipse> finally i can use php 7 syntax!
1557[15:21:12] <jelly> relipse, don't forget to do cleanup,
as documented in stretch release notes. apt-get autoremove, and
removing obsolete packages
1558[15:21:18] <jelly> !obsolete
1559[15:21:18] <dpkg> If you remove a repository from your
sources.list (e.g. removing <dmm>), then you should check what
packages you have installed from the other repository. Synaptic and
aptitude have a "Obsolete and Locally Created Packages"
list. Or, "aptitude search ~o". Note this doesn't
include packages that exist in the repo at a different version to
the one you have installed; see <not available> <list
repositories>.
1560[15:21:30] <relipse> do you have a link to the release notes
1561[15:21:35] <jelly> !stretch release notes
1562[15:21:36] <dpkg> The release notes for Debian 9
"Stretch" are at
replaced-url
1564[15:22:25] <jelly> (the same document is linked in /msg dpkg
jessie->stretch ; when you're ready for stretch->buster,
read also /msg dpkg stretch->buster)
1594[15:37:40] <relipse> if you can just use other sources,
what's the shame in that?
1595[15:38:14] <haba713> azeem: yes, i did. it works ok. thanks!
1596[15:38:30] <raekuul> shame? none. guarantees of compatibility
and stability? also none
1597[15:39:00] <ratrace> right. and it's not about that. the
reason 8.0 won't end up in bullseye is the reason you
can't use 8.0 even from another repo, at least not initially
1600[15:39:42] <ratrace> game over really, transitions are frozen
1601[15:39:56] <jelly> !bullseye freeze
1602[15:39:56] <dpkg> The first steps of the <freeze> to
release Debian 11 "Bullseye" will start on January 12,
2021.
replaced-url
1603[15:40:45] <jelly> !literal bullseye freeze
1604[15:40:45] <dpkg> "bullseye freeze" is
"<reply>The first steps of the <freeze> to release
Debian 11 "Bullseye" will start on January 12, 2021.
replaced-url
1605[15:41:52] <jelly> dpkg, no, bullseye freeze is
<reply>the freeze for Debian 11 'bullseye' has begun
on 2021-01-13.
replaced-url
1646[16:02:47] <jim> separate /var, because logs go there, and
the debian package cache goes there too
1647[16:02:58] <ratrace> omarek_: clearly we can't know how
big your projects will become and what your future needs will be .
as a rule of thumb, 10G suffices for installing a desktop with
mostly used programs, and then some. I wouldn't go below 20G
personally.
1648[16:03:38] <omarek_> jim: Not currently, but I may need to
run a couple of docker containers. I have a couple of years of
experience.
1649[16:03:56] <ratrace> I also wouldn't separate out any
directory unless: a) needed different mount options for them, b)
needed bigger/other storage media
1650[16:04:03] <jim> in what languages?
1651[16:04:09] <jelly> omarek_, separate /var is very nice on
servers. Home use? Dunno, are you planning on having huge mysql or
postgres databases
1652[16:04:18] <omarek_> jim: Python and recently Rust.
1653[16:04:32] <omarek_> I'm moving away from web
development.
1655[16:04:42] <jim> ratrace, also, you're protecting / from
overfilling
1656[16:04:44] <jelly> separate /usr is senseless
1657[16:05:11] <jhutchins> I find a seperate /var isn't as
useful as people think. If /var is full and logging stops, many
programs will stall or freeze the system.
1658[16:05:18] <jim> it's where most of the content of
packages go
1659[16:05:18] <omarek_> I typically had separate /, /home, and
swap partitions. I can make a separate /var. How big /var makes
sense?
1660[16:05:20] <dvs> jelly: agreed
1661[16:05:26] <ratrace> jhutchins: +1
1662[16:05:39] <ratrace> and also most important ones not to
overfill are under /run
1663[16:05:45] <cybercrypto> omarek_: Standard debian partition
is pretty much good for general development purposes. You may want
to separate /home and /var to make it better. I am running 18GB for
root fs /home=200GB and /var=32Gb. Absolutely no issues for my home
needs.
1664[16:06:03] <jhutchins> omarek_: I strongly recommend a single
partition until you establish where your storage is going to grow.
1665[16:06:11] <omarek_> Not planning to run real servers, more
like dev environment and proof of concept stuff.
1666[16:06:14] * ratrace seconds that
1667[16:06:58] <dvs> omarek_: if you're not running a server
then I wouldn't make a separate /var partition.
