[00:19] <h2oly> turned out, putting it in legacy only was the only way to force it to live boot that drive
[04:37] <hwittenborn> I'm attempting to run some things under fakeroot for some testing, but the permissions are staying under my user for some reason.
[04:38] <hwittenborn> For example, I've tried running `fakeroot -- touch file`, but checking the permissions for the file with `ll` is showing the ownership to be my user still.
[04:38] <hwittenborn> Am a doing something wrong?
[04:44] <rfm> hwittenborn, fakeroot just makes files created in the environment appear to be owned by root to commands in the environment. once the environment is exitted the illusion disappates.
[04:46] <rfm> hwittenborn, try running just "fakeroon" (which starts a shell in the fake environment.  In that shell, rouch a file and ll it, see it's supposedly owned by root.  now exit that fakerood shell with ^d and ll the file again.
[04:47] <hwittenborn> Yep, that's working.
[04:48] <hwittenborn> How would I create a Debian package then? Would `dpkg -b` not work with fakeroot then?
[04:50] <rfm> hwittenborn, since the archivers that dpkg calls (I don't know offhand if it's ar, tar, or cpio) would run inside the fakeroot it would see the ownership of files created as root, and write that ownership into the archive.
[04:50] <rfm> hwittenborn, but the deb would need to be de-archived as real root for the files to be created as root...
[04:52] <hwittenborn> So the files in an archive don't change (or just set correctly rather) while it's in an archive?
[04:55] <rfm> hwittenborn, the archive (it's ar, by the way) is a serialized form of the files included, it has header information that says what the file should b enamed, what ownership and permissions it should have, and the actual data.  when the archive is extracted, the archiver reads that, creates s file with the specified name and asttributes, and copies the data into it.
[04:57] <hwittenborn> That makes enough sense.
[04:57] <hwittenborn> I think I got all the info I need for now, thanks!
[04:58] <illuminated> is there a splinter channel for the ubuntu channel that is offtopic?
[05:00] <toddc>      #ubuntu-offtopic
[05:00] <matsaman> /msg alis list *ubuntu-o*
[05:01] <rfm> also #ubuntu-discuss for stuff which is offtopic here (not being support) but still about ubuntu in general
[05:03] <matsaman> weird distinction
[05:03] <illuminated> thanks everyone
[05:31] <Guest65> anyone know where i can post a Mysql question?
[06:19] <bodiccea> there was a change in /etc/kernel/postinst.d/apt-auto-removal in 21.04. Before, I was able to keep more kernels by editing the line 'previous_version="$(echo "$debverlist" | sed -n 2p)"' with something like "sed -n 2,5p". Apparently the new script is totally different. How could I do the same ? As I use my own kernels (usually 2), I need to keep 3 Ubuntu kernels plus those 2, that is 5 total.
[07:19] <CoolerX> Hi
[07:19] <toddc> CoolerX: welcome
[07:21] <CoolerX> I am trying to install rbenv using apt-get
[07:21] <CoolerX> sudo apt-get install rbenv
[07:22] <CoolerX> https://github.com/rbenv/rbenv#using-package-managers
[07:22] <CoolerX> after installation they tell us to run a doctor script to check if rbenv was installed properly curl -fsSL https://github.com/rbenv/rbenv-installer/raw/main/bin/rbenv-doctor | bash
[07:23] <CoolerX> this detects multiple rbenvs
[07:24] <CoolerX> https://bpa.st/CYMQ
[07:24] <CoolerX> I uninstalled it following the instructions given here https://github.com/rbenv/rbenv#uninstalling-rbenv
[07:24] <CoolerX> doing a sudo apt-get purge rbenv
[07:25] <CoolerX> and then reinstalled
[07:25] <CoolerX> it's still saying there are multiple
[07:25] <CoolerX> why does apt-get install into both /bin/rbenv and /usr/bin/rbenv ?
[07:27] <enyc> CoolerX: thats' not down to apt-get, thats' down to usrmerge
[07:28] <enyc> CoolerX: https://wiki.debian.org/UsrMerge
[07:28] <toddc> rbenv 1.1.2 is the latest you can specify the version by apt install rbenv-1.1.2   someone else can enlighten us on the locations but that may be normal
[07:28] <enyc> CoolerX: or better  https://www.freedesktop.org/wiki/Software/systemd/TheCaseForTheUsrMerge/
[07:28] <enyc> point being  probably /bin is link to /usr etc.
[07:29] <toddc> correct but to verify that is above my skill set
[07:29] <geirha> ls -ld /usr/bin /bin
[07:34] <geirha> the weird doctor script could've detected that by testing the found executables with [['s -ef operator. [[ /usr/bin/rbenv -ef /bin/rbenv ]] would be true, which means it's the same file
[07:38] <CoolerX> enyc, no
[07:38] <CoolerX> I checked with ls -l
[07:39] <CoolerX> lrwxrwxrwx 1 root root 26 Jul 23  2018 /bin/rbenv -> ../lib/rbenv/libexec/rbenv
[07:40] <CoolerX> lrwxrwxrwx 1 root root 26 Jul 23  2018 /usr/bin/rbenv -> ../lib/rbenv/libexec/rbenv
[07:40] <CoolerX> they link to different executables
[07:40] <geirha> CoolerX: ls -ld /usr/bin /bin
[07:40] <CoolerX> https://bpa.st/JXDA
[07:41] <geirha> so /bin is symlinked to /usr/bin, that means /bin/rbenv is the same as /usr/bin/rbenv, which is /usr/lib/rbenv/libexec/rbenv
[07:41] <CoolerX> hmm
[07:42] <geirha> also see   readlink -e /bin/rbenv   and   readlink -e /usr/bin/rbenv
[07:42] <CoolerX> I don't think so
[07:42] <CoolerX> https://bpa.st/W7UA
[07:43] <CoolerX> those are separate files
[07:44] <geirha> but both of them point to the one under /usr/lib
[07:44] <CoolerX> geirha, hmm https://bpa.st/LHBA
[07:44] <CoolerX> geirha, /lib is symlinked too?
