55[01:31:18] <impermanence> forgive the ignorance and also
please ignore the specific name: I'm just curious about the
syntax: is something like this allowed: apt-get install
terraform-0.12@some_company
56[01:31:36] <impermanence> in other words is thing@other_thing
allowed and if so what does this mean?
57[01:31:48] <impermanence> it's referencing something at
some repo?
59[01:33:08] <sney> not with an @ but you can use a /, like
with a system that had backports enabled you could do 'apt
install linux-image-amd64/buster-backports' to get a newer
kernel than the one in stable
60[01:33:40] <impermanence> sney: basically I'm looking
for the analog to this: apk add thing@other_thing
61[01:33:45] *** Quits: n4dir (~n4dir@replaced-ip) (Remote host closed the connection)
62[01:33:49] <impermanence> I just don't know what that
means in terms of debian.
63[01:34:24] <sney> anything that you tell apt to install has
to already be configured in /etc/apt/sources.list though. there
isn't a global database of sources or anything.
64[01:34:53] <sney> the term is repository, or repo.
65[01:35:30] <impermanence> sney: I'm aware yes. I'll
ask in alpine I guess.
71[01:46:22] <Urk> jmcnaught: This is the result of my
permissions check
replaced-url
72[01:46:45] <Urk> Am I correct that a lack of write capability
for both / and user is the problem?
73[01:47:21] *** Quits: Filo (~filohuhum@replaced-ip) (Remote host closed the connection)
74[01:47:21] *** Quits: HeXiLeD (~grumpy@replaced-ip) (Remote host closed the connection)
75[01:47:33] <Urk> As you may recall, Libreoffice is read only,
and will not let me save to disk. Earlier it prompted me to edit it
before altering a document, but when complete it won't let me
save to disk.
76[01:47:52] *** Joins: Filo (~filohuhum@replaced-ip)
81[01:50:31] <jmcnaught> Urk: use "groups" command to
see which groups you are in. Use "ls -l /path/to/file" to
see the permissions on a file.
82[01:51:13] <Urk> Is what I did with namei -m /media/to/path
ok?
83[01:53:02] <jmcnaught> Urk: sure, it helps show if you have
permission to navigate the path to the file. But you should also
check the permissions on the file itself.
225[04:52:33] <dgriffi> is anyone here using Pine with local
~/Maildir?
226[04:52:42] <dgriffi> or Alpine?
227[04:52:50] <dvs> !anyone
228[04:52:51] <dpkg> Please do not ask if anyone can help you,
knows 'something' or uses 'some_program'.
Instead, ask your real question. (If the real question _was_
"does anyone use 'some_program'?" ask me about
<popcon> instead.) See <ask> <ask to ask>
<polls> <search> <sicco> <smart questions>.
229[04:53:24] <dgriffi> I got it to show me the inbox, but not
any of the folders.
233[04:56:18] <dgriffi> How can I make Alpine (with the maildir
patch) show me the mail folders in ~/Maildir as it does if
they're stored in MBOX format in ~/mail?
234[04:56:41] <dgriffi> I can get it to show me INBOX stored
there, but nothing more.
235[04:56:50] <qing> s there something like
nautilus-extension-xterm ?
260[05:27:40] <Urk> Am I correct that the media file is in /user
because /user is part of the file name?
261[05:29:15] <jmcnaught> Urk: /user/ is part of the path to the
mounted filesystem because it was probably automounted by udisks2 or
similar which mounts things under /media/username/
262[05:29:23] *** Quits: riff-IRC (~riff2@replaced-ip) (Quit: PROTO-IRC v0.73a (C) 1988 NetSoft - Built on 11-13-1988 on
AT&T System V)
272[05:35:19] <Urk> I am not sure. I am logged in as user, and
tried to open and save a file in Libreoffice.
273[05:36:11] <Urk> Libreoffice is coming up as read only.
However, initially there was an optional edit button displayed on
the first page, and by the time I was done editing the edit option
was gone.
