[01:51] <Donavan01> if I want to provide a file server via web access to users whats the best combo of programs/services are best practices for linux ... windows I know but no linux
[01:52] <matsaman> not something a sensible Unixy sysadmin would do, but
[01:52] <matsaman> you can get webUIs for, for example, vsftpd
[01:52] <retran> Donavan01, there's lots of "home cloud" software for that
[01:53] <retran> Donavan01, are you just wanting to provide a web-based client to files on a system? no prob. lots of stuff for that
[01:55] <Donavan01> im not looking for home setup eventually I hope to expose this thing to the world ...working on a business with a freind and one of the aspects is we need to have a file drop for people to upload things and download them if I were using windows it would be super easy for me but not so much linux.... should I effectively build a database and just let the database be accesable or is it better to have linux do it at the system level
[01:58] <retran> Donavan01, then you need help from experts
[01:59] <retran> Donavan01, not just random ppl on IRC
[01:59] <Donavan01> im sure and they often hang out in here
[01:59] <retran> Donavan01, but more help than can be feasibly provided over chat
[01:59] <retran> I'm such an expert, but it's unfeasible to do this over cheat
[01:59] <retran> chat
[02:00] <Donavan01> im not looking for a how to im just wanting to know which is the best route something like an ftp with ssl or something more database driven
[02:00] <retran> neither
[02:01] <Donavan01> I just dont know the names of things to start looking them up and looking up file server and linux gives you way too much
[02:01] <Donavan01> retran ok what is the best option here
[02:02] <retran> there's no one route, you have to spec out a software suite
[02:03] <Donavan01> suggestion for this ... this is what I mean I dont know the products in linux ive been stuck in windows land most of my career not for lack of trying on my part
[02:03] <retran> there's no "products" to do what you're talking about out of the box
[02:03] <retran> you have to assemble stuff together custom
[02:04] <retran> hire a programmer
[02:04] <retran> or learn to program
[02:06] <retran> Donavan01, maybe I'm misunderstanding
[02:06] <retran> Donavan01, I think you might be fine with a personal cloud
[02:14] <retran> Donavan01, what I would highly recommend, instead of doing this yourself, or anything custom, is to get a Google G-Suite account
[02:14] <retran> Donavan01, or office365
[02:14] <retran> Donavan01, they're SO CHEAP and good
[02:14] <retran> Donavan01, they have both email and file sharing and document sharing all comes with their basic account
[02:15] <Donavan01> no your not misunderstanding  I intend to make these machines of mine person clouds for my clients I want system level control over the files and a way to show it to them in a web format... I will level what ever langauge is needed (already got a bunch under my belt that are far less useful)   I just dont know what the linux world looks at as the best for things like scripting is java king, python, php ... is ftp on if im using SSL with it or is there
[02:15] <Donavan01> a better solution
[02:15] <mihael> How do I install traceroute on Ubuntu 18?
[02:15] <retran> Donavan01, sounds like you're the wrong person to be setting this up for clients
[02:15] <retran> Donavan01, it's not some trivial thing
[02:16] <Donavan01> I just cant afford the licensing for windows server so I need to the linux route I just dont know what is best over here and googling isnt giving me any clue
[02:16] <retran> Donavan01, it's more than just downloading (or buying) the right software
[02:16] <Donavan01> retran I have 25 years of system admin experience I think I can handle new technology
[02:16] <retran> Donavan01, you have to know how to configure it
[02:16] <retran> Donavan01, why are you in here asking these kind of basic questions like you don't have a clue then?
[02:16] <Donavan01> I just havnet done much in linus which is why I am asking
[02:17] <retran> Donavan01, from the type of questions you're asking, and the words you are using, I doubt you can handle it
[02:17] <retran> no offense
[02:17] <retran> you might be qualified for the business side
[02:17] <retran> just not actually doing it :)
[02:17] <semitones> You gotta start with gentoo
[02:17] <retran> semitones, lol!!!
