18[00:10:17] *** Quits: invra (~invra@replaced-ip) (Remote host closed the connection)
19[00:10:17] <rekishi> invra: my buster box gives software full
speed up, but when running speedtest or any user DL from outside
network, it throttles to 4Mb
20[00:10:18] <sney> rekishi: that other guy was right that my
idea was a shot in the dark, but if you're having network
throughput issues even when nothing is running, it still suggests a
lower level issue.
21[00:10:50] <rekishi> sney: i checked iftop, and transmission
is outputting at 35Mb (which I pay for)
28[00:13:37] <rekishi> mason: so the entire system was on
hardrives on a dell workstation ported to a new barebones chenbro,
on the previous hardware I had no issues with throttling or
transmission choking out other processes, new hardware -- new
issues, which do not persist in other operating systems
29[00:13:54] <rekishi> sney: are there buster compatability
issues with 5.4?
42[00:18:58] <invra> sney: wah i typed /me gives a beer to
sney.. and all of a sudden weechat crashed, screen crashed and i got
kicked out of aws ssh istance... lol wtf
106[00:54:41] <annadane> (note that that's su - and not su
and yes there is a difference
107[00:54:43] <annadane> )
108[00:54:45] <annadane> !buster su
109[00:54:45] <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...). To approximate the previous behaviour, put
"ALWAYS_SET_PATH yes" in /etc/login.defs. See
replaced-url
168[01:41:08] * annadane has heard some horror stories about btrfs,
people prefer zfs
169[01:41:21] <EdePopede> Gigglebyte: does the chromium-sandbox
(iirc) really run SUID root? i just looked into the .deb, and was
left with the impression it needs it.
170[01:41:42] <somiaj> zfs has only been in linux for a while,
and at one point I thougth there was some push to make btrfs the
core filesystem, but that didn't happen.
171[01:41:51] <hierbat> thank you, it is interesting to see
what's out there but I'll stick with the default install
for now then!
172[01:42:05] <hierbat> yeah seems like btrfs has been around
for 12 years too
173[01:42:21] <somiaj> with lvm, you can get some of the volume
capaiblities of zfs/btrfs as well, and have some additional
features.
230[02:38:04] <velix> Interesting, Debian doesn't have
/etc/dir_colors, but directly `eval`s the output of dircolors. When
I need to change the colors for directories, I need to use `sed`. I
think, this is a bug.
231[02:39:44] *** Quits: jinmiaol2 (~jinmiaolu@replaced-ip) (Remote host closed the connection)
287[03:10:01] *** Quits: leorat (~leorat@replaced-ip) (Remote host closed the connection)
288[03:10:14] <ksk> ..And there have been lengthy discussions
about if https actually protect you in anyway, let alone hide which
packages you are downloading (it might now, to my knowledge)
291[03:14:00] <dragonfly28362> indeed i just installed
apt-transport-https now. to use the https how do i have to upgrade?
with apt the ordinary way? apt upgrade?
292[03:15:19] <dragonfly28362> if i use synaptic package
manager, witch i prefer, i believe it is https automatic right?
319[03:29:04] *** Quits: citypw (~citypw@replaced-ip) (Remote host closed the connection)
320[03:30:28] <x09x> there's a solution to this but the
codebase of apt must be refactored and many distros & systems
will break, so yeah that's why Debian still uses http
321[03:32:42] *** Quits: kaivai (~kaivai@replaced-ip) (Remote host closed the connection)
324[03:34:50] <somiaj> that is a bit old, and maybe as more and
more mirrors get valid certs (with letseyncrypt this is now less of
a pain and issue than it use to be), maybe it will be enabled by
default, I was just sharing some info about why debian has stuck
with http, and the safeguards they have in place, and https is
avilable for mirrors which support it
341[03:55:45] <cyveris> x0n: Who gives a shit if you're
installing metasploit? No one cares about you, or anyone else that
might be installing metasploit.
349[03:56:52] <dpkg> A backport is a package from a newer Debian
branch, compiled from source for an older branch to avoid dependency
and <ABI> complications.
replaced-url
363[03:59:47] <cyveris> ksk: There is on oftc, according to
debwiki
364[03:59:57] <ksk> kk.
365[04:00:16] <somiaj> cyveris: are the matainer of the package,
or someone else making a backport?
366[04:00:22] <cyveris> Someone else.
367[04:00:54] <kline> cyveris, i do think its relevant that some
people may not want others on the path between them and the mirror
to know *what* packages are being downloaded
368[04:00:59] <kline> and reasonable
369[04:01:03] <nvz> I just had a random curiousity about
something.. I'm guessing programs like Transmission
(bittorrent) do this because they're downloading things
non-sequentially.. but it seems like it reserves the space
beforehand.. but I dont really understand fully the how/why of it
and it confuses me when trying to figure out if I have enough disk
space and torrents are downloading taking up space they dont occupy
370[04:01:16] <somiaj> cyveris: Have you ever been a matainer?
anyways, I'm looking up the details, but as a non matainer
backport I think it is a bit more work.
371[04:01:17] <nvz> I'm not really even sure if its
allocating it all up front or what
372[04:01:29] <cyveris> somiaj: I have never been, sadly. :/
373[04:01:43] <cyveris> I might hop over to oftc and inquire
there.
374[04:01:57] <ksk> nvz: it most likely just pre-allocated for
every file. look for an option in the client?
375[04:02:16] <nvz> yeah that'd make it even more
complicated
376[04:02:35] <somiaj> cyveris: I know the process to get
packages into debian better than backports, so I am seeing if I can
find any useful information, otherwise I'll point you someplace
you might be able to answer that question.
377[04:02:44] <cyveris> Thanks!