1668[16:07:10] <jhutchins> As examples, a server that
under-allocated /home because it was a web server, but it had
multiple clients serving pages out of /home, or another that
didn't allocate enough /var and was running mail and web pages
from /var/replaced-url
1669[16:07:11] <jelly> omarek_, honestly not making the /boot too
small is more important than having a separate /var or even /home :
1670[16:07:13] <omarek_> I currently have 15/20GB full on my /
partition.
1671[16:07:17] <jelly> for a home system
1672[16:07:18] <omarek_> Right after apt-get celan
1675[16:08:18] <dvs> omarek_: However, 20GB is a little worrisome
if it includes /var when upgrading to a new release.
1676[16:09:00] <omarek_> I think it's quite tricky to resize
existing partitions?
1677[16:09:16] <jhutchins> I can't count the times someone
made a seperate /boot, then did an upgrade with /boot un-mounted.
You may ponder the results.
1678[16:09:18] <jelly> omarek_, just set up a 10GB / and 10GB
/home on LVM, leave the rest of the VG free, then grow them later.
Avoid partitions.
1688[16:10:24] <jim> omarek_, quick primer on lvm...
1689[16:10:26] <ratrace> the partitioning industry hates that
simple trick. click here for more info!
1690[16:10:29] <jelly> omarek_, you can't migrate an
existing system
1691[16:10:38] <jim> with partitions, you make partitions using a
partition tool, format them with a filesystem directly, and mount
the partition with the filesystem. with LVM, instead of putting
filesystems in partitions directly, you put 'LVM physical
volumes' in partitions, then you make a 'volume
group' (which is just a list of physical volumes), and put
physical volumes in the volume groups, then you can make
'logical volumes', and these are what you would format and
1692[16:10:38] <jim> mount, and they got allocated from a volume
group.
1695[16:12:11] <jelly> omarek_, well, the migration is
"format new disk with 1 partition for /boot 1 for LVM, set up
LVs for / and swap (and optionally for /var and /home) and restore
all data to new fs structure and set up boot loader"
1696[16:12:13] <omarek_> So LVM is solid enough for everyday use?
1697[16:12:20] <jelly> absolutely
1698[16:12:29] <jim> has been for like 20 years
1699[16:12:50] <jim> quite stable
1700[16:12:50] <omarek_> I never personally met anyone using it,
which is why I'm a bit cautious.
1701[16:13:10] <jelly> d-i still has dumb defaults about
partition and LV sizes.
1753[16:49:06] <slop> Question - I recently tried gnome again and
leaning back towards XFCE - what do YOU prefer? Gnome has more bloat
(e.g., games etc) than I care of. I know they can be uninstalled but
just being picky.
1776[16:54:33] <EdePopede> i like to have the option to change
things as i need them. already says much about something specific
without even naming it.
1777[16:54:37] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
1855[17:26:26] <EdePopede> n4dir, that's all something for
the other pc for the moment. but as soon as i allow the first
"apps" to run inside my browser i don't really care
anymore about what get used somewhere else ;)
1872[17:30:37] <EdePopede> > just ran a test session on the
other PC after the first one seems to run like forever, killed the
other processes (tracker-miner-fs, tracker-miner-apps, applet.py,
agent) manually, waited for checked after one and a half hours now,
and the systemd tree is gone.
1873[17:30:46] <EdePopede> this was from... some days ago
1874[17:31:07] <EdePopede> i think this time it is still running
1875[17:31:08] <n4dir> i do recall xfce right after boot 70 MB
Ram usage. Window Managers could have 60. And that was not in the
nineties
1876[17:31:23] <n4dir> i fail to see what more xfce does than it
does 10 years ago
1877[17:31:56] <jhutchins> n4dir: That's why we like it.
Quality has improved though.
1878[17:32:21] <n4dir> yeah, it being not that experimental as
far changes are concerened is very good
1879[17:32:24] <freem> n4dir: maybe your answer is "it now
uses gtk3"
1880[17:32:39] <n4dir> i checkd version? 16 ? whatever is in Sid
right now. I didn't see any difference. ha ha
1881[17:32:51] <EdePopede> yep, all still there. last 2 times i
started killing them one by one to see if some monster daemon will
then terminate, one time it did, but not the other time.
1900[17:41:18] <freem> > can a pentium machine with 1g ram,
able to run all the versions from wheezy to buster?
1901[17:41:18] <freem> I have an older computer, with less than
200Mio RAM, CPU ~700MHz, debian 9 + LXDE worked fine there last time
I checked.