274[05:36:40] <Urk> Don't recall having this problem with
Libreoffice before the hard drive enclosure was moved to the other
system for a couple of days.
275[05:36:45] <jmcnaught> Urk: so you have read permission on
the file but not write permission. Can you run "ls -l
filename" on the file?
276[05:38:38] *** Quits: n4dir (~n4dir@replaced-ip) (Remote host closed the connection)
278[05:40:18] <Urk> jmcnaught: This is the results of ls-l
Assuming my understanding is correct, the file appears to have
permissions since all files in the pdq group have rwx.
replaced-url
285[05:43:57] <jmcnaught> Urk: that does not show any
libreoffice files. I am asking for the output of "ls -l
/media/user/56e844af-585d-477e-887b-a1931858c6b6/pdq/LIBREOFFICEFILE
but I do not know the name of the file you trying to open.
286[05:44:03] *** Joins: jerry (~jerry@replaced-ip)
287[05:44:06] *** Quits: sklei4 (~sklei4@replaced-ip) (Quit: I fell asleep.)
288[05:46:12] *** Quits: yans (~yans@replaced-ip) (Quit: chaos is the only true answer)
294[05:51:44] *** Quits: koniu (~koniu@replaced-ip) (Remote host closed the connection)
295[05:52:07] <jmcnaught> Urk: is
/media/user/56e844af-585d-477e-887b-a1931858c6b6/pdq/libreoffice the
file you are trying to open? The file's name is
"libreoffice", really?
297[05:53:40] <Urk> jmcnaught: Not sure, but skeptical I have
the right path. I would think something like this would be more
likely in the /usr/bin
298[05:54:12] <jmcnaught> Urk: you are obviously confused. What
is the name of the file you are trying to open?
299[05:55:04] <Urk> libreoffice is the file I am trying to open,
and apparently it is the right file name because it opened the
program at the command line.
302[05:55:46] <jmcnaught> Urk: I thought the problem you were
having was libreoffice was opening a file in read-only mode, would
not let you edit or save the document. What is *that* file's
name?
303[05:58:05] <Urk> libreoffice is the file that I am having
trouble with. What I did, I took a copy of a contract for a
different guest, and tried to save it to a folder for the guest that
needs a contract, and with the hopes of editing it. It is still
coming up as read only, though there is a popup at bootup
authorizing edit, selecting it doesn't give me "save"
privileges.
316[06:04:06] <Urk> jmcnaught: Here is an interesting situation.
Even though another instance/date of the program has write
privileges, any attempt at saving with the files name 03-12-2021,etc
ends up with Error saving the document
05-25-2020_Residential-Lease-Agreement:
317[06:04:06] <Urk> Object not accessible.
318[06:04:06] <Urk> The object cannot be accessed
319[06:04:06] <Urk> due to insufficient user rights.
323[06:05:24] <Urk> jmcnaught: It would have to be in the
/media/user/56e844af-585d-477e-887b-a1931858c6b6/pdq/ since all of
the files are in /pdq
324[06:05:41] <Urk> However, at login I logged in under /user
325[06:05:59] <themill> spinningCat: you might want some context
there... I'm guessing that you're running apt with a
broken PATH
326[06:06:57] <jmcnaught> Urk: the name of the directory does
not matter. I think you need to read about and understand Linux/UNIX
users, groups, and permissions. It is pretty clear that the
filesystem mounted at /media/user/… used to be /home on
another computer, and the directories and their files are owned by
different users.