378[04:02:55] <nvz> like right now the drive I have mounted on
/home on that machine is down to like 21G used.. between active and
pending downloads I probably have near 100GB to download
380[04:03:21] <ksk> utilize `ncdu' to see where all the
space has gone ;)
381[04:03:30] <ksk> most likely "ncdu -x /home"
382[04:03:31] <nvz> I'm almost certain it'll all fit
on there.. but what I'm not certain is, wht has been allocated
already and what hasnt
383[04:03:31] <kline> nvz, "torrents taking up space they
dont [*yet*] occupy". technically it should be easier as you
dont have a moving target. if you have 20GB on your torrent
partition, and 1 torrent in flight currently at 2GB used, will your
new 10GB torrent fit?
386[04:04:10] <nvz> I know exactly where the space has gone :P
387[04:04:17] <kline> nvz, if the in-flight torrent in 5GB, then
yes. if it's 15GB, then no. if the torrent client preallocates
the 15GB is needs upfront for that torrent, then you have an
accurate representation immediately
388[04:04:27] <nvz> what I dont grasp is space reserved and not
yet occupied
389[04:04:45] <kline> i mean, it's just the opposite of a
sparse file:
replaced-url
390[04:05:16] <nvz> kline: it seems like thats whats occuring,
that when the torren starts to download it pre-allocates and file
managers show the space as no longer available.. for the entire
torrent
391[04:05:41] <nvz> where it gets complicated is when you have
two active, two queued but already started to download some, and two
not yet started..
392[04:05:46] <kline> but the most trivial thing to do is if you
have a 10GB torrent file coming down, to open a file, write 10GB of
0s, then update by overwriting each chunk as it arrives
393[04:05:48] <nvz> trying to figure out whats what
394[04:05:59] <kline> very naive though
395[04:05:59] <nvz> I know it'll all fit cause I trust my
calculations before I started all this
396[04:06:16] <nvz> but what I dont like is not being able to
look at it and tell whats going on :P
397[04:06:21] <nvz> and not really understanding why
398[04:06:24] *** Quits: Tom01 (~tom@replaced-ip) (Remote host closed the connection)
399[04:06:33] <nvz> I vaguely grasp whats happeneing..
400[04:07:09] <kline> remember that the torrent client isnt just
guarding against you messing up
401[04:07:31] <somiaj> cyveris: This is what it appears to me,
first you should be able to matain and support that package through
the life of buster, this means being able to update it as testing
gets updates and security fixes. After that you should file a
wish-list bug and ask the matainer (maybe provdie them whave you
have done). If the matainer doens't want to matain the
backport, you can then choose to do so, and it
402[04:07:32] <kline> pre-allocating disk space means that
things like log-files will get errors for running out of disk space
that the torrent client knows it will need
403[04:07:37] <somiaj> appears the process for a non-matainer is
similar to any other package and you get to use mentors.debian.net
to upload it to and get a sponsor.
404[04:07:44] <EdePopede> if it's possible to write a
specific block of the file then it would definitely make sense.
"start at $offset and write these $size blocks".
405[04:07:53] <kline> it would be pretty frustrating if a
torrent failed because a log file grew oversized
406[04:07:55] <kline> EdePopede, you can
407[04:08:14] <somiaj> cyveris: here is the basic guide lines
for getting a package into debian via a mentor,
replaced-url
408[04:08:16] <nvz> kline: yeah I get why it'd be
beneficial, just not so much the mechanism at play reserving space
ilke that.. only reserve I'm familiar with is the FS level
reserved blocks for root..
409[04:08:28] <EdePopede> i had already midnight commander
telling me that the disk is full right before the copy finished. not
nice.
410[04:08:30] <cyveris> somiaj: Thanks a ton!
411[04:08:32] <nvz> I also dont really understand the method
used by this client well enough
412[04:08:50] <nvz> I been just assuming it reserves the whole
amount upfront when any data from that torrent first hits the disk
413[04:08:52] <kline> easiest thing to do is go and look at the
source of your favourite client
414[04:09:05] <kline> thatll tell you the actual truth
415[04:09:08] <nvz> yeah.. probably not :P
416[04:09:17] <kline> what client is it?
417[04:09:30] <nvz> pretty sure transmission is not only written
in C, which I barely grok, but its got so many damn pieces..
418[04:09:37] *** Quits: annadane (~annadane@replaced-ip) (Remote host closed the connection)
419[04:10:04] <lwp> nvz, in general the torrent protocol is
randomly downloading small bits of the total file, not sequential at
all, so it makes sense to reserve the entire space at the outset
420[04:10:19] <nvz> lwp: I know, and I said as much on the
outset
437[04:12:46] <kline> it's used here in
"tr_sys_file_preallocate()"
438[04:13:26] <nvz> kline: yeah well I'll have to
investigate it more.. but the dialogue helped me think more on it
439[04:13:39] <nvz> didnt even contemplate the possibility its
doing this per-file
440[04:13:49] <nvz> which if thats the case, thats much harder
to keep track of
441[04:14:03] <kline> im not really sure what you're trying
to achieve
442[04:14:16] <nvz> its trivial to know its gonna pre-allocate
the entire torrent.. I got the torrents total size and size
downloaded so far at a glance..
443[04:14:33] <nvz> but if its randomly during downloading
creating and pre-allocating new files..
444[04:14:41] <nvz> thats just a f'n mess
445[04:15:02] <nvz> I have no way of easily telling when these
chunks of data are of a new file that will be pre-allocated
446[04:15:09] <kline> is this causing you an actual problem,
though?
447[04:15:25] <nvz> not yet.. other than having to watch the
shit too damn closely
448[04:15:32] <kline> too closely for what?
