=== robbins is now known as abyss [00:33] i am thinking in another concept i called: OFUSCATED SMOKE SCREEN (new security sysinfo paradigma) [00:35] anyone know how to connect to a samsung tv? It has wifi, and macs have a built in method called airplay. But my google-fu is coming up short for ubuntu [00:35] So anyone know why pulseaudio wouldn't see my soundcard while alsa does. [00:37] chibill[m]: you could try this to get more info: pulseaudio -k; pulseaudio -vvvvv [00:37] at least that might work. It might not if ubuntu auto-restarts pulsea [00:39] Waiting on my stuff to load, launched the default DE to see if LXDE is somehow busted. [00:51] Hm it sees the card but isn't making it available. [00:52] Qubes OS is on the right path [00:52] most isolated programs context is the right decision to QA(quality assurance) like stability, security, performance, privacy, symetric/assimetric processing and finnally scalability [00:52] i would like to add another security info concern, how to implement OFUSCATED SMOKE SCREEN in addition to a isolated vm's environment [00:52] lets think some about this new paradgima/techniqué [00:52] At least it sees it if I run it as root, as a normal user even after killing pulseaudio with pulseaudio -k it says the daemon is running. === meyou_ is now known as meyou [00:55] isapgswell: you may like to read about EROS-OS and Fucshia [00:56] sarnold good idea [01:01] sarnold it is too difficult to think QUANTUM energy, performance and scalability without the QubesOS/PureOS and even Tails contribs [01:05] sarnold repeating: quantum security analisys, energy "offer", , performance and scallability [01:05] hail allan turing !!! [01:18] After rebooting I found pulseaudio wasn't running, as my volume control could not connect to the daemon. This is the log from trying to start the daemon. https://pastebin.com/TCMLPHMD [01:20] chibill[m]: maybe try some of the stuff here: https://askubuntu.com/questions/1056153/pulseaudio-not-working-daemon-already-running-and-no-permission-for-home-folder [01:35] Now its not seeing the USB sound card at all in Pulseaudio. [01:36] If I launch it with sudo it can see it but no applications can use pulseaudio. [01:48] How can I know whether lightdm is using x or wayland in the first place, and configure it to ensure it uses X? [01:51] kandinski: ' ps -efly | grep Xorg' . [01:54] Bashing-om: thanks! ok, so maybe I need to run barrier with --display :0 [01:55] there are a bunch of lightdm processes, and it's a bit confusing [01:55] kandinski: Can not say - no xperience with barrier. [01:56] well, the ps output pointed out that lightdm was using display :0, so... [01:56] (can't hurt) [01:56] right? /usr/lib/xorg/Xorg -core :0 -seat seat0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch [01:57] in this case, barrier is launched as an accessibility optional keyboard by lightdm-gtk-greeter [01:57] kandinski: ' echo $XDG_SESSION_TYPE ' work for you ? [01:57] can't do that while under lightdm [01:57] the user session is X, and that's fine [01:58] it's the dm/greeter that's giving me trouble, so I can log into the user session [01:58] Bashing-om: does this make sense? [01:59] (yes, $XDG_SESSION_TYPE is set and it works for me under my user, to answer your question) [02:10] kandinski: Maybe ' lightdm --show-config ' will give a hint on which config file to edit ? [02:11] Bashing-om: I think I have the correct one, there's a GUI even. [02:38] Bashing-om: oh fudge. Snap-installed applications have sandboxed file acess permissions, so barrier from snap has issues creating files under /run [02:38] GROAN [02:39] (letting you know in case you'd like to have a mental alert for this kind of issues, since you've been nice enough to help me, for which I'm grateful) [02:41] kandinski: Not been of the help level that helps :( - but pleased you know what is not happening. And yep - that ^ is but one of the reasons why I avoid snaps on the desktop. [02:42] this is well, five yaks under my current need-to-do task, but at least I have some visibility [03:03] anyone know of a UI text file comparison program that I can open by selecting to files in a file browser like dolphin and right clicking? [03:05] samian: should be able to do that. I use a plugin for nautilus that provides the menu-item-to-compare-diff-the-files. should be a way then for dolphin. maybe a plugin/extension. *shrug* [03:08] samian: look at FSlint to see if it has the options you want [03:09] surely yes, fslint is a grandpappy utility with many options. =) [03:35] yes I use dolphin [03:41] samian: https://wiki.archlinux.org/title/Dolphin#Compare_files === genii is now known as genii-core [04:00] i would like to disable multithread to access my internet digital open banking [04:00] exclusive read/write buffer to my app running [04:00] is it possible? [04:00] on ubuntu [04:11] 1 mouse click is symetric input [04:11] 2 mouse clicks is assymetric input [04:11] i want to exclude USER INPUT IRQ EVENT from smp and thread [04:12] start/stop event needs quickly ininterrupt processing [04:16] mouse and keyboard [04:16] neutral list [04:17] skip mice/keyboard from smp and threading [04:19] processing [05:50] is this a good place to discuss snapd? [05:54] x0n^, this is a support room, use #ubuntu-discuss for non-support discussions, or #ubuntu-offtopic for less focused discussions === EriC^ is now known as EriC^^ [06:37] Good morning, I'm having issues with ecryptfs private directory: I have to manually run the "ecryptfs-mount-private" command a good few times before the folder is actually mounted. Is there a way to fix it? [07:01] RupertEverton: Are you using Ubuntu Desktop and do you have auto login enabled? [07:02] also are you using a password or something else (fingerprint) to log in? [07:03] I'm using Ubuntu desktop and I do have auto login enabled. I dont need the private directory to automount or anything, I just need it to mount when I issue the command. But as of