[07:45] <geirha> ah yes, that too
[07:46] <CoolerX> why does ls -l list it as separate files?
[07:46] <CoolerX> https://bpa.st/W7UA
[07:46] <enyc> geirha: all these years i did'nt know about readlink!
[07:48] <geirha> add -i to ls to see the inode number
[07:52] <CoolerX> geirha, hmm https://bpa.st/6FHQ
[07:53] <CoolerX> the inode number is the same
[07:53] <CoolerX> so if it is the same exact file why doesn't it show the l flag in the mode ?
[07:53] <CoolerX> lrwxrwxrwx
[07:53] <geirha> yes, which means they're the same file. If they were copies, they would've had different inode numbers
[07:53] <CoolerX> l for link right?
[07:54] <CoolerX> it should have also showed it in blue
[07:54] <geirha> CoolerX: because the file itself is not a symlink, but a path component further up the chain is, Mainly /lib is the symlink to /usr/lib
[07:54] <CoolerX> with the arrow pointing to the actual location
[07:54] <CoolerX> geirha, hmm
[07:54] <CoolerX> that's confusing
[07:55] <Sayona> Hi, is a way to create a shortcut  to change keyword language in ubuntu 20.04?
[07:55] <geirha> that's where readlink -e comes in. It goes through all path components and follows the symlinks
[07:55] <CoolerX> if you symlink a directory everything inside it should also be a symlink
[07:59] <geirha> that's not how it works
[08:04] <geirha> Sayona: There should be checkmark under Settings -> Keyboard to include a keyboard layout changer on the top panel
[08:11] <Sayona> geirha, thanks
[08:11] <CoolerX> geirha, how does it work?
[08:11] <CoolerX> if a folder is a symlink to a different folder
[08:12] <CoolerX> what about the stuff inside the folder?
[08:12] <CoolerX> what about the stuff inside the symlinked folder?
[08:13] <EriC^^> CoolerX: a symlink is just something that points to the actual dir, like it has /path/to/actualdir and you go there when you access it
[08:14] <geirha> to figure out the "real" path to the file, you have to check each directory compononent individually, which is what the readlink -e command does for you
[08:14] <CoolerX> hmm
[08:15] <CoolerX> I find it odd that ls doesn't say anything when accessing something inside a symlinked folder
[08:15] <CoolerX> but it does say something if you ls the symlink itself
[08:16] <CoolerX> that creates the situation where you have to ls all the parent folders of the file to know whether your path is real or just a symlink
[08:16] <CoolerX> or use readlink I guess
[08:16] <CoolerX> readlink -e /lib/foo/bar
[08:19] <geirha> in most cases you're only interested in the file, not whether there are symlinks along the way
[08:47] <mrvdv> I can send email fine from my server to say eg. *@gmail.com but i can't get any email to my server from *@gmail.com.. i get a email back saying 550 Unrouteable address.. is there something i can change for this /
[08:50] <illuminated> sounds like the username you're sending to from your gmail account doesn't exist on your server...
[08:50] <mrvdv> it does exist as its going through VESTA
[08:51] <mrvdv> i got my VPS provider to fix it but i would like help so i can fix it myself with some members on here
[09:09] <doug16k> I just ran tune2fs -l /dev/nvme0n1p2 and it says: "Last checked: Sat Dec 31 19:56:00 2016" and "Maximum mount count: -1" and "Check interval: 0 (<none>)"
[09:09] <doug16k> do I really have to set that up?
[09:10] <doug16k> that means I haven't fsck my / partition for years, right?
[09:13] <doug16k> Lifetime writes:          60 TB
[09:14] <doug16k> 740TB to go
[10:42] <fling> Hello.
[10:43] <fling> How to have zfs 2.0 on a longterm ubuntu?
[10:44] <cbreak_> you could use dkms or so
[10:44] <fling> will apt auto upgrade the module on kernel upgrade?
[10:45] <fling> also is not zfs 2 available from backports?
[10:48] <Guest88> I am learning to create a private apt repository server and then install the packages to ubuntu client machines from that server.. Can someone point me to documentation link on this
[10:49] <fling> which package provides zfs module? zfs-initramfs ?
[10:51] <tango_uniform_xr> would it work to use apt pinning and install the zfs-dkms package from future release?
[10:52] <fling> tango_uniform_xr: idk? would it?
[10:53] <fling> tango_uniform_xr: will not apt upgrade break it?
[10:53] <tango_uniform_xr> i'm just throwing some ideas, you would also need zfsutils 2.0 i believe
[10:55] <tango_uniform_xr> just be aware that tracking packages from future releases won't be supported, and can cause some issues
[10:57] <fling> # apt-file find /lib/modules/5.8.0-55-generic/kernel/zfs/zfs.ko
[10:57] <fling> linux-modules-5.8.0-55-generic: /lib/modules/5.8.0-55-generic/kernel/zfs/zfs.ko
[10:59] <fling> tango_uniform_xr: can't I get the module and userspace from backports somehow?
[11:00] <tango_uniform_xr> i don't see zfsutils-linux in the backports
[11:02] <MiguelAngel> Hi everyone, I just found a bug on Ubuntu Server installer, how should I report it? Or help with the code to Merge Request?