449[04:15:35] <nvz> cause I dont know enough to know if I can
just leave it
450[04:15:51] <lwp> my favored solution to this problem, is to
have WAY more empty disk space than I will need, and then not worry
about it.
451[04:15:52] <nvz> I dont know if I'm out of space or if
my original calculations were correct
452[04:15:54] <kline> are you trying to micro your disk space by
moving files on and off a filesystem that youve overcommitted?
455[04:16:40] <nvz> na, I'm just downloading new stuff
I'm gonna put on my file server eventually.. but my file server
is like 90mi away and here I only have a 1TB drive thats almost full
now :P
457[04:17:03] <kline> the preallocation means that you're
not likely to accidentally overcommit yourself
458[04:17:19] <nvz> I could if its doing it per-file as it gets
to each file
459[04:17:21] <ladderff> Hello, possible regression upgrading to
buster on a macbook pro 5,5: no /dev/video0 for the built-in camera.
Wasn't a problem on stretch!
460[04:17:25] <kline> if you download 5x10GB torrents onto a
40GB drive, itll tell you in advance
461[04:17:35] <nvz> if its doing it as I thought, per-torrent
every time.. thats predictable
462[04:17:59] <EdePopede> <nvz> space reserved for files
not yet written.. thats a foriegn concept to me <-- that's
just like you reserving some space in the house before bringing in
all the furniture
463[04:18:06] <nvz> I then know my 21GB remaining is 21GB
remaining minus the remaining undownloaded portions of any torrents
464[04:18:07] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
465[04:18:24] <kline> nvz, just pretend that all your files
download instantly
466[04:18:29] <nvz> cause right now I got 3 torrents with
progress.. but only two are downloading one is queued
467[04:18:29] <kline> (just you cant use them until later)
468[04:18:42] <nvz> so I am not sure if all 3 of those are
accounted for or what
469[04:19:21] <EdePopede> the torrents are quarantined until
everything is in place
470[04:19:38] <nvz> cause I had put a roughly 5 and 10 GB
torrent in there as well.. those have no progress so far.. and
should fit in the 21GB or so remaining easily
471[04:20:04] <nvz> but if any of the not yet downloaded files
in the current two torrents or queued partially downloaded torrent
are still not accounted for..
473[04:21:42] <EdePopede> nvz, i'd check right before and
after starting it. if it happens everytime i'd assume it would
also happen next time you don't check it.
475[04:22:22] <nvz> nm it seems to be treating the pre-allocated
files like they're actually there so I guess its fine.. I can
click a directory its downloading and go to properties and see it
says 6GB used even though the torrent has only downloaded 1GB so far
476[04:22:36] <nvz> so I can actually check with the filemanager
and see the allocated space
477[04:22:38] <kline> if its a 6GB file in the torrent
478[04:22:44] <kline> and it's 6GB "on disk"
479[04:22:48] <kline> and only 1GB has downloaded
480[04:22:57] <kline> you can trust its only ever going to be
6GB
481[04:22:57] <EdePopede> 5 to go?
482[04:22:58] <nvz> I thought it was only showing me that the
space was no longer available, but not showing it actually in use
483[04:23:09] <nvz> but I guess its showing both
484[04:23:21] <nvz> which is fine.. as long as I know how to
tell whats happening
486[04:23:59] <EdePopede> does the client have its own
bookkeeping on the reserved space? so that if something fails it
would delete this partial file and free all the space (or continue
later, whatever)
487[04:24:10] <nvz> kline: I mean clicking on the parent
folder.. any good torrent with more than a file typically is inside
a directory
488[04:24:38] <nvz> EdePopede: that I wasnt sure of either..
hence the current issue where I have things I've started but
then queued
489[04:24:46] <nvz> are they garbage collected when inactive?
490[04:25:06] <nvz> but now I got a better handle on it to
figure it out I guess
491[04:26:07] <EdePopede> as long as the fs doesn't have a
"reserved" flag for the blocks but only "used"
then i see no other option than letting the client handle this.
492[04:26:09] <nvz> I oft micromanage torrenting.. I'll
turn crap up as far as number of peers and simultaneous downloads
then turn it down.. sometimes start things above my simultaneous
downloads when one torrent is showing little bandwidth usage.. etc.
493[04:26:18] <nvz> I just wanna know how to go about this when
space is an issue
494[04:26:46] <nvz> up to now I had no way of knowing for sure
out of paused, queued, downloading torrents exactly what was
accounted for already and what wasnt
495[04:27:01] <nvz> now I got a better idea and know how to
figure it out :P
496[04:27:20] <kline> nvz, best would be to ask the
community/developers of your client
497[04:27:23] <nvz> idk what made me think the file managers and
stuff wouldnt see the space in use
498[04:27:25] <kline> they will know the actual behaviour
499[04:27:26] *** Quits: n3p (~n3p@replaced-ip) (Remote host closed the connection)
500[04:27:33] <nvz> cause they sure show it not available
anymore
501[04:27:34] <EdePopede> maybe downloading a segment, tagging
the block as work in progress, data transfer, tagging as used.
502[04:27:36] <kline> i can tell you that you can ask linux to
preallocate you files
503[04:27:46] <nvz> I just couldnt grasp how to find usage of a
file thats not there yet :P
504[04:28:04] <nvz> now I know I can just click the parent
directory and see its usage.. it should equal the total torrent
505[04:28:16] <nvz> despite how much the torrent client said is
downloaded
506[04:28:31] <EdePopede> yeah, midnight commander also has a
Preallocate Space option
509[04:29:02] <nvz> it dodnt seem to me like it worked like say
chrome/chromium where it just creates a file then fills it up
510[04:29:11] <nvz> calls em .crdownload until they're
finished
511[04:29:18] <EdePopede> gah
512[04:29:27] <nvz> seemed like it was doing some black magic
shit to me :P
513[04:29:29] <EdePopede> or .part like youtube.dl
514[04:30:27] <nvz> but mind you, I was one of those kids who
when they were teaching us base 10 in grade school.. you got the 1s,
10s, 100s, 1000s place.. you put a number in this place and.. blah
blah blah
515[04:30:36] <EdePopede> there was a user once in firefox with
a networked /home and the browser was downloading to /tmp first and
then transferred it over the network. not really effective.
