=== chris14_ is now known as chris14 === asarandi2 is now known as asarandi === Vercas1 is now known as Vercas [02:11] so [02:11] yup [02:19] did some ubuntu updating. 22.04.2 updated all packages and rebooted. seems like all network-type requests are defaulting to ipv6? `ping google.com` shows PING google.com(lga34s34-in-x0e.1e100.net (2607:f8b0:4006:80b::200e)) 56 data bytes and never gets a response, but `ping 8.8.8.8` works fine. any updates what could've changed? [02:19] not just ping, everything, but using that to test [02:20] thought it was dns at first, but seems like dns is up and running and I get valid IPs back, or, at least, `nslookup google.com` (and other domains) looks the same as it does on my OSX-laptop [02:23] alright boys i finally got steam to show [02:24] got rid of everything and just downloaded the deb from the steam website [02:24] even had to delete the .steam folder [02:25] new steam UI is way better for proton games [02:25] crankharder: is there any chance you put something like disable_ipv6=1 into your grub.cfg and it was overwritten when kernels were installed? [02:26] goddard: :D // https://arstechnica.com/gaming/2023/06/valve-gives-steam-its-biggest-update-and-redesign-in-years/ . [02:27] yeah its nice [02:27] like you can now just select a "non-steam" game from the UI and easily launch and install it [02:28] i was getting the exact same FPS with Diablo 4 as i was on Windows [02:28] sarnold: unlikely. looking... [02:28] kicker is Linux is using way less memory [02:28] so my 16 gigs of memory was getting all eaten up by windows and stuff [02:28] sarnold: well, i guess it's not there now. but i really doubt I edited anything boot-related on this system [02:30] crankharder: well, dang, it was always a bit of a long shot [02:33] crankharder: from within a networkmanager connection editor, there's an ipv6 settings tab; you could change the 'method' from automatic to disabled; that'd probably do the trick [02:33] ubuntu-server, no ui [02:34] aha [02:36] crankharder: /etc/gai.conf; look for "precedence ::ffff:0:0/96"; also $ man gai.conf and familiarize yourself with RFC 3484 [02:47] crankharder: I thought about suggesting what xar suggested, but wasn't sure if it'd be better to just not get ipv6 on the interfaces at all; try dhcp6: false and accept-ra: false from https://netplan.readthedocs.io/en/stable/netplan-yaml/ as needed === JanC_ is now known as JanC === realivanjx6 is now known as realivanjx === ord is now known as quem [07:18] morning [07:41] Hi all. [07:41] I've got "odbcinst: symbol lookup error: odbcinst: undefined symbol: iniOpen" error installing msodbcsql17. [07:41] Tried many different packages but didn't work [07:41] Different versions, I mean [07:42] unixodbc is installed [07:42] And it works === realivanjx5 is now known as realivanjx [09:04] Is the live media just gonna refuse to boot if it can't mount the system drive? :/ [09:04] Oh no nevermind just took a while and it tried to run an fsck for no good reason [09:22] So now I think ubiquity is just kind of completely stuck "Sahtop [09:23] Saving installed packages and top isn't really showing any activity, does it need to run as root even though it doesn't ask for it? [09:24] Is there a better way to install Ubuntu to an existing partition and try to get it to boot semi-manually? [09:24] Lartza: maybe start by telling the room what exactly you are trying to achieve. what you did so far and what are the problems/errors you see? [09:27] I'm trying to install Ubuntu into an existing GRUB system, LVM inside LUKS, started ubiquity with ubiquity -b and didn't tell it to format the volume (it was already ext4) and now it's kind of just sitting. I mostly see errors as "Home directory not accessible" [09:27] APT is aparently locked though I can't install iotop... it could be super slow? [09:28] the installer shpuld be able to create an encrypted luks system during the install process [09:28] why this complicated way? [09:28] Because the system already has Arch [09:28] Dual boot [09:29] i dont think that is a supported combination [09:29] Well no, of course not realistically, but I don't know why the installer wouldn't just install regardless [09:29] because you use it in a away it was not tested for [09:30] so errors and bugs can happen [09:30] I mean I'm telling it to install to a partition that it sees, and it's not [09:30] But yes, in a way sure [09:30] I'll test