Achylles | how to install docker in devuan? | 00:02 |
---|---|---|
ham5urg | One customer has a running Devuan with zero downtime for 3 years now. | 00:02 |
Achylles | and translate-shell | 00:02 |
brocashelm | maybe miyo would be a good way to introduce devuan to someone not familiar | 00:02 |
brocashelm | oh, server? | 00:03 |
ham5urg | Yes. A SMB server. | 00:04 |
brocashelm | nice | 00:05 |
ham5urg | Achylles, I use LXC and circumvent docker. | 00:06 |
argie | Hello people. Anyone noticed a change in `apt show` on daedalus? (2.2.4+devuan1 amd64) It seems it only outputs the short description, not the full one available in control files | 00:52 |
argie | Particularly at /var/lib/dpkg/status, it seems the full descriptions are available, however `apt show <package>` only shows the short description | 01:01 |
Tenkawa | do you have an example? | 01:05 |
Tenkawa | of a pkg you used to test with? | 01:05 |
Tenkawa | or is daedalus the pkg? | 01:05 |
argie | Nope, the package apt is version 2.2.4+devuan1 | 01:06 |
argie | Let's say the package apt, if you look for "Package: apt" in /var/lib/dpkg/status there's the full description in the "Description:" field | 01:06 |
argie | However, when running `apt show apt` all the Description: field shows is "commandline package manager", this is, the short description | 01:07 |
Tenkawa | remember daedalus is also still in devel so I am not 100 % surprised... let me debootstrap a chroot and take a look real quick | 01:08 |
Tenkawa | ok getting the pkgs now | 01:15 |
Tenkawa | wow that was a slow dload today | 01:33 |
Tenkawa | installing my vm now | 01:33 |
argie | lol, thanks for testing for such an "issue" | 01:36 |
Tenkawa | nah I'm just really curious myself | 01:37 |
Tenkawa | I like checking out things like this | 01:37 |
Tenkawa | which release did you start at? I am starting at Chimaera though | 01:39 |
Tenkawa | (to compare against) | 01:39 |
argie | I might have started at beowulf, and have been migrating to chimaera (when it was testing) and recently migrated to daedalus | 01:40 |
Tenkawa | I'm only getting Description: field shows is "commandline package manager" too | 01:40 |
Tenkawa | in Chimaera | 01:41 |
argie | Allright | 01:41 |
argie | So at least it's reproducible | 01:42 |
Tenkawa | found it | 01:42 |
Tenkawa | try this | 01:42 |
Tenkawa | sudo dpkg-reconfigure locales | 01:42 |
Tenkawa | set your languages | 01:43 |
Tenkawa | pick your preferred output language | 01:43 |
Tenkawa | then try it again | 01:43 |
Tenkawa | now its outputting the whole paragraph for me | 01:43 |
Tenkawa | it was doing just Description-en: before which was the giveaway | 01:44 |
argie | Well, I reconfigured it with my two languages and set en_US.UTF-8 as default | 01:44 |
Tenkawa | the locales weren't generated yet | 01:44 |
argie | And still there's just the short description | 01:44 |
Tenkawa | hmmm... mine shows it now | 01:45 |
Tenkawa | I picked all 3 english though | 01:45 |
argie | Well, there surely is some weird config in my part, I got to say | 01:46 |
Tenkawa | for en_us to generate | 01:46 |
Tenkawa | locales have been a weird thing in the past for Linux in general | 01:46 |
argie | I put some variables to my other language, specifically LC_MEASUREMENT, LC_NUMERIC, LC_TIME, and LC_MONETARY | 01:46 |
Tenkawa | ahhh | 01:46 |
Tenkawa | yeah that could overlap | 01:46 |
argie | All of those ones to es_AR.UTF-8, the rest en_US.UTF-8 | 01:46 |
argie | Lemme comment them and see if it fixes it | 01:47 |
Tenkawa | after you logout/login regenerate it again | 01:47 |
Tenkawa | you should see how tough this is with doublebyte languages | 01:48 |
Tenkawa | you have a few special characters I know | 01:48 |
Tenkawa | heheh | 01:48 |
argie | :P | 01:49 |
argie | Well, I logged out and it from my tty, reconfigured locales with the special LC_* from /etc/default/locales commented (added en_GB.UTF-8 as well), and tried `apt show apt` | 01:50 |
* Tenkawa worked on multilanguage support in the 90's | 01:50 | |
argie | Still short description | 01:50 |