now, it just mounts whenever it wants after i issue the command a good few times [07:04] what do the logs say? [07:04] where can i find them? [07:04] RupertEverton: You probably need to unlock your keyring first, I had that issue before when enabling auto login. [07:04] journalctl [07:07] mgedmin I'll restart and try again fresh and see what the logs say [07:08] 👍 [07:16] Alright this time it opened on the first try, could be that i removed the inegration with pam ? [07:18] can you describe what happened? you reboot, the desktop autologins, you open a terminal, type ecryptfs-mount-private, it works? [07:18] were there any password prompts? [07:19] rebooted, autologged, opened the Private directory and ran the Access-Your-Private-Data.desktop script [07:19] got prompted for my login pass as usual [07:20] tbf it was also slow at unmounting before, while now it worked instantly [07:20] so i think it was the integration with pam that was causing issues [07:20] just to be clear, this is what im referring to: pam-auth-update --force [07:21] hm, unlikely [07:21] should i reboot and give it another try? this time i'll type the command directly in terminal [07:22] see if anything changes [07:22] no reason, it does Exec=/usr/bin/ecryptfs-mount-private [07:23] with Terminal=true [07:23] yeah i know [07:23] but at least with a reboot we can see if it was a one time thing or not [07:24] the PAM configuration for ecryptfs-utils does two things: (1) automount the Private directory on login, if you use a password to log in [07:24] (2) re-wrap your encryption passphrase with the new password whenever you change your login password [07:25] there's no way PAM could cause a slow unmount [07:25] the slow unmount was really the secondary issue tho, i was mostly concerned with the slow and inconsistent mount [07:25] maybe some applications were keeping files open inside ~/Private (tracker indexer? but isn't Private excluded from desktop indexing by default) [07:26] the slow mount is more interesting, and it would be helpful to see what's going on in journalctl if you reproduce it [07:27] to reproduce it i need to reboot. after it's mounted the first time it's usually quite quick to remount [07:28] not that ecryptfs logs much, I see only one message on login [07:29] "gdm-password][3298]: pam_ecryptfs: Passphrase file wrapped" [07:29] yeah i was looking just at that [07:30] actually [07:31] gnome-screensaver-dialog[15908]: pam_ecryptfs: seteuid error [07:31] --- [07:31] umount.ecryptfs[6820]: Failed to find key with sig [***]: Required key not available [07:35] mgedmin I'll reboot a few more times with pam integration enabled and disabled -- brb === EriC^ is now known as EriC^^ [07:52] mgedmin now this is weird, it worked every time with and without pam. The only difference shown in the logs is that when pam is enable i have the following three logs: [07:52] systemd-modules-load[47x]: Module 'ecryptfs' is built in [07:52] systemd[108x]: Perhaps try the interactive 'ecryptfs-mount-private' [07:52] lightdm[109x]: Perhaps try the interactive 'ecryptfs-mount-private' [07:52] yeah [07:53] afaict the pam_ecryptfs: seteuid error is purely costemtic [07:53] *cosmetic [07:53] https://bugs.launchpad.net/ubuntu/+source/gnome-screensaver/+bug/1085706 [07:53] Launchpad bug 1085706 in gnome-screensaver (Ubuntu) "pam_ecryptfs: seteuid error" [Medium, Triaged] [07:53] also, lightdm? gnome-screensaver? what ubuntu version are you using? [07:53] ubuntu budgie 21.04 [07:54] I'm also very surprised that unmounting needs the secret key [07:54] you can use `keyctl show` to see if you have any keys added to the the kernel keyring [07:55] I see two user: keys here (as an ecryptfs user) [07:55] AFAIU one's used to encrypt filenames, and the other one to encrypt file contents? [07:55] looked at /usr/bin/ecryptfs-mount-private (which is a bash script) a bit [07:55] yes i also have two keys [07:55] the way things work, it asks for your password, then uses that password to unwrap the encryption key and stuff it into the kernel keyring [07:56] nothing should be randomly removing keys from the kernel keyring [07:56] so I cannot explain the "umount.ecryptfs[6820]: Failed to find key with sig [***]: Required key not available" [07:57] i got two consecutive errors like that one with different key ids, so i'd imagine those were the two keys stored in the keyring [07:59] it would be intresting to see what keyctl show shows when that error happens [08:06] speaking of ecryptfs, it's deperecated, and I wonder about alternatives [08:06] fscrypt? [08:06] is it actually? [08:07] i knew it's no longer employed on ubuntu installers in favor of luks [08:07] https://en.wikipedia.org/wiki/ECryptfs says it's deprecated, but the citation leads to a 404 [08:09] Hi [08:10] I need help please with: Installing ubuntu 20.04 server: how to delete a partition? https://unix.stackexchange.com/questions/652999/installing-ubuntu-20-04-server-how-to-delete-a-partition [08:10] tbh the ecryptfs website doesnt seem to have been updated since 2020 [08:14] uluntu: the error is produced here: https://github.com/canonical/subiquity/blob/2c7ec18aa870b8982972fc7c761a3ac34e5b0484/subiquity/common/filesystem/actions.py#L293 [08:14] the logic of it escapes me [08:16] https://bugs.launchpad.net/subiquity/+bug/1868249 [08:16] Launchpad bug 1868249 in Ubuntu on IBM z Systems "subiquity installer does not allow to delete existing partition" [Medium, Confirmed] [08:21] can linux (ubuntu) perform BOTH machine-based and user-based 802.1x authentication? [08:22] so first machine-based to put the computer into a network and then user-based whenever a user logs in? [08:24] @mgedmin: I read https://github.com/canonical/subiquity/blob/2c7ec18aa870b8982972fc7c761a3ac34e5b0484/subiquity/common/filesystem/actions.py#L293 How should I Use that? Thanks. [08:24] the bug link is more useful [08:25] afaict your choices are (1) erase entire disk and create a new partition