what happens if I do tell it to format the volume [09:31] Is there a supported dual boot scenario? Install Ubuntu first then shrink the LUKS and create another LUKS for Arch? [09:33] the installer only supports encryption as a single boot system [09:34] there is the manual partitioning that you tried. but thats not tested of course because it includes endless test cases [09:34] Yeah so basically the answer is "yes", I can't really get the installer to do anything crazy and will have to manually install Arch next to Ubuntu :) [09:34] Well yes to the latter question, not necessarily that it's still "supported" === ord is now known as quem [09:35] you can spawn a shell at any time and try to fix the steps yourself [09:36] Yeah if I can figure out what is going wrong with ubiquity, I have opened the luks so it does see the partitions just something goes wonky [09:37] this is just another unsupported way but: i do my jammy installs via debootstrap atm :) [09:38] I mean it wouldn't be the first time I did something like that, I once installed Arch from inside a running Debian VPS :P [09:38] And Arch install likely is already somewhat similar to debootstrap, haven't looked === soix is now known as nocturnally [09:50] Ubiquity actually claims it installed, now I just need to figure out how to boot it or look into alternative methods :O [11:32] heh [11:32] solved my nvidia-driver-530 issues under secure boot on lunar [11:34] turns out that if you use the Software & Updates preference panel's Additional Drivers pane to install the 530 drivers, it properly recognizes that the installed linux-signatures-nvidia-6.2.0-23-generic package has the 530 kernel modules sigs [11:35] Whereas if you attempt to install it manually with 'sudo apt-get install nvidia-driver-530' that isn't properly recognized and it wants to install dkms to rebuild the nvidia kernel modules and manually sign them [11:36] Actually, I think I may have seen this before [11:36] where, under secure boot, upgrading between nvidia releases on the command line doesn't always honor the presence of the sigs for the nvidia kernel mods [11:37] super-cool howarth_ ! Thanks! [12:06] Weird that it made a difference [12:06] One would have thought that the software and update panel would just effectively do a simple 'sudo apt-get install nvidia-driver-530' [12:10] Install: libxnvctrl0:amd64 (510.47.03-0ubuntu1, automatic), libnvidia-common-530:amd64 (530.41.03-0ubuntu2, automatic), libnvidia-fbc1-530:amd64 (530.41.03-0ubuntu2, automatic), libnvidia-gl-530:amd64 (530.41.03-0ubuntu2, automatic), libnvidia-extra-530:amd64 (530.41.03-0ubuntu2, automatic), nvidia-compute-utils-530:amd64 (530.41.03-0ubuntu2, automatic), nvidia-driver-530:amd64 (530.41.03-0ubuntu2), [12:10] linux-modules-nvidia-530-6.2.0-23-generic:amd64 (6.2.0-23.23, automatic), libnvidia-encode-530:amd64 (530.41.03-0ubuntu2, automatic), linux-modules-nvidia-530-generic-hwe-22.04:amd64 (6.2.0-23.23), nvidia-utils-530:amd64 (530.41.03-0ubuntu2, automatic), xserver-xorg-video-nvidia-530:amd64 (530.41.03-0ubuntu2, automatic), linux-objects-nvidia-530-6.2.0-23-generic:amd64 (6.2.0-23.23, automatic), pkgconf:amd64 [12:10] (1.8.1-1ubuntu2, automatic), libnvidia-decode-530:amd64 (530.41.03-0ubuntu2, automatic), nvidia-kernel-common-530:amd64 (530.41.03-0ubuntu2, automatic), nvidia-prime:amd64 (0.8.17.1, automatic), screen-resolution-extra:amd64 (0.18.3, automatic), nvidia-settings:amd64 (510.47.03-0ubuntu1, automatic), pkg-config:amd64 (1.8.1-1ubuntu2, automatic), libnvidia-cfg1-530:amd64 (530.41.03-0ubuntu2, automatic), [12:10] pkgconf-bin:amd64 (1.8.1-1ubuntu2, automatic), nvidia-kernel-source-530:amd64 (530.41.03-0ubuntu2, automatic), libpkgconf3:amd64 (1.8.1-1ubuntu2, automatic), libnvidia-compute-530:amd64 (530.41.03-0ubuntu2, automatic) [12:11] history.log in apt shows that [12:11] my guess is that the command line installation might not be automatically installing linux-modules-nvidia-530-generic-hwe-22.04:amd64 [12:11] hence the attempt to use dkms to rebuild the kernel modules [12:12] when 'sudo apt-get install nvidia-driver-530' is used instead [12:13] which might make sense as I don't think the hwe stuff is as carefully integrated [12:13] dependency wise [12:15] Actually, it is odd that they are still on 530 [12:15] as I think 535 become a stable release from that already and 530 got left as beta [12:15] https://www.phoronix.com/forums/forum/linux-graphics-x-org-drivers/nvidia-linux/1392385-nvidia-535-54-03-linux-driver-released-with-vulkan-updates-dma-buf-v4-wayland-protocol [12:16] they haven't added 535 to mantic yet [12:16] but it is on https://launchpad.net/~graphics-drivers/+archive/ubuntu/ppa [12:24] hi there people. I have a theme applied on ubuntu 22.04. can this not be applied system wide. for example, the theme is not applied to firefox and software centre [12:28] vershan: maybe this is helpful. im not sure it is the latest information but you can continue from there: https://www.omgubuntu.co.uk/2020/06/change-snap-app-theme [12:29] @ravage, will have a look thanks [13:12] Is there a command I can use to automatically and programatically remove all old kernels? [13:13] Or do I have to make a whole bespoke script to detect which version I'm running, which ones are installed, and then remove the old ones? [13:17] bahamat: on a default installatipn "sudo apt autoremove" will keep the active kernel and one version before that [13:17] ravage: Nope, that doesn't work. [13:18] then you manually installed kernel packages without the meta package that takes care of removing the old ones [13:19] I never explicitly installed a specific kernel version. [13:28] bahamat: pastebin dpkg --list | grep linux-image please [13:29] lotuspsychje: linux-image-5.4.0-144-generic and linux-image-5.4.0-152-generic, but that's all [13:29] Here's what I've done so far: [13:30] install system, dist-upgrade, remove linux-headers-generic, linux-headers-*, linux-firmware [13:30] bahamat: you only have 2 kernels installed and you want to uninstall 1 of them? The default in ubuntu is to keep the last kernel as a backup [13:30] bahamat: what release of ubuntu is this? [13:30] 20.04 [13:31] why did you just install 20.04 over 22.04? [13:31] also, why are you removing the headers and linux-firmware? [13:31] Is it not still LTS? [13:31] Because I'm building a cloud image and 99% of deployments won't use either of them. [13:32] All the hardware is virtual, so it doesn't need any firmware [13:32] bahamat: 20.04 is still LTS, but why install it over 22.04 which will give you longer support and more modern kernels and software [13:32] and almost nobody will be doing kernel development or compiling anything that needs to link against the kernel. The few who do can install it manually. [13:32] cause it has hitler's birthday right in the version number [13:32] please dont TomyLobo [13:33] TomyLobo: please take that mess elsewhere [13:33] oof, what a reaction to a joke [13:34] Because I'm creating cloud images and my users need all available LTS versions. [13:35] yeah that's a better reason than mine :) [13:35] If ubuntu-18 was still LTS, I'd create one of those too. [13:36] what are you using to create your cloud images btw? [13:37] last time I used packer and I was wondering if there are other tools for the same purpose [13:45] Well, since autoremove won't work, bespoke script it is. [13:46] bahamat: autoremove works. By default it keeps the last kernel, which is the exact state you're in [13:46] by last you mean latest or the last one that booted? [13:47] TomyLobo: the previously installed kernel [13:47] i agree with leftyfb , ubuntu keeps the last few kernels and you got only 2? [13:48] Any way to modify that behavior? [13:49] bahamat: whats your endgoal exactly? only keep 1 kernel active? [13:49] From the experience of my users, there's no "previous" kernel. They get an image with a currently working kernel. That's the "first" one. They won't need a previous one, and if they do for some reason they can install it manually. [13:49] thats not how the ubuntu system works, and pretty unsafe situation if you dont have previous kernels to fallback to [13:50] And purging the old kernel saves 300MB in the image size. That will reduce the image size by about 25%, so it's worth it. [13:50] lotuspsychje, read up, bahamat is making cloud images. the fallback is not needed [13:50] and it's invalid anyway, since the machine is fresh there is no "known working" kernel to fall back to [13:51] lotuspsychje: It's a base image for a cloud environment. If the installed kernel is broken I'll fix it before publishing the image. There's no need for a fallback kernel at all. [13:51] bahamat: so just write a script that deletes the non-running kernel [13:51] bahamat: also, install the hwe kernel [13:51] leftyfb: Yeah, that's what I ended up with. [13:51] you'd