1[00:00:15] <mutante> alexrelis[m]: also i would recommend to
replace Google Authenticator with Authy app. it will work whereever
Google Authenticator works but you can have a backup which is
encrypted with your own key. so you can move to another device if
you lost your phone
2[00:00:37] <mutante> (and have backup codes/restore
passphrase saved elsewhere)
54[00:38:57] <alexrelis[m]> cluonbeam: Hmm.. But isn't the
whole point of OTP the ability to authenticate with a secondary
device if that makes sense? If somebody has access to your master
key, then they have access to everything. If you made sure your OTP
was only stored on your phone (or a single device) and only made
backup codes on paper, then if an attacker had access to your
password manager, they would also need access to your phone, or
55[00:38:57] <alexrelis[m]> access to wherever you stored those
recovery codes to get in.
62[00:49:04] <cluonbeam> alexrelis[m]: But they don't have
access to everything. The Bitwarden DB requires 2FA to decrypt.
Without my Yubikey, you have squat.
63[00:49:09] <cluonbeam> Yubikey AND passphrase.
64[00:49:54] <cluonbeam> So, sure, take my phone. Good luck.
You'll need it.
65[00:50:27] <alexrelis[m]> Hmmm.. I might need one of those
Yubikeys then, haha.
66[00:50:45] *** Quits: fax (~quassel@replaced-ip) (Remote host closed the connection)
96[01:14:05] <sney> when you get that error, you can hit alt-f4
to see the info console, where it might give you more useful details
as to why the installer couldn't create the filesystem.
98[01:15:00] <sney> if you don't want a raid, that's
fine, but it shouldn't stop you from installing to sda
normally. your partition layout is a little oddball but should be
fully supported.
99[01:15:50] <sney> the detected raid devices in your
screenshot are not being changed by the installer so they have
nothing to do with the problem
100[01:16:21] <oxek> 'wipefs' would be my tool of
choice
101[01:16:52] <sney> that's nice, but wipefs is not
included in the installer environment and this is an installer
problem.
102[01:18:29] <sney> and in any case: according to those
screenshots, sda is a 240GB ssd with no pre-existing raid metadata,
and sda is the only disk where changes are being made. the raid is a
red herring, and irrelevant
110[01:21:29] <sney> dka: how many physical disks does this
system have?
111[01:21:33] <dka> 1
112[01:21:42] <oxek> my understanding is that there are only 2
storage devices - the ssd and the flash drive. The ssd contains old
RAID metadata at the beggining of the drive, which debian installer
uses to display those RAID entries, and then tries to create MBR
& partitions.
113[01:21:45] <sney> sigh, so the raid members are partitions on
the disk.
114[01:21:46] <oxek> this will of course fail
115[01:21:59] <oxek> hence, the ssd needs to be wiped
116[01:22:10] <dka> how
117[01:22:19] <sney> just the beginning, which wipefs would do,
if you had it in that environment.
118[01:22:24] <oxek> dka: are you in the installer enviroment
right now?
119[01:22:31] <sney> you should be able to select each of those
entries in partman and delete them through the menus.
120[01:22:35] <dka> yes
121[01:22:47] <dka> i already tried
122[01:22:55] <dka> how can i exit alt+f4?
123[01:23:01] <sney> alt-f1
124[01:23:06] <oxek> dka: run blkdiscard /dev/sda
125[01:23:11] <dka> how can I run think?
126[01:23:17] <sney> alt-f2 and f3 would take you to consoles
where you can run things.
127[01:23:41] <dka> blkdiscard: cant open /dev/sda device or
resource busy
128[01:24:38] <oxek> reboot the machine and run that command
again
129[01:25:04] <oxek> before you reach the partitioning step of
the installation
130[01:27:01] <dka> blkdiscard /dev/sda now hang
131[01:27:20] <oxek> give it time
132[01:27:25] <dka> and succeeded without any output
133[01:27:26] <oxek> about a minute
134[01:27:35] <oxek> the disk is now blank
135[01:27:40] <oxek> you can proceed with the installation
209[01:51:23] <cluonbeam> Android also supports use via NFC, so
you could get the 5C if you're fine with USB-C only, or the 5
and use NFC for your Android and USB for everything else.
210[01:52:06] <cluonbeam> They're SUPPOSEDLY coming out
with a USB-C+NFC version "eventually(tm)".
211[01:52:32] <another> double pro tip: alias ipa='ip
-color=auto -brief address'
212[01:53:07] *** Quits: pkx (~pkx@replaced-ip) (Remote host closed the connection)
213[01:55:54] <oxek> another: nice tips, do you have a list
somewhere?
237[02:03:04] <jhutchins> So how many systems are you going to
log in to that don't have bash, if not as the default shell,
then at least as available?
238[02:03:04] *** Quits: FuzzyByte (~weechat@replaced-ip) (Remote host closed the connection)
240[02:04:28] <jhutchins> When the number of your $HOME
directories starts to approach four figures, keeping everything as
close to the installed defaults starts to seem like a good idea.
250[02:09:23] <jhutchins> oxek: I know of people who have a
shell script on some git server that goes through and does all of
their customization.
251[02:10:15] <jhutchins> Years ago though, I was asked to sit
down at a workstation and do a fairly simple task, and I was so used
to all of my shortcuts and customizations it took me 15 minutes just
to get to the task.
252[02:11:02] <jhutchins> It depends on what kind of environment
you work in. If you're in the industry, three years is a long
time.
253[02:12:19] <jhutchins> Then there was the time I was tasked
to do a skills assesment, and the keyboard had been left set to
Hungarian. (Didn't get that one.)