516[04:30:45] <nvz> then like 3rd grade came around and they
dropped this 100,000 thing on me and I just couldnt grasp it
517[04:30:53] <nvz> cause 100 comes before 1000
518[04:30:59] <EdePopede> heh
519[04:31:00] <nvz> hundred-thousand sounded ridiculous to me
522[04:31:28] <EdePopede> because you're english
language-wise
523[04:31:44] <EdePopede> germans are used to it, 13 already is
three-ten :P
524[04:31:57] <nvz> yeah and that means we were never taught not
to try make sense of the language
525[04:32:09] <nvz> which is a futile struggle
526[04:32:22] <nvz> we drive on parkways and park on driveways
527[04:32:30] <EdePopede> if you *really* want to do this you
don't get around its history
528[04:32:35] <EdePopede> heh
529[04:33:49] <gZen0n> Hi, I need to move/copy/delete several
files recursively, all formatted with a same pattern (^FSCK[0-9]*)
but located in diferent directories with undefined depth. I can list
them with "find . | grep mypattern", but how could I work
on them in the same time ? I want to copy them for bkup, delete
those which are 0o sized and rename others. Thanks
530[04:34:18] <nvz> find supports an exec option
531[04:34:25] <nvz> there is no need to pipe with find
532[04:34:46] <nvz> I am no find guru though so I will not
comment further on the specifics
533[04:34:59] <gZen0n> Ok thanks
534[04:35:37] <nvz> but really to someone who is savvy with
find, its a trivial oneliner, and fwiw #bash is often quite handy
for such questions
535[04:35:38] <EdePopede> like: -exec rm {} \;
536[04:35:56] <EdePopede> or + instead of \; if it's ok to
do it in parallel
537[04:36:34] <nvz> they dont seem to mind questions related to
shell utilities as they're just like libraries of sorts
extending the language..
538[04:37:04] <nvz> either way folks in #bash would all be more
savvy with use of the shell for such things if you can't get
help here
545[04:39:12] <gZen0n> @EdePopede, Yes, I was going to use
exiftool to get the extension for renaming files, does it seems ok ?
546[04:39:18] <nvz> gZen0n: na, if you're on debian its a
debian question as much as a bash one.. but what I do is ask myself
is my question specific to one or another.. and in this case the
answer is no.. so the next question is, where are folk likely to
know/help the most?
547[04:40:31] <nvz> like I personally while I do not condone
support of non-debian distros, I do think armbian debian versions
can be supported cause they use our packages unmodified.. their
repos contain only kernel, driver type stuff.. so if your question
isnt hardware/kernel related, its applicable and reproducible here
548[04:40:39] <EdePopede> gZen0n: filetype recognition? gives
you more info than just file(1) i guess, no idea if this is really
needed
549[04:40:49] <nvz> they literally use our repos.. they dont
even mirror them
550[04:41:04] <nvz> their repos have like a dozen or so packages
thats it
552[04:41:44] <nvz> though on the other hand I dont really
condone kde support even on pure debian in here :P
553[04:41:58] <nvz> cause most of us just dont use it, and
you're better off in #debian-kde :D
554[04:43:03] *** Quits: jinmiaol2 (~jinmiaolu@replaced-ip) (Remote host closed the connection)
555[04:43:09] <gZen0n> @EdePopede, as the pattern implied, these
are recovered files after a ddrescue. and "file
--extension" answers ???, while "exiftool -T" answers
right extensions
556[04:43:16] <nvz> just gotta know both what will suit your
situation best as well as respect the supporters.. this question was
a case where it was just as applicable one place as another but one
place (here) is rather idle and only a subset of the folks here will
have firm grasp of advanced shell usages
562[04:45:13] <nvz> gZen0n: fwiw there is also globstar and
extglob you can ask about in #bash or lookup.. they're shell
modes that expand the capabilities of globs to allow things like
recursion
563[04:45:33] <nvz> I've used it before but its not strong
enough in my toolkit to know it off the top of my head
564[04:45:36] <gZen0n> @EdePopede, Yes, i'm reading file
manual, and I see mime-type option, but the output format is not
"raw"
565[04:45:51] <EdePopede> xsx i got a hit! mpv-shot0001:
jpeg/jpg/jpe/jfif
566[04:46:05] <nvz> with an extended glob doing recursion you
don't need find per-se
567[04:47:16] <nvz> thats one of the odd bits about bash.. is
that it has all these shell modes that change the syntax and
behavior of the language :P
568[04:47:39] <EdePopede> btw, "find | grep". if the
output of ls is not to be trusted, what about find?
569[04:47:45] <nvz> you manipulate them with shopt if
you're not familiar
570[04:47:50] <gZen0n> @nvz I didn't expect #bash was more
active, godod to know
571[04:48:17] <nvz> gZen0n: yes both #debian channels have
gotten slower in recent years except during certain times of the day
572[04:48:32] <nvz> I sit in #bash too among others, and its far
more active
573[04:49:07] <gZen0n> @EdePopede, find is recursive in whatever
depth
574[04:49:25] <EdePopede> gZen0n: it was about weird filenames
iirc
575[04:49:41] <EdePopede> like they could include \0 or line
breaks
576[04:51:10] *** Quits: rare_energy (~quassel@replaced-ip) (Quit: who knows)
579[04:52:47] <gZen0n> Well thank you for your suggestions, I
will then write a script ( I was hoping to find a "one-line
command", as I didn't script for a while :P)
612[05:04:42] <nvz> -s1,8 are most common for commands 1 is
user, 8 is root, 5 is config files also.. beyond 1,5,8 the rest is
more programmer stuff :P
613[05:05:58] <EdePopede> are the others ones still in use? with
n,p,s added or other stuff?