layour, or (2) switch to a shell on ctrl+alt+f2 and use a different partitioning tool (parted, or cfdisk) [08:25] But how to use it? [08:25] I like cfdisk but I think it's deprecated/unmaintained === sharperguy_ is now known as sharperguy [08:42] @mgedmin: How the subiquity bug report helps? === SpydarOO7 is now known as Spydar007 [08:50] uluntu: reading it seems to confirm that your choices are those mgedmin listed. so that's reassuring, i'd say. [08:51] I think so. [08:51] uluntu: is you use case that you need to retain some partitions but need to install ubuntu to the same disk? [08:53] @tomreyn: yes. I had wanted to delete that partition, create a new one to replace the deleted one and install ubuntu server on the new created one. [08:54] so i guess just destroying that partition outside of the installer GUI is your easiest otion then [08:54] btw. there's also #ubuntu-server [08:57] @tomreyn: Thank you for letting me know the channel #ubuntu-server [09:00] you're welcome, uluntu. i also doubt they'll have a different answer - it's an unfixed bug. but i can be wrong. === EriC^ is now known as EriC^^ [09:04] i thought you could ssh into ubuntu by default when you first start up? [09:06] as soon as an ssh server is installed, sure [09:07] i don't think the desktop installer installs on by default [09:11] Correct the desktop does not have ssh installed by default [09:13] I Just tested and it is not installed on ubuntu server 20.04 by default either [09:13] i think ubuntu server lets you choose to install one on the last but one installation screen [09:15] correct but not as default how hard is sudo apt-get install openssh-server after install otr the check box option either way [09:15] i had the client but not the server so i could only ssh out. [09:16] i installed openssh server and everything worked fine, but now i cant' ssh in anymore [09:16] most of the VM templates have it installed for user but not root so even those you have to add a user before ssh into a box [09:18] sail0r: what error message are you getting if any? [09:20] aallllllso having wreleess kb problems.. ihink i remember having the same issue last time i installed ubuntu [09:20] no response [09:20] Rare for ssh to fail either way first test is to verify user and password and IP assuming local network on the server [09:21] I seem to mess up my notes all the time [09:21] i installed xrdp and did one mod and thats all i can think that would cause an issue. but not sure why that would prevent ssh sessions from getting through [09:22] can you ping the server? [09:23] well i wasnt sure if you have to change a config on / for all users or if it just auto config the user account config [09:23] user always [09:24] well, im actually on it now... wanted to see if i missed something before going to bed [09:24] does "ssh localhost" from the machine itself work? [09:24] yes [09:25] that's a good sign then :) [09:25] have you got a firewall installed, eg ufw? [09:25] no firewawll [09:25] fresh install everythings still closed up [09:27] are you able to visit https://www.grc.com/x/ne.dll?bh0bkyd2 from the machine? [09:27] it's an online port scanner and from the "common ports" tab you'll be able to see if ssh is open from the greater internet [09:30] shouldnt it be available on my intranet? [09:30] they're on the same network [09:31] host n client [09:32] ah. if it's lan you should be golden, hmm... [09:33] what's the local ip address? [09:35] ssh username@local-IP-address [09:36] yup, worked he first time but now it gets closed [09:38] could you paste the output of "ssh -vv local-ip" to https://pastebin.ubuntu.com/ ? [09:38] couuld this command have done anything to it?"echo "gnome-session --session=ubuntu-2d" > .xsession [09:38] " [09:41] https://pastebin.ubuntu.com/p/gJcj83gMfw/ [09:41] sail0r: no echo command does not change anything related to the network [09:41] i cant recall right now if i logged in using the ip the first time or the hostaname [09:42] you need to replace "local-ip" with you local ip address :) [09:42] which you can find from "ip a s" or the network settingsconnections guis [09:42] toddc: i was wondering if it was the output sent to the .xsession command that might have changed some config or somethin [09:43] francis: in the host file or the etc ssh config? [09:44] where is that script reading the debug info from to get that reply? [09:44] neither. take a look at https://itsfoss.com/check-ip-address-ubuntu/ [09:44] no echo will just verifiy a output of the opertion in this case gnome [09:44] once you have you r local ip address (let's say 192.168.1.5), then you "ssh -vv 192.168.1.5" [09:45] toddc; thanks [09:46] as francis said hostname adds a few more issues work with a ip static is preferred then deal with a hostname second as that willl involve you router [09:46] i had already done that, i just cant remember if i was trying to use the hostname or not. ill have to try in the morning. although it would be ideal to have it work for the hostname. [09:48] well i didn't expect that my device would get the same ip every tie anyway, and wouldn't the hostname be prefereble for that anyhow [09:48] ? [09:48] default on raspbian and lmde seem to work with hostnames [09:49] you router may not be updating hostname to ip that thaks a while to update if any changes happed [09:50] so for testing ip does not need a router ---better till everything works and is stable [09:51] hrrmm, i guess that makes sense. im passing out so ill try again later. thank for the help guys [09:51] ip goes direct to the server vs host name gomes from the router and has to be updated on both machines to find each other [09:52] in simple terms === EriC^ is now known as EriC^^ === wode is now known as wodes === entuland_ is now known as entuland === EriC^ is now known as EriC^^ [11:04] is lightdm prefered over gdm3 ? [11:05] by who? [11:05] ubuntu desktop uses gdm3 [11:06] I'm sure there are some users somewhere that prefer