271[02:42:31] *** Joins: Louis (~Louis@replaced-ip)
272[02:42:31] *** Quits: Louis (~Louis@replaced-ip) (Excess Flood)
273[02:44:28] <jhutchins> CyberManifest: Unless you've done
something strange with your system you shouldn't have to.
274[02:44:53] <CyberManifest> jhutchins: I'm building a
frankenmachine
275[02:45:08] <CyberManifest> jhutchins: so in a way I have done
something strange with my system
276[02:45:17] <jhutchins> Ok. I was just going to ask how
you'd dug this hole, but there it is.
277[02:45:28] *** Joins: Louis (~Louis@replaced-ip)
278[02:45:29] *** Quits: Louis (~Louis@replaced-ip) (Excess Flood)
279[02:45:49] <jhutchins> Let's try this, it might have a
clue:
280[02:45:51] *** Joins: Louis (~Louis@replaced-ip)
281[02:45:52] <CyberManifest> jhutchins: honestly I don't
have any problems presently, I just wish to add that key and in the
process learn about it
282[02:45:52] <jhutchins> !ssb
283[02:45:52] *** Quits: Louis (~Louis@replaced-ip) (Excess Flood)
284[02:45:53] <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>.
285[02:46:10] *** Quits: kreyren (~kreyren@replaced-ip) (Remote host closed the connection)
286[02:46:13] *** Joins: Louis (~Louis@replaced-ip)
287[02:46:14] <jhutchins> !key
288[02:46:14] <dpkg> Install the latest debian-archive-keyring
package to get the latest GPG key used to sign the Debian archive.
295[02:47:40] <CyberManifest> jhutchins: first I'm not
wanting to add back ports, second, I've already installed
debin-archive-keyring package but I'd like that Official Key
and not the Archive
296[02:48:01] <jhutchins> There's at least one factoid that
has the command string to download and add the key, but I'm not
sure what it is and I don't want to spew 20 factoids to find
it.
297[02:48:44] <CyberManifest> !gpg key
298[02:48:50] <jhutchins> CyberManifest: I know, I was hunting
for the random factoid that has both steps. Hit cluonbeam's
link.
300[02:50:44] <CyberManifest> jhutchins: cluonbeam, again
that's in regards to standard archive keys I'd like to add
the encrypted key
replaced-url
301[02:51:17] <cluonbeam> Read the manual, and follow the steps.
302[02:51:28] <cluonbeam> Apply what you've learned.
Critical thinking. You want to learn, I'm not going to hand you
the answer.
303[02:51:55] <CyberManifest> cluonbeam: that manual refers to
specific key numbers and not an encrypted .asc
304[02:52:24] <cluonbeam> I know exactly what it refers to, but
everything you need is in that document.
305[02:53:01] <CyberManifest> cluonbeam: which man ?
306[02:53:47] <cluonbeam> Read through the very brief section
that you were linked to. There's only two commands listed.
307[02:53:59] <CyberManifest> cluonbeam: that's a wiki not
a man
308[02:54:10] <CyberManifest> thanks for trying to confuse me
more
309[02:54:21] <cluonbeam> Again, I know that. But "there
are two commands listed" means that the man page for ONE of
those commands has what you're looking for.
310[02:54:51] <cluonbeam> You come in here stating that you
literally intend to build a frankenbox, which is unsupported in this
channel, and that you want to learn. So I'm nudging you as much
as you ought to be, and you're complaining.
311[02:55:03] <cluonbeam> I could take the channel hard-line
rules route and say "frankenboxes are unsupported, have a nice
life."
312[02:58:45] <CyberManifest> what an egregious obfuscated mess;
also *unsupported* and here I thought there was camaraderie in
freedom of choice in the Linux community
313[02:59:00] *** Quits: FuzzyByte (~weechat@replaced-ip) (Remote host closed the connection)
314[02:59:44] <CyberManifest> guess I'll switch to arch
where things are easier and better supported and more helpful
332[03:07:44] <cluonbeam> I am giggling like a schoolgirl at the
idea that that kind of handholding is going to be offered in
#archlinux of all places.
333[03:07:45] <CyberManifest> For anyone else needing help with
the same question I had; I found my answer at:
replaced-url
334[03:08:02] <cluonbeam> Wow, so you CAN help yourself. Will
wonders never cease?
335[03:08:18] <CyberManifest> cluonbeam: their wiki is a better
handholding than the obfuscated mess that debian is
336[03:08:28] <CyberManifest> cluonbeam: yeah and without a man
337[03:08:42] <cluonbeam> "obfuscated" ==
"can't figure out man apt-key"
338[03:08:54] <CyberManifest> cluonbeam: so again you were less
than helpful
339[03:08:56] <cluonbeam> Or even read it.
340[03:09:12] <CyberManifest> apt-key is deprecated
341[03:09:27] <CyberManifest> if says so during use
342[03:09:30] <CyberManifest> even*
343[03:10:06] <CyberManifest> cluonbeam: it's obvious you
don't have a clue
344[03:10:36] <cluonbeam> It's obvious we have a very
different definition for the word "obvious." ;)
345[03:10:38] <cluonbeam> Have fun.
346[03:12:08] <cluonbeam> I figured that that option would have
been easier for you to follow than managing a gpg keyring in
/etc/apt/trusted.gpg.d, but... that's what I get for trying to
do something nice.
366[03:23:52] *** stefanc_diff_ is now known as stefanc_diff
367[03:24:30] <oxek> I don't want to read either, but when
I end up having to do it, I'd rather save others time by
telling them what I learned.
368[03:24:52] <oxek> this simple 'mechanical'
knowledge should be distributed
369[03:25:00] <cluonbeam> The guy came in demanding to be
handheld while stating openly that he was doing something to his
system that we don't support.