Tenkawa | well darn | 01:50 |
Tenkawa | yeah I'm getting this package provides...... | 01:51 |
Tenkawa | 3 lines long | 01:51 |
Tenkawa | thats long descript right? | 01:51 |
argie | The "Provides:" section? | 01:51 |
Tenkawa | just a sec | 01:51 |
Tenkawa | starts with apt-transport-https | 01:52 |
Tenkawa | then is like 15 lines long | 01:52 |
Tenkawa | let me scp it out... can you pastebin yours? | 01:53 |
argie | I think I'm talking about a different part of the command | 01:53 |
Tenkawa | I'm going to try to pull mine out of the vm | 01:53 |
argie | All of the Provides:, Depends:, Recommends:, and so on sections are not part of the description | 01:53 |
argie | https://www.debian.org/doc/debian-policy/ch-controlfields.html#description | 01:54 |
Tenkawa | can you pastebin what you are seeing? | 01:54 |
argie | https://paste.debian.net/1209713/ | 01:55 |
Tenkawa | yes.. thats correct | 01:56 |
Tenkawa | from debian unstable: | 01:56 |
Tenkawa | Description: commandline package manager | 01:56 |
Tenkawa | Version: 2.3.8 | 01:57 |
Tenkawa | thats its current description | 01:57 |
argie | Let me share you what I found on /var/lib/dpkg/status so you can get what I mean | 01:57 |
Tenkawa | ok | 01:57 |
fsmithred | I see a lot more in ceres and beowulf | 01:57 |
argie | https://paste.debian.net/1209714/ | 01:58 |
fsmithred | yeah, like that | 01:58 |
argie | See the longer description after line 21? That's what doesn't show up on `apt show` | 01:58 |
fsmithred | I don't have daedelus yet, and chimaera is turned off. | 01:58 |
fsmithred | but I'm pretty sure we haven't done any work on daedelus yet. We're still finishing chimaera. | 01:59 |
Tenkawa | argie: yeah... thats what started showing up when I did the locales | 01:59 |
Tenkawa | thats what locales fixed for me | 01:59 |
argie | What I find weird is that daedalus and chimaera share the same version of the apt package | 02:00 |
argie | Even with ceres, it seems, by what I find on pkginfo.devuan.org | 02:00 |
Tenkawa | before it just said Description-en: then nothing | 02:01 |
Tenkawa | all those lines were gone | 02:01 |
Tenkawa | for me | 02:01 |
Tenkawa | once I generated my locales it looked better | 02:01 |
fsmithred | argie, that's the version I would expect. We have been building packages for unstable, and they have been migrating to chimaera. That migration got turned off a few days ago. | 02:02 |
fsmithred | so daedelus will have whatever is in ceres now | 02:02 |
argie | Anyways, Woah, wait | 02:02 |
argie | I just changed daedalus from chimaera on my sources.list URL, did apt update | 02:02 |
argie | And now the full description shows up! | 02:03 |
Tenkawa | fsmithred: hey... which is "more widely available?" pkginfo or deb? | 02:03 |
Tenkawa | the site urls | 02:03 |
fsmithred | package repository site? | 02:03 |
Tenkawa | yeah | 02:03 |
fsmithred | deb.devuan.org is round-robin | 02:03 |
Tenkawa | ok good | 02:03 |
Tenkawa | thought so | 02:04 |
fsmithred | pkgmaster.devuan.org is our main server, mirrors sync to that. Please don't use it. It's slow. | 02:04 |
argie | I think I found the issue | 02:04 |
fsmithred | if deb.d.o is slow, pick a fast mirror from the list. | 02:04 |
argie | At least on my mirror, devuan.dcc.uchile.cl, daedalus doesn't seem to have Translation-en | 02:04 |
Tenkawa | argie: yeah that would do it | 02:04 |
fsmithred | yeah, it's not really ready yet. I wouldn't mess with it now. | 02:05 |
fsmithred | ceres would be safer, I think. | 02:05 |
argie | I'll just have to wait then, no big deal | 02:05 |
* Tenkawa has major MNLS flashbacks.... | 02:05 | |
Tenkawa | lol | 02:05 |
argie | lol, whats MNLS? | 02:06 |
fsmithred | MNLS? | 02:06 |
Tenkawa | multi national language support | 02:06 |
argie | Oof, indeed | 02:06 |
Tenkawa | what I did in the 90's for telecom | 02:06 |
Tenkawa | translation software | 02:06 |
argie | Oh no, the 90s, codepages and standars everywhere | 02:07 |
Tenkawa | yep | 02:07 |
Tenkawa | indeed | 02:07 |
Tenkawa | now to refill my drink... cheers for now... | 02:08 |
Tenkawa | bbialw | 02:08 |