636[05:23:01] <velix> It's damn hard to work with
debian.org, since the pages automatically switch to my native
language, which is often translate a bad way.
676[06:22:03] <colemickens> Is there something wrong with the
current Buster live images? Gnome-shell is completely unusable in a
qemu/kvm VM. I use this same config regularly with other
debian-based distros running gnome-shell without issue/
689[06:35:26] <colemickens> oh sorry, I thought you asked if I
was running unstable for some reason.
690[06:35:56] <colemickens> Yeah, the CPU gets pegged to 80%
basically the entire time and the shell just sort of hangs at some
point. Not just slow but all input ceases to work. Yet, I used Tails
a while back and it worked fine in the same environment.
712[06:56:13] <rottentomatoes> colemickens: in low resources pc
you must rely on lots of swap wich is trouble if is live, lots of
ram is greatfor performance cpu come in second place
719[06:57:41] <colemickens> Yup, worked like a charm
720[06:57:42] <rottentomatoes> gnome is heavier
721[06:58:07] <colemickens> I understand. But I don't
understand why gnome-shell would work fine on a downstream of
debian, but not in debian. I doubt the Tails forlks are
perf-patching gnome.
778[08:10:06] <pingfloyd> colemickens: there's also contrib
and non-free which are optional
779[08:11:25] <pingfloyd> colemickens: it contains the one,
because that's the most floss one. floss as possible by
default, and then the user can decide how much they want to taint
their installation.
780[08:11:42] <colemickens> let me be more clear
781[08:11:47] <colemickens> the live disk is missing the
security repo
782[08:11:54] <colemickens> is that intentional, or am I
mistaken?
783[08:11:57] <pingfloyd> oh
784[08:12:10] <pingfloyd> if it's stable, you should add it
788[08:13:13] <pingfloyd> that is, we get our security updates
there, through upstream
789[08:13:22] <colemickens> hm, okay. I'll add this to my
list to circle back on, I was under the impression I was on a stable
disk. (seems low pri, I assume it's correct after an install)
795[08:14:26] <pingfloyd> I suspect it is an oversight, or the
image being used, isn't very well handled.
796[08:15:42] <pingfloyd> in short, it's a bad idea to not
have the security repo with stable, because of how updates are
maintained between the branches.
797[08:16:42] <pingfloyd> i.e., new stuff gets added to sid and
if it passes, trickles down into testing. Stable happens after
testing is frozen and moved over.
798[08:17:21] <pingfloyd> so less of an issue on the rolling
side of it, but huge one on the stable release side.
799[08:18:29] <pingfloyd> stable is extremely dependant on
security backports (what the security repo is full of) in
maintaining a decent level of security
800[08:19:19] <warsoul> pingfloyd a good sources.list how it
should look
802[08:20:34] <fury__> trying to install debian in a bhyve VM.
I've mounted the installer ISO, and for some reason I'm at
a bootloader. the prompt says "OK ". very little/confusing
google results, since googling the word "OK" is nearly
impossible, and I don't know what this is other than I can
guess it's the bootloader. how do I load the kernel and boot
into the installer?
817[08:25:17] <fury__> I am at a bootloader, and need to know
how to load the kernel. really simple, impossible to google. all I
need to know is how to load the kernel.
818[08:25:27] <fury__> pingfloyd: yes, I've used this image
before. it's fine.
819[08:25:36] <pingfloyd> fury__: then this must be an issue
with your hypervisor
820[08:25:39] <fury__> again, just need to know how to load the
kernel. nevermind why I'm at a bootloader.
821[08:25:52] <pingfloyd> fury__: possibly your setting for the
VM
822[08:26:00] <fury__> generally, the kernel needs to be loaded
manually for bhyve.
823[08:26:14] <pingfloyd> bootloader is what loads the kernel
824[08:26:21] <fury__> I just need to know the command to load
the kernel. again, really simple, impossible to google. I'll
wait for somebody who knows how to load the kernel manually from the
installation ISO.
825[08:26:35] <fury__> let's not get into discussions that
aren't relevant
826[08:26:43] <mathgrad> try 'linux vmlinuz-*'
827[08:26:50] <mathgrad> dunno if it'll work
828[08:27:05] <fury__> linux is an "unknown command".
that's probably a grub thing, this prompt says "OK "
829[08:27:07] <pingfloyd> fury__: does your machine use EFI?
830[08:27:50] <pingfloyd> or I should say the VM
831[08:28:09] <pingfloyd> I think you're rather SOL if it
doesn't
832[08:28:51] <fury__> the vm works fine generally. I'm at
the bootloader. let's not talk about VMs. I just need to know
how to load the kernel at an "OK " prompt.
833[08:28:52] <rottentomatoes> isnt it a freeBSD issue?
replaced-url
834[08:29:01] *** debhelper sets mode: +l 1240
835[08:29:26] <fury__> again, I just need to know how to load
the kernel from an "OK " prompt, and I can handle the
rest.
836[08:29:35] <fury__> it is not a freebsd "issue"
837[08:29:44] <fury__> it's not an issue at all, just need
to load the kernel from the bootloader.