327[06:06:59] <spinningCat> dpkg -i package i did themill
328[06:06:59] <dpkg> package package i did themill is already
installed
329[06:07:22] <spinningCat> so i try to install package
330[06:07:31] <themill> spinningCat: so that would mean
you're running dpkg with a broken PATH
331[06:07:51] <spinningCat> i am in the path that has the
package
332[06:08:05] <themill> PATH is an environment variable
333[06:08:49] <spinningCat> i mean i download google-chrome
package and i go to download/ and execute dpkg package
334[06:08:51] * themill returns to wild guessing in the absence of
actual data
335[06:08:54] <themill> !buster su
336[06:08:55] <dpkg> In buster, su no longer overrides PATH by
default, requiring that you use "su -" or "su
-l" for login shells (which is not really a new thing at
all...). See
replaced-url
338[06:09:46] <themill> (also, "apt install ./foo.deb"
is better than "dpkg -i foo.deb")
339[06:09:52] <Urk> jmcnaught: Here is the message I get
replaced-url
340[06:10:30] <spinningCat> themill, that fixed the error export
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin i
am not sure it is valid approach
341[06:11:06] <Urk> jmcnaught: The data I am trying to access
has been on the same hard drive enclosure that was used on both
laptop so I am not sure how this could have happened.
342[06:11:09] <themill> spinningCat: the better approach is what
dpkg sid.
343[06:11:24] <themill> or just use sudo
344[06:11:32] <Urk> In both cases, I accessed the files by going
into the area "other locations".
357[06:25:46] <Urk> Here is a more accurate path, but still
indicates no permissions.
replaced-url
358[06:26:50] <themill> you need to read that output first, I
think
359[06:29:35] <abff> Urk do you see how its trying to read
multiple files? ls thinks you asked for two files because one of
your directories has a space in it, and spaces are how command line
utilities seperate arguments
362[06:31:46] <abff> whenever a file or a directory has a space
in it's name you should do one of the following: put a
backslash \ before the space 'ls -l /your/file/is\
in/here.odt', or encapsulate the path/file in quotes ' or
"
363[06:32:06] <abff> now bash differentiates the two quotes and
I forget how
364[06:32:16] <abff> so I usually opt to use the backslash
366[06:32:36] *** Quits: wicknix (~slaxxx@replaced-ip) (Remote host closed the connection)
367[06:33:27] <Urk> abff: I fixed the space problem, but still
get the same error
replaced-url
368[06:34:01] <jmcnaught> Urk: are you using tab completion to
help type in the file name?
369[06:34:23] <Urk> jmcnaught": yes
370[06:35:27] <Urk> And I have been using home, and end to edit
the commands in front of the file name.
371[06:36:41] <themill> that is clearly not true. One of them
has a space "David Borders" and the other has a dash
"David-Borders"
372[06:36:57] <jmcnaught> Urk: did you rename the directory that
had a space in it, or did you just replace the space with a dash in
the command you used?
373[06:37:08] <themill> *snap*
374[06:37:30] *** Abrax is now known as \
375[06:37:39] *** \ is now known as Abrax
376[06:37:53] <Urk> jmcnaught: I placed a hyphen in the name so
that the file was renamed.
385[06:41:15] <Urk> There doesn't seem to be a problem with
the other folders so I have no idea how this happened. I have been
working with the hard drive enclosure with no problem prior to now.
386[06:43:37] <Urk> I think something is wrong with the folder.
The files works fine if it is removed from the folder.
397[06:50:10] <Urk> jmcnaught: now that I understand your
question better, I placed hyphens in the name, and didn't
rename the directory. The directory can not be renamed because of
permissions problem.
399[06:52:35] <themill> perhaps you meant "namei" not
"name -i"; if you're running these commands with a
file path that you've just made up by putting a - in there
(without actually renaming the directory), then they're not
going to work
406[06:56:37] <themill> (you might want to ask it to do so in
that case)
407[06:57:04] <Urk> themill: yes, but I get an error "no
such file or directory" for the David Borders directory. I have
no idea why this error is showing up.
replaced-url
408[06:57:36] <themill> Urk: would that be because it's not
"David-Borders" it is "David Borders"?
412[07:00:26] <Urk> themill: If I remove the hyphen then only
the last name is recognized. And I did try it to see what happened,
and get the same error either way.