lightdm [11:06] note that modern gnome desktop requires gdm for some features to work correctly (fast user switching? screen locking? I don't remember all the details) [11:06] trying to install xserver and it asked whether to install gdm3 or lightdm [11:07] instaling xfce4 [11:27] ac5tin, gdm3 is a cut-down GNOME DM so makes sense for a GNOME desktop... makes less sense with XFCE so I'd use whatever Xubuntu does [11:42] I'm having a no-sound issue -- in settings i see the bar coloring up when i play a video but there's no actual output. i tried to force-reload alsa, remove pulseaudio config and reboot, reinstall alsa-base and pulseaudio. alsamixer shows that the speakers are not muted and at full volume. the following is the output of "pulseaudio" [11:42] E: [pulseaudio] pid.c: Daemon already running. [11:42] E: [pulseaudio] main.c: pa_pid_file_create() failed. [11:44] is anyone good with wireless connection on printers? [11:44] dunno, the office mfp is wireless and works fine here [11:45] all I had to do is use its touchscreen to connect to the right wlan with the right wpa key [11:45] I connect my printer to the router and it works fine [11:45] I get as far as filling in the password, but then it beeps in protest [11:47] It`s a Brother printer, I am installing new drivers on a Windows machine at work, and hoping my laptop detects it [11:47] Ubuntu should detect things like that? [11:48] I think it`s mostly a wireless issue, but making things work between PCs and printers, etc [11:49] at one point it was easier to print from ubuntu than from a windows laptop -- ubuntu would autodetect the printer, we never figured out how to make it work on windows [11:49] but wifi setup comes first, and ubuntu can't help here [11:50] Ubuntu is usually very smooth with interactions [11:50] (ignoring bugs, which are, alas, not yet all eradicated) [11:50] A usb cable would probably work right a way, but I need the wireless [11:54] dragon2: does gnome systemsettings recognize your printer? [11:56] no printer detecrted [11:58] dragon2: maybe you should journalctl -f and plug it with usb first, to see what it does [11:59] mdns-scan or avahi-browse -a can tell you if the printer is connected to your network and is broadcasting mDNS announcements [12:01] I'm getting 10 of these for my printer, plus an extra 3 from cups on our local server [12:11] what is the best way to set DNS servers permanently for the whole system? [12:11] have your dhcp server hand them out? [12:13] you can configure static DNS in netplan [12:13] (or whatever network configuration tool you use) === war is now known as Guest3818 [12:15] Hi folks [12:15] mgedmin not possible in systemd-resolv? [12:16] systemd-resolved takes upstream DNS configuration from systemd-networkd or network-manager, depending on which one you use [12:17] I guess the simplest option is to replace the /etc/resolv.conf symlink with a plain file that has your desired DNS server [12:17] but then you lose systemd-resolved's features like caching and DNSSEC support === ada_ is now known as ada === ada is now known as ada_ [12:22] mgedmin i applied conf here https://linuxize.com/post/how-to-set-dns-nameservers-on-ubuntu-18-04/ but still not affected [12:22] ah so it's a desktop and you use network manager [12:22] yes [12:23] and what does systemd-resolve --status say? [12:23] DNS Servers: 192.168.1.1 [12:23] hmm [12:23] have you tried disconnecting and then reconnecting? [12:24] Is it possible to filter a repository so that I get only one package from it? [12:25] mgedmin restarted the systemd-resolv === s1__ is now known as s1 [12:25] bailsman: yes, see man apt_preferences [12:26] Found this https://help.ubuntu.com/community/PinningHowto is that the right approach? [12:27] yes-ish [12:27] I wouldn't touch APT::Default-Release, it'll cause pain on upgrades [12:28] the other bit, /etc/apt/preferences, is the thing I mentioned earlier [12:28] I want to install prometheus-node-exporter from Groovy so that I can use tls. But that is the only package I want to pull/update from groovy. I realize that this is 100% at my own risk and there is a decent chance cross release packages won't even work at all or malfunction in weird ways. [12:28] https://manpages.ubuntu.com/manpages/impish/en/man5/apt_preferences.5.html [12:28] finding a ppa that backports the thing you want is better [12:28] if there's no ppa you can try to create one and do the backport yourself [12:30] all you need is pull-lp-source prometheus-node-exporter groovy; cd prometheus-node-exporter-*; dch -i and adjust the version number to be same as in groovy with a ~ubuntu20.04 suffix (assuming you're backporting to focal) and set the release to focal [12:30] then debuild -i -S, cd .., dput *.changes [12:30] oh and you need to have a gpg key associated with your launchpad account and a ppa created in launchpad [12:31] and I omitted the ppa:yourname/ppa argument from dput [12:31] Thing is that I don't want to watch prometheus github, if there's a security issue or anything, I want it to auto update or at least flag that there's an update. If it breaks completely after the update, that's fine, maybe I'll rebuild it myself the.n [12:31] well, hmm [12:31] why ubuntu still doesn't fully encrypt the disk including /boot yet? [12:32] I suspect you'll get library compatibility issues if you try to install a groovy package on focal, but eh, why not try it out in docker or lxd or something [12:40] mgedmin renderer in netplan should be netowrkd or networkmanager? [12:40] it depends on which one you use [12:41] if you don't understand it, don't change it [12:41] It actually worked in Docker! Does that mean this ISNT a major footgun? Also: don't I need the default release, even though it messes with upgrades, just to make sure I don't accidentally upgrade another package? [12:42] Is there an apt preferences only method? I did the pinning for prometheus-node-exporter [12:43] you pin everything in groovy