argie | o/ | 02:09 |
argie | I just really came to see if you knew about this issue, so I'll be off too | 02:09 |
argie | Happy devuaning or smth | 02:09 |
fsmithred | :) | 02:09 |
ham5urg | Does anyone uses suspend-to-disk? In the last 10 years I did not used it. But who knows, maybe it's similar fast as suspend-to-ram. The only point which stood against it, that the window-manager like gnome does not have a menu for it (only suspend-to-ram). | 10:09 |
lts | The big use case for suspend-to-disk would be to avoid systemd's "a stop job is running". I haven't had use for it without systemd | 10:17 |
ham5urg | I like to suspend without bothering about the battery (laptop) | 10:20 |
ham5urg | No mention about systemd (which itself is undesired) | 10:21 |
ham5urg | I would like to test the speed of suspend-to-disk (hibernate) with a new SSD I built into a laptop. | 10:22 |
freemangordon | I am using it, on ubuntu though :) | 10:22 |
freemangordon | and yes, it is a laptop with ssd | 10:22 |
freemangordon | 1TB intel nvme SSD | 10:23 |
ham5urg | freemangordon, how fast does the laptop go to sleep/awake? | 10:23 |
freemangordon | 16GB of ram, it takes seconds to hibernate and resume | 10:23 |
ham5urg | Does it depend much onto the RAM-size? I have 64GB. | 10:23 |
freemangordon | it should depend, or rather on the amount of ram that's actually used, IIUC | 10:24 |
ham5urg | Yes | 10:24 |
ham5urg | Is it more of "30 seconds" or "3 seconds" in most of your cases? | 10:25 |
freemangordon | keep in mind I have a dedicated swap partition | 10:25 |
freemangordon | it is more of '15 seconds' to hibernate and '5 seconds' to resume | 10:25 |
freemangordon | those 5 seconds does not include machine BIOS time | 10:26 |
ham5urg | Ok, that is a bit of time but ok. | 10:26 |
freemangordon | which is 20-30 seconds until it shows all the logos etc :) | 10:26 |
freemangordon | is it? I would say this is fast | 10:26 |
freemangordon | but yeah, maybe my expectations are low | 10:27 |
ham5urg | I would prefer a dedicated partition or file for hibernation. Anytime a machine of mine used a swap partition it went into state of disfunction. | 10:30 |
ham5urg | Never used a swap partition for the last 10 years in production. | 10:30 |
onefang | swapspace FTW. It resizes a swap files as needed. | 10:51 |
GyrosGeier | I use suspend-to-disk daily | 11:13 |
GyrosGeier | here at $company, the last person to leave turns off the power for the lab | 11:13 |
GyrosGeier | so all my desktop boxes suspend to disk when the power button is pressed | 11:14 |
gnarface | last i heard there's a hard minimum on the swap size of 100% of physical ram size for hibernate ("suspend to disk") to work | 11:15 |
gnarface | ham5urg left already but 64G might indeed be a hit to boot times | 11:16 |
gnarface | though it seems like there should be some way to compress and store it faster | 11:16 |
gnarface | in theory anyway | 11:16 |
ham5urg_ | No, I'm here, but sometimes my connection broke. Don't know why | 11:17 |
GyrosGeier | it depends | 11:18 |
GyrosGeier | my TalosII has a gzip coprocessor | 11:18 |
GyrosGeier | that would probably be a speedup when writing to SSD | 11:19 |
GyrosGeier | everyone else, no | 11:19 |
ham5urg_ | Talos2 is a nice machine. | 11:20 |
GyrosGeier | indeed | 11:30 |
GyrosGeier | I don't have suspend working on that one yet though | 11:31 |
guest74 | Hi. Since some time I have a problem with some keys suddenly not being registered. | 13:00 |
guest74 | What could cause the issue, how to check it, and fix it? | 13:01 |
djph | bad keyboard? try with a new keyboard. | 13:05 |
fsmithred | turn it over and shake the bread crumbs out of it | 13:30 |
Guest56 | fshmitred it isn't because of their contamination. It is a software related problem. It happens the most often for the left mouse button. | 16:12 |
Guest56 | After a restart, it works again. | 16:12 |
brocashelm | is there a workaround to get orphaner working again? | 17:32 |
brocashelm | i tried downgrading to 1.7.33 (last version with it), but i ran into some conflicts with "which" and "tempfile" not being found | 17:33 |