847[08:32:11] <pingfloyd> that sounds like a pretty ridiculous
limit
848[08:32:29] <jim> fury__, I googled for 'boot loader witn
"ok" prompt', and there's a good chance
it's the freebsd boot loader
849[08:33:18] <pingfloyd> 'sysutils/grub2-bhyve is required
to run Linux or any other guests that need a Grub bootloader.'
850[08:33:29] <jim> I can't go down the path of which one,
I don't know what all the versions there are
851[08:33:37] <pingfloyd> 'sysutils/bhyve-firmware is
required to run UEFI guests'
852[08:33:40] <fury__> yeah that's what I'm using
853[08:33:48] <pingfloyd> looks like those are you two options
here
854[08:34:09] <fury__> yeah I'm a bit further along than
that
855[08:34:20] <fury__> looks like this might be a task for
tomorrow hah
856[08:34:22] <fury__> thanks anyway
857[08:34:27] <pingfloyd> fury__: beyond that, you should
probably consult their support resources. This is hardly a debian
issue, let alone even a linux issue.
858[08:35:00] <fury__> didn't come here for any of that -
just came here to ask how to load the kernel.
859[08:35:28] <pingfloyd> likely a specific issue with bhyve
860[08:35:41] <fury__> and have tried more than a few times to
steer the discussion away from bhyve/freebsd/et al
862[08:36:11] <pingfloyd> fury__: how to boot a kernel directly,
is dependant upon bhyve
863[08:36:24] <fury__> but so far other than mathgrad suggesting
something that didn't work for me, have only received input on
things I didn't ask about
864[08:36:26] <fury__> no, it isn't
865[08:36:30] <pingfloyd> fury__: the facilities it provides to
do so
866[08:36:34] <pingfloyd> fury__: yes it is
867[08:36:35] <fury__> I'm asking how to boot the linux
kernel from the linux bootloader
868[08:36:43] <fury__> that is the entirety of my question
869[08:36:48] <fury__> and I'll again steer the discussion
away from bhyve
870[08:36:58] <pingfloyd> it's not a debian issue
871[08:37:14] <fury__> it is, because again, I'm not asking
about anything related to anything but linux
872[08:37:19] <pingfloyd> do you go to the pet store for work on
your car?
873[08:37:22] <fury__> I am asking how to load the linux kernel
from the linux bootloader
874[08:37:26] <fury__> which part do you not understand
875[08:37:32] <pingfloyd> because you like to give your pets
rides?
876[08:37:37] <pingfloyd> fury__: you're not getting it
877[08:37:54] <fury__> how do I boot the linux kernel from the
linux bootloader
878[08:37:59] <fury__> literally the only question I'm
asking
879[08:38:04] <fury__> I can take care of the rest
880[08:38:16] <fury__> I do not have a single question about
bhyve or freebsd
881[08:38:20] <warsoul> pingfloyd i have a debian9 vps how can i
upgrade to debian 10
882[08:38:24] <fury__> believe me I know how to boot a freebsd
kernel from the freebsd bootloader
883[08:38:31] <fury__> my question has absolutely nothing to do
with that
884[08:38:34] <warsoul> without loosing the data
885[08:38:34] <jim> fury__, can we zoom out for a moment, what
is it you're trying to do?
886[08:38:34] <pingfloyd> this isn't freebsd
887[08:38:39] <pingfloyd> fury__: get it through your head
903[08:41:24] <dpkg> #debian is for Debian users and
Debian-related questions. We aren't experts on other
distributions, and it gets frustrating when people from other
distributions ask for help in here. Use your common sense and ask in
a channel where they know your distribution.
904[08:42:00] <fury__> pingfloyd: you're not paying
attention. you're so hung up on being right, that you're
not bothering to read what I'm saying. I gotta go, thanks
anyway.
915[08:54:12] <EdePopede> pingfloyd: i'd blame the lockdown
at least partially. people getting frustrated, going nuts, kids
looking for some fun or what they consider being some.
916[08:54:38] <EdePopede> and the idiocracy level rising in
general
1049[11:16:47] *** Quits: grumble (~grumble@replaced-ip) (Quit: The bacteria inside Neil Armstrong were the first
non-human life to land on the moon. And Neil Armstrong was their
space suit.)
1062[11:31:29] <jim> my personal view, I run debian because I
know debian devs work their ass off, and through that and through
their analytic ability, they make debian a very high quality dist
1069[11:36:22] <jim> does that have to do with the quality of
suse? I don't know, but I don't think so... quality of a
dist depends on how trustable each person is, and ultimately how
trustable each software item is
1080[11:48:26] <brutser> jim: i just wanted to install veracrypt
and since that's not in the official debian packages, i could
either add some repository or download .deb from the veracrypt site
and install it
1081[11:49:02] *** Quits: conta (~Thunderbi@replaced-ip) (Quit: conta)
1082[11:49:03] <brutser> that opensuse site has many repositories
for debian as well
1083[11:49:25] <brutser> anyway,, i decided to just grab the .deb
file and install it
1096[11:56:16] <jim> brutser, usually, before the next revision
of a dist is released, the dev/maintainers of that dist will test
the packages together to make sure they all work together; a failure
could mean either the software that failed doesnt' get
included, or, it delays the release until it can get fixed
1109[12:01:58] <ratrace> brutser: separte what you want run once
into a script, then in .xsessionrc [ -f /path/to/that/script ]
&& <execute it> [ -f /path/to/that/script ] &&
rm /path/to/that/script
1110[12:02:17] *** Quits: dez (uid92154@replaced-ip) (Quit: Connection closed for inactivity)
1111[12:02:27] <brutser> ok
1112[12:04:22] *** Quits: tuxmania (~tuxmania@replaced-ip) (Remote host closed the connection)
1138[12:19:34] <shtrb> using rsyslog 8.1901 (debian stable ) ,
how do I properly use properties to select messages from something
which arrive from a different server , I tried
replaced-url
1162[12:48:05] <shtrb> trek00, thanks ! it will stop (I think)
but I wish to use the variables (such code will also take localhost
messages into it's own folder which I do not wish to do )
1163[12:48:12] *** Quits: Tobbi (~Tobbi@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
1169[12:49:40] <shtrb> server side , I get the logs , I just wish
to have remote messages in their own folder and not in the general
logs (but not to change the local logs on the server in any way )
1170[12:50:09] <shtrb> I was working with
replaced-url
1171[12:50:43] *** Quits: conta (~Thunderbi@replaced-ip) (Quit: conta)
1185[13:05:07] <brutser> ah yes, that is because the conky config
file is in old format i need to convert that to the new format
1186[13:05:16] <tomreyn> brutser: notify-send (line 55) is an
on-screen notification utility. you're possibly missing a
relevant message there because it's not installed
1263[13:41:48] <simplicius> but it seems to be corrupted I cannot
see the bar at the bottom of the screen when playing
1264[13:41:52] <knob> tomreyn, wow, that's the best way
I've seen it described. Because usually, if one mentions to
change the port, people start screaming that it's not security.