to some very low priority like -1, and then pin prometheus-node-exporter to something higher than the default pin (which is 500) [12:43] Beautiful! Thank you === anicet_ottflo is now known as AlbertLarsan68 === war is now known as Guest4669 === anicet_ottflo is now known as AlbertLarsan68 [13:34] xrandr works with mutter??? [13:35] or autorandr [13:36] xrandr works with X11 ... no matter what window manager you use [13:36] ogra humm [13:37] i disabled xrandr via gsettings [13:37] it break my wallpaper rendering [13:51] daemon/service is assynchronous and no arbritary [13:51] agent/client is synchronous and arbritary [13:54] mouse_click/mouse_release, dual IRQ?? [13:55] duo IRQ? [13:55] same event? [13:57] what about thread listening transition events [13:57] sorry, thread waiting transitions terminating [13:59] isapgswell, are you aware that you are dumping completely context-less stuff into the channel since two days ? it would really help if you phrased proper questions in full sentences, described what you have tried which lead to which errors etc etc just dumping phrases like "fabookle mobatle or not ?" into the channel wont help you or us to help you [14:00] ogra :-D [14:01] I think is better just to type /quit and go to /server irc.freenode.net instead of pulling up with jerks ;X like Unit193 and there French hating Americans. [14:01] ;X [14:01] Of the IRCOPS from this net [14:02] Diving the nets [14:18] How can I reload the v4l2 kernel module? For some reason my webcam locks up sometimes and I have to reboot. I want to see if just reloading the kernel module works [14:21] magic_ninja: is that an integrated webcam? [14:21] tomreyn, no, it is a usb webcam. [14:21] if not, just pull the usb cable and reconnect it, maybe to a different usb port. [14:21] is there a way to reset a usb port via software? [14:22] most likely, but i dont know it [14:22] beyond this it is very likely the module it uses is "uvcvideo" [14:22] fair enough. It actually isn't even showing up on usb. [14:23] regarding powering usb ports off/on ... see https://stackoverflow.com/questions/18765725/turning-off-a-single-usb-device-again [14:24] (via the bind/unbind sysfs nodes) [14:25] Unplugging and plugging back in works, but only sometimes. [14:25] It worked this time, but has failed the other dozen or so times it happened. [14:25] I kind of wonder if perhaps something changes after I apply updates that causes it. [14:32] https://askubuntu.com/questions/645/how-do-you-reset-a-usb-device-from-the-command-line should work, it sends a USBDEVFS_RESET ioctl to the device. i assume that's also possible through sysfs, but not sure how === bsmith093 is now known as Guest2667 [14:48] What user does cloud-init run as? I've done `su - myuser` in a bash script I'm giving it and it says it's myuser (verified with $(pwd) in the script) but everything it creates etc is chowned to `root`... why? [14:48] also stuff like $USER and $HOME aren't set [14:56] tsujp, perhaps more a question for #ubuntu-server [14:57] Cheers ogra [15:44] Are there any issues with running Ubuntu and Xubuntu with the same /home directory? [15:45] bl: i would *assume* not. [15:45] unless it's two different kernels accessing it at the same time [15:46] oh, that's a good point. i hadn't thought of it like that. i was thinking for a single machine [15:47] if you'll do multi-boot (booting only one at a time) or switch DEs during login then i assume it will be fine. [15:48] but.. that's theoretical [15:48] what are DEs? [15:48] desktop environments [15:48] gotcha [15:48] i.e. gnome-shell/metacity (ubuntu) and xfce4/xfwm4 (xubuntu) [15:49] we'll it'll be an interesting experiment [15:49] yea [15:49] just don't expect it to be easily reversible [15:50] and remember that by now (21.04) home is ecrypted by default [15:51] a situation where multiple kernels might access /home at the same time would be if the storage (ie. /home) is on network storage and multiple machines can access at once [15:51] or virtual machines? [15:52] ioria: thanks, i didn't realize that was the case [15:53] bl if you ask me, it's just not a good idea (can lead to conflicts, i mean the programs settings in the hidden folders and such ) [15:54] ioria: sounds like this just isn't a use case considered by the Ubuntu/Xubuntu teams [15:54] bl: a situation where multiple kernels might access /home at the same time would *not* be a problem is that home was e.g. an NFS export. it *woould* be a problem if those were e.g. virtual machines both accessing the same storage device directly. [15:55] bl: most people will choose one desktop environment and keep using that. [15:56] what's your use case? [15:59] We have done mutible DE's at our installfest with no issues to show the differences . It works and we have had no issues but I might not call that stable or well tested [15:59] Hello. I want to install Ubuntu bit without a predefined desktop, and then build up a i3-wm based one, package by package. Is it better I start from the desktop ISO, or is the server ISO fine too? I prefer building UP instead of removing packages. [15:59] tomreyn: oh, is that because NFS will serialize the IO? also i guess my use case is just wanting to be able to use either distribution. i have used Ubuntu for years and i like a lot of things about it but i think Xubuntu is supposed to be a bit lighter weight [16:00] ratrace: desktop kernal build and change to the new DE and programs [16:01] bl especially if you use different versions of the two OSs (Xub 18.04 and Ubu 21.04); in this case, you'll have different versions of the same programs stroring its settings [16:01] toddc: I don't understand what you're saying. I'm primarily asking if I can build up to a desktop of choice, starting with headless server ISO [16:02] ioria: that's a good point, i was definitely planning on keeping the same version ie. 20.04 [16:03] bl: i don't actually know how exactly NFS or other network file systems ensure