422[07:05:21] <themill> bruteforce would be « find
/media/user/56e844af-585d-477e-887b-a1931858c6b6/pdq/Documents/Crescent-City-Property/Tenants/2021/
-name 03-12-2021_Residentia-Lease-Agreement.odt -exec namei -m {} +
»
423[07:06:40] <Urk> I don't think the path is right because
III is missing from the end of David Borders. However, I still get
command not found when running diagnostics.
425[07:08:02] <Urk> I got in, and there is a lack of write
permission. The path was not right because the folder is actually
David Borders III and I was only searching for David Borders.
426[07:08:12] <themill> *sigh*
427[07:08:29] *** ChanServ sets mode: +o eir
428[07:08:29] *** eir sets mode: -qo $a:Abrax eir
429[07:08:34] <themill> you were asked several times if the path
was correct and if you had been using tab completion
430[07:09:13] <Urk> themil: maybe I am not as smart as I am
supposed to be . . . haha
434[07:13:24] <Urk> themill: At this point I have no idea what
is wrong. Users and groups have read and write permissions, but it
hasn't changed anything.
435[07:13:42] <themill> pastebin or it didn't happen
437[07:17:10] <Urk> themill: I get a new error now, and it says
the folder is locked. I'm wondering if there was more than one
problem with this folder? I will research on how to get rid of the
lock file tomorrow, and just worry about finishing the agreement
tonight.
499[09:01:16] <NetTerminalGene> guys, pulseaudio uses about 5%
CPU after launching firefox (i set firefox to load lots of site at
launch) after restarting pulseaudio service, it turns back to normal
(doesn't consume CPU)
522[09:27:04] <ratrace> NetTerminalGene: based on information
you gave, I'd say it's not hard to imagine loading
"lots of sites at launch" would do something with PA that
caused it to look like it's consuming 5% of CPU
523[09:27:53] <ratrace> and when you say it wasn't like
that "before"... before what?
524[09:30:44] <NetTerminalGene> ratrace: guess what
525[09:31:01] <ratrace> what
526[09:31:26] <NetTerminalGene> what do i mean with
"before"? guess what
533[09:34:25] <NetTerminalGene> ratrace: go and do better
things, please
534[09:34:40] <ratrace> also, you could strace pulseaudio by pid
and see what it's doing. I don't know PA internals, but
running that as an example now, I can see it spends most of its time
ppoll()'ing for events. slower when there's nothing
playing audio, and faster when something (firefox) is
535[09:34:59] <ratrace> NetTerminalGene: yup, will do. thanks
for playing.
536[09:35:10] <NetTerminalGene> playing?
537[09:35:39] <NetTerminalGene> you're not helpful
682[12:50:54] <aypea[0]> so I just upgraded to bullseye and I
have a weird issue with named. as soon as it starts systemd tells it
to shut down. can anyone help me with where to look? I'm
running out of ideas and we searches are being less than fruitful
687[12:54:48] <ratrace> aypea[0]: I'd start with the system
logs. journalctl to see if systemd is complaining about things, and
you can configure bind for explicit logging
690[12:57:31] <aypea[0]> ratrace: heh. only thing it says is
'systemd[1]: named.service: Succeeded.' after it happily
killed it. before that not a peep of an error from either it or
bind.
691[12:57:53] <aypea[0]> infact systemd is issuing an rndc stop
as part of the start. it's being very polite about it
692[12:58:42] <ratrace> surely there's something about it
in the logs. doesn't make sense systemd would actively call
rndc to stop it
704[13:06:19] <aypea[0]> ratrace: ok. fixed. you actually helped
but I can't point to a specific thing you did. just all of it
made me think of one more thing. :) it appears that I had customised
ExecStart via an /etc/systemd override and whilst it worked well
under buster, under bullseye it broke as per description.
716[13:11:48] <aypea[0]> I think it may have been seeing that it
couldn't load the DNSSEC for the root zone (.) that I saw from
journalctl -u bind9.service (but didn't see in syslog land cos
I'm blind).