1902[17:41:24] <freem> jim: ^
1903[17:41:35] <freem> so your pentium with 1Gio RAM is probably
good
1904[17:41:45] <freem> now, that implies not running a
webbrowser, ofc
1905[17:41:52] <n4dir> yeah, similar experiences here. Ram from
256 to 1 Gig of Ram, all worked fine
1906[17:41:58] <jim> EdePopede, probably the applet talks to a
print server rather than an individual printer directly... so if it
installs a print server, it might install such an applet
1907[17:42:04] <n4dir> besides ... you have guessed it ... modern
web-browsers
1908[17:42:08] <freem> I also need to upgrade it to debian 10...
1910[17:42:41] <EdePopede> not sure why i'm using a DE
myself, maybe because WM only was a mess of one or the other kind
before i started usind KDE long ago. and having that panel thingy
for the network is nice, i don't think i ever knew how to use
the cli for what it does
1911[17:42:52] *** Quits: earthundead (~earthunde@replaced-ip) (Remote host closed the connection)
1912[17:42:55] <freem> someday. Maybe. I mostly keep that oldie
because it still have old hardware, like RS232, PCI, SCSI, etc
1913[17:43:22] <n4dir> EdePopede: well, you could start the xfce
panel in any WM (or at least quite some)
1914[17:43:31] <EdePopede> jim: makes sense, should check it.
does the installer offer "print server"? can't
remember, may have to check the screenshots. D:
1915[17:43:34] <n4dir> fluxbox ! DO IT !
1916[17:43:39] <freem> EdePopede: well, I started using CLI a
lot, and almost only that, when I started using a tiling WM and zsh.
Without them, it's a pain.
1917[17:44:05] <EdePopede> before kde there was fvwm?(2|95)
1918[17:44:28] <freem> I still like my claws, speedcrunch,
zenmap, etc , though
1920[17:45:26] <EdePopede> one of the tasks (not highest priority
though) is to get a gtk free desktop.
1921[17:45:35] <n4dir> uff
1922[17:45:37] <freem> good luck with that
1923[17:45:40] <EdePopede> *nod*
1924[17:45:49] <freem> do you have a good, lightweight, mail
client?
1925[17:46:05] <freem> and I only use ncurses applications when I
can really find nothing better
1926[17:46:18] <freem> so, alpine or mutt are not really good
alternatives to claws here :)
1927[17:46:44] <jim> EdePopede, loooong before kde :)
1928[17:47:05] <freem> a tool like gparted would also be nice,
but this I can live without, I don't resize or move partitions
everyday after all
1929[17:47:07] <EdePopede> ncurses isn't the worst. for text
things i don't really see the need for gtk, sometimes it's
just convenient. copying, resizing, this kind.
1930[17:47:35] <n4dir> someone spoke about lxqt ? the other day.
1931[17:47:36] <EdePopede> jim: iirc kde was the first de i ever
used
1932[17:47:37] <freem> it's not the worst, but it's far
from ideal
1933[17:48:03] <n4dir> no idea if you could get away from gtk by
such choices though
1934[17:48:07] <freem> the only ncurses applications I really use
are vim and aptitude, basically. Aptitude is a tool I often want to
rewrite, and vim is not really ncurses
1935[17:48:21] <EdePopede> n4dir: another entry on my list, i was
really waiting for something lightweighted based on Qt. which i
prefer over Gtk anyway.
1936[17:48:43] <n4dir> EdePopede: the guy who spoke about it
considered it more than just ok, without fanboyism.
1937[17:48:53] <EdePopede> htop and mc are also open here
constantly.
1938[17:48:55] <n4dir> but that is really just hearsay
1954[17:51:25] <n4dir> and quite some non gtk/qt i seem to recall
1955[17:51:34] <freem> antix?
1956[17:51:38] <n4dir> yup
1957[17:51:39] <EdePopede> even the default settings for the open
dialog are horrible. and i have to set them every time to be
different. and then there's a single mini menu for the selected
file and the dialog itself.
1958[17:51:56] <freem> firefox's UI sucks.
1959[17:52:12] <freem> tbh, most webfuckers UI sucks...
especially when they are mainstream
1960[17:52:21] *** CommunistWolf is now known as CapitalistWolf
1961[17:52:28] <jhutchins> freem: That must be why millions of
people use firefox.
