[00:47] that'sgoing to break everything [00:47] snapis the debian snap installer folder [00:48] stop messing with things until you get a bac kupapp [00:49] sudo apt install how to install Back in Time on ubuntu [00:50] it'sin discovertoo [00:51] kdeorgnome [00:52] launch discover [00:52] orsoftware [00:53] depending on your distro [00:53] when you get it configured run a back up then you can break things [00:54] I got arch and broke it 15 times during installation [00:54] it was fun fixing stuff [00:54] best way to learn [00:54] in a safe VM environment [00:58] rofl [00:59] I work with stuff everyday that can break your stuff where it can't be fixed casein point Tensorflow [01:00] and nobody in #tensorflowwill talk about it [01:13] we should have speech recognition in about a week. I got it working before tensorflow locked my models with a checkpoint save instead of the way I do it. Just cleaning up behind that event [01:14] it definitely works === chris14_ is now known as chris14 [01:55] Query: joke [01:55] 1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step [01:55] [{'intent': 'joke', 'probability': '0.9999943'}] [01:55] JENNY: It takes a lot of balls to golf the way I do. [01:57] whoever thought AIs can't understand human emotions was totally wrong [02:04] jason: As a sidenote, this channel is usually used only for technical support for Kubuntu. There is #ubuntu-offtopic for general chat though. === jason is now known as jason_bourne === freeone3000 is now known as freeone3001 === guiverc2 is now known as guiverc [03:20] Hi! I'm trying to get Ubuntu to boot using a wireless USB dongle. I have the drivers installed and everything works *after* boot (`route -n` is showing all routes go through the wlx iface) but during boot, systemd seems to need a network before networkmanager comes up [03:22] `systemd-analyze critical-chain` shows a simple tree, `graphical -> multi-user -> ssh -> network -> NetworkManager -> dbus -> sockets -> snapd -> sysinit -> snapd.apparmor -> apparmor -> local-fs -> microk8s -> multipathd -> systemd-remount-fs -> systemd-journald -> system -> .slice`. This is ... a bit odd, since I would think microk8s would depend [03:22] on network and local-fs instead of the other way around [03:23] What's the best way to restructure this so that I have wifi early enough for my other boot units to recognize it? [04:28] E: Unable to locate package [04:30] E: Unable to locate package error [06:40] hola [06:41] hola [06:41] 33 [06:41] 33 === EriC^^_ is now known as EriC^^ [06:50] 33?? [06:52] can we help you cole_ ? [06:53] I'm trying to upgrade from lunar to mantic but do-release-upgrade refuses because it incorrectly thinks I have too little space available in /boot. If I pre-installed the latest kernel from the mantic repo, would it skip that check? [06:53] thats not reccomended anymore dabbler_ lunar is eol [06:54] hola [06:54] holaaaa [06:54] zorra [06:54] 33 [06:54] !es [06:54] En la mayoría de los canales de Ubuntu, se habla sólo en inglés. Si busca ayuda en español entre al canal #ubuntu-es; escriba "/join #ubuntu-es" (sin comillas) y presione intro. [06:54] dabbler_: sounds odd, what is the space in /boot? [06:54] could you pastebin "df -h" ? [06:55] lotuspsychje: yes, it's EOL. hence my wanting to upgrade out of it [06:55] dabbler_: upgrading from an eol release is risky due !usn security issues, clean install would be wiser [06:55] if you pastebin the error it's giving you it might give a clue as to if the idea would work [06:56] EriC^^: It is odd. it's a bug seemingly specific to a non-default btrfs option used in my /boot. It reports half the actual amount of space. [06:57] oh [06:59] EriC^^: here's the error: https://pasteboard.co/c0c00d891bNJ.png [07:01] the relevant line of `df -h`: /dev/sda2 256M 72M 178M 29% /boot [07:02] actual numbers are double (except the %, obvs) [07:04] lotuspsychje: I appreciate the warning, but it's a risk I'll have to take [07:04] its your system dabbler_ : ) [07:07] dabbler_: is it easy to expand the space? [07:07] I wish :( [07:08] how about reversing the non-default option? or some workaround so it reports correctly? [07:10] unforunately it's an option that can only be set when the fs is created. I think it's mixed_groups [07:11] without that option there's too much overhead space wasted for multiple kernels to fit in the 512 MB fs [07:12] i see [07:13] can you think of any clever ways of remounting it that would make