Yet "reduce the noise" is a good way to describe it. I
like it.
1267[13:42:21] <knob> simplicius, but does it actually run?
1268[13:42:45] <cyveris> knob: When people talk about changing
the port, they're usually presenting it as a security measure.
You're right, it's not. But can it reduce log volumes?
Absolutely.
1269[13:43:10] <simplicius> yes it run but they told me
"there's either something else wrong in your environment
or your binary"
1270[13:43:14] <knob> cyveris, exactly. That reduction is
probably worth it.
1271[13:44:14] <cyveris> knob: For most people, yeah. I'm a
weirdo - I send spurious auth data and IDS events to my
employer's SOC. They use it for training for new analysts.
1322[14:18:56] <Exagone313> Hi, it seems debian buster kernel
doesn't have CONFIG_CGROUP_HUGETLB enabled, which is required
for an app I want to use (cilium in kubernetes). Are you aware of a
custom kernel having that config, or do I need to build a kernel
manually (which wouldn't be nice to automate)? Thanks for your
help.
1332[14:28:12] <brutser> hi, i install the following for audio:
sudo apt install --no-install-recommends -y pulseaudio pnmixer
pavucontrol << but pnmixer is giving some issues, what are
good alternatives?
1368[15:11:11] <trek00> Exagone313: you just download the kernel
version you like (the same of debian for example), copy then modify
the config and compile with make deb-pkg to create a nice debian
package
1372[15:13:42] <jaakkos> networkd seems to remember interface
Name= assigned from .link files after reboot, even if those files
are long gones. How does it remember them? I'd like to get back
to original persistent network names
1395[15:29:57] <trek00> shtrb: try with the if-then, as the
documentation says: rsyslog v7+ does no longer support BSD-style
blocks for technical reasons. So it is strongly recommended not to
use them.
1416[15:46:54] <shtrb> trek00, It seems that I didn't read
correctly the documenation for the compare operator , the doc says
"Compares the “value” string provided and the
property contents. These two values must be exactly equal to match.
The difference to contains is that contains searches for the value
anywhere inside the property value, whereas all characters must be
identical for isequal. As such, isequal is most useful for fields
like syslogtag or FROMHOST, wh
1417[15:46:55] <shtrb> ere you probably know the exact
contents." but I used a property !
1420[15:48:35] *** Quits: Anderson69s (~Anderson6@replaced-ip) (Remote host closed the connection)
1421[15:49:21] <Exagone313> trek00: thanks I'll guess
I'll build kernel, software in question may not be easy to
patch. after installing built packages, do I need to hold kernel
packages to make sure these aren't replaced by repo packages?
1557[17:26:21] <zodd> I am trying to cast to a chromecast device
from a Buster laptop. Regardless of fw settings or even fw off the
caster (starting with the easiest: chromium) the device (TV) is
never found. A windows laptop has no issues; same for android phones
1558[17:26:36] <zodd> am I missing a kernel module or ... ?
1559[17:27:22] *** Quits: TomyWork (~TomyLobo@replaced-ip) (Remote host closed the connection)
1617[18:00:19] *** Quits: nickodd (~nickodd@replaced-ip) (Remote host closed the connection)
1618[18:00:29] <annadane> or ##linux, but generally, people on
irc lurk, so you should probably just wait for an answer in
#armbian, cross-posting isn't a great habit to get into
1619[18:00:37] <annadane> unless i guess you need an answer *now*
1620[18:00:44] <abrotman> and it's OT here
1621[18:02:12] <epsilonKNOT> nah, not in too much of hurry
1622[18:02:19] *** Quits: M1zuki (~m1zuki@replaced-ip) (Remote host closed the connection)
1623[18:02:29] <epsilonKNOT> we used dd like so: dd
if=armbian...img of=/dev/mmcblk0 status=progress
1624[18:02:43] <epsilonKNOT> but then when we screened using
screen 1500000
1625[18:02:53] <epsilonKNOT> it just showed some weird symbols
1648[18:17:50] <vlt> Hello. How can I predict the name of the
first eth device to prepare /etc/network/interfaces? On one machine
I got "enp1s0", on another "ens3" ...