370[03:25:02] <oxek> so that people can spend time on
'harder' problems instead
371[03:25:10] <cluonbeam> The CORRECT path would have been to
tell him that what he's doing is unsupported, bye.
372[03:25:14] <cluonbeam> So, let's move on.
373[03:25:19] <oxek> oh, I didn't read that far
374[03:25:35] <cluonbeam> CyberManifest | jhutchins: I'm
building a frankenmachine
375[03:26:05] <oxek> to be fair, what is the best place to ask
for help with frankendebians?
376[03:26:09] *** Quits: alaskabear93 (~alaskabea@replaced-ip) (Remote host closed the connection)
377[03:26:14] <cluonbeam> ##linux?
378[03:26:21] <oxek> yeah, you're right
379[03:26:25] <somiaj> most likely no one wants to support that
399[03:44:40] <CyberManifest> since apt-key is deprecated what
is the gpg equivalent of 'apt-key list' when I do
'gpg --list-keys' it made a pubring.kbx but didn't
list any keys even when I ran it a second time... neither did
'gpg --list-secret-keys'
400[03:46:31] *** Quits: thiras (~thiras@replaced-ip) (Remote host closed the connection)
405[03:50:42] <rocketmagnet> hi everyone, there is a feature to
move a window with the alt key, my debian mate has no settings for
this in the shortcuts. how is the other program called where i can
change such settings ? they are in a hirarchy like org.gnome. etc
but i can't recall the name
577[08:22:22] <kline> diogenes_: see also my post from august:
replaced-url
578[08:23:25] <diogenes_> lol Socrates looks nice, even after
the poison :)
579[08:24:02] <n4dir> at the top wiki.debian.org mentions three
drivers : *mach64 and *video-r128 and *radeon. radeon seems to be
used. So i guess i need one of the other two, but how to figure out
which one?
580[08:24:11] <kline> lets make a slight adjustment to the canon
and call it nightshade:
replaced-url
605[09:08:18] <ym> Hi. I have a little issue with pppd. It
starts at bootup ok, journal also says everything is fine - local
and remote addresses allocated, but "ip a l dev ppp0"
command doesn't shows any addresses allocated and internets
aren't pingable. Workaround "systemctl restart
networking" works, but I want to figure out, what's the
problem. Any suggestions?
606[09:09:25] *** Quits: fax (~quassel@replaced-ip) (Remote host closed the connection)
607[09:15:32] <ym> Interfaces configuration is this:
619[09:32:00] <shtrb> Good morning , does anyone have a clue how
to solve "usb_modeswitch_dispatcher[34997]: Dispatcher for
'usb_modeswitch'; usually run by udev, not intended for
interactive operation" message loop ?
639[09:48:44] *** Quits: mandeep (uid394387@replaced-ip) (Quit: Connection closed for inactivity)
640[09:49:14] <shtrb> Anyone who still remember udev rules can
say if ATTRS{idVendor}=="XXX",
ATTRS{idProduct}=="YYYY", RUN+="usb_modeswitch
'%b/%k'" is wrong for modern udev/systemd (buster and
up) ?
648[09:55:50] <shtrb> ym, didn't see your previous
question, but you do know that you might be using networkd and as
such that file is not handeld at all ?
649[09:56:05] *** Quits: akp55 (~akp55@replaced-ip) (Remote host closed the connection)
662[10:05:20] <shtrb> ym, if you didn't answer that before
, does using pon rostelekom (you should have file
/etc/ppp/peers/rostelekom and proper chap-secrets for it) work for
you ?
665[10:06:56] <ym> I have /etc/ppp/peers/rostelekom and proper
chap-secrets for it. The issue is that my configuration works only
after systemctl restart networking and not after startup, though
logs telling me everything just fine.
681[10:21:35] <ym> Seems like replacing "auto
rostelekom" to the top of interfaces file solved the issue I
described earlier. Thanks everyone for suggestions.
742[11:40:23] <melissa_> I want the freedom to not install
updates for years. If I have my apt sources pointing to testing, and
I don't install updates for, say, 3 years, is there a
significant chance my system breaks?
743[11:40:23] *** Quits: unlike (~deb@replaced-ip) (Remote host closed the connection)
744[11:40:45] *** Quits: deb (~deb@replaced-ip) (Ping timeout: 240 seconds)
746[11:42:05] <ratrace> melissa_: all software has bugs, sooner
or later. if you're lucky and can run 3 years with no _need_ to
apply an update, then all is well. but if (more likely) you end up
needing an update, you're setting yourself for big fail unless
you run stable.
747[11:42:19] <melissa_> And what if I have them pointed to
Bullseye, a few more releases come out, and I follow the regular
upgrade path of doing apt-get update / system-upgrade and change my
sources to the next release until I arrive at the new stable. Is
that more or less likely to give problems?
748[11:42:51] <ratrace> !xy
749[11:42:51] <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
750[11:43:55] <melissa_> I want to install GNU on a computer
that I might not use for years, and then start using again, with
little maintenance
751[11:44:34] <ratrace> you're more likely to achieve that
with _stable_ rather than testing
752[11:45:05] <JPT> If you never connect it to the internet, it
is at least /somewhat/ safe to stick with one installation and never
update it.
753[11:45:27] <JPT> If you're going to connect it to the
internet, i would always urge you to keep it up to date
754[11:45:30] <ratrace> as current stable becomes oldstable, 3
years from now you should be able to start usign it again and apply
only bugfix and security updates. _sometimes_, but that's rare,
they might introduce incompatibilities, esp. over longer time spans,
but that's unlikely
756[11:46:38] <melissa_> How so? Because at one point Bullseye
will become stable. So if I install that (and point sources to
bullseye rather than testing), won't it just take longer until
my system gets outdated?