[11:03] <tango_uniform_xr> trying to install zfs-dkms zfsutils-linux from hirsute wants to upgrade libc6 and other stuff, so now that i see that i don't recommend it
[11:03] <guiverc> MiguelAngel, `ubuntu-bug subiquity`  (you didn't say release; but you file using package name which I've assumed is subiquity)
[11:04] <MiguelAngel> It's on the newest version of the package from stable, 20.04.02 as I remember
[11:04] <guiverc> MiguelAngel, https://help.ubuntu.com/community/ReportingBugs
[11:04] <MiguelAngel> Thank you very much guiverc
[11:05] <MiguelAngel> So here we can't do merge requets with solving code? It's all done by you?
[11:05] <guiverc> MiguelAngel, an alternative is always upstream (https://github.com/canonical/subiquity) - it's a Canonical product so it'll be same people
[11:06] <MiguelAngel> Perfect, I'll check it and will try to do the MR. Thank you very much! Appreciate it!
[11:39] <fling> which media to use for a vm install?
[11:40] <rbasak> fling: use a cloud image, not an "installer"
[11:40] <fling> I want one for headless with matching in-kernel to userspace versions
[11:40] <fling> zfs-0.8.3-1ubuntu12.6
[11:40] <fling> zfs-kmod-0.8.4-1ubuntu11.1
[11:40] <fling> this is not good ^
[11:40] <fling> rbasak: where to get the cloud image from?
[11:42] <rbasak> fling: hopefully your vm platform has an integrated solution. multipass, lxd, vagrant all have something built in
[11:42] <rbasak> Raw images are avialable from http://cloud-images.ubuntu.com/releases/
[11:42] <rbasak> But they need to be configured for non-cloud platforms like a simple VM.
[11:43] <fling> rbasak: I have to use an image with zfs on root
[11:43] <kraiskil> Just updated to 20.04. The program 'i' disappeared - would anyone remember the package name? Googling for "ubuntu i" doesn't really work.
[11:43] <kraiskil> (apt install i, or apt search i don't help either)
[11:43] <tango_uniform_xr> fling the only installers that do zfs on root are the desktop ones, if you're doing another setup you need to do manually
[11:45] <fling> tango_uniform_xr: ok, can you send me a link to such installer?
[11:48] <fling> The one I'm running has non-matching versions -> zfs-0.8.3-1ubuntu12.6 <-> zfs-kmod-0.8.4-1ubuntu11.1
[11:48] <tango_uniform_xr> https://ubuntu.com/download/desktop
[11:48] <tango_uniform_xr> i think it only does desktop installs
[11:50] <fling> desktop install is fine I can just drop the packages later
[11:50] <fling> where to get older 20.04 images?
[11:51] <rbasak> fling: just because the package version strings don't match doesn't mean they're incompatible. The .6 and .1 to me indicate that they've been extensively patched.
[11:51] <fling> rbasak: kernel module and userspace are 5 months apart and zfs has no stable abi
[11:52] <fling> rbasak: and this mismatches caused pools to get silent corruptions all the time
[11:52] <rbasak> fling: maybe they've been patched to be compatible. Have you checked?
[11:52] <fling> rbasak: which is especially bad for versions prior zfs 2
[11:52] <rbasak> Stable Ubuntu LTS releases support multiple kernel versions at once.
[11:52] <rbasak> So there might have to be things going on that go beyond the package version string you see.
[11:53] <rbasak> You can't draw any conclusion unless you actually look at the contents.
[11:53] <fling> rbasak: these are not package versions, this is what `zfs --version` returns
[11:54] <rbasak> Looks to me that what you've pasted are package versions. If you're getting that from --version, it's probably been arranged that way during a package build.
[11:54] <tango_uniform_xr> fling i'm on focal and for me, `zfs` is not a real package
[11:54] <rbasak> Like I say, check the contents.
[11:54] <rbasak> No point arguing about labels.
[11:54] <fling> tango_uniform_xr: they are not package. kernel module comes from linux-modules package or something
[11:55] <fling> tango_uniform_xr: and userspace from zfsutils or something
[11:55] <fling> I don't remember the actual names
[11:55] <fling> rbasak: how to check the contents? What am I looking for there?
[11:56] <fling> stable abi was planned for zfs 1.0 which has never been released and this is 0.8
[11:57] <rbasak> fling: grab the sources, apply quilt patches, and inspect the results. What you're looking for? That's up to you. You're complaining about an incompatibilty. Presumably you'll be able to demonstrate that one exists in the sources.
[11:57] <fling> rbasak: are you saying I should look for versions in the sources?
[11:58] <rbasak> fling: no, that'd still be hypothetical.
[11:58] <fling> rbasak: I'm sorry
[12:01] <mrvdv|znc> haha be sexy
[12:01] <mrvdv|znc> :D
[12:01] <jailbreak> !ot | mrvdv|znc
[12:02] <TJ-> !info iprint | kraiskil
[12:02] <TJ-> !info iprint | kraiskil 18.04
[12:02] <TJ-> !info iprint | kraiskil xenial
[12:02] <TJ-> !info iprint xenial | kraiskil
[12:03] <TJ-> oh!!
[12:03] <TJ-> !info iprint bionic | kraiskil
[12:03] <jailbreak> nice spamming of the bot, TJ-.
[12:03] <TJ-> jailbreak: I got there eventually!
[12:03] <jailbreak> xD
[12:03] <jailbreak> fling, you want 20.04.2 images?