blockhead | is there a way to find out when Chimaera is scheduled to become Stable? I checked forums and devuan.org and could not see any information regarding that | 17:56 |
brocashelm | blockhead: according to the devs, chimaera will be released sometime this coming month so as to not be overshadowed by bullseye | 17:59 |
golinux | It will be announced "when it's ready" | 17:59 |
brocashelm | you can already switch to chimaera if you're interested by just changing your sources.list | 17:59 |
golinux | Indeed | 18:00 |
golinux | Many have already done so | 18:00 |
brocashelm | it's stable and not that far away from ceres (unstable) in terms of package versions | 18:00 |
brocashelm | it's on kernel 5.10 | 18:00 |
* blockhead smile t "when its ready" - I totally get that :) | 18:00 | |
golinux | The hang up is mostly in writing documentation at this point | 18:00 |
brocashelm | making sure the colors match up ;) | 18:00 |
blockhead | my devuan install is already at chimaera. I had to do that for some reason that i forget :p | 18:01 |
blockhead | lol at colors :D | 18:01 |
golinux | Hopefully, that is taken care of. | 18:01 |
brocashelm | i use ceres, lol | 18:01 |
* golinux is still on jessie | 18:01 | |
brocashelm | :O | 18:01 |
* golinux does not like messing with the machine | 18:02 | |
brocashelm | it's also a good way to delay the unnecessary upstream decisions | 18:04 |
brocashelm | like with deprecating which | 18:04 |
blockhead | golinux: you mean the hardware? | 18:04 |
brocashelm | now you have to use command -v (?) | 18:04 |
* golinux is not well suited to the digital age . . . any aspect of it. | 18:05 | |
golinux | Aaaaand we're off-topic . . . | 18:05 |
blockhead | i just wanted to make sure things were all OK with chimaera, Debian switched over on the 14th and the delay made me a little bit worried | 18:06 |
brocashelm | also, keep in mind that chimaera/bullseye is also where wicd shits the bed; python2/3 problems | 18:07 |
brocashelm | so if that's your way to connect to the internet, the only other gui alternatives would be networkmanager and connman | 18:07 |
golinux | amesser said he was going to have a look at it. | 18:07 |
* golinux keeps fingers crossed | 18:08 | |
fsmithred | delay? | 18:08 |
brocashelm | yeah, i know there's been talks about patching python | 18:08 |
blockhead | i'd assumed that devuan syncronized to debian release times? | 18:08 |
brocashelm | hopefully it succeeds and it can be yet another reason to switch over | 18:08 |
fsmithred | nah. jessie was two years late, ascii was a year late and beowulf was almost as late. | 18:09 |
golinux | blockhead: No, we are always out of synce a bit. | 18:09 |
fsmithred | we are way ahead of our normal schedule this time. | 18:09 |
blockhead | ohhhhhhh, i did not know that | 18:09 |
fsmithred | probably a few weeks, not a few months | 18:09 |
brocashelm | devuan forks 200+ packages and has some exclusive packages (like the cinnabar themes for beowulf) on top of syncing with debian (except where systemd and related packages are BTFO) | 18:10 |
blockhead | good to know. i am reassured | 18:11 |
brocashelm | the forks are usually edits due to high systemd dependency | 18:11 |
brocashelm | like with network-manager and openvpn, whose packages have to be devuanized first | 18:11 |
blockhead | when I used BLFS, I figured out that most systemd dependencies are imaginary. recompile with the right ./configure options and POOF, problem goes away | 18:12 |
brocashelm | but the syncs are usually close enough in my experience (i update 2-3 times a day) | 18:12 |
blockhead | 2 to 3 times a day? wow. Recently i've been updating every one to two days and I wondered if that was too much | 18:13 |
fsmithred | n-m does not have to be devuanized. I think that's because we now have elogind. | 18:13 |
fsmithred | ok, I see the chimaera version of n-m is devuanized. | 18:14 |
fsmithred | but not beowulf | 18:14 |
blockhead | elogind ... i see a "d" on the end and get concerned :p | 18:15 |