that concurrent writes are possible, but i'm sure you can read up on this online. [16:03] why it's wrong to set dns server in /etc/resolv.conf? [16:03] ratrace: you can but the server kernal is not optimised for a desktop envoroment so it would be better to change A-B [16:06] toddc: it's actually the same kernel [16:07] yeah I think it's the same linux-image-generic [16:07] ratrace: i think both approaches can work. the server installation will have less software installed, and some different defaults, whereas the desktop installation will have more software installed you may want to remove. [16:07] a third approach would be debootstrap or the like [16:07] tomreyn: right. I want a very minimal setup for i3-wm and steam, nothing else. [16:08] yeah debootstrap, I have setup scripts but for servers. hence me wanting to start with the server ISO [16:08] there's also the legacy debian-installer for some releases which could create small customized installations === qskwood_ is now known as qskwood [16:10] so start with subiquity and install network-manager afterwards, and replace the systemd-networkd configuration and remove netplan [16:10] this is probably far from a full list of the differences between a ubiquity and a subiquity installation. [16:11] tomreyn: you are correct 12.04 and newer are the same I learnid somthing today [16:19] I don't really plan to use network manager but networkd. thanks for confirming tho, I'll manage. Just wasn't sure if ubuntu is "still" the same big repo, and it only depends which packages you install. [16:24] I set dns=none to NetworkManager conf but it still overwrites /etc/resolv.conf, any idea? [16:33] hi [16:33] I cannot install Linux Mint with a amd dedicated gpu [16:36] Guest9956: this is a Ubuntu support channel. We don't support Linux Mint here. [16:37] Guest9956: you can find them on the Spotchat server [16:38] |Mint [16:38] !Mint [16:38] The Ubuntu channels can only provide support for Ubuntu and its official flavors, since other distributions and derivatives have repository and software changes. So please use their dedicated support venues, for example: Linux Mint (#linuxmint-help on irc.spotchat.org), Kali Linux (#kali-linux), and LXLE (#lxle) [16:38] Guest9956: ^ [16:38] I have a USB connected Raspberry Pi on my Ubuntu laptop. The laptop is and needs to be connected to wifi, sometimes 10. network, sometimes 192.168 types, depending on the coffee shop I am in. When I do an ip a, each device does see the other. The pi has a 16.254 local address for the usb. In order to connect the two though I have a problem. dhcp allocated a 10. address to my wifi, but no address to my usb0 device. I can force [16:38] dhcp to do that with dhcpcd but then I lose access to my wifi. How should I be doing this. I would add that the Pi too has wifi. When I do use dhcpcd on the Ubuntu machine I can indeed connect to pi.local [16:46] Is it possible to override font sizes for specific applications? [16:47] on the communication question that I asked. I solved it - just needed to tell dhcpcd to probe BOTH interfaces === LjL^ is now known as LjL [16:51] rroblopx [16:51] d [16:52] when I connect a device Ubuntu allocates it a name such as enx16ac4bd157b9 - how does it determine that name and why? === pizzaiolo is now known as pizza [16:59] jpmh, because the new network interface naming protocol, https://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ [16:59] it is a breeze when MAAS or cluster is setup [17:00] oerheks: TY - I will go and read that. But why can it notjust give the same name each time I connect the same device [17:02] oerheks: that document implies that it does do what I want it to do and it doesn't - what am I missing here? [17:05] jpmh, create a static naming in /etc/udev/rules.d/70-persistent-net.rules, some old reading https://itectec.com/ubuntu/ubuntu-network-interface-name-changes-after-update-to-15-10-udev-changes/ [17:05] maybe #ubuntu-server can answer more than this // [17:06] oerheks: TY agaun -now reading that === denningsrogue2 is now known as denningsrogue [17:21] so this is an out there question: I want to organize and tag a group of files on my harddrive, but without changing the file names. Any suggestions? I've been using folders and links, and maybe I should just use a spread sheet. I dunno [17:29] Does anyone know how to adjust scaling to support HiDPI screens? Spotify is barely usable as it's too small -- followed the suggestions here but none worked https://community.spotify.com/t5/Desktop-Linux/Linux-client-barely-usable-on-HiDPI-displays/td-p/1067272 [17:40] nvm, solved the above with menulibre [17:54] how can I get apt update and install to work for hirsute/impish in docker? it would appear I need to add one or more keys. [18:05] JamesB192: works for me with nothing special [18:07] have a strange issue. I unplugged my computer (Asus MB). then the system would not start up. had to reset the bios, then the system would boot, but hand during booting. so I booted to an older kernel and it works. Now I am trying to figure out why the newer kernel will not boot. running5.8.0-53-generic, but5.8.0-55-generic will not boot. Not sure what is going on [18:09] I have tried update-grub and various things, but nothing seems to fix it. [18:10] quantum: what Ubuntu version? [18:10] Not sure what to try next. right now if I boot with recovery mode it hangs after saying ok to reached target local encrypted volumes [18:10] 20.04 [18:12] since I can boot to the older kernel version if feels like something got corrupted with the newer one. I feel like it is something with the NVIDIA drivers, but am not sure [18:12] you could try reinstalling that kernel> sudo apt install --reinstall linux-generic-hwe-20.04 [18:13] I tried switching the nvidia driver while booted in the running5.8.0-53-generic, but don't know if that switched it for the other kernel. either way it didn't boot still [18:13] Are you using encryption? [18:13] don't