764[14:02:05] <abff> Franciman I've been using mednafen,
it's similar with no gui but the documentation is plentiful and
configuring key presses and controllers is easy
832[14:41:43] <abff> Franciman: if you want more gui interface,
I recommend retroarch, it 's a whole emulation ecosystem with a
gui and it's own package manager (specifically for emulators)
834[14:44:46] <abff> the downside to retro arch is that it has
-so many- features it's probably outside of a single
system's needs, which is why I like mednafen I can easily just
create a shortcut keyboard command to boot up whatever game I'm
playing right now
886[15:31:50] <festerdam> I have a thinkpad E540, whose touchpad
sometimes gets jittery and seems to be detecting touches when there
are none. Also happens on Windows. It often gets solved by cleaning
the pad with alcohol. I think it could be sebum building up on the
surface. Is there a way to change the touchpad's tolerance to
input?
887[15:32:06] <Franciman> abff, I know it, thanks, yeh I wanted
to stay lite enough :P
888[15:32:13] <Franciman> I just wanted to play some megaman
games ahahah
932[16:16:39] <Budd> Is there a way to create a network
configuration for a wildcard interface? Use case: When I plug a USB
adapter in (any adapter, any port), I want to configure it for my
mesh network.
934[16:17:01] <betuxy> Hey guys, after disabling the serverside
encryption on my nextcloud, apache doesnt start anymore and gives me
this error: /usr/sbin/apachectl: 99: ulimit: error setting limit
(Operation not permitted), i already set the ulimit (8192) in
limits.conf according to a website but that didnt help. DO you have
any ideas?
935[16:17:26] <betuxy> ulimit -n gives 819, the whole thing is
running in an systemd-nspawn container
936[16:17:58] <ratrace> Budd: wildcard no, but you can configure
allow-hotplug NIC
960[16:25:00] <Budd> ratrace: Sure, the hotplug part works fine,
but it seems the current/future naming scheme is by bus location, so
if the interface shows up in a new place, it won't be
recognized (and my needs are simple; same statuc ipv6 address on all
interfaces)
975[16:28:02] <Budd> I haven't learned about
ssytemd-networkd; I'll read up.
976[16:28:04] <ratrace> Deknos: then what's the problem?
977[16:29:08] <ratrace> Budd: also... could be possible to bind
the interaface by MAC via udev, now that I think about it, so you
can keep using it with interfaces(5)
979[16:29:58] <ratrace> Budd: something like a custom rule like
this: SUBSYSTEMD=="net", ATTR{address}==" mac address
here", NAME="wifi0_or_whatever"
980[16:30:08] <ratrace> %s/SUBSYSTEMD/SUBSYSTEM/
981[16:30:28] <Ark74> Anyone with experience on control/rules
files, I'm trying to custom build sugar-artwork with sbuild,
but for some reason even when it's set as dependencies python
3, it will try to use python 2 and failing
982[16:30:37] *** Quits: j7k6 (~j7k6@replaced-ip) (Remote host closed the connection)
987[16:32:34] <Budd> ratrace: thanks; I'm trying to set up
something that would work for any MAC. A quick read of
systemd-networkd looks like it has a [Match] section that might do
everything I need.
994[16:34:54] <Budd> ratrace: I'm setting up machines that
just slap a /128 IPv6 address on every interface and run babeld for
routing.
995[16:35:24] <Budd> So any onboard NIC uses the same settings.
996[16:35:29] <ratrace> wh... why /128. standard is /64 subnet
997[16:36:01] <ratrace> or you mean... on single machine?
998[16:37:32] <Budd> Each physical machine has a unique /128
that it uses for all its interfaces. Often includes an IBSS link,
which doesn't behave like a regular subnet (some nodes
can't see each other).
999[16:38:54] <ratrace> wait... what do you mean uses a /128 for
all its interfaces? you're assigning the same IPv6 to all the
NICs on the machine?
1000[16:39:01] <Deknos> ratrace, i just found out about it, after
wanting to use it.