[12:03] <TJ-> kraiskil: I found it in a 18.04 chroot with " apt-file search -x '.*bin/i$' "
[12:04] <fling> jailbreak: I got 20.04.2 image, looking for older ones, hoping to find one with matching zfs versions
[12:04] <fling> rbasak: you get silent hypothetical pool corruptions this way :>
[12:05] <kraiskil> TJ-, thanks. Guess its gone then from 20.04 :(
[12:05] <fling> rbasak: non-matching versions are incompatible by the default uless stated otherwise
[12:05] <fling> rbasak: the stable abi is still have to be implemented for this to be the case
[12:05] <tango_uniform_xr> fling where are you seeing mismatched versions? i am on 20.04 and i see the same for my kernel module and zfsutils-linux
[12:06] <TJ-> kraiskil: see https://launchpad.net/ubuntu/+source/iprint/+changelog
[12:06] <fling> tango_uniform_xr: `zfs --version` on both livecd and on installed system
[12:06] <TJ-> kraiskil: note "Deleted in cosmic-release (Reason: (From Debian) ROM; dead, trivial, many alternatives; Debi...) "
[12:06] <tango_uniform_xr> zfs-0.8.3-1ubuntu12.9
[12:06] <tango_uniform_xr> zfs-kmod-0.8.3-1ubuntu12.7
[12:07] <fling> tango_uniform_xr: this is what I would like to have
[12:07] <tango_uniform_xr> did you update the system?
[12:07] <fling> tango_uniform_xr: no, have livecd running now
[12:08] <rbasak> fling: I'm just saying that the code you're running doesn't necessarily match what you think you see in the version numbers. Stable distributions patch to fix things and the version number just represents the patch base, not the final result. I don't know what the situation is with zfs. I'm just saying that the version number is the incorrect label to use for comparison in any stable distribution.
[12:09] <fling> rbasak: ok
[12:09] <rbasak> A simpler example is when distros patch for security issues, and users come along and tell us that we're vulnerable because we haven't updated to the latest upstream release. It's the same point.
[12:09] <fling> rbasak: also the kernel version is 5.8 but 0.8.4 is only compatible with up to 5.6
[12:10] <tango_uniform_xr> fling i've had no issues making zfs pools from the ubuntu installer, as well as my own system
[12:10] <tango_uniform_xr> one i made from installer using "zpool create" is still running several months later
[12:11] <fling> tango_uniform_xr: you can't detect the silent corruptions you get from the mismatching versions
[12:11] <fling> tango_uniform_xr: there is no tool for this
[12:11] <fling> tango_uniform_xr: they will not appear like checksum errors in zpool status
[12:12] <fling> tango_uniform_xr: silent corruptions usually cause file size to not match the actual data length you read from the files, zero padding, zeroed files, random crashed, unreadable objects in snapshots and unimportable pools
[12:13] <fling> tango_uniform_xr: the issues don't always hit you right away and you still need an external tool for comparing checksums if you can't notice it in general use
[12:13] <tango_uniform_xr> i'm not sure, but for what it's worth the pool i said i made was a bootable one, it's still booting and everything checks out
[12:14] <fling> ignorance is bliss
[12:14] <kraiskil> TJ-, lol. "many alternatives" - guess it is time to learn new tools then :D
[12:15] <rbasak> Maybe there's something going on here that you're missing, and everyone else is actually OK and there aren't widespread reports of unexplained corruption because users generally don't have a broken setup like you think they do?
[12:18] <tango_uniform_xr> the zfs-dkms package isn't used by default
[12:19] <tango_uniform_xr> sorry, i got something confused
[12:19] <fling> rbasak: corruptions are expected but not reported a lot because noone is comparing checksums
[12:26] <fling> tango_uniform_xr: do you have the livecd image you used for the install?
[12:35] <tango_uniform_xr> fling i'm about to look at the desktop installer right now
[12:40] <tango_uniform_xr> fling upon further inspection, it seems that the desktop installer is using the HWE kernel, and that is why you see the different versions there
[12:41] <fling> tango_uniform_xr: ok, looks like I will have to perform the backup plan and to somehow install the older kernel version and then recreate pools from scratch
[12:42] <fling> Also where should I bugreport this to?
[12:42] <tango_uniform_xr> fling something could go like, install to VM > install zfsutils on VM > install the desired system from your VM, but you will have to go through manual install like https://openzfs.github.io/openzfs-docs/Getting%20Started/Ubuntu/Ubuntu%2020.04%20Root%20on%20ZFS.html
[12:44] <fling> tango_uniform_xr: it will probably be simplier to just fix the system I have by matching versions and recreating the pools
[12:47] <tango_uniform_xr> fling looks like you can report bugs here https://github.com/canonical/ubuntu-desktop-installer/issues
[12:47] <fling> thanks
[12:48] <tango_uniform_xr> fling but really it would be against the HWE kernel if you think they shouldn't use a different zfs module version
[13:01] <sweb> how tel squid to use another proxy(Socks5) ?
[13:05] <jailbreak> sweb, I'd like to know that, too.
[13:12] <TJ-> tango_uniform_xr: I'd think it more likely that the zfs utils package should be part of HWE to match the HWE kernel
[13:13] <tango_uniform_xr> TJ- maybe that too
[13:13] <TJ-> tango_uniform_xr: that's what HWE is for, and why kernel and Xorg libraries are part of it
[13:14] <tango_uniform_xr> yeah i know about HWE, that's how i recognized it was using HWE kernel with just a uname -r :p
[13:15] <sweb> ok, in maas how to restart squid process, seems it's not controlled by systemd.
[13:17] <leftyfb> sweb: you might be better off asking in #maas
[13:26] <wedr> Hi, need help with fixing this weird artifact issue. https://i.imgur.com/QMsBn9i.png     I have to copy/paste text inside a text editor in order to read the messages.