[02:17] <semitones> If you can master that
[02:17] <semitones> You can do anything
[02:17] <retran> stahp that
[02:18] <derpadmin> semitones, evil
[02:18] <Donavan01> becasue what is true in windows land doesnt always hold true in linux world and I was trying to find out if what I thought was best practices might have been a glarring flaw because of the difference in the files systems for example
[02:18] <retran> Donavan01, this has nothing to do with Windows vs. Linux. you aren't getting it
[02:18] <retran> just stop saying "windows land" and stuff, for a start
[02:18] <retran> get that out of your head
[02:18] <retran> you keep naming off services and technologies
[02:19] <retran> but have yet to present a cohesive problem for any of us to actually solve
[02:19] <pavlos> mihael: does "sudo apt install traceroute" work?
[02:19] <semitones> OMV might be nice for what you're trying to use but it's also finicky
[02:20] <Donavan01> yeah it kinda does when im talking about file storage yes I understand the the web component isnt however last time I set up a linux web server was about 12 years so im not up on what the best programs still are is it samba ... I dont know which is why im asking a broad question and giving examples
[02:20] <pavlos> mihael: I dont have an 18.04 to test
[02:20] <semitones> Donavan01: check out the open media vault project - they have an awesome forum with a bunch of great guides too
[02:20] <derpadmin> Donavan01, just setup sftp in your ssh config, change the home dir of the users that will upload files,  and point your ngnix instance to these folders, adjust permissions accordingly, password protect if needed
[02:21] <derpadmin> *sftp chroot jails
[02:21] <derpadmin> maybe setgid on the folders, so the content will be created with proper group permissions
[02:22] <derpadmin> you see retran, wasn't that hard
[02:22] <mihael> pavlos: No it doesn't. "E: Package 'traceroute' has no installation candidate"
[02:23] <retran> derpadmin, lol
[02:24] <Donavan01> derpadmin, semitones ... thank you
[02:24] <retran> derpadmin, I feel for his poor customers
[02:25] <pavlos> mihael: I started a VM to test ... wait a bit
[02:25] <derpadmin> check this Donavan01 https://200013.net/doc/LINUX/ssh+sftp+sshfs/chroot_+_force_sftp.txt
[02:29] <pavlos> mihael: I was able to install on 18.04 server
[02:30] <pavlos> mihael: do you have universe repo enabled?
[02:32] <mihael> how do I enable that?
[02:32] <mihael> so it is disabled by default?
[02:32] <pavlos> mihael: You can enable it in the gui, let me find out
[02:33] <mihael> I enabled it via: `sudo add-apt-repository universe`
[02:34] <pavlos> ok,do a sudo apt update first then try to install traceroute
[02:38] <pavlos> it should install version 2.1.0
[02:41] <pavlos> https://imgur.com/a/qm3cifI
[02:51] <neteffect> hello im having trouble running fsck, can't get it to work during reboot with touch /forcefsck, won't work at command line, won't umount during reboot, parted wouldn't umount either.  How do I run a fsck?
[02:56] <neteffect> i had a power outage, is fsck the only thing for checking integrity?
[02:56] <pavlos> neteffect: some details, are you on 20.04?
[02:56] <neteffect> 18.04
[02:56] <neteffect> .5
[02:57] <pavlos> shutdown -rF now will reboot your system and run an fsck
[02:57] <neteffect> ok i'll try that ty
[03:00] <pavlos> take a look at your /etc/fstab, the last column is to check fs for the root / partition is should be 1 (not 0)
[03:00] <neteffect> oh i'll do that now
[03:01] <neteffect> it is 1
[03:01] <pavlos> example: UUID=778636d8-22cc-476f-b801-ff9f5ac9ef5d /               ext4    errors=remount-ro 0       1                    <------------------------------- see the one
[03:01] <neteffect> but i reboot and it's still 1
[03:01] <neteffect> yeah it says 1
[03:01] <pavlos> it will remain one
[03:01] <neteffect> oh
[03:02] <pavlos> but if you touch /forcefsck, after reboot and fsck, that filw will be gone
[03:02] <pavlos> file*
[03:02] <neteffect> i never see fsck during reboot
[03:05] <neteffect> its like it doesn't run
[03:05] <pavlos> I think you can pass -V for verbose output ...