1001[16:39:03] <Budd> Works great for installing a data
logger/camera outside - it doesn't have to be in range of any
particular machine; as long as it's in range of any other
machine.
1024[16:50:04] <Budd> I agree it would be unusual (and probably
broken) to use the same /64 address on two interfaces, but when each
node is a router in a mesh there's no clean cluster I can treat
as a /64 subnet.
1030[16:56:27] <Budd> I realize this is getting off-topic, but I
didn't have to do anything odd; static /128 addresses for each
interface (no static gateway), run babeld, and Bob's my uncle.
1100[18:11:51] <rudi_s> (Of course you could download the .deb
manually, verify its checksum and install it. But that has the same
effect as manually adding non-free to your sources.)
1103[18:13:17] <oerheks> without adding, you will not get updates
1104[18:13:51] <n4dir> the clear advantage of having the non-free
repo enabled is that for the given package you will get updates. And
i see no harm in having it enabled, as long you install a non-free
package anyway
1105[18:14:16] <n4dir> ups, oerheks already said it
1205[19:59:07] <bifunc2> Hi. Debian has reproducible builds (at
least for 99.9% of everything). No other distro has it! So how can
anyone serious about OSS recommend any other distro to me? Without
reproducible builds, "open source" is meaningless because
nobody knows the binaries are coming from the source code
they're supposed to be coming from!
1206[19:59:18] <bifunc2> Do I have this figured out right?
1208[20:00:06] <bifunc2> They keep talking about all sorts of
distros, Arch and whatnot, but it's all only "open
source" in name only. If you only get final binaries,
what's the point?
1209[20:01:56] <sney> that's the idea, though it's
still a relatively new concept, and afaict debian isn't at
"99.9% of everything" either as it's still a work in
progress:
replaced-url
1210[20:02:15] *** Quits: _aeris_ (~aeris@replaced-ip) (Remote host closed the connection)
1212[20:02:59] <bifunc2> hardly any "open source"
fanboy talks about this issue and in practice they just download
compiled images. why are they so trusting? this has been going on
for decades now. truly strange isn't it?
1217[20:07:09] <dg1727> Is there possibly a recent regression in
the kernel relating to this?
1218[20:07:22] <sney> !based on debian
1219[20:07:22] <dpkg> Your distribution may be based on and have
software in common with Debian, but it is not Debian. We don't
and cannot know what changes were made by your distribution (compare
replaced-url
1220[20:08:04] <EdePopede> bifunc2: someone was mentioning iLinux
somewhere recently, had a look at their website and forum. looks
more like a windows on linux kernel, at least he sounds as he
wouldn't care.
1221[20:08:06] <sney> debian's kernel images are as close to
upstream as possible, so if there's a driver issue in debian,
it's almost definitely present in the kernel.org source as well
1222[20:08:08] <petn-randall> bifunc2: Gentoo is a thing. While
reproducible builds are cool, and I fully support it, "open
source" still has meaning even without it. I mean, we've
been doing it like that for the last 30+ years.
1223[20:09:18] <petn-randall> dg1727: File a bug with your OS,
and if it also affects Debian, they'll probably help fixing it
on Debian, too.
1224[20:09:21] <bifunc2> petn-randall what i think is strange is
everyone thinks "like that" is just fine and they've
been thinking that for decades
1225[20:09:33] <dg1727> sney: The person who raised that GitHub
issue claimed to be running Debian Buster. I can try other support
avenues.
1226[20:09:49] <bifunc2> in practice, is it easier to detect
unwanted stuff in a binary than i imagine?
1227[20:09:54] <bifunc2> that could explain it maybe
1232[20:10:54] <petn-randall> bifunc2: Because doing reproducible
builds is *hard*, and it's only become more relevant with cloud
build servers. And yes, you can disassemble binaries and analyze
their behaviour.
1313[21:18:29] <tomreyn> "Each alternative has a priority
associated with it. When a link group is in automatic mode, the
alternatives pointed to by members of the group will be those which
have the highest priority."