[14:03] <Franciman> hi all
[14:03] <Franciman> does anybody use microsoft teams on ubuntu?
[14:04] <leftyfb> Franciman: you'll need to contact Microsoft for support with their application
[14:04] <Franciman> I don't need support
[14:04] <Franciman> I just want to know other users experiences
[14:05] <BluesKaj> Hi all
[14:05] <Franciman> in gnome, can you do screen sharing?
[14:05] <Franciman> gnome over wayland
[14:27] <signofzeta> Yes, I use Teams on Linux.  the whole point of an Electron app is to ensure feature parity across platforms, but they somehow failed at that.  the windows and macOS versions are ahead of us in features, but what is there does work perfectly.  i haven't tried screen sharing, though.
[14:28] <signofzeta> but I've been shared at, and it's worked.
[14:28] <Xavierdarkness>  
[14:28] <Franciman> signofzeta: do you use gnome on wayland?
[14:28] <signofzeta> yes
[14:28] <Franciman> thanks
[14:28] <signofzeta> I'm on Ubuntu 20.10, where Wayland is the default
[14:29] <Franciman> I have no problem with anything, too
[14:29] <Franciman> but I can't screen share
[14:30] <signofzeta> I haven't tried that, sadly, and I'm not at home today to test.
[14:31] <rapha> if "netplan apply" brings up a static interface, but does not give it the IP that's configured in the YAML, what's still missing?
[14:38] <leftyfb> rapha: you still have dhcp enabled in that same yaml or a different one?
[14:44] <rapha> leftyfb: eno4 has dhcp: true, and it must be so. eno3 has dhcp: false and a static ip, and is working properly. the if in question is eno1, which has dhcp: false and a static ip (and gw & ns set explicitly) but is not getting configured by netplan apply.
[14:44] <rapha> fwiw i now found a thread on askubuntu.com that claims a reboot can sometimes be the only way ... so began a reboot, but it usually takes 10-15 minutes (datacenter hardware)
[14:47] <locsmif> Does anybody know why mysqlrouter's AppArmor profile is so broken?
[14:53] <locsmif> It just won't start under Hirsute
[14:53] <locsmif> Until AppArmor is completely disabled, then it works just fine.
[14:54] <locsmif> Honestly, after seeing everything broken by AppArmor, I absolutely despise it by now
[14:56] <locsmif> The amount of time bug reporting and tweaking AppArmor profiles.. I no longer think it's worth it, tbh
[15:12] <catphish> how might i run a script on a terminal (instead of login) at boot?
[15:13] <rapha> leftyfb: rebooting definitely makes it work.
[15:14] <leftyfb> catphish: https://www.howtogeek.com/687970/how-to-run-a-linux-program-at-startup-with-systemd/
[15:15] <catphish> leftyfb: interesting i'm on that page, i'm not sure if that's the right answer, at least i'm not sure how to give it a terminal, but i think i can make that do what i want
[15:15] <leftyfb> catphish: "give is a terminal"?
[15:16] <catphish> i want to run a process in a tty instead of running login
[15:16] <catphish> so that it displays on the screen at boot
[15:16] <leftyfb> catphish: what is this for?
[15:17] <catphish> leftyfb: specifically, i'm trying to build a computer that boots, runs a speed test, and displays the results
[15:18] <leftyfb> catphish: then run this in a GUI window the pops up at boot after auto-logging in
[15:18] <catphish> i wasn't planning to tun a GUI, would prefer just to use a tty
[15:19] <leftyfb> catphish: said tty would need to be auto-logged in. Which is a horrible idea
[15:20] <catphish> why?
[15:29] <leftyfb> catphish: this might help you:   exec 1>> $LOG < /dev/tty2 > /dev/tty2 2>&1
[15:30] <leftyfb> catphish: $LOG being the output of whatever you want to be displayed on the terminal
[15:31] <catphish> leftyfb: thanks, i've just been looking at this approach: https://bbs.archlinux.org/viewtopic.php?id=129509
[15:31] <catphish> (running it through systemd,which has an option for tty apparently)
[15:49] <catphish> got it working, i needed to do 2 things 1) create a systemd service with StandardInput=tty StandardOutput=tty TTYPath=/dev/tty1 2) reconfigure logind not to spawn on any ttys
[15:51] <leftyfb> catphish: why that last part? Couldn't you just add a clear ; speedtest-cli to your command?
[15:52] <catphish> leftyfb: it doesn't appear to work at all when login is running, i'm not sure why not
[15:52] <leftyfb> ah, because it's probably trying to bind to it as opposed to just displaying to it
[15:52] <catphish> i suspect that's the case
[15:52] <fling> Is there a packagelist for `apt-mark manual` for a minimal system?
[15:53] <fling> I want to remove all the desktop stuff and keep it really minimal
[15:53] <catphish> in any case, this seems to me to be the "correct" way to do it, without auto-login or console redirection, i'm happy anyway :)
[15:59] <fling> tango_uniform_xr: also how to replace this HWE kernel with the regular one properly?
[16:03] <tango_uniform_xr> fling are you trying to edit ubuntu installer?
[16:05] <fling> tango_uniform_xr: nah, shrinking the system, going to recreate the pool and put the system back
[16:06] <tango_uniform_xr> you can install the desired kernel, remove hwe one, update grub config with `update-grub`
[16:09] <fling> tango_uniform_xr: what is the package name?
[16:10] <tango_uniform_xr> fling `linux-image-generic-hwe` you can install the new kernel (do that first unless it's not booted) then you can install `linux-image-generic-hwe`
[16:11] <fling> tango_uniform_xr: I have linux-generic-hwe-20.04
[16:11] <tango_uniform_xr> sorry, `linux-image-generic` is what i meant to put at the end there
[16:11] <fling> ok
[16:13] <Tuor> Hi, is it possible to install java 7 or even java 6 on Ubuntu 20.04 (even if it's only a tar.gz extracted in /opt )?