it appear larger without interfering with the upgrade process? [07:15] maybe a union mount? [07:16] i wonder what would happen if you unmount it [07:17] just momentarily when it's checking the space or for the whole upgrade process? [07:17] just when it's checking [07:18] but then i wonder if apt would fail later on the kernel install [07:18] I'm happy to try it, but I'm not sure how to get the timing right [07:19] It seems cleaner to first try to pre-install the right kernel beside my existing one and see if it skips the check [07:19] maybe try first your idea to install the kernel manually and see how that goes, who knows it might just not complain a the start [07:23] EriC^^: do I want the latest from mantic or mantic-updates? [07:24] not sure which it tries to install [07:24] !info linux-image-generic mantic [07:24] linux-image-generic (6.5.0.28.28, mantic): Generic Linux kernel image. In component main, is optional. Built by linux-meta. Size 10 kB / 16 kB. (Only available for amd64, armhf, arm64, ppc64el, s390x.) [07:24] you want the above dabbler_ ^ [07:24] thanks [07:26] that's weird. https://packages.ubuntu.com/mantic/kernel/linux-image-generic shows 6.5.0.27.27 [07:27] in mantic-updates it's .28.28 [07:28] I guess !info combines the two then. I'll add both to my sources.list [07:31] ok [07:33] it wants to upgrade linux-libc-dev :-/ [07:35] salve [07:35] EriC^^: would that cause trouble given that I'm running a 6.2 kernel right now? [07:38] blender canal ??? [07:40] spiderjoker: https://developer.blender.org/docs/handbook/communication/chat/ ? [07:41] thanks [08:02] dabbler_: sorry back, no idea [08:06] I have a question regarding packaging of .deb. Will the directories inside my package be created automatically if they do not exists? Or do I need to specifiy shell commands in DEBIAN/preinst ? [08:10] nomicon: maybe the #ubuntu-devel channel might be able to help better with that === pushkarnk1 is now known as pushkarnk [09:48] wtheck is going on with Ubuntu's apt repositories, im getting timeout on apt-update on my github actions runners -> Could not connect to archive.ubuntu.com:80 (185.125.190.39), connection timed out [09:49] i've rerun the thing 5times already [09:53] Jesus: https://status.canonical.com/ no wonder.. ppa.launchpad.net Major Outage, next time i should probably subscriibe to this page's rss feed [10:01] vincejv: if you like, we got an rss stream of canonicals status in #techrss [10:02] nice thanks lotuspsychje appreciate it, don't know about this [10:13] When using ZFS, if the zpool is set to ashift=0 then what is the default value? === root is now known as Guest6369 [10:28] Hm. That page has some translation errors [10:36] apparently some third party thing, so not translatable via launchpad. They probably just run it through google translate === zer0bitz_ is now known as zer0bitz [10:44] we're seeing issues downloading packages from both security.ubuntu.com and archive.ubuntu.com [10:45] the initial SYN-ACK response comes quickly, but once the HTTP GET is sent things just stall from the archive side. [10:48] eg. downloading http://security.ubuntu.com/ubuntu/pool/main/g/glibc/libc-bin_2.31-0ubuntu9.15_amd64.deb from 91.189.91.83:80 [11:22] guys please check this bug and if you are affected, please join it so it's get more people and eventually more attention by the developers: https://bugs.launchpad.net/ubuntu/+source/gnome-shell-extension-appindicator/+bug/2039543 [11:22] -ubottu:#ubuntu- Launchpad bug 2039543 in gnome-shell-extension-appindicator (Ubuntu) "Miss placed Hexchat icon when switching workspaces" [Undecided, Confirmed] [11:22] it happens to different apps, not just Hexchat btw [11:23] we're no longer seeing issues reaching the archives. [11:23] lang, I confirm it's still not fixed until today in 24.04 with the latest update [11:24] ice9: this was wrt. downloading packages from (archives|security).ubuntu.com [12:14] Hi, I've got a PC they want me to try and fix in my internship, I'm at home right now and don't have access to it but I just wanted to ask for some advice about the issue. The PC shows in /etc/os-release that it has jammy (22.04 LTS) on it currently and I found some old eoan (19.10) repositories which must have been left over from an upgrade in /etc/apt/sources.list which I removed and kept only [12:14] jammy repositories. Apparently after an upgrade it started getting a kernel panic on boot https://x0.at/2y2L.jpg from that I'm