1314[21:18:50] <tomreyn> yes, it's abitraryily chosen
values, it just matters hwich one is higher and which one is lower.
1315[21:19:12] <impermanence> thx
1316[21:19:25] <tomreyn> the quote above is from the
update-alternatives(1) manual
1317[21:19:29] <tomreyn> !man
1318[21:19:29] <dpkg> man is, like, a little guy that knows
everything about Linux... just do "man commandname" and he
will teach you all about commandname. You can search for the
appropriate man page with "apropos term". Enable <tab
completion> in your shell to fill in the man page names. Man
pages are available online too, ask me about <dmp>. See also
<docs> <grounding>.
1365[21:54:29] <petn-randall> impermanence: 1,2,3,4,5 would not
be optimal, because if you want to add your own in between,
you'd have to change all the other around. Other than that,
it's just highest number wins.
1366[21:54:59] <impermanence> huh.
1367[21:56:19] <petn-randall> That's why back in the day,
when you programmed on the C64 in BASIC, you numbered every line in
increments of 10, to allow some room for bugfixes ;)
1390[22:20:29] <tcurdt> I need to restart dhclient (during
automation with ansible) as I change the config ... there is no
dhclient service, instead there are services for the invididual
interfaces that differ on the various machines
1391[22:20:33] <tcurdt> it seems the most standardize way would
be to restart the networking service
1401[22:27:59] <jmcnaught> tcurdt: I have found that interfaces
marked auto get a dhclient that is started/owned by
networking.service, but interfaces marked allow-hotplug get a
dhclient process that is owned by ifup@eth0.service (or whatever
interfaces name)
1405[22:31:21] <petn-randall> tcurdt: You can run `ifdown
<if> && ifup <if>` to bring up the interface
with the new config. But if that's your link you're
connecting with, that's of course risky.
1406[22:32:05] <petn-randall> tcurdt: One safer way is to just
reboot instead, or use something like network-manager that is more
robust/flexible in that regard.
1407[22:32:21] <tcurdt> it seems like with the auto interfaces
dhclient is still coming from ifup@eth0.service for me
1429[22:39:10] <dpkg> From Debian 6.0 "Squeeze"
onwards, the use of the networking init script to reload networking
configuration («/etc/init.d/networking restart») is
discouraged. Instead, perform the actions you really want. Re-apply
changes from /etc/network/interfaces: ifdown eth0; ifup eth0. Remove
all iptables rules: iptables -F. See also <mdblows>.
1430[22:39:36] <petn-randall> IIRC the only way to restart the
network with ifupdown is bringing the interface down and up again.
But IIRC it's easier to do with network-manager.
1431[22:40:16] <petn-randall> jmcnaught: That's the one!
I've been poking dpkg for the last few minutes, it slipped my
mind ...
1432[22:40:40] <petn-randall> jmcnaught: 2:0! Best of five? ;)
1433[22:43:00] <jmcnaught> petn-randall: that factoid hasn't
been edited since 2014, so maybe things are different now. From what
I understood it's better to "ifdown eth0"; edit
/e/n/i; then "ifup eth0" if you need to make changes,
because ifup and ifdown act on the contents of /e/n/i which does not
necessarily reflect the current state of the interface.
1434[22:43:20] <jmcnaught> petn-randall: restarting networking
probably works most of the time though
1436[22:45:22] <petn-randall> jmcnaught: That was my thought,
too. But IIRC newer ifupdown saves state now, so edit /e/n/i, ifdown
eth0, ifup eth0 should work now. Would even be interesting to test.
1440[22:47:21] <tcurdt> dealing with interfaces themself like
this isn't really particular universal though ... fine for
interactive - but feels like more hassle for automation
1441[22:47:34] <tcurdt> I guess I just stick with work around
1442[22:47:48] <tcurdt> *my
1443[22:49:31] <petn-randall> tcurdt: I use the ansible nmcli
module for that, and just ignore ifupdown for the most part.