think so, but it seemed to say o.k. to that line [18:14] [o.k.] Reached target Local Encrypted Volume [18:15] that will reinstall the 5.8.0-55-generic kernel? [18:16] is there a way to reinstall all the NVIDIA stuff too? [18:17] https://askubuntu.com/questions/1319889/ubuntu-20-04-stuck-at-ok-reached-target-local-encrypted-volumes ran into the same symptoms as you did [18:17] quantum: It should install it as that should be the kernel pulled in by the meta package [18:18] yeah that looks like where mine stopped [18:18] So whatever it was going to check next failed [18:19] yes, is there a script it is following [18:21] he says something about getting stuck at gnome display manager, that sounds about right [18:21] quantum: see if nvidia is installed for the -55 kernel in> dkms status [18:21] which seems like it is something with the graphics drivers to me [18:22] vidia, 460.80, 5.8.0-55-generic, x86_64: installed [18:22] is there a way to --reinstall it though if it got corrupted or something [18:23] sudo dkms install nvidia/460.80 -k 5.8.0-55-generic [18:24] yes it is installed apparently for the 55 kernel, would this work 'sudo apt install --reinstall nvidia-driver-460' work? [18:25] It may attempt to reinstall [18:25] or should I try with dkms uninstall then reinstall [18:26] quantum: I don't think you want to uninstall [18:27] well, if I do the reinstall it my screw up my currently running kernel so that doesn't seem like a good choice, unless I downgrade the driver then do an install and a reinstall after booting to the older version I guess [18:27] I mean if I do a reinstall it may screw up my currently running system [18:28] well, let me try a reboot as is since I did the kernel reinstall and see what happens. I will hopefully be back [18:28] quantum: it will just tell you already installed on that kernel, just tried with a different dkms driver [18:28] yes dkms says it is sintalled for both kernels [18:28] the kernel reinstall may have triggered dkms to build new [18:29] ok, be back after a reboot hopefully [18:33] o.k. I think the apt install --reinstall linux-=generic-hwe-20.04 seemed to fix it [18:34] seemed to book up with 5.8.0-55-generic o.k. now [18:34] thanks [18:34] good [18:34] still very strange, no idea what my MB did to it [18:35] wonder if there is a glitch with my NVMe [18:36] Could have been some odd error with the kernel install [18:37] oh, well, guess I know how to fix it now if it happens again === mybalzit1h is now known as mybalzitch [18:48] hi. I'm running Focal live USB ISO env... I'm in "Try out ubuntu" mode because I want to debootstrap the installation. The live env no longer has the ability to set up non-en keyboard layout? [18:48] So, my wifi will just completely stop working [18:49] I have to reboot to fix it [18:49] when I do, the logs are gone [18:49] this is a snapshot of the logs before I rebooted: https://i.imgur.com/6wfaZBq.png [18:50] also, I did export all the logs before I rebooted... so maybe it's in there? [18:50] there's just a ton of files and I don't know which to look in === stwalker1ter is now known as stwalkerster [18:53] I need some help. I have full-disk encryption and auto-login (because why password twice), but although gnome-keyring-daemon is included in startup applications, the GNOME keyring is NOT unlocking automatically, nor prompting to unlock... [18:53] can anyone help me figure out why my wifi just stalls requiring a reboot all the time? [18:53] To make matters worse, goa-daemon is starting, not seeing an unlocked keyring, and thus reporting that there are no online accounts [18:54] Which in turn causes Evolution to clear its settings. Every. Single. Boot. [18:54] is one supposed to do a apt-get dist-upgrade without breaking Ubuntu [18:54] How do I fix this little issue? [18:54] bumblefuzz: https://ubuntuforums.org/showthread.php?t=2354328&p=13614520&#post13614520 [18:54] quantum: in theory, although any given package update *might* theoretically break something. ;) [18:54] But it doesn't, say, go from 18.04 to 20.04 or anything nuts like that [18:59] o.k. but was trying to figure out if it is something typically done. it sounds like there is generally no reason to do it [18:59] tomreyn: Hello. Do you remember me? [18:59] quantum: Basically, no. You can roll things back in the worst case. [18:59] I mean...roll back an individual package. [18:59] But, 10 years, never had a dist-upgrade break my user OS. [19:00] jeremy31 did you look at the picture I posted? [19:00] I don't think that's the problem [19:00] it stalls because of keys and encryption [19:00] ok thanks [19:01] bumblefuzz: Are you using TKIP encryption [19:01] no [19:02] I can check the logs but I don't know where to find the messages that popped up before I rebooted [19:02] I exported everything [19:02] so it should be there [19:03] bumblefuzz: on a Live ISO? [19:03] no [19:03] older kernel logs should have the errors [19:04] bumblefuzz: run the 2 commands at https://gist.github.com/jeremyb31/a2bee9856d8c13f42f1835bc31bf9480 in terminal and post the termbin URL after the second command is done. [19:05] https://paste.ubuntu.com/p/VSmZmN7jqG/ [19:08] is there a bot factoid for (focal's) apt/sources.list ? [19:09] bumblefuzz: try> echo "options iwlwifi 11n_disable=1 swcrypto=1" | sudo tee /etc/modprobe.d/iwlwifi-opt.conf [19:09] jeremy31https://termbin.com/0vas [19:10] bumblefuzz: > sudo iwconfig wlp0s20f3 power off [19:10] will that not shut off my wifi? [19:11] bumblefuzz: no, just disable power management, change the router to channel 6 [19:12] ratrace, no [19:12] and the first command? [19:12] swcrypto? [19:12] sudo cp /usr/share/doc/apt/examples/sources.list /etc/apt/sources.list [19:12] basicly for any ubuntu version. [19:12] bumblefuzz: According to an archlinux thread, that might fix it [19:12] oerheks: k, thanks [19:12] Huh, didn't know that one. [19:12] all you need is to add universe probably [19:13] Unit193, found @ https://ubuntuforums.org/showthread.php?t=2455131 [19:13] I use deb822 formatting, so it'd be different. But that's great, I'll have to remember that if someone else asks. [19:14] Unit193, maybe i should ask for this solution to be added as factoid [19:15] bumblefuzz: The router might need a reboot as one line in the results is unusual> Authentication Suites (2) : PSK unknown (4) [19:16] bumblefuzz: also try with firewall disabled [19:24] jeremy31 so now we just wait and see if it happens again [19:24] ? [19:25] bumblefuzz: That swcrypto doesn't take effect until a reboot [19:25] and what does that do exactly? [19:26] enables software encryption if I remember [19:26] bumblefuzz: lets change that> echo "options iwlwifi 11n_disable=8 swcrypto=1" | sudo tee /etc/modprobe.d/iwlwifi-opt.conf [19:27] and reboot? [19:27] 11n_disable=1 actually disables wifi-n, =8 enables agressive TX, reboot [19:31] jeremy31 anything else? [19:32] not that I can think of, unless there is an update for the router [19:32] not that I see [19:33] I'll just wait and see [19:33] thanks for your help! [19:33] ok [19:34] Anyone using firefox here? [19:35] I use firefox everywhere [19:35] cbreak, do you have that infamous firefox 89 now [19:35] If you have [19:35] Do you see the downloads button? [19:36] yes, and yes. [19:37] 89 on Mac OS, Ubuntu, Android [19:37] I see a download button on the first two [19:37] Like the one you get a dropdown of current downloads from if you click it right? [19:37] to the right of the url bar [19:37] yes [19:37] just like before, but with a different icon. [19:37] cbreak, its suddely missing in mine [19:38] right click toolbar, select customize... [19:38] and drag it back, if you find it there [19:38] or find it in the overflow menu and drag it out [19:38] Yeah its not there either [19:38] In theCustomize toolbar page [19:38] cbreak, whats the overflow menu? [19:38] the thing that looks like a right shift [19:38] >> [19:39] I don't see thsat [19:39] I don't see that [19:39] oh wtf [19:40] cbreak, I just noticed it [19:40] What in the name of fucking fuck is this new icon [19:40] it's down ... load? [19:40] Its a pie chart [19:40] pikapika: Can you keep the language a bit cleaner here? Thanks. [19:40] A white circle with a blue slice [19:40] wtf [19:40] Unit193, these new "ui developments" make it hard to but I'll avoid [19:40] here it looks like an arrow pointing down [19:40] !language | pikapika [19:40] pikapika: 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 [19:41] cbreak, it used to be a downward arrow [19:41] I get frustrated with firefox too, yeah. [19:41] with a kind of dish below it [19:41] Its now just a pie chart icon [19:41] it changes to a progress circle while you download stuff [19:41] I think? [19:41] cbreak, thats a new thing [19:41] like a radial progress bar [19:41] I always had one or two paused downloads [19:41] well... not a bar. A progress pie chart [19:41] But it didn't change the icon [19:41] I canceled it [19:41] And now its back to the down arrow [19:42] It was a pie chart before [19:42] Which it wasn't even in this partial download stage in the earlier ui [19:42] well, grats, you found your button. It was just cosplaying [19:42] I'll have to investigate how to keep the same Down arrow in both states now [19:42] Goddamn ui designer idiots [19:42] having progress is nice. [19:43] you could enable text labels :D [19:43] in the good old times, text was all we had. These new-fangled "graphical iconographics"... don't need 'em [19:44] Unit193: he's been told SEVERAL times about his bad language and continues to swear. [19:45] Yeah but like its very unintuitve for the downloads icon to change to a completely different thing depending upon state [19:45] might not be intuitive, but it's useful. [19:46] it is probably possible to skin it to some degree... [20:17] hey all. i've got a fresh install of 20.04 lts. when i run snap-store from terminal I get this: https://pastebin.com/j7hkMALj [20:18] ive attempted to reinstall snap-store as well as libgtk3.0 with no change [20:18] has anyone else run into this? [20:18] why run snapstore from terminal? [20:18] it won't launch from gnome. just hangs and closes [20:18] was trying to troubleshoot why [20:20] tried to reinstall? or proper update; sudo apt update && sudo apt dist-upgrade [20:20] i was attempting to avoid reinstallation. i've done apt-get update and upgrade [20:21] ill try dis-upgrade [20:21] dist* [20:22] 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. [20:23] snap remove snap-store and snap install snap-store again should fix it [20:23] i have attempted that as well [20:23] no change [20:24] oerheks: have you tried with snap remove --purge [20:24] i think i have. but can't hurt to try again [20:25] sorry wrong tag [20:25] its all good [20:26] same error on output after running a purge === genii-core is now known as genii [20:27] @arif-ali https://pastebin.com/j7hkMALj [20:27] i cannot run it from terminal too, so try from the menu again [20:28] it opened in the task bar, hung and closed [20:29] is snap store any different than gnome-software? im giving this laptop to a 13 year old to learn on [20:29] if not I just won't worry about [20:29] https://status.snapcraft.io/ [20:29] tbh, I get an error too, never tried snap-store before [20:30] ah. their site shows intermittent access [20:30] maybe i've over complicated this too much [20:31] is snap store suffering side effects of Fastly? === compuguy7 is now known as compuguy [20:48] spartanturtle, nope, the snap store stopped using fastly a while ago [20:49] (not related) === bsmith093 is now known as Guest5733 === entuland_ is now known as entuland [21:34] in focal, how can I tell systemd that nullmailer shall only run when any network interface with a default gateway is up? === ghostcube_ is now known as ghostcube === h1pot is now known as BeneditoSalvador === R2robot_ is now known as R2robot === Guest4669 is now known as war