1683[18:29:37] <amprxc> Hello, I'm using debian 10. I'm
trying to hide the output in the terminal when I execute this
command (from my script) : timeout "$_deauth_time_in_sec"
aireplay-ng -0 0 -a $_bssid -c $_station_to_deauth $_interface_name
& > /dev/null 2>&1. But I still see the deauth
messages. How can I hide its ? Thanks
1684[18:29:56] <warsoul> trek00 is a tcl for a eggdrop
1685[18:30:34] <trek00> amprxc: put the & sign after the
redirection
1699[18:39:37] <LtL> warsoul: note, you may have to specify
locations of two tcl files when compiling. It doesn't always
find them, use 'find' to locate them and the syntax to use
when using ./configure .. it will tell you the syntax, very easy.
1910[20:46:28] <sney> missing firmware for the wifi card
1911[20:46:30] <sney> !firmware live
1912[20:46:31] <dpkg> Unofficial <live> images - containing
non-free Debian <firmware> packages - for Debian 10
"Buster" as a live OS are available at
replaced-url
1913[20:46:36] <sponix> Henry151: if you haven't already --
might try the "non-free" it has firmware/drivers not
included in the regular installer
1914[20:46:39] <Henry151> it's broadcom bcm43xx stuff so i
installed firmware-b43-installer
1915[20:47:04] <sney> b43 is a mess and can be a pain to get
working manually. I think the firmware live iso would be a quicker
approach
1920[20:47:25] <Henry151> so i don't really want to start
from a blank one again
1921[20:47:47] <Henry151> anything other than
firmware-b43-installer that i might try that might work?
1922[20:47:53] <sney> !b43
1923[20:47:54] <dpkg> b43 is a Linux kernel driver supporting
PCI/PCIe wireless LAN devices with Broadcom chips. Firmware is
required, ask me about <b43-fwcutter> to provide.
replaced-url
1924[20:48:03] <sney> the wiki page should list a few options
1925[20:48:57] <Henry151> thank you.
1926[20:49:35] <Henry151> and, side question, if i install a
firmware package, is there any way to get debian to "see"
it, without rebooting?
1927[20:49:51] <sney> reload the module in question
1928[20:49:55] <Henry151> like, from the wiki i see there is
b43legacy and so i just installed that, do i have to reboot to see
if it worked?
1929[20:49:58] <Henry151> oh
1930[20:50:01] <Henry151> ok
1931[20:50:09] <sney> so for b43 you would do 'modprobe -r
b43;modprobe b43'
1999[21:39:38] <rozenglass> f8e4: totally possible if it's a
public instance. The internet is a hostile place, tons of bots are
roaming the internet all the time trying to catch easy pray, someone
who might have a username of "admin" and a password of
1234 or something.
2000[21:40:09] <rozenglass> sorry, didn't notice I was
replying to a very old message
2005[21:42:55] <rozenglass> Generally, I think it is recommend to
use fail2ban to automatically block IPs doing too many bad ssh login
attempts, and disable password login and use an ssh key instead
2006[21:43:00] *** nsegkos is now known as nksegos
2009[21:44:14] <sney> vps providers often offer a configurable
firewall that sits outside the vm, so you can limit the ip range
that ssh is even allowed from
2068[22:25:38] <Exagone313> Hi, I followed
replaced-url
2069[22:25:38] <Exagone313> can I build the "common"
package? Thanks for your help. PS: Since this might sound like an XY
problem, my issue is that I need a kernel with
`CONFIG_CGROUP_HUGETLB=y` which is not set in debian kernel.
2071[22:27:49] <Exagone313> as a last resort, I'll build
upstream kernel "normally" without build a debian package
(I just hope I can use debian's config file I got instead of
fiddling with make menuconfig) :/
2076[22:31:25] <Exagone313> also, I need kernel headers to build
module using wireguard-dkms, I seriously hope this can be done, or I
need to rebuild kernel with upstream patch :/
2102[22:37:49] <sponix> Exagone313: I'm not familiar with
what you are trying to do. But I would recommend bringing down the
sid source and seeing if that options does it for you
2103[22:39:03] <Exagone313> I think I need someone that do kernel
builds using debian build system (debian/rules.gen)
2106[22:39:26] <sponix> Exagone313: I followed this to build
rtorrent from sid
2107[22:39:27] <sponix> !ssb
2108[22:39:28] <dpkg> First, check for a backport on
<debian-backports>. If unavailable: 1) Add a deb-src line for
sid (not a deb line!); ask me about <deb-src sid> 2) enable
debian-backports (see <bdo>) 3) apt update; apt install
build-essential; apt build-dep packagename 4) apt -b source
packagename 5) dpkg -i packagename-ver.deb To change compilation
options, see <package recompile>; for versions newer than sid
see <uupdate>.
2109[22:39:28] *** Quits: chambln (~user@replaced-ip) (Remote host closed the connection)
2122[22:46:10] <Exagone313> sponix: it's required by some
kubernetes-related software I want to use. on another new machine I
used ubuntu and it got that config, but I can't easily
reinstall that one which is "stuck" with debian (although
I'd prefer to use debian everywhere)
2159[23:00:57] <Exagone313> I don't want to use a mailing
list to find out the issue xD
2160[23:01:07] <trek00> :)
2161[23:01:47] <Exagone313> the only time I used a mailing list
was for asking a specific curl question, and in one day I received
so many mails that I just unregistered and gave up
2170[23:06:13] <Exagone313> Now I have another little issue. That
setup is for a weird server where I can't configure the
bootloader, it actually loads the hosting provider's kernel,
BUT by chance supports kexec on boot, which I use to load a local
kernel. The issue is that I need a stable path to vmlinuz and
initrd, I don't want to have to maintain a symlink manually. Is
there a way to add a postinst hook that will create that symlink? I
found that there is a "linux-update-symlinks
2171[23:06:13] <Exagone313> " tool that exists, but I just
miss the hook that calls it. :)