1450[22:51:45] <jmcnaught> petn-randall: I just tested by booting
a VM with enp1s0 in /e/n/i configured for DHCP. I edited /e/n/i to
set enp1s0 to static, did "ifdown enp1s0 && ifup
enp1s0" and it does reconfigured the interface, but it leaves
dhclient running.
1451[22:52:26] <tcurdt> petn-randall: well, the networking is
slightly different for installing on a cloud image vs e.g. vagrant
... hence I would ideally leave the interfaces alone altogether
1452[22:52:57] *** Quits: basicmiracle (uid213868@replaced-ip) (Quit: Connection closed for inactivity)
1453[22:53:42] <tcurdt> it doesn't even restart/reload it?
1454[22:53:45] <tcurdt> jmcnaught:
1455[22:53:59] <tcurdt> --^
1456[22:54:45] <jmcnaught> tcurdt: I have seen this happen to
people in #debian where they changed /e/n/i from dhcp to static, but
have not rebooted yet, and they had a left-over dhclient process
hanging out periodically reconfiguring the interface
1519[23:43:06] <tomreyn> oh web server, not web browser, ignore
me
1520[23:43:12] <petn-randall> :)
1521[23:43:15] <HelloShi1ty> Unit193: that dpkg didn't
return nothing either
1522[23:43:36] <petn-randall> HelloShi1ty: Ok, what makes you
come to the conclusion that you have one?
1523[23:44:10] <HelloShi1ty> but I have a service hosted in my
laptop that I can access via a web site, so this web site must be
being served by some web server
1524[23:44:23] <HelloShi1ty> and I didn't installed apache
nor nginx
1525[23:44:33] <petn-randall> HelloShi1ty: Ok, what service?
1526[23:44:45] <sney> 'ss -lp |grep http', if it's
a regular webserver listening on 80/443
1528[23:45:10] <Unit193> That's what I was going with, but
don't remember the ss switches as well as netstat.
1529[23:45:37] <HelloShi1ty> that return nothing either
1530[23:45:50] <sney> I can always remember l for listen and p
for process. there might be a way to filter without grep
1531[23:45:51] <HelloShi1ty> but this service is listening in a
port other than 80
1532[23:45:59] <petn-randall> Unit193: LOL, I was just looking up
the man page of ss, because netstat isn't installed on most
Debian installations nowadays. :D
1533[23:46:07] *** Quits: ax56234 (~NickServ@replaced-ip) (Remote host closed the connection)
1534[23:46:09] <sney> ok, then ss -lp |grep whateverportnumber
1535[23:46:21] <petn-randall> `netstat -npl` from my heart, but I
had to look up `ss -lp`.
1536[23:46:56] <wuffi600> HelloShi1ty: do it as root then you can
see also the process/services names.
1537[23:46:58] <HelloShi1ty> netstat -tnlp | grep 80 this returns
nothing either
1538[23:47:18] <sney> why would you grep for 80 when you just
told us it's not listening on 80? are you even reading the
things you type
1539[23:47:24] <oerheks> how do you tell; this service is
listening in a port other than 80
1540[23:47:45] <HelloShi1ty> but if I change the '80'
for th port it is listining in
1541[23:47:48] <HelloShi1ty> I get some output
1542[23:47:57] <HelloShi1ty> but it doesn't tell me nothing
about a web server, I guess
1543[23:48:13] <HelloShi1ty> is 'node' a web server?
1544[23:48:13] <sney> it will tell you the name of the process
that is listening on that port
1545[23:48:18] <sney> it sure can be.
1546[23:49:22] <HelloShi1ty> Can this 'node' be related
to node.js ?
1547[23:49:47] <sney> yes
1548[23:50:03] <HelloShi1ty> and is this node.js a web server?
1549[23:50:16] <sney> nodejs has a http module that can act as a
webserver, yes.
1550[23:50:41] <HelloShi1ty> hum, so it must be that