[16:13] <fling> tango_uniform_xr: how to also update initramfs?
[16:14] <tango_uniform_xr> fling `update-initramfs -u` to update latest one or `update-initramfs -ukall` to remake all of them
[16:15] <fling> tango_uniform_xr: update-initramfs: Generating /boot/initrd.img-5.8.0-55-generic
[16:15] <fling> tango_uniform_xr: but 5.8.0 is hwe which I uninstalled
[16:15] <fling> should I remove them by hand first?
[16:16] <fling> rm worked :>
[16:18] <fling> what is open-vm-tools and open-vm-tools-desktop doing?
[16:26] <fling> Why are there libs in `apt-mark showmanual` list?
[16:40] <fling> >> Adding boot menu entry for UEFI Firmware Settings
[16:41] <fling> which command for this? ^
[16:42] <jeremy31> fling: I think it is part of update-grub
[16:42] <rcr> how would i list the apt pacakges available for a single `/etc/apt/sources.list.d/foo.list` ?
[16:43] <fling> jeremy31: thanks :>
[16:45] <fling> how to prevent linux-image-generic from updating?
[16:47] <fling> or linux-modules specifically?
[16:50] <jeremy31> fling: why block linux-modules?
[17:00] <mra90> how to disble icons cascade
[17:01] <mra90> I want them do be next to each other even if they are differne instances of the same program like a terminal for example
[17:01] <lotuspsychje> mra90: you can tweak your system with dconf-editor mostly
[17:02] <lotuspsychje> like the way icons behave when you click on them
[17:03] <lotuspsychje> mra90: but not sure, the option you are looking for works on gnomes dock
[17:04] <mra90> I don;t see such an option anywhere
[17:04] <mra90> very anoying
[17:04] <lotuspsychje> mra90: maybe docky can do it
[17:05] <fling> jeremy31: because there is this 'zsys' package which is taking auto snapshots
[17:05] <fling> jeremy31: and I want to prevent it from running zfs commands on userspace upgrades
[17:07] <lotuspsychje> mra90: check org/gnome/shell/extensions/dash-to-dock/click-action
[17:11] <mra90> ok thanks
[17:17] <pi0> does one need to specify cores in ubuntu
[17:17] <pi0> i mean i installed it
[17:19] <leftyfb> pi0: huh?
[17:23] <pi0> i have a 6 core cpu
[17:23] <pi0> on my laptop
[17:23] <pi0> using blender
[17:24] <pi0> not sure if i have to specifiy the cpu for blender to take advantage of my cores
[17:25] <leftyfb> pi0: https://blender.stackexchange.com/a/87035
[17:32] <pi0> hmm
[17:44] <nyuszika7h> how do I get Chrome to detect system dark mode, so websites using `prefers-color-scheme: dark` styles will display in dark mode? it works fine on Windows 10, but won't detect system dark mode on Ubuntu 21.04
[17:52] <bewees> can someone who doesn't run kde can try to install `apt install kde-plasma-desktop|grep gnome` and paste that for me
[17:57] <francis> bewees: on focal that gets a solitary "pulseaudio-module-bluetooth | bluez-alsa orion-gtk-theme tk | wish *gnome*"
[18:48] <GQCGPKnmgGGJ> LuRIng Fr3EnoD3 usErS herE i5 impuDEnt tHEfT
[18:52] <fling> which dir to mount efi partition before updating grub?
[18:52] <bewees> francis: thanks
[18:53] <fling> is it /boot/efi?
[18:58] <Ky0sh1r0> hi, i wrecked my gnome-keyring, which isn't that a big thing. but, when starting over with an empty one, I cannot connect a google account via gnome online accounts, it stalls at loading "accounts.google.com". I also found a hint saying the same problem appears on epiphany, which is true on my system 21.04. Is this a new problem on 21.04? It worked since 16.04 until 20.10 so far I could remember. (as posted on freenode for a
[18:58] <Ky0sh1r0> moment)
[18:59] <fling> hmm I'm running update-grub but on reboot I still see the old kernel entry there
[19:02] <Guest26> I am making my own version of Ubuntu, called Moobuntu, and I will need work on it.
[19:02] <Guest26> My preview of Moobuntu: https://jpcode05.github.io/moobuntu/
[19:07] <DODO11> salut
[19:07] <DODO11> Lantizia
[19:08] <DODO11>  vous parlez le français
[19:10] <Lantizia> non
[19:24] <Guest76> how many people actually use irc?
[19:24] <oerheks> Guest76, how is that related to ubuntu support?
[19:24] <Mekaneck> !ot | Guest76
[19:26] <sarnold> Guest76: it's hard to know for sure but hundreds of thousands of people is likely https://netsplit.de/
[19:26] <oerheks> i would expect 'how many use ubuntu' :-D
[19:27] <oerheks> also those stats are not available.
[19:27] <fling> got ubuntu booting yay!
[20:05] <dj> quit
[20:37] <ice9> ubuntu 21.04 randomly crash when left without activity and i have to hard reset the laptop; any idea?
[20:40] <illuminated> are you hitting an OOM situation, you think?
[20:40] <ice9> illuminated, not at all, i have only the browser opened with 5 tabs and the device is left like that
[20:41] <illuminated> nfi then, sorry
[20:41] <Alabalistic> ice9: how old is the laptop and last deep cleaning?