1962[17:52:30] *** CapitalistWolf is now known as CommunistWolf
1963[17:52:39] <EdePopede> jhutchins: there are no real
alternatives
1964[17:52:49] <freem> jhutchins: millions of people have always
used windows. Do you think windows UI rocks?
1965[17:52:57] <n4dir> jhutchins: or as there simply is no
alternative?
1966[17:53:03] <EdePopede> freem: compared to Gtk? :P
1977[17:54:33] <EdePopede> i have this Borland C++ 4.0, the docs
include some MS documentation (help files).
1978[17:54:37] <freem> jhutchins: there's always something
good to say to everything. For example, it can be used. Painfully,
yes, but it can be used, so that's a good thing.
1979[17:54:59] <EdePopede> some of them are hard to find in the
hierarchy, but i've read good things in them back then.
1981[17:55:50] *** Lord_of_Life_ is now known as Lord_of_Life
1982[17:55:58] <EdePopede> fileman wasn't that bad, only it
was not needed anymore after norton commander didn't switch the
platform. so it was killed.
1983[17:56:13] <freem> when I was still a windows user, I used
replacements for almost everytools there. It was before vista when
they introduced that weird thing taking height on wide screens...
1984[17:56:23] <EdePopede> at least it was definitely better than
dosshell, if anyone remembers.
1987[17:57:05] <EdePopede> freem, sure. VGA driver, Total
Commander, IrfanView, some text editor, netscape/opera. and then it
was usable :)
1988[17:57:07] <n4dir> whenever i run in Windows i install all
apps i use in Linux. Once that weekend is over, i wonder why bother
running Windows at all
1989[17:57:09] <freem> well, it was a bit like mc and the likes,
so people can build themselves an idea even nowadays
1990[17:57:22] <freem> EdePopede: add guipro/hulk, too
1991[17:57:33] <EdePopede> huh? don't know this one
1992[17:57:36] <freem> 7zip was a better file explorer, too
1994[17:58:19] <EdePopede> but i had someting sitting in the
corner of the screen with buttons, some system infos, looked a bit
futuristic. norton dashboard maybe
1995[17:58:21] <freem> basically, it added the $mod+click to
move/resize windows, and other nix-like thingies
1996[17:58:47] <freem> since when you start using them, it's
hard as hell to go back on standard windows
1998[17:59:20] <freem> "When some tools exists to move the
windows by pressing a key and clicking with the mouse ; HulK provide
a set of features to help the management of the windows."
replaced-url
1999[17:59:33] <freem> it's part of my survival toolkit,
even if I didn't had to use it since ages
2017[18:03:45] <freem> I mean, MS-DOS 6 and above were usable.
The thing they put in winXP... was more a torture tool
2018[18:03:54] <EdePopede> and windows adapted some of the
features over the decades. it only takes like forever. virtual
desktops, panels, desktops bigger than the physical resolution.
2019[18:03:57] <azeem> that sounds off-topic?
2020[18:04:20] <Azrael_-> i want to run postgresql 9.6 side by
side with postgresql 11 on buster, is this possible by just install
both of them at the same time or do i have to do something special?
2022[18:04:50] <azeem> Azrael_-: it is possible, Debian just
ships only one major version of Postgres
2023[18:04:54] <azeem> !postgresql
2024[18:04:55] <dpkg> PostgreSQL is a robust, mature and
full-featured relational database implementing a large subset of
<SQL>. Supports views, stored procedures, triggers and
scripting. Packaged in Debian, read
/usr/share/doc/postgresql-common/README.Debian.gz ("for the
impatient").
replaced-url
2033[18:08:27] <Azrael_-> that was the suggestion back then from
this channel when i needed 9.6 due to an old software incompatible
with the newer versions. didn't know it better back then. the
old software is still running but want to try out a new software and
don't want this an old postgres because in the end i'll
try to get rid of the old postgres and don't want to have to
migrate anything again
2034[18:09:01] *** Joins: conta (Thunderbir@replaced-ip)
2114[19:14:57] <zorrow> hey, i'm trying to figure out why my
vm can't reach the internal dhcp (virt-manager with
libvirt-qemu). for whatever reason, every single vm doesn't get
an ip, but if i ifdown enp1s0 it tells me it releases ip
192.168.122.xxx
2118[19:19:06] <jim> GNU\colossus, this is an older pentium, it
probably needs pae, but I'm not sure of that (the kernel he has
running is a 686-pae, not 64 bits, 32 bits)
2124[19:24:35] <petn-randall> jim: You can't directly
upgrade from wheezy to buster, you need to go wheezy → jessie
→ stretch → buster. Also, IIRC Intel 80586 support was
dropped somewhere on the way, so you won't be able to upgrade
all the way.