757[11:47:02] <ratrace> and overtime as oldstable becomes
oldoldstable, like today it's still possible to run wheezy or
jessie, and there's that "debian LTS" project or
whatsitcalled where certain contributors keep maintaining
oldoldstable with _some_ security fixes
758[11:47:24] <ratrace> melissa_: becuse right now bullseye is
not stable and it may change until freeze and release in
incompatible ways
759[11:47:46] <ratrace> also, if ou install bullseye _now_ and
then reactive that in 3 years, I don't think there will even
exist an upgrade path
760[11:48:09] <melissa_> I see
761[11:48:39] <ratrace> most notable upcoming change is
depreciation and removal of python2 for example
762[11:49:10] <ratrace> ("upcoming" in terms of what
bullseye will have compared to buster)
763[11:49:48] <melissa_> nooo
764[11:49:51] <melissa_> don't remove python 2
765[11:50:12] <ratrace> it's dead. bereft of life. it
ceased to be. it's an ex python.
766[11:50:41] <nkuttler> python2 is gone from sid
767[11:50:58] <nkuttler> shouldn't remove it from a stable
system though
776[11:54:03] <gry> melissa_: it is a wiki, if you sign up
<replaced-url
777[11:54:07] <ratrace> melissa_: I don't think they
specifically recommend to use contrib and non-free, versus not using
them at all. That's just an example sources.list
779[11:54:41] <melissa_> Yeah but if people copy/paste that,
they will add contrib and non-free even if they didn't have it
before
780[11:55:25] <ratrace> and nothing bad will happen with that.
also people should NOT copy&paste from internet into terminals
wihtout understanding the code they're pasting
781[11:55:40] <flayer> i've been doing it wrong all these
years!
782[11:56:05] <melissa_> So would this upgrade path work? Buster
-> LTS, and after June 2024, do an upgrade path to whatever will
be stable by that time?
783[11:56:30] <ratrace> I don't think skipping stable
releases is a supported upgrade path
784[11:56:48] <melissa_> I mean the regular upgrade path,
without skipping any
785[11:56:55] <ratrace> so you will have to go from buster to
bullseye to debian-next-next
793[12:00:46] <brutser> I have put a line in fstab to mount an
SD card, this works fine, but when the SD card is not present, it
makes boot wait 150 seconds or so for the SD card, is there some way
I can tell fstab to not bother when SD card is not there? or should
I do this not with fstab if i want this?
794[12:01:18] <ratrace> brutser: yes, use "nofail"
option in the fstab line
795[12:01:42] <brutser> ah ok :) so nofail says, if it's
not there, don't bother?
796[12:01:59] <ratrace> right
797[12:02:14] <brutser> ok thanks!
798[12:02:49] *** Quits: brutser (574149f0@replaced-ip) (Remote host closed the connection)
872[12:34:48] <jelly> melissa_, LTS is mostly just and
extenstion of security support for a release. going with buster and
then buster-lts, you still have the usual buster->buster+1
release upgrade path
876[12:37:20] <jelly> (and offtopically, ubuntu seem to have
some issues with that, going from 18.04 LTS to 20.04 LTS is STILL
not ready, even though 18.04.1 was out last month)
888[12:39:49] <jelly> and canonical giving up on their Unity
thing which was rather nice
889[12:39:50] <ratrace> and more ontopically, I've always
had success with debian stable upgrades. there might be issues with
backported packages, and third party repos.
896[12:42:51] <jelly> going sysvinit->systemd from 7 to 8 was
a lot less problematic than introducing udev 4->5
897[12:43:22] <jelly> 7->8 was mostly upstream stuff, apache
2.2 -> 2.4 changes and such
898[12:43:54] <ratrace> there's of course more chance of
something going wrong, the more diverse you made your installation
in term of packages. but for ubuntu, I've had issues with base
installation upgrades, ie. install 16.04 -> upgrade ->
do-release-upgrade -> boom
899[12:44:22] <ratrace> wheezy to jessie surprised me too. I was
expecting issues due to init change, yes, but it was just fine
900[12:46:08] <ratrace> I wonder if python2 removall will mess
up buster->bullseye upgrade
901[12:46:48] <jelly> you don't have to remove obsolete
packages if you don't want to
910[12:55:06] <no_gravity> You can configure git diff colors
with values like #ff0000 for red etc. But those don't work when
inside screen. Any ideas why and how to make them work?
934[13:08:13] <ratrace> bots here really are stupid... they
should be made to trigger on explicit keywords. bug #... and not
just #... dpkg reacting to ! only, not every sentence starting with
dpkg (which is rather not uncommon in a Debian support channel with
"dpkg2 being the primary package manager)
935[13:08:49] <ratrace> no_gravity: I'm not 100% sure but I
think with 256 color terminals you don't have the option of
specifying random rgb value
936[13:09:10] <no_gravity> ratrace: Why does my terminal outside
of screen support it then?
937[13:09:26] <diogenes_> Greg Kroah-Hartman said tha Debian
runs the world.
938[13:09:44] <ratrace> no_gravity: I don't know.
approximation? or do oyu actually get 4 billion colors?
939[13:10:35] <no_gravity> ratrace: I would have to make
screenshots with slightly different colors and compare them in gimp
...
940[13:11:21] <no_gravity> Let me do i t...
941[13:11:23] <no_gravity> Let me do it...
942[13:11:24] <ratrace> no_gravity: right. well, I don't
know. I misread your original question thinking you need just
regular color support, but I'm not familiar with supplying full
32-bit range of colors, sorry.