guessing the root cause is that something went wrong during the upgrade involving glibc. I mounted the root partition and chrooted in to it, checked the installed glibc version and tried to upgrade again. I don't have the output of these commands (forgot to take pics before [12:14] leaving) but basically, from what I recall, the installed glibc was 2.30 and when trying to upgrade or install anything using `apt` it complained about glibc, I think it said it was too old and that the packages required a newer version. I understand glibc is kind of annoying/hard to install manually to some other version system wide, I've never had to deal with this, I've been using debian and the [12:14] updates have been smooth so any advice on what to try? [12:14] I know we could just get the data we need off from the root partition and reinstall but I wanted to try resolving this. [12:22] I think if I compiled, or more likely, cross-compiled glibc since that system will probably be missing some build dependencies and then try to install that glibc into /lib/x86_64-linux-gnu it would be able to possibly boot but that's probably not gonna go smoothly and apt doesn't fix the problem with `apt` [12:22] and that doesn't fix* [12:26] Xeroine, http://security.ubuntu.com/ubuntu/pool/main/g/glibc/ ; livecd -> chroot -> download required pkgs; install them [12:27] oh right I can just download the package [12:27] thanks, I'll try that [12:27] ok [12:56] i was looking at something on an forum.opensuse post that eluded to the existance of # systemctl status rtkit-daemon [12:57] the solution to the problem required modifying /etc/security/limits.conf and I assume that those changes will require a restart in the real-time scheduling daemon. [12:57] is there an analagous daemon in ubuntu and if so, how to restart it and/or make the changes to `limits.conf` active? [12:57] https://forums.opensuse.org/t/help-please-music-player-daemon-and-real-time-scheduling-alsa-audio-output/153962 [12:58] it should be preinstalled ... the package is called rtkit [12:59] (but afaik it is only used by pulseaudio, newer ubuntu uses pipewire, not sure that actually needs that daemon still) [12:59] ice9: that bug already got a lot of affects, see the bug daniel refers toward contains the upstream bug [13:03] ice9: did you file bug #2062538 ? [13:03] -ubottu:#ubuntu- Bug 2062538 in gnome-shell-extension-desktop-icons-ng (Ubuntu) "Clicking the top bar shows desktop icons above app window" [Undecided, New] https://launchpad.net/bugs/2062538 [13:12] Hi all [14:07] hi [14:19] nice [14:19] ok [14:19] ko [14:37] ogra_: i'm using pulse for things. [14:39] ogra_: but i don't think i'm using it on that machine per se... mpd's got it's own pulse output module and that's being set to a sink on my laptop over TCP so I don't think the local pipe/pulse server gets involved. [14:39] i was just looking to see if there was a way to address the error i was seeing in the mpd output. [14:39] well, then rtkit should also only apply to your laptop [14:40] in terms of any ALSA effects such as the OP was talking about in the post I assume, however, MPD is still throwing errors on the server side [14:40] event: RTIOThread could not get realtime scheduling, continuing anyway: sched_setscheduler failed: Operation not permitted [14:41] if you changed limits.conf i think it only applies after a reboot ... [14:41] ok [14:41] then i'll look after i reboot [14:41] I still keep getting 14.46 E: Failed to fetch http://archive.ubuntu.com/ubuntu/dists/jammy-updates/main/binary-amd64/Packages.gz File has unexpected size (2018947 != 2018948). Mirror sync in progress? [IP: 91.189.91.83 80] [14:41] even tho ubuntu canonical health says everything's operational?? [14:41] thank you [14:42] it's happening on my github runners when doing apt-update [14:43] vincejv, perhaps the people in #launchpad knwo more [14:43] do i need to get a ubuntu pro subscription to get this thing working? lol [14:43] indeed not [15:11] how to install Davinci Resolve on Ubuntu 24.04 since some packages names have changed: [15:11] selecting 'libapr1t64' instead of 'libapr1' [15:11] selecting 'libaprutil1t64' instead of 'libaprutil1' [15:12] and there is no package with name 'libasound2' [15:14] ice9: Davinci Resolve probably won't support Ubuntu 24.04 until after its released. Right now it's still in beta. Perhaps fall back to 22.04 and then upgrade once 24.04.1 is out? [15:19] ice9, Also, that's a