[20:42] <ice9> Alabalistic, it's a brand new laptop with fresh installation of ubuntu
[20:43] <Alabalistic> display goes off and don't weakup when you open it
[20:44] <ice9> no, the display is on but it freeze on the BIOS wallpaper
[20:45] <Alabalistic> uname -r
[20:45] <Alabalistic> what is the output
[20:45] <ice9> 5.11.0-22-generic
[20:45] <Alabalistic> and let's see if your kernel supports the great hardware you have
[20:46] <Alabalistic> what is your video card
[20:46] <Alabalistic> sudo lshw  # will list all the specs
[20:46] <ice9> intel and nvidia
[20:46] <leftyfb> ice9: try disabling suspend/hibernate and any power saving features
[20:47] <bewees> can someone who doesn't use kde-plasma or gnome in ubuntu please paste all of `apt install kde-plasma-desktop` :-)
[20:47] <Alabalistic> I use Kubuntu with Plasma so can't realy enter in gnome
[20:47] <Alabalistic> bewees:  no idea what do you mean
[20:48] <Alabalistic> nvidia driver is the main suspect for ice9
[20:48] <bewees> Alabalistic: can you check if you have this package installed: gnome-control-center or gnome-menus
[20:49] <bewees> `dpkg -l|grep gnome-control-center`
[20:49] <Alabalistic> gnome-control-center-data/hirsute,hirsute 1:3.38.5-1ubuntu1 all
[20:49] <Alabalistic> gnome-control-center-dev/hirsute,hirsute 1:3.38.5-1ubuntu1 all
[20:49] <Alabalistic> gnome-control-center-faces/hirsute,hirsute 1:3.38.5-1ubuntu1 all
[20:49] <Alabalistic> gnome-control-center/hirsute 1:3.38.5-1ubuntu1 amd64
[20:49] <bewees> ok, so not installed
[20:50] <Alabalistic> bewees:  the output of your command gives null
[20:50] <bewees> for some reason my ubuntu-server tries to install those gnome packages when i try to install kde :(
[20:50] <Alabalistic> I did apt list. | grep but was wrong
[20:51] <Alabalistic> i can check my jump box there is kubuntu on it
[20:53] <Alabalistic> https://termbin.com/daf8
[20:53] <Alabalistic> this is | grep gnome bewees
[20:58] <bewees> thanks Alabalistic, you've got those packages also installed, why?
[20:59] <Alabalistic> it's a kubuntu minimal install with ssh and teamviewer noting else
[20:59] <bewees> according to apt-rdepends it kde-plasma shouldn't be dependent on those packages https://dpaste.org/T80S
[20:59] <bewees> really weird, i wonder what i'm missing
[21:36] <robertglick[m]> Hello
[21:52] <jamesp> Over in #ubuntu-offtopic, I am discussing plans for my first Linux Distro, Moobuntu.
[21:53] <goddard> where does gnome shell save the openvpn settings?
[21:54] <Alabalistic_> try  /etc
[21:54] <Alabalistic_> by default all settings are in /etc
[21:55] <Alabalistic_> otherwise do ls -la in your home directory and see the hidden folders, starting with dot
[22:07] <goddard> not finding anything
[22:08] <sarnold> goddard: /etc/NetworkManager/system-connections/ perhaps?
[22:10] <goddard> i dont know if gnome devs know this but how the hell can you remove a file chooser text file
[22:10] <goddard> i literally have no way to make the field empty
[22:11] <matsaman> oh yeah they've really been making those fields worse
[22:11] <sarnold> can you rephrase the question?
[22:12] <goddard> well enter a file in the file chooser filed for a key or cert in vpn
[22:12] <goddard> you cant remove it
[22:16] <goddard> then their password prompts are a modal that takes over the whole screen
[22:17] <goddard> what if my password is saved in a password manager?
[22:17] <goddard> i have to know its gonna ask me for a password so i  can copy it
[22:17] <goddard> crazy dumb
[22:18] <goddard> i like gnome but these are a little too much
[22:20] <Alabalistic_> plasma baby
[22:20] <Alabalistic_> love it
[22:20] <goddard> yeah plasma is great
[22:22] <goddard> really both are awesome
[22:22] <goddard> plasma needs to refine more and gnome needs to refine less
[22:22] <goddard> but its cool in the power section of settings
[22:23] <goddard> looks like it has new power options
[22:44] <webchat44> This is infuriating. Systemd keeps taking over more and more shit. I don't have the NTP commands to sync time anymore, clock is wrong, and systemd claims there;s an NTP service going
[22:44] <webchat44> what the fuck.. how do I just sync the fucking time
[22:44] <webchat44> the man page has all sorts of options about how to show that systemd has taken over, but I don't see a sync option.
[22:44] <webchat44> it enables and disables NTP, but it says its enabled, and it clearly aint synced
[22:45] <webchat44> $ timedatectl
[22:45] <webchat44>                Local time: Tue 2021-06-22 18:40:47 EDT
[22:45] <webchat44>            Universal time: Tue 2021-06-22 22:40:47 UTC
[22:45] <webchat44>                  RTC time: Tue 2021-06-22 22:40:47
[22:45] <webchat44>                 Time zone: America/New_York (EDT, -0400)
[22:45] <webchat44> System clock synchronized: no
[22:46] <jamesp> this is just one of the reasons so many people hate systemd
[22:50] <webchat45> Let us try this again. How do I sync the time? See this manpage? https://manpages.ubuntu.com/manpages/bionic/man1/timedatectl.1.html It literally does not say how to force an NTP synchronization. You can turn it on or off, but it says absolutely zip about forcing a sync. You know, people with multiboot frequently have clocks that are hours off...
[22:51] <oerheks> multiboot, windows probably does that, there are many pages howto force windows or ubuntu to keep the same clock setting
[22:51] <oerheks> what have you found?