1024[14:06:24] <SPARC-Corgi> Meh ppc port is kinda glitchy
don't think it would matter much anyways. Probably just gonna
try a raid card if I can find one
1085[15:33:23] <dvs> blurkis, #debian-next on OFTC
1086[15:34:38] <blurkis> dvs: thanks. (But I solved it, for some
reason kde started a mysql instance as $user and installing mariadb
would not install without me personally killing that instance.
1112[15:58:22] <n4dir> though the output of lspci also also has:
01:05.1 Display controller: Advanced Micro Devices, Inc. [AMD/ATI]
RS480 [Radeon Xpress 1150] (Secondary)
1124[16:05:55] <ElDiabolo> I have 2 desktop systems where I just
did an upgrade. On one of the thunderbird is on 68.12.0-1~deb10u1,
on the other it is 68.10.0-1~deb10u1. Both are amd64. How can I find
the cause.
1165[16:33:32] <oxek> imo this is how debian sources.list should
look like by default
replaced-url
1166[16:33:49] <ElDiabolo> oxek, I just checked my online system,
this has security updates :-).
1167[16:33:52] <shtrb> Does anyone have a clue where kdeconnect
store it's sms db (even if it store locally) ? tried "find
. -name *kdeconnect* -print | grep -vvv hexchat | xargs" which
gave me ./.cache/kdeconnect.app ./.cache/kdeconnect.sms
./.config/kde.org/kdeconnect.app.conf ./.config/kdeconnect
./.config/kdeconnect/MACID/kdeconnect_pausemusic
./.config/kdeconnect/MACID kdeconnect_runcommand
./.config/kdeconnect/MACID/kdeconnect_sendnotifications
./.local/share/kpeoplevcar
1168[16:33:53] <shtrb> d/kdeconnect-MACID but none of these
location seem to store anything that I could identify as some kind
of a db.
1172[16:36:07] <somiaj> buster-updates is more optional than
security-updates though, it just gives you access to some packages
before they are merged in a point release.
1173[16:36:56] <oxek> 'optional' but if you don't
have buster-updates, then you're on old ca-certificates and
thus have lots of web breakage due to that major root CA expiration
1199[17:00:51] <chrisw> Just upgraded a machine to buster, and it
seems to now be requesting (and getting) a new ip address via dhcp
every few seconds. Any ideas what might have happened and how I
could fix this?
1200[17:02:18] *** Quits: amlchief (~amlchief@replaced-ip) (Remote host closed the connection)
1204[17:10:26] <somiaj> chrisw: what do you use to configure your
network?
1205[17:11:11] <chrisw> @somiaj: not sure I understand the
question, the dhcp server is a UDMP Pro with a static ip assigned
for this device, which was working fine up until the buster upgrade
1206[17:11:22] <chrisw> UDM Pro, even
1207[17:11:59] <chrisw> also just noticed the root fs has come up
read only, which isn't ideal... wonder why that might be?
1208[17:12:37] <somiaj> debian provides multiple tools, the
interfaces file, network-manager, wicd, and other ways to configure
the network. Which one are you using?
1211[17:13:12] <somiaj> hmm, it could be the read only part of
the file system causing the issue since dhcp can't write its
lease file. I would track down why that is happening. Did it fail an
fsck?
1212[17:13:20] <chrisw> @somiaj - no idea, I'm afraid, had
this machine for years, have always just let upgrades do their
thing, how would I check?
1213[17:13:22] <somiaj> might want to reboot into a live system
and fsck the file system.
1214[17:13:42] <chrisw> dmesg now showing anything obvious for
the fs, where else should I be looking?
1261[17:41:50] <oxek> linux-image-amd64 depends on
linux-image-4.19.0-10-amd64, and if you do 'apt show
linux-image-4.19.0-10-amd64' it will tell you it is 4.19.132
1268[17:48:37] <oxek> you only need a signed kernel if you have
enabled secureboot on a UEFI system, and you can sign kernels
yourself and put your key into the UEFI
1269[17:48:49] <oxek> on a bios system, all of that is irrelevant
1307[18:26:03] <chrisw> @oxek - cool, thanks, ifconfig has always
"just worked", so use it, will try and re-train to `ip a`
now :-)_
1308[18:26:44] <shtrb> Is using virt-manager to manage a dozen or
so vms on a laptop is a waste of time and I should be using ovirt
instead or it's not waste of time /resources ?
1309[18:26:47] <chrisw> hurrah! A real shell, not the crap
HP's iLO offers...
1310[18:27:43] <ratrace> shtrb: it's just a frontend to
libvirt and qemu. if it works for you and satisfies your needs, no
need to use anything else.
1313[18:28:10] <ratrace> personally I prefer using qemu-system on
the command line, but that's just my preference. at the end of
the day, it's the same qemu in the background
1337[18:33:07] <oxek> so if they no longer care about it, or
provide funding to the devs, it's essentially dead, no?
1338[18:33:16] <ratrace> virt-manager is the defacto linux-native
response to virtualbox. cockpit is RHEL specific control panel. if
RH abandons it and other distros/contributors don't pick up....
well...
1353[18:37:35] <chrisw> okay, so why would `systemctl restart
networking` result in my interface being DOWN in `ip a` output?
1354[18:37:42] <ratrace> oh yes, that.... indeed I forgot about
it
1355[18:38:07] <JyZyXEL> "to boost their near-term
finances" at the expense of their future it sounds like :P
1356[18:38:16] <ratrace> chrisw: because restarting
networking.service does ifdown+ifup. if your NIC is not configured
in e/n/i or errors out, that will happen
1363[18:39:34] *** Quits: magyar (~magyar@replaced-ip) (Quit: Riding the split)
1364[18:39:39] <ratrace> chrisw: networking.service controls
ifupdown framework which is configured by "e/n/i" which is
usual abbrev for /etc/network/interfaces
1369[18:41:39] <chrisw> @shtrb - nothing magic here, simple two
physical nics, which are still showing as eth0 and eth1, even after
the stretch to buster upgrade
1370[18:42:01] <chrisw> (which seems to have gone badly wrong
somewhere - the system boots up with a read-only root partition, and
I'm still trying to find out why...)