golinux | Many have a "d" phobia | 18:16 |
fsmithred | yes, blockhead. It's a fork of systemd-logind | 18:18 |
brocashelm | for systems that don't use systemd as init | 18:18 |
mason | blockhead: seatd is a potentially more-acceptable alternative that's being batted around | 18:19 |
mason | blockhead: see: https://git.devuan.org/devuan/seatd | 18:19 |
mason | More acceptable in this case means "it's not a hunk of systemd renamed" | 18:20 |
blockhead | understood | 18:20 |
fsmithred | you can do without that stuff if you use a window manager and don't expect user to reboot/shutdown/mount | 18:21 |
blockhead | would you beleive I exit the wm, log out from the console and then log in as root to shut down? | 18:22 |
fsmithred | yeah, that's one way to do it | 18:23 |
mason | blockhead: sudo or su might speed that | 18:23 |
fsmithred | I usually set up sudo nopasswd for reboot/shutdown | 18:23 |
* Xenguy holds out a crucifix, upside down... | 18:24 | |
fsmithred | lol | 18:24 |
blockhead | ? | 18:24 |
Xenguy | :) | 18:24 |
fsmithred | because I'm insecure | 18:24 |
fsmithred | any of my users (that's me) can shut the thing down. | 18:24 |
Xenguy | I don't reboot much, so when I do, I'm in the habit to have a root virtual console always open, plus in X I have a root terminal running screen as well, in one of my workspaces | 18:26 |
Xenguy | root is always ready to do battle, just in case ; -) | 18:26 |
blockhead | :) | 18:27 |
ham5urg | When I try to start UngoogledChromium with a second Monitor attached via USBC-dock, I get this error: http://paste.debian.net/1209802/ | 18:38 |
ham5urg | If the monitor is detached, all works fine. | 18:38 |
fsmithred | ham5urg, I'm not getting that error, but that's with second monitor on vga port. | 18:57 |
fsmithred | simple fix would be to create /etc/machine-id, I think. | 18:57 |
fsmithred | or maybe symlink to /var/lib/dbus/machine-id which should be regenerated every boot. Check /etc/default/dbus for that. | 18:59 |
fsmithred | I think they're both 32 characters | 18:59 |
Tenkawa | thats a hardware acceleration error | 19:06 |
Tenkawa | it is trying to treat that doc like a hardware accel capable device (which its not) | 19:06 |
Tenkawa | er dock | 19:07 |
Tenkawa | thats at least what I surmise from my investigation | 19:10 |
fsmithred | that makes sense. I don't get the machine-id error when I tried it with vga | 19:12 |
fsmithred | and I don't have usbc dock to test | 19:12 |
ham5urg | I'm trying to repeat the error. Sometimes it can be repeated, sometimes not. | 19:12 |
Tenkawa | ham5urg: forget the error.. does the monitor work? | 19:13 |
ham5urg | Tenkawa: Yes. | 19:14 |
Tenkawa | then whats the problem? | 19:14 |
Tenkawa | those messages are inconsequential | 19:15 |
ham5urg | Sometimes - and I don't know why - I get the problem with ungoogled-chromium | 19:16 |
ham5urg | But it's working now for quite a while | 19:16 |
Tenkawa | what "problem" though is what I am asking. | 19:16 |
ham5urg | Ungoogled-chromium won't start if this happens. | 19:17 |
Tenkawa | ok.. so this is inconsistent. it happens sometimes but not always? | 19:18 |
ham5urg | Yes | 19:18 |
Tenkawa | with it being on a dock connected video that is going to be very hard to troubleshoot | 19:18 |
ham5urg | Yes | 19:19 |
ham5urg | I works now, today it did not worked for some hours, now it works. Will see what will happen in coming days. | 19:19 |
Tenkawa | when it does start having problems again save the /var/log/messages syslog and kern.log files on that day | 19:20 |
ham5urg | Ok | 19:21 |
Tenkawa | should be some indication there what possibly triggered the event | 19:21 |
Tenkawa | it could be a term called a "race event" where its just pure timing and things overlapped | 19:21 |
Tenkawa | 2 things fought for the same spot at the same time and one won | 19:22 |
Tenkawa | hard to say without logs and/or seeing it when it happened | 19:22 |
Tenkawa | cheers for now all.. back to preparing for next incoming storm here | 19:24 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!