[22:52] <webchat45> The fact that there is not a single manpage describing a command analogous to ntpd -gg. Just tell me what command to give so systemd does the proper thing.
[22:53] <webchat45> *ntpd -gq, excuse my typo. This is nuts. Why is this so hard? Why does systemd make all this 100x more obscure and convoluted?
[22:54] <webchat45> ServerFault answers say that systemd literally can't do that. https://serverfault.com/questions/948974/force-systemd-timesyncd-to-sync-time-with-ntp-server-immediately
[22:55] <oerheks> ... ntp is not standard installed. and no need for it, AFAIK
[22:55] <oerheks> and does that survive reboto?
[22:55] <oerheks> c/reboot
[22:55] <webchat45> NTP has been standard issue on every nix box I've used for decades
[22:55] <webchat45> It works great, when it's there, because it'll fix itself, keep time discipline, etc
[22:56] <oerheks> Ubuntu by default uses timedatectl / timesyncd to synchronize time, you are free to choose something else
[22:56] <webchat45> oerheks, also, if you think there's no need for NTP in the world, you're probably not qualified to even care why it exists or understand the various issues that arise in trying to keep a somewhat linear looking time scale in a universe with moving objects
[22:57] <sarnold> timedatectl stuff works fine for most folks, but feel free to install chrony or ntpd if you'd rather
[22:57] <webchat45> sarnold, I'm pissed they replaced something that worked great with something that lacks basic functionality and doesn;t actually work.
[22:57] <webchat45> I had enough trouble getting here, because I had to manually set my clock
[22:58] <oerheks> your actual issue was an other OS that hijaacks your time control. but sure, rant away
[22:58] <webchat45> oerheks, no, windows isn't allowed to modify the system clock.
[22:58] <webchat45> I turned my damn computer off for a few hours.
[22:58] <webchat45> That shouldn't cause massive headaches.
[22:58] <sarnold> replace the cmos battery?
[23:00] <leftyfb> webchat45: first off, settle down. second, did you try using timedatectl yet?
[23:00] <webchat45> Or maybe technology was invented decades ago to make that a moot point, and it's infuriating that we're regressing in functionality so badly
[23:00] <webchat45> leftyfb, yes, and I linked to the man page
[23:00] <webchat45> leftyfb, it literally cannot force a sync.
[23:01] <webchat45> Funny though, it claims "NTP service: active"
[23:02] <leftyfb> yes, an NTP service is active. It does not say "ntpd" is active
[23:02] <webchat45> Yes, and any reasonable NTP service comes with a force sync command.
[23:03] <webchat45> This is a completely trash implementation
[23:03] <leftyfb> webchat45: is this for a server or workstation?
[23:03] <webchat45> a workstation, and basically every cryptographic network protocol shits itself as soon as it realizes the time is wronfg
[23:03] <webchat45> so I had to manually set it just to get *here*
[23:04] <leftyfb> webchat45: please watch the language and attitude. We are trying to help and you are doing nothing but going off the rails on how much you hate the situation instead of working with us to resolve it
[23:04] <webchat45> That's the kind of thing that keeps anyone besides people with infinite patience or prior tech industry experience from wanting to use linux. Turn it on, and the first thing you get are SSL errors about the clock?
[23:05] <leftyfb> webchat45: if timedatectl does not meet your requirements, then disable it and install and setup ntpd
[23:05] <webchat45> leftyfb, fine where do I file feedback then?
[23:05] <webchat45> If the answer is , "we don't support that", i'd like to leave feedback
[23:05] <oerheks> to windows, that hijaacks the clock?
[23:05] <webchat45> I can always install, compile, code up something else myself. But WHY is this so hard? it's one of the most basic functions of a computer.
[23:06] <webchat45> It leaves a nasty taste in my mouth every time systemd turns up somewhere new
[23:06] <leftyfb> webchat45: https://github.com/systemd/systemd/issues
[23:07] <webchat45> no, to Canonical, for using such a crap implementation
[23:07] <webchat45> poettering is entirely about the EWONTFIX
[23:07] <leftyfb> webchat45: stop. Go install ntpd and be done with it. Then report the bug to the project itself, not to a distribution which includes it.
[23:08] <webchat45> leftyfb, No. I want to file feedback that I believe the project should choose differently
[23:08] <webchat45> Is that somehow too offensive for you?
[23:08] <leftyfb> webchat45: asking ubuntu remove systemd isn't going to happen
[23:08] <cbreak> why would you want to remove systemd?
[23:08] <webchat45> Did I say that?
[23:08] <cbreak> it's clearly superior to all alternatives
[23:08] <webchat45> You've suggested that there are other NTP services that work with systemd
[23:08] <leftyfb> cbreak: please stop, you're not helping
[23:08] <webchat45> Why can't we have one of those?
[23:08] <sarnold> webchat45: 'ubuntu-bug /usr/bin/timedatectl'
[23:09] <webchat45> sarnold, thank you.
[23:09] <oerheks> Change timedatectl / timesyncd to synchronize time back to ntpd ...
[23:09] <cbreak> systemd can start what ever thing you want
[23:09] <webchat45> Finally, someone who can answer a direct question rather than saying, "you're holding it wrong."
[23:09] <cbreak> it's quite flexible
[23:09] <oerheks> one could try ..
[23:10] <oerheks> i would start with debian then :-D
[23:28] <webchat45> there, filed.. now to go deal with my actual issue, instead of politics. JFC...
[23:28] <webchat45> By the way, attacking people who ask for help, by saying, "
[23:28] <webchat45> "No one needs it" is just being a complete ..well you know.