1372[18:42:52] <chrisw> the read only root file system seems to
cause the dhcp client to repeatedly renew its leases, which ends up
configurin the UDM pro that's that dhcp server
1373[18:43:23] <ratrace> chrisw: if your rootfs mounts RO, there
could be a filesystem corruption issue
1374[18:43:24] <chrisw> I can remount the fs rw just fine, which
is also confusing... where/how do I configure the kernel boot
command line?
1375[18:43:32] <ratrace> assuming of course it isn't
misconfigured
1376[18:43:47] <chrisw> ratrace: don't think so, I suspect a
misconfigure from upgrading stretch to buster
1377[18:43:49] <ratrace> chrisw: ah maybe you're missing a
fstab entry for /
1378[18:43:53] <shtrb> In order to still have ethX you should
have disabled predictable interfaces ( via kernel argument ) , if
you have network fs that fail to load you might get into RO case
1379[18:44:14] <chrisw> @ratrace - I don't think so there
either, everything is there and fine, just mounted ro
1380[18:44:31] <shtrb> *or via udev rules
1381[18:44:41] <ratrace> chrisw: then comb the logs, I'm
sure there's a reason filed there somewhere
1382[18:45:34] <chrisw> @shtrb - I followed the stuff in
replaced-url
1391[18:47:58] <chrisw> @shtrb - right, that answers that: udev
has defs for eth0 and eth1 :-)
1392[18:48:13] <shtrb> chrisw, you can disable it (if you wish )
1393[18:48:19] <somiaj> chrisw: stil sounds like you need to boot
into a live system and fsck your file system first
1394[18:48:49] <ratrace> chrisw: in some cases rootfs is mounted
RO by initramfs, I don't remember if it's default or not
(I have custom config, so can't check for default), and then
one needs a fstab entry to remount rw. if you have that entry, then
there must've been an error somehwere which must've been
logged
1395[18:49:02] <shtrb> chrisw, and you might have done some
systemd magic that might had broken your boot process.
1396[18:49:23] <oxek> shtrb: the name might be offensive to
someone
1413[18:51:15] <chrisw> @ratrace - okay, so I wonder what changed
as part of the stretch to buster upgrade? I did notice some stuff in
the console relating to a volume group not being found, but
it's still there, just mounted readonly
1414[18:51:47] <ratrace> somiaj: it's in final generated
grub.cfg
1415[18:52:16] <chrisw> what's weird is if I run
update-grub, I get `/usr/sbin/grub-mkconfig: 253:
/usr/sbin/grub-mkconfig: cannot create /boot/grub/grub.cfg.new:
Directory nonexistent`, but I boot through a grub menu, so
where's that coming from?
1416[18:52:26] <somiaj> oh there it is, yea that is where I found
it, so must be something that should be remounting it int he boot
process
1417[18:52:46] <shtrb> ratrace, I was wrong , it's was less
crazy -
replaced-url
1439[18:57:49] <chrisw> looking at the start of dmesg, I see [
0.000000] Command line: BOOT_IMAGE=/vmlinuz-4.19.0-10-amd64
root=/dev/mapper/zuul--vg-root ro quiet - where is that coming from?
1440[18:57:50] <shtrb> chrisw, generated via os-probe and
grub-update
1442[18:58:30] <ratrace> somiaj: I ran into that few years ago
and remembered to always mention root in fstab (for filesystems that
use fstab), but things may've changed since, in newer
initramfs-tools
1443[18:59:39] <somiaj> might be if you don't have an fstab
entry you need a .mount unit file (Since those are basically
equivlant). I have never not had rootfs in fstab, so explains why I
havn't ran across such a thing
1444[18:59:55] <chrisw> @shtrb - right, but like I said above,
grub-update can't even find the directory it wants to write
to...
1445[19:00:27] <somiaj> chrisw: output of ls /boot please (use a
patebin)
1446[19:00:43] <ratrace> somiaj: I install debian with
debootstrap, and it was in the early testing periods where I ran
into that
1447[19:00:54] <MTecknology> I know creating a symlink in this
way is the "wrong way", but I also don't understand
how libraries work well enough to know how to solve this issue
properly-
replaced-url
1448[19:00:58] *** Quits: nickodd (~nickodd@replaced-ip) (Remote host closed the connection)
1449[19:01:00] <MTecknology> original_path to diverted_path, the
library is found.
1450[19:01:02] <shtrb> chrisw, I'm not sure about your setup
but : 1. you can info at the UEFI, 2. you may have info already
written in /boot
1451[19:01:02] <chrisw> @somiaj : `initrd.img-4.19.0-10-amd64` is
all that's in there
1452[19:01:02] <ratrace> and yea, either fstab or a proper .mount
(as fstab gets converted to .mount units anyway)
1453[19:01:27] <shtrb> is your /boot also on lvm ?
1454[19:01:30] <chrisw> @shtrb - I'm afraid I don't
know what `you can info at the UEFI` means
1455[19:01:32] <ratrace> MTecknology: are you not installing
steam from regular debian package?
1456[19:01:32] <somiaj> chrisw: did /boot use to be a seperate
partition?
1457[19:01:49] <chrisw> @shtrb - yes, I think /boot is on lvm
1458[19:02:03] <shtrb> chrisw, typo , part of your boot
infromation is stored at the UEFI (efivars ?) .