[03:06] <pavlos> am trying to force my VM to do that ...
[03:06] <neteffect> i can't ever get it to run
[03:06] <neteffect> dunno  how to pass -V to touch /forcefsck
[03:08] <neteffect> i see gparted has a check, can't get that to work
[03:09] <pavlos> if you're running on sda1 you cannot unmount it. You have to run from another place to unmount sda1 and fsck
[03:09] <neteffect> k
[03:10] <pavlos> gparted will not let you unmount the running system
[03:10] <neteffect> oh
[03:10] <pavlos> can you boot off an live iso, then fsck your disk?
[03:11] <neteffect> that is pretty good idea
[03:11] <pavlos> live iso will be in ram (try ubuntu), your disk sda1 or sda2 can be fsck'd
[03:13] <neteffect> +
[03:13] <neteffect> ok
[03:25] <neteffect> pavlos, hey thank you, live cd was much easier
[03:26] <pavlos> glad it helped
[04:13] <bumblefuzz> so, I updated ubuntu today
[04:13] <bumblefuzz> and no I have no audio'
[04:13] <bumblefuzz> can anyone help?
[06:14] <DarkTrick> I removed tracker-miner-fs from /lib/systemd/system but it still gets called at startup. How can I make it stop working?
[06:17] <cucumber> hi there everyone
[06:17] <cucumber> whenever I install a new operating system there is a power cut like the whole city loses its power
[06:18] <cucumber> It feels like someone tries to inject the newly installed operating system with a malware that needs a forced system reboot
[06:18] <cucumber> how well protected is ubuntu againsts such attacks?
[06:18] <cucumber> and how can I detect those?
[06:18] <cucumber> I mean aside from the usual NSA and other kinds of backdoors (just kidding)
[06:19] <cucumber> :)
[06:19] <Maik> cucumber: you know this channel is for support only, this sounds more offtopic to me
[06:19] <cucumber> Maik
[06:19] <cucumber> Yes I do
[06:19] <jdgr> cucumber: Uh...
[06:19] <Maik> well then
[06:19] <cucumber> And I am really looking for ways to prevent tampering
[06:20] <jdgr> Have you tried doing the install while in a massive lead vault with the machine plugged into a power generator?
[06:20] <cucumber> I mean on the surface it might look like I be trolling, but I really am not
[06:20] <jdgr> All your packets need to be printed onto paper, analyzed and hand entered into the machine
[06:20] <cucumber> jdgr I don't have a power generator
[06:20] <Maik> take it offtopic please
[06:21] <cucumber> ok
[06:21] <cucumber> alright
[06:21] <cucumber> :(
[06:21] <Maik> thanks :)
[06:29] <cucumber> no Maik thank you for being as tolerant as possible I understand
[06:29] <cucumber> no worries
[06:29] <Maik> :)
[06:29] <cucumber> but generally speaking how can an os be protected against exploits?
[06:29] <cucumber> not just ubuntu, but any OS
[06:29] <lotuspsychje> !discuss | cucumber
[06:29] <lotuspsychje> not here cucumber
[06:29] <cucumber> ok thanks
[06:42] <nubonix> hey looking for an autohotkey alternative... i would like to write a program that i can type a short sequence of words...suchas selenium_boiler_plate and it would return text from a file that i have written, and this would be available to any application suchas gedit, any ideas?
[06:43] <nubonix> so basically an alias anywhere
[07:26] <jelly> on ubuntu 18.04, when I Alt-Space to get the window properties menu, how do i select a setting to change without mouse use?  Cursor keys work but neither Enter nor Space select
[07:26] <jelly> only left mouse click seems to work
[07:30] <guiverc> jelly, enter selects for me  (I'm not on 18.04/bionic though sorry)
[07:50] <jelly> probably got fixed
[07:51] <jelly> guiverc, which release are you on?