2125[19:25:25] <sney> "older pentium" could also refer
to a handful of budget chips from the mid 00s
2126[19:25:36] <sney> (thanks for being consistent, intel!)
2151[19:41:30] <zutat> anyway. the package descriptions also
mention "modern PCs" and "older PCs". maybe the
pae version also drops support for ancient hardware
2152[19:42:15] <jim> jelly, yeah, he's only got 1 g... is an
upgrade from wheezy to buster (not directly of course) likely to
work?
2153[19:42:29] <jim> we finally got the archive url working
2226[20:40:58] <MikeDebian> guys.. most of the times after using
a behringer UCA222 audio interface my headphones stop working as
they should... only after several reboots / shutdowns, plugging and
unplugging, it eventually comes back working (with debian 9). now
with debian 10 it also eventually comes back to work but then
without doing anything (just closing and opening a few simple apps
to listen to the audio, such as video app, vlc, etc) it stops
working
2227[20:41:08] <MikeDebian> then I hear some clicks on my
headphones from time to time
2228[20:41:39] <MikeDebian> can't see anything wrong with
pavucontrol and alsamixer
2229[20:41:46] <MikeDebian> headphones are recognized
2230[20:41:50] <MikeDebian> and selected
2231[20:41:56] <sney> if that's with pulseaudio, try this:
replaced-url
2232[20:42:02] <sney> (it improved things with my focusrite as
well)
2233[20:43:16] <MikeDebian> sney, i'll have a look
2234[20:43:48] <MikeDebian> btw, which version of focusrite do
you have?
2247[20:54:22] <MikeDebian> sney, it didn't fix my problem..
i rebooted with headphones unplugged... put some audio to play with
"videos" app it works on speakers, plug the headphones, it
works with the headphones. then close the videos app, open chrome,
visit youtube and play some video, I hear a click on the headphones
and no more sound again
2249[20:54:54] <MikeDebian> it was working nicely before using
the behringer audio interface... this is so frustrating... happened
with debian 9 as well (not with debian 8 I beleive)
2269[21:15:35] <lowin> Hello. I have setup initramfs dropbear on
my system to allow unlocking crypt_root remotely. but after I run
"cryptroot-unlock" it unlocks the root partition and
closes the connection. there are still some non root partitions that
I would have to unlock after root. Is there a way to do that?
2270[21:16:29] <petn-randall> lowin: You can always ssh in and
unlock those then.
2271[21:16:56] <lowin> I can't because sshd doesn't go
up before mounting fstab
2272[21:17:58] <lowin> I should probably switch to a keyfile
unlocking mechanism after mounting root
2297[21:38:00] <MikeDebian> im trying to stop pulseaudio to start
it in debug (verbose) mode to check what might be happening but it
auto spawns every time the process is killed
2298[21:38:09] <MikeDebian> pulseaudio does not seem to have any
systemd service
2302[21:39:42] <jmcnaught> MikeDebian: try "systemctl --user
stop pulseaudio.service". There is also a pulseaudio.socket
user unit that you might also need to stop.
2319[21:48:53] <jmcnaught> MikeDebian: you could use
"systemctl --user --full edit pulseaudio.service" and
change the ExecStart= line to add --verbose, then restart
pulseaudio, and the messages will be in its journal
("journalctl --user --unit=pulseaudio.service")
2327[21:52:21] <MikeDebian> I didn't need to do that
though.. being fast enough before it starts
"/usr/bin/pulseaudio --start --log-target=syslog" did the
trick
2335[21:58:54] <jhutchins> MikeDebian: Under SysV there were
often config files where you could specify options for system-level
programs. Look in /etc/defaults, maybe dpkg -l <pulse audio
package> . Look in /usr/sare/doc/pulse*
2366[22:22:15] <alexrelis[m]> slop: I use Element with the IRC
bridge.
2367[22:22:52] <slop> Thanks - I noticed irssi hasnt been updated
since ~2019 or so. was trying to determine if its still supported
2368[22:23:06] <alexrelis[m]> A lot of people advise not to but
bridging has some benefits. One of which is the ability to view
message history for when you're not online.
2369[22:24:17] <apollo13> you do not need bridging for that
though, any bouncer lets you do that as well
2370[22:24:52] <slop> ok going to switch from Irssi, the lack of
updates concerns me :D (paranoia!)