1459[19:02:13] <somiaj> chrisw: and somehow after the reboot
/boot isn't being mounted correctly, because you should have
more than the init image in there
1460[19:02:22] <chrisw> Not really sure why I'm on lvm,
other than that some ancient debian upgrade probably told me I
needed to do it
1461[19:02:23] <shtrb> chrisw, I don't know how /boot on lvm
can normally boot into a working envoirment
1462[19:02:28] <MTecknology> ratrace: I installed it from the
.deb they provided. The reason for the location of that steamui.so
file is that /usr/bin/steam unpacks a bunch of stuff in the home
directory and executes from there.
1463[19:02:38] <ratrace> MTecknology: "they"?
1464[19:02:48] <MTecknology> steam? valve?
1465[19:03:04] <ratrace> MTecknology: steam is a regular package
on debian. use that. it's just a launcher anyway, and it
installs latest from valve
1466[19:03:07] <ratrace> ,v steam
1467[19:03:08] <judd> No package named 'steam' was
found in amd64.
1468[19:03:18] <chrisw> @somiaj - indeed, something has gone very
screwy as a result of the stretch / buster upgrade - that said, this
machine has been upgraded for probably a decade at least
1469[19:03:23] <ratrace> ,v steam:i386
1470[19:03:24] <judd> No package named 'steam:i386' was
found in amd64.
1471[19:03:31] <ratrace> oh how do you operate this thing....
1492[19:09:34] <MTecknology> chrisw: fyi- in irc land, we follow
$nick: $message. Highlights don't always work for twitter-style
mentions and if you follow the convention, you can use tab-complete
for it. such as mt<tab>
1515[19:15:10] <chrisw> anyway, bit dismayed to see /boot is ext2
- is that normal?!
1516[19:15:20] <shtrb> ratrace, p.d.n is down :-(
1517[19:15:57] *** Joins: metro (~metro@replaced-ip)
1518[19:16:05] <ratrace> many orbits ago, people would put /boot
on ext2 to... I don't know... save a few cycles of performance
once per boot by not haivng a journal there
1519[19:16:10] <chrisw> @shtrb - sorry to be pedantic, but in
case anyone is able to fix that, it's missing a dns entry,
whether the service is up or down we can't currently tell...
1520[19:16:43] <chrisw> ratrace - plausible then, okay, so how
come if I ls /boot, there's just the one kernel image there?
1522[19:16:53] <chrisw> I definitely see a grub menu with a bunch
of entries on it...
1523[19:17:14] <MTecknology> I still put /boot on ext2. I was
told journaling is counter-productive for boot things.
1524[19:17:38] <ratrace> MTecknology: you were told wrong
1525[19:18:26] <somiaj> chrisw: almost sounds like your fstab was
corrupted during the upgrade, since both / and /boot aren't
there, might just need to remount everything properly and fix the
fstab file
1526[19:18:28] <shtrb> ratrace, the non ext/fat32 /boot UEFI
support is a recent thing
1528[19:19:38] <chrisw> somiaj - fstab looks fine (well, to me,
am I missing something?):
replaced-url
1529[19:19:48] <ratrace> shtrb: well if one wants to use ext2 for
/boot, that's fine I suppose. the journaling and full features
of ext4 don't in any way counter the productivity of /boot
1530[19:20:14] <ratrace> boils down to preference, and not a
technical advantage
1532[19:20:49] <MTecknology> ratrace: It turned out to be very
easy to get steam:i386 installed, but because I'm running
frankian, the nvidia-driver-libs-i386:i386 package it said I should
install is proving quite painful. :/
1533[19:21:14] <MTecknology> The joys of bleeding edge
hardware... I have no external monitor support without the nvidia
driver from experimental.
1534[19:21:17] <ratrace> wellp... had to be a reason we warn
against frankendebians :)
1535[19:21:55] <shtrb> It's a great time to drop all pins ,
and move to the safety of testing :D
1551[19:24:45] <ratrace> yeah buster backports is all about
440.100
1552[19:25:20] <MTecknology> I guess the solution is to say screw
the external monitor and drop back to 440. External monitor support
is still pretty janky anyway.
1553[19:25:33] <chrisw> somiaj: So, during startup, I see `Volume
group "zuul-vg" not found`, but nothing in dmesg - where
might that be coming from? Interesting to see that
`/dev/mapper/zuul--vg-root` ends up being mounted, is that double
dash significant?
1554[19:26:04] <chrisw> and still, the mystery of where the grub
menu that I see during boot is actually coming from?!
1555[19:26:42] <somiaj> well if it can't find your volume
group, it won't find the volumes to mount, that coudl be the
cause of the issue
1557[19:27:12] <chrisw> somiaj - okay, but what is it that ends
up being mounted then? where does volume group config live?
1558[19:27:14] <RhinoCodes> convice me to use Debian
1559[19:27:30] <somiaj> it could be, names have to match exactly,
my lvm knowledge is really limited as I don't use it on my
personal systems, and it has just worked on the few servers I have
1579[19:31:01] <shtrb> somiaj, I doubt anyone would ask users
(the same as it was systemd as a whole ), it would be the Debian
community who would decide on that subject
1580[19:32:11] <MTecknology> I'm in aptitude and pressed a
key... but I don't know which key. The package is now showing
as yellow and "iw". I know the i means install. Does
anyone happen to know what the w means?
1586[19:35:26] <shtrb> somiaj, I think most users can't vote
(who are not DM / DDs)
1587[19:35:28] <somiaj> MTecknology: from dpkg,
Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
-- it appears @ is triger-awaiting
1593[19:37:44] <MTecknology> interesting; can you enlighten me to
where you found @=w?
1594[19:37:58] <MTecknology> and @=trigger-awaiting?