[07:53] <guiverc> groovy
[08:46] <echoSMILE> Is there any reason my Weather panel item stop working showing forecast data ?
[08:46] <echoSMILE> https://docs.xfce.org/panel-plugins/xfce4-weather-plugin at Ubuntu 19.10
[10:02] <thetoblin> I'm having some real trouble getting Cuda to work with Conda. When trying to use Cuda through python code, I get the error "libNVVM cannot be found".I have looked into every post and search result I can find on the topic, such as this one: stackoverflow.com/questions/…. But nothing I try seems to work.Does anyone have any idea what the cause is,
[10:02] <thetoblin> and how to fix it? I'm guessing it has to do with adding some export lines to .bashrc, but I can't seem to get them right
[10:02] <thetoblin> https://stackoverflow.com/questions/48385686/cuda-library-nvvm-not-found
[10:34] <thetoblin> does anyone have any insights into how to get cuda to work with conda? I get the error "libNVVM cannot be found", and no search result on the matter seems to be able to assist me
[11:15] <martti> hm. this is weird
[11:17] <martti> so I changed a dns entry, and restarted the supervisor services serving our api backend. thing is that the dns entries now should have different content, but they didn't.
[11:19] <Habbie> martti, then the caching must be somewhere else
[11:19] <martti> I even restarted systemd-resolved and it still didn't work. only restarting supervisor after restarting resolved did the trick, but now I wonder how systemd services appear to have a localized dns cache?
[11:20] <martti> which seems very surprising to me.
[11:24] <martti> hmm.
[11:24] <martti> maybe when I restarted resolved the new dns settings weren't actually updated on wherever my hosts get their dns
[11:28] <martti> according to https://github.com/systemd/systemd/issues/940 dig even circumvents nss/resolved completely, meaning that my attempts to debug the issue, too, may have been nonsensical.
[11:40] <emOne> ubuntu help out pls
[11:40] <emOne> I am trying to run a compile script on an exfat drive
[11:40] <emOne> the script fails due to failed symlinks on exfat
[11:40] <emOne> apparently exfat doesn't support symlinks
[11:41] <emOne> is it possible to simulate symlinks so I can run this compile script without finding or ordering another HDD just for this one task
[11:42] <martti> tmpfs should support symlinks. that means that you mount a ram disk to work on. if you have enough memory, harware-wise
[11:46] <lukegb> Out of curiosity: who decided to make downloading Ubuntu Server require 5 clicks from the homepage
[11:46] <lukegb> including two clicks on very small blue "Option 2" and "Option 3" links?
[11:59] <EriC^> lukegb: that guy "ubottu"
[11:59] <EriC^> he's the mastermind
[12:00] <lukegb> damn ubottu! *shakes fist*
[12:07] <BluesKaj> Hi all
[12:07] <Hejkki> hi
[12:07] <Hejkki> any idea on how can i get umlauts and other special characters working with my remote irssi?
[12:08] <Hejkki> when i type a with two dots, i see 2 question marks, and when i push enter, i see it correctly, but i bet none of you see it correctly?? Ã¤
[12:09] <Hejkki> i am using gnome-terminal with full utf-8 support enabled
[12:09] <Hejkki> i used to have the special characters working while using Gentoo
[12:10] <Hejkki> so i think (but might be wrong) that the problem is local, not at the server side.....?
[12:12] <martti> Hejkki: set up your remote LANG/locale-genamajick
[12:13] <Hejkki> and how will i do that?
[12:14] <martti> https://askubuntu.com/a/89983
[12:17] <martti> the problem might be local, though, yes
[12:17] <Hejkki> :)
[12:28] <conjo> Ping me testing setup on done thanks
[12:28] <conjo> fone
[12:29] <mar77i> come again?