question to ask Davinci Resolve support channels [15:19] is there any reason i couldn't send a command from THE nfs/sshfs server to THE nfs/sshfs client to tell it to umount -l everything as part of a shutdown script so that my guake stops hanging everytime i reboot my server? [15:20] well... `sudo umount -l -a -t nfs,sshfs` [15:28] I didn't know a server could issue such commands over nfs. Assuming it does, sending the `-l` is lazy unmount, which says unmount when the resource isn't busy. If you have an active application leveraging something from that resource, it's not going to unmount [15:28] no, i sent it over ssh [15:28] that makes more sense [15:28] it worked insofar as i mounted something, issued the command, and guake didn't crap out [15:29] nothing was using the resource [15:29] I don't know why guake would care about a mount. seems rather odd [15:29] think i need to add -f to that list then? === alucardromero3 is now known as alucardromero [15:29] i don't know, don't care, just don't want my it to happen anymore!! [15:30] this seems the path of least resistance, plus, the server's going down anyway, might as well try to unmount them becasue i never remember to do it. [15:30] right now I'd try removing the `-l` from the options first... if it still hangs, then maybe try with `-f` [15:30] still getting hangs, then it's an issue with guake [15:30] well, i think i went the other way... i started using it without anything and added -l [15:31] Kind of suspect it's an issue within guake [15:32] what i've been doing: forget to unmount, reboot, guake locks up, i have to launch a terminal, maybe can see what i'm doing, if not, have to drop to a TTY, then tell it to umount -l... [15:32] yeah, well, since i can't fix guake, again, path of least resistance. [15:34] other than part of your remote command doing a force kill on guake processes, and the restarting it... guake has lots of strange behaviors and hangs [15:34] I stopped using it when it would hang when I issued a reboot command to the system [15:35] yeah, i think i looked for a replacement for 30 minutes but it does most everything i want so i've satisficed on it. [15:35] an issue for another time. [15:35] sorry couldn't be of more help :( [15:35] but i can put an ssh command in a systemd shutdown script? [15:36] that;s really what i'm after...and how to ignore errors... [15:37] almost wonder if something like ansible might be something better suited to helping you manage these things [15:38] https://docs.ansible.com/ansible/latest/collections/ansible/posix/mount_module.html [15:39] oh... [15:39] his is because services are stopped on shutdown in pretty much the opposite order as they were started – if your unit needs networking to be available, it should have After=network.target, so that it gets started after network, but stopped before network. [15:40] that would be an important thing [15:42] nework.target is wrong, this is just the networking stack (mainly loopback) but wont tell anything about devices being online [15:44] (IIRC there is an "network-online.target" ...) [15:45] https://superuser.com/questions/1701476/how-can-i-have-my-service-be-the-first-service-to-begin-stopping-on-system-reboo [15:46] so there's another target besides network.target? [15:48] yes [15:48] what about just using runlevel6.target [15:49] because network.target only means all protocols are loaded and your loopback is up (so i.e. dbus can use it etc) [15:50] there is a network-online.target you should be able to use that will only fire once there is a working connection [15:50] i understand what you're getting at, but if i run the ssh command at the highest runlevel, then it should be executed first as the shutdown goes down the runlevles? [15:50] systemd hs never and doesnt use runlevels ... you should use shutdown.target for things you want at shutdown [15:56] never? [15:56] $ ll /usr/lib/systemd/system/runlevel* [15:56] lrwxrwxrwx - root 5 Apr 01:38 /usr/lib/systemd/system/runlevel6.target -> reboot.target [15:57] anyway, i found what you were talking about with the network-online.target here: [15:57] https://askubuntu.com/questions/1449198/running-a-script-at-shutdown-before-networking-goes-down [15:57] and i can patch this together, thanks for the tip. [15:58] 👍 === sotaoverride is now known as Guest2477 [17:27] k1tten === pah is now known as pa === Biapy4 is now known as Biapy === guiverc2 is now known as guiverc [23:01] interesting