need to reboot in the middle, though, so you're not running the kernel you're about to throw out :) [13:52] just to be safe [13:52] might be optional [13:52] probably should make this your last stept [13:52] -t [14:04] What's the difference between hwe and hwe-*-edge? [14:06] a pre-release of the coming hwe kernel [14:06] currently there should be no difference [14:06] ravage: There's not, which was what confused me. [14:07] in some cases it can be handy to test out HWE earlier [14:08] specialy if you need a specific HWE kernel number/series [14:08] Yeah. But in my case, I won't need that. [14:10] Looks like this will do the trick for me: apt-get -y purge $(dpkg --get-selections | grep ^linux-.*-5.* | grep -v $(uname -r) | awk '{print $1}') [14:11] Does 22.04 have an hwe kernel yet? [14:12] yes [14:12] currently 5.19 [15:16] why not grep -E '^linux-.*-[0-9]+.*' === Fisher244594 is now known as Fisher24459 === NotEickmeyer is now known as Eickmeyer [16:22] how do I change the resolution of the login screen? [16:29] zeratul: xorg or wayland? and you want it to be lower or higher? [16:44] alkisg, xorg and I want to change both the resolution and the aspect ratio [16:48] from 1024x768 to 1920x1080 === tristan123 is now known as tristan_ [17:02] zeratul: I think that's controlled by gdm. At one point it was configured by a gdmrc file. [17:23] zeratul: what's the output of: xrandr | nc termbin.com 9999 [17:25] hi all [17:26] i am having a problem connecting to webpages just wondered if anyone can help [17:26] What kind of problem? [17:27] i click on a webpage and it just buffs and i dont get the page that with wifi and ethernet [17:27] So you can't access any internet page at all, with any browser? [17:28] no [17:29] Is that pc connected over ethernet now? Are you on that pc, or in another one? [17:31] after awhile it comes up with try again. as it times out [17:32] What's the output of this command? host www.google.com [17:32] (run that in a terminal,in the affected pc) [17:38] Hello, if I set up cron jobs with `sudo crontab -e`, they will be executed by root, right? [17:39] Because I have a cron job that does `git pull` at some point. I get the error "fatal: detected dubious ownership in repository at", whereas if I `cd` that directory then `git pull`, it doesn't complain [17:39] Sorry, if I `sudo git pull` it doesn't complain, if I just `git pull` it complains [17:41] Bardon: yes, but why not do the opposite, run `crontab -e` and make everything owned by the user, not by root [17:44] Because I wouldn't have the permissions to do what I want. (This script pulls from some git repo, that corresponds to a static website that I host, builds it, then moves it to the folder where nginx loads it) [17:44] alkisg: But if jobs created with `sudo crontab -e` are run by root, then why does it not complain when I `sudo su` then `git pull`? [17:45] justache, where is the gdmrc file located? [17:45] (It wasn't `sudo git pull`, but `sudo su` then `git pull`, my bad, even though I'm not sure it makes a difference) [17:46] Bardon: because then everything is owned by root. The problem is mixing root with non-root [17:48] I'm sorry I don't understand. At what point am I missing root and non-root? [17:49] "fatal: detected dubious ownership in repository at" sounds to me like you created the git repository as user, not root [17:49] Why do you need to run git as root,why don't you want to run everything as the user? [17:51] correct, I created it as a user, not root [17:52] What I don't understand is when I'm logged in with that user, I sudo su, then git pull and it doesn't complain! [17:53] And I don't run everything as the user because the user can't move the generated website to the folder that nginx expects [17:53] and I guess I can't use sudo for only one command in this script because it'll ask for a password, which I can't provide safely since it's run by cron [17:54] Usually nginx runs as www-data. You can run a cronjob and clone from git as that user, www-data. Not the "bardon" user. [17:54] I suppose I could give write permission to that user, but I thought it was safer to only allow root to write in that folder [17:54] alkisg: Right, that would be a solution :) [17:55] But still, this bugs me, why does the script, when run by sudo crontab gives a warning when gut pull from the root account doesn't! [17:55] Meh, matrix receives the messages after a while, I guess