1595[19:38:19] <somiaj> MTecknology: @ was my typo, I found that
Status= line at the top of dpkg -l | haed
1596[19:38:23] <somiaj> dpkg -l | head
1597[19:38:23] <dpkg> ii | head 2.9-1.2 somiaj's private
warez collection
1598[19:39:00] <somiaj> I just know that always lists the
possible status a package can be in, seems this is related to
dpkg-trigger, and that package is installed but waiting some triger
to finish configuring
1610[19:44:44] <MTecknology> The curses interface is super duper
fantastic for working out dependency issues. It's a bit
confusing at first, but you can select/view one package, then use +
to select a specific version, if it shows red, then there's a
dependency that needs to be resolved. You can let aptitude try to
auto-resolve it or press enter on that version to see package info
for that specific version and
1611[19:44:50] <MTecknology> it'll highlight in red the
problems. Then you can expand those, choose to install, and if
it's red, you can dig even deeper.
1619[19:53:43] <chrisw> somiaj - well, progress... installed
ntfs-3g, which meant I could get the last two mounts in the fstb I
posted working, (I'd systemctl mask'ed them...) and now
everything comes up rw
1632[20:00:30] <MTecknology> The message that says ~"has
files in directory, not removing" should tell you which path.
1633[20:01:36] <Ede|Popede> MTecknology: i used aptitude some
years ago, but for dependencies there's also debtree. make it
create a graph together with dot and you can even keep them for the
next time.
1634[20:01:58] <gribouille> MTecknology, the package was removed
a long time ago
1637[20:03:37] <MTecknology> You could maybe re-install/re-remove
and it should spit out that message again. Otherwise- I'm not
sure that it's possible.
1731[21:10:22] <sney> message #215 in that thread covers it
pretty succinctly.
replaced-url
1732[21:11:08] <sney> if Oracle continues to refuse to provide
security support for their own previous versions, then those same
versions can't have security support in debian stable releases.
therefore they wont be in stable or stable+1
1739[21:16:18] <oxek> oracle provides support for 6.1 6.0 and
5.2, so there has been progress in that too
1740[21:18:26] <oxek> and... virtualbox installation from oracle
repo failed...
1741[21:20:39] <oxek> ¯\_(ツ)_/¯
1742[21:20:41] <sney> oracle has a long history of gradually
making life harder for free software projects. They did it with
virtualbox, they did it with mysql, they did it with java, etc...
1743[21:20:56] <oxek> I know oracle is bad
1744[21:21:34] <sney> the people making those decisions for
debian are probably just not very inclined to bend over backwards to
package virtualbox. particularly when all of the libvirt/qemu/etc
stuff is available without any oracle-isms.
1796[22:08:30] <karlpinc> Well, I could use aptitude install....
and if there's depenencies it always asks and I can say no. But
that's kinda lame. There must be a way to "just ask".
1797[22:08:51] <oxek> apt -s install
1798[22:09:08] <jmcnaught> karlpinc: "aptitude search
'?provides(virtualbox)'" (short for ~Pvirtualbox)
will show you which real packages provide a virtual package.
It's none for virtualbox on buster for me.
1807[22:10:59] <karlpinc> oxek: (Why are apt options documented
only on the apt-get man page? Or) did you mean apt-get?
1808[22:11:10] <oxek> karlpinc: I did mean apt
1809[22:11:24] <oxek> apt is the user interface to apt-get,
apt-cache, etc.
1810[22:11:34] <karlpinc> oxek: Ok. (Just another reason to be
unpleased with "apt". ;)
1811[22:12:02] <jmcnaught> Virtual packages are used when
there's more than one package that could fill that role (think
apache2 or nginx both provide httpd allowing other packages to
agnostically depend on 'httpd'. You can't install a
virtual package directly, you need to explicitly select one to
install.
1812[22:12:03] <karlpinc> oxek: Well, but not aptitude. (I
assume.)
1889[23:42:56] *** mirrorbird is now known as SleepyJoe
1890[23:43:18] <oxek> sney: what I said earlier was mistaken -
oracle does not provide support for vbox 5.2 nor 6.0, support was
ended in July 2020, hence only vbox 6.1 (the latest release) is
supported.
1891[23:43:58] <oxek> I can see why that would be an issue for
debian
1892[23:44:33] <jhutchins> I thought there was a vbox component
or driver or something that wasn't free as well.
1893[23:44:52] <oxek> version 6.0 cut support for 32bit hosts
(and that's why the latest 5.2 download is still offered), and
6.1 cut support for software virtualization (hence 6.0 download is
still offered).
1894[23:45:10] <oxek> 6.1 will receive long term support
1895[23:45:18] <oxek> so 6.1 could be suitable for debian
1896[23:45:30] <oxek> (no big cuts in functionality are planned)
1897[23:45:47] <oxek> jhutchins: vbox could still be in contrib
& non-free
1898[23:45:54] <oxek> just like rar and others
1899[23:45:55] <jhutchins> I was realised that the free VMWare
thing wouldn't handle 64b software about five years ago.
1900[23:46:11] <sney> jhutchins: oracle's licenses are
always crappy, but if security support is available then it can
still be in non-free. but it isn't.
1904[23:47:08] <sney> and oracle might be saying that
they'll LTS 6.1 right now, but we know from experience that
they might just change their minds with no warning
1905[23:47:20] <sney> it's not my call but I wouldn't
do it
1910[23:48:55] <oxek> vbox is the only "open-source"
option on windows, no?
1911[23:49:13] <sney> yeah, the worst thing about oracle is that
their software is usually pretty useful. so we can't just
disregard them and their products completely.
1918[23:56:29] <somiaj> isn't windows WSL or whatever it
called an option (not open source) but comes with windows -- though
I think it is mostly for chroots/console like setups