the bug at https://github.com/matrix-org/matrix-appservice-irc/issues/1724 is still unresolved ... :/ [17:55] -ubottu:#ubuntu- Issue 1724 in matrix-org/matrix-appservice-irc "Silent message drop since the last maintenance" [Open] [18:00] Bardon: to discover what's different, at the top of your script, run: ( env; pwd ) >/tmp/script-$$.log [18:00] You'll see that some environment variable is different, e.g. HOME not set or different cwd [18:01] Hi [18:04] alkisg: Ah ok, I see :) [18:06] alkisg: Καλησπέρα! [18:10] I got lost in grep. I have a file and more less it looks like this: https://pastebin.com/tjub8K3B Basically it contains answers to a test. It's usually a single answer (one letter) like on the first line in the link. But sometimes it shows answers to multiple choice question like in the second line. Here's what I did: cat second_step.txt |awk {'print $2" "$3" "$4" "$5'}. I only need those letters. Using awk I [18:10] eliminated line numbers and now I wanted to use dots to distinguish single from multiple (separated by a coma). I'm looking for grep (or anything) that will give me this: cat second_step.txt |awk {'print $2" "$3" "$4" "$5'} |grep everything until it hits the dot. [18:11] Sorry for the off topic, but I tried so many things.... [18:16] I mean if someone can come up with the idea to do it leaving the first row and omiting the first dot(the one after line number) that's even better, but not that important. [18:17] socket: send a bigger example e.g. 10 lines of your initial and final files [18:22] alkisg, I had to edit it out. But those lines all look the same. The might have more text (hence awk up to fith comlumn), but it's everything up to the first dot in each lines that I need and the rest needs to disappear. [18:23] | cut -d. -f1 [18:23] but i think you mean the second dot [18:24] there's also #bash btw, and similar channels for other shells [18:26] tomreyn, exactly that!! [18:26] what's f1 here? [18:27] I mean if you have any idea how to do this output: 29. A, B or 29(no dot) A, B [18:27] that's even more interesting. [18:28] It's `29 A, B ` ? [18:30] socket: -f1 selects the first "field". the "field" ends where the delimiter character (given by '-d') is found. see the man page, it's simple enough. [18:30] 29 A, B [18:31] oh that first dot is optional, then cut isn't useful there. [18:31] printf "28. A, B. Text.\n29 C. More text.\n" | sed 's/^[0-9][0-9][.[:space:]]([^.])./\1/' [18:31] That gives "A, B" for the first line, and "C" for the second, not sure if that's what you want [18:32] printf "28. A, B. Text.\n29 C. More text.\n" | sed 's/^[0-9][0-9]*[.[:space:]]*\([^.]*\).*/\1/' [18:32] "not sure what you want" is certainly the main issue here [18:32] Resending as matrix probably borked it :) [18:35] I got what I wanted, I appreciate it, guys. The rest of it is at this point academic (for me) and I asked out of curiosity. The original line starts with a number and the dot follows that number. I "awked it out" because I didn't know how to omit the first dot to use the second one as a delimiter where I finish the line. [18:38] for an old Intel MacBook Air, should I install Ubuntu in EFI or old Bios/MBR mode? The old method should not require an EFI partition but does it also has some drawbacks? [18:38] Ubuntu would be the only OS on it, so no dual boot with macOS or Windows [18:39] whatever works [18:39] it doesnt matter one way or the other [18:40] there are no implications for signed kernels etc.? [18:40] afaik this device does not have secure boot :) [18:41] well you cant verify signatures on BIOS [18:41] well if it odnst have secure boot [18:41] then its irrelvent [18:45] Another small thing is that UEFI can do firmware updates. Not sure if that applies to MacBooks, I've never even seen one of them :) === ord is now known as quem === drigy2 is now known as kopachke === ord is now known as quem [20:38] Hi, I'm a new user who installed ubuntu on an old Windows laptop last week. It was working until today. No matter what I do, my laptop won't load ubuntu. I tried running a boot repair by following instructions here: https://linuxconfig.org/ubuntu-22-04-not-booting-troubleshooting-guide [20:38] It ran the boot repair but when i restarted my laptop was still not booting [20:38] I have a pastebin url. Adding it here now [20:39] https://paste.ubuntu.com/p/Yj73M6RKNV/ [20:40] Is this chat the correct place to get help? [20:40] Yes [20:40] thanks jeremy31 [20:41] what do you recommend I do to fix the boot? [20:42] More details: it's an old Toshiba laptop that I bought in 2014. I wiped Windows during my ubuntu installation [20:43] When I try to boot up, I just see the Toshiba screen forever. Before when it was working properly, I would see the ubuntu icon on the bottom of the Toshiba screen [20:44] Right now it shows no ubuntu icon [20:44] I am not a boot issue expert [20:45] darn, you're the only person replying tho [20:46] should i try Askubuntu.com? [20:46] how long did you wait for it to boot before giving up? [20:47] Guest60: Is there any setting in the BIOS to select an UEFI file to boot? [20:47] I have uefi selected already. I tried changing it to csm once and that failed, so I changed it back to uefi [20:48] first boot could take a while w little ram [20:48] Guest60: if you hold shift do you get grub? [20:48] will try that now [20:48] if you hold it right after you power the pc on [20:52] holding shift doesn't do anything. First it takes me to a setup screen with BIOS settings. When I exited, I tried holding shift but it just took me to the Toshiba screen [20:52] i tried holding shift before the bios screen too [20:52] didn't do anything [20:53] Normally ubuntu loads within a few minutes, but today it just stayed stuck on the Toshiba screen for 20 minutes [20:53] I had to force shutdown or unplug it to try again [20:53] Guest60: try tapping 'esc' instead of shift [20:53] ok [20:54] same result [20:55] if that doesnt work try pressing f12 for the one time boot menu and see what's there and if there's an option to browse for an uefi file [21:00] with newer versions of ubuntu, is firefox only a snap package now? [21:01] tried to upload a pic but i guess it doesn't work [21:01] Ubuntu repositories only offers Firefox in snap, yes syphyr [21:02] hm thanks [21:02] Guest60: imgur.com [21:02] f12 took me to a screen with options USB, HDD/SSD, ODD, LAN, and a settings icon [21:02] Guest60: ok, do you have a live usb you can boot to troubleshoot further? [21:02] But it got faster start-up time in recent updates if that's what worries you syphyr [21:02] yeah, i was thinking it would be slow [21:03] does this link work? https://imgur.com/a/CmVC1JA [21:03] and require a lot more disk space [21:03] I have my bootable usb that i used to install ubuntu [21:04] it was what I used to try the boot repair tool [21:04] but boot repair didn't fix the issue :( [21:04] I got the screen saying it was repaired but the laptop still wouldn't boot up [21:07] Guest60: ok, boot up the live usb [21:10] do i click 'Try or Install Ubuntu'? [21:15] heh [21:15] Click compile kernel [21:15] weird, the default firefox installed with a deb file shares the same config as the version of firefox installed by snap [21:16] all the same plugins are thre [21:16] How do you keep ahold of what is snap and what is apt etc [21:16] EricC^^ I'm not on this screen: https://imgur.com/a/kuGH64x [21:16] Do I click 'Try Ubuntu'? [21:16] *now [21:17] Whatdo you wanna do [21:17] Wipe your hard drive and install it, or try it and keep your data? [21:17] I wanna try fixing my boot before i reinstall [21:18] cuz my installation was working for a week [21:18] All your precious data is in /home? [21:18] Backup home and format your hard "fucking" drive and do a clean install [21:19] As long as your important files are safe [21:19] i don't have much saved here. It's just that I expect to eventually run into the same issue if I reinstall the same way i did before [21:19] ok [21:19] i already did a full wipe during this installation [21:19] Whats the issue [21:20] my laptop won't boot up. It gets stuck on the Toshiba screen [21:20] It used to have an ubuntu icon on the bottom, so I knew it was loading that [21:20] now it doesn't anymore [21:20] BIOS-settings? [21:21] Here's my pastebin url from my boot repair tool: https://paste.ubuntu.com/p/Yj73M6RKNV/ [21:21] You need to be logged in to view this paste. [21:22] repasting my earlier msg: I'm a new user who installed ubuntu on an old Windows laptop last week. It was working until today. No matter what I do, my laptop won't load ubuntu. I tried running a boot repair by following instructions here: https://linuxconfig.org/ubuntu-22-04-not-booting-troubleshooting-guide [21:22] Use some fuckin' pastebin that works [21:22] It ran the boot repair but when i restarted my laptop was still not booting [21:22] it's an old Toshiba laptop that I bought in 2014. I wiped Windows during my ubuntu installation [21:22] When I try to boot up, I just see the Toshiba screen forever. Before when it was working properly, I would see the ubuntu icon on the bottom of the Toshiba screen [21:23] ok [21:23] idk how any of this works. I'm new to ubuntu [21:23] boot repair offered to post stuff to a pastebin and gave me that link [21:23] Ubuntu works like magic, just rm -rf /, everything [21:24] What is the problem u are having [21:24] He already told us basically [21:25] Ok so just re-install the Linux [21:25] Did you change something, before the Ubuntu-icon disappeared? [21:25] On such an old device, I reccomend a lighter-OS DE, such as Ubuntu-mate [21:25] i only updated the desktop screen [21:25] What does that mean [21:26] What did you do exactly [21:26] the desktop image. I updated it from the default jellyfish to another image that's in the ubuntu package [21:26] ok [21:26] Well that's not your problem then [21:27] So does Ubuntu still work like normal [21:27] laptop isn't loading ubuntu at all. It's defaulting to the Toshiba screen [21:28] Reinstall a lighter OS [21:29] Guest60: yes, try ubuntu [21:29] ok [21:29] Ubuntu isnt light [21:29] Its a lag-hog [21:29] due to the GNome environment or Unity or whatever [21:29] Try xubuntu [21:29] be elite [21:30] Guest60: once it loads open a terminal and type 'sudo efibootmgr -v | nc termbin.com 9999' and share the link it gives [21:30] You really don't know your actual problem and we specialists don't now either [21:31] If an icon disappears from a desktop, it sounds like Windows [21:31] dude are you serioues [21:32] literally thousands of Linux OSs have desktop icons [21:32] fuck! are people devolving or something [21:33] We are deratarding [21:33] flouride confirmed [21:33] yea, this was an old windows 8 laptop. I wiped windows to install ubuntu [21:34] thousands of Linux OS have desktop icons, but only in Windows do they disappear like so the user believes he should be in a mental institution [21:34] Eric^^ - I got the url https://termbin.com/hcq9 [21:35] ugh, so this happens often to ppl who install ubuntu on Windows machines? [21:35] what the fuck [21:36] Guest, I suggest a light Linux OS [21:36] Ubuntu is kinda heavy [21:36] !language | SIMP_King [21:36] SIMP_King: Please avoid any language that may be considered offensive, including acronyms and obfuscation of such - also see https://wiki.ubuntu.com/IRC/Guidelines || The main channels are English only, for other languages, see https://wiki.ubuntu.com/IRC/ChannelList [21:36] whattepussulickinfuck [21:36] @ignore teear [21:36] !ignore teear [21:37] and this room is for Ubuntu support questions. if you want to discuss other topic use #ubuntu-discuss for Ubuntu related topics. everything else in #ubuntu-offtopic . thaks [21:37] i wanna see if there's a fix here first before I try installing a different os from scratch :/ [21:37] Guest60: give EriC^^ a moment [21:38] from what i see the bootorder looks wrong [21:38] What distro you installed? [21:38] the version? [21:38] Guest60: why the zfs? [21:39] ubuntu 22.04 LTS [21:39] Guest60: try "sudo efibootmgr -o 0,1" [21:39] I put in a boot order of usb first so it'd actually read my bootable usb [21:39] maybe that already fixes it [21:39] if not come back :) [21:39] heh [21:39] Guest60: the boot repair info shows the hard drive is using zfs [21:40] should I wait for EriC before I try that? === madmax__ is now known as madmax [21:40] yes you can wait if you are not in a hurry [21:41] i'll try what he comes back with first [21:41] Is there a correct way to fix this warning? "Fontconfig warning: "/etc/fonts/conf.avail/53-monospace-lcd-filter.conf", line 10: Having multiple values in isn't supported and may not work as expected" [21:41] Guest60: can you or someone upload the boot-repair log to some pastebin, it's asking me to login to view the paste [21:41] https://paste.ubuntu.com/p/Yj73M6RKNV/ [21:41] it would save alot of time to see everything in one place [21:42] how do i do that? [21:42] EriC^^: https://dpaste.com/H2QQRN33N [21:42] thanks ravage [21:43] Guest60: type "sudo mount /dev/sda1 /mnt" [21:44] this fixes most font errors: sudo fc-cache -f -v [21:46] i typed that in but idk what it did [21:46] Guest60: then "sudo mkdir -p /mnt/efi/microsoft/boot" [21:46] terminal just took me to the next line [21:47] fuckin' Microsoft [21:47] Guest60: then 'sudo cp /mnt/efi/ubuntu/shimx64.efi /mnt/efi/microsoft/boot/bootmgfw.efi' [21:50] done [21:51] there's a space between .efi and /mnt, right? That's how i entered it [21:51] Yes [21:51] Guest60: also run ravage 's command, "sudo efibootmgr -o 0000,0001,0002,0004" [21:53] Guest60: then reboot, and try to hold shift to get grub so we know at least grub is booting, if it doesnt work try f12 again and choose the hdd/ssd option [21:53] ok [21:55] thanks, oerheks, but that did not fix it [21:57] i'm now on a screen that says "grub>" [21:57] nice [21:57] type 'normal' and hit enter [21:58] click 'ubuntu' or 'advanced options for ubuntu'? [21:58] ubuntu [21:59] it's loading ubuntu now [21:59] great [21:59] will i have to hold 'shift' and type 'normal' every time i turn it on? [21:59] no [22:00] when it boots, try running 'sudo update-grub' for good measure [22:02] oerheks, the issue is line 8 and 9. https://pastebin.com/TNMbF0K3 [22:03] like maybe create two separate familys? [22:04] updating grub now [22:04] syphyr: you can try to reset the file to defaults [22:04] sudo -- sh -c 'mv /etc/fonts/conf.avail/53-monospace-lcd-filter.conf /root/53-monospace-lcd-filter.conf.bak; wget -qO /etc/fonts/conf.avail/53-monospace-lcd-filter.conf https://termbin.com/glso' [22:05] EriC^^ Done. Do I restart again now? [22:05] Guest60: sure give it a shot [22:05] no shift this time [22:08] it's loading ubuntu normally now [22:09] tysm EriC^^ :D [22:10] Guest60: no problem [22:10] thanks ravage [22:10] that file looks like its fixed [22:10] when i do gpg --list-keys i can see my key that i used to encrypt this file.... now when i do gpg --decrypt file.txt > decrypt.txt its telling me: gpg: decryption failed: No secret key [22:14] that was exactly what I needed, just had to expand it out into two separate [22:17] I have an even older Windows computer that i installed the same version of ubuntu on. Hopefully this one doesn't end up with the same prob [22:17] I've saved the code so i can try that out if this other machine has problems booting [22:18] EriC^^ you are a lifesaver [22:20] Guest60: does the other computer still have windows? dont just run the code if it does cause it would overwrite its bootloader [22:21] the other computer was installed the same way. I wiped windows and installed ubuntu with the same bootable usb [22:21] i'm typing on it now [22:21] really the problem wasn't windows itself, toshiba's uefi seems finicky, what's the other pc's make? [22:21] this is an HP computer [22:21] oh ok [22:22] i think we bought it in 2011 [22:22] ah hp also have finicky uefi, for those, i get by with disabling the windows entry and not the bootmgfw.efi workaround thing [22:23] you'd run sudo efibootmgr -v and get it's number e.g 0004 for windows boot manager blabla, then run 'sudo efibootmgr -A -b 0004' for instance to disable it, then ubuntu should boot [22:23] -a -b 0004 makes it active again in case you ever want to revert [22:26] Install Ubuntu in BIOS mode on something from 2011 [22:26] ah, ok. I'll save this and try it if the hp comp has a problem [22:28] i have no clue what mode it was on when I installed it :/ Hopefully it's ok [22:43] hi i am running an old (19.10) ubuntu live usb and trying to write either 18.04 ubuntu or 20.04 ubuntu to a dvd but "disk image writer" application shows the dvd burner (destination) as greyed out and not selectable. can anyone help please? [22:52] Guest22: brasero might be what you need [22:53] I've also just noticed this error when starting snap's firefox: https://pastebin.com/Pjf79Z04 Firefox still starts up and seems to work ok. Is this an issue with my custom 5.15 kernel? This command works fine "systemd-run --user --scope echo hello" [22:55] older threads mention a grub boot parameter: systemd.legacy_systemd_cgroup_controller=1 [22:55] but legacy_systemd_cgroup_controller is not in my kernel source [22:56] maybe i can just ignore? [23:06] I dont think the log I posted was complete. here is a better one: https://pastebin.com/DV0MbuuG [23:07] it looks like it fails with permission errors, but then afterwards the kernel accepts it === AKTheKnight7 is now known as AKTheKnight [23:46] custom 5.15 kernel? [23:57] yeah i built my kernel with latest lts [23:57] but i dont think its kernel related [23:59] from that log, it looks like systemd initially fails with permission errors, but right afterwards the kernel allows it