[12:29] <conjo> This client look bare as on android lols
[12:29] <conjo> Thanks
[12:40] <Koopz> so... i read that swap is basically 1:1 reflecting the RAM
[12:40] <Koopz> meaning my swap should have the same size as my RAM, aye?
[12:42] <Koopz> unless that information i found on the internets is outdated and really true, i can't understand why my system has 32 GB of RAM while swap is reporting 2 GB
[12:43] <Koopz> i mean if it's true it's totally understandable why my nagios constantly reports 0% swap left
[12:43] <mar77i> uh, swap != ram
[12:44] <derpadmin> Koopz : it used to be true, but at 32, I would not put 64G ram
[12:45] <derpadmin> Koopz, swap is opportunistic, I personally disable it when I can on servers
[12:46] <Koopz> so comparing several servers i have a 16 GB RAM machine with 4 GB swap, the previously mentioned 32 GB RAM machine with 2 GB swap, my mail server with 32 GB RAM and 16 GB swap and a webserver with 64 GB RAM and 32 GB swap
[12:47] <Koopz> looks like my dev machine is misconfigured
[12:47] <derpadmin> it won't hurt you either
[12:48] <derpadmin> you can use fileswap as well
[12:48] <derpadmin> so that way, you can skip the swap partition
[12:48] <Koopz> in theory swapping would only be needed if my system's load requires more RAM than physically available right
[12:48] <derpadmin> and come back and add a swap file if needed
[12:49] <derpadmin> Koopz, incorrect, it is opportunistic, it will move less frequent access to swap
[12:49] <derpadmin> but you can change the behavior with "swapiness"
[12:49] <derpadmin> to get exactly what you describe
[12:50] <derpadmin> swapiness 0 if I recall
[12:51] <Koopz> i should rather find out how to limit this graylog instance since it apparently takes everything it wants...
[12:52] <derpadmin> Koopz, maybe cgroups, not sure, but yeah, I would go with that too
[13:18] <Caim2f> Hello :)  Nice to meet you ! Glad to be here :P  I've been a long time macOS/windows user but just switched recentely. I would appreciate any sort of guidance please :D  I seem to be having some problem with the rendering not sure how to put my finger on it but something feels very off on ubuntu 20.04 :(   I'm using a RTX2070 and have 2 27" 1080p
[13:18] <Caim2f> monitors.  When the window is minimized the rendering seems okay but when it's maximized the text and sub elements seem a bit skewed if that makes sense ? Like they are a bit more wide than normal
[13:18] <Caim2f> Is there a setting I'm missing somewhere ? Looked in nvidia panel but couldn't find anything
[13:21] <Texas> Caim2f: What do your display settings look like?
[13:21] <Caim2f> I think it might be just Chromium that's bugged though
[13:21] <Texas> Firefox works fine?
[13:22] <Caim2f> Yes
[13:22] <Texas> Let me check some things on my machine
[13:22]  * Texas uses FF when on ubuntu
[13:22] <Caim2f> Also in Nvidia X Server settings in X Server Display Configuration I see this bug
[13:22] <Caim2f> where I have the 2 monitors
[13:23] <Caim2f> If i maximize the window the 2 subelements don't scale properly
[13:24] <Texas> Caim2f: Have you rebooted your OS since making these configuration changes?
[13:24] <Caim2f> Yes. Been using this for almost one week :D
[13:25] <Texas> Caim2f: And this has been happening the whole time you had it? And only a chromium issue?
[13:26] <Texas> hrm
[13:26] <Texas> chromium works fine for me maximized...
[13:28] <Texas> Caim2f: What nvidia driver do you have installed?
[13:28] <Caim2f> Yes . MAybe I'm just imagining things ?
[13:28] <Caim2f> 450.66
[13:28] <Caim2f> Can you see this ? https://share.icloud.com/photos/0ubtef-seisG8GqMBsS10bFFg
[13:30] <Texas> Caim2f: Yes, I can. Looking now.
[13:32] <Caim2f> Should I override Antialiasing in nvidia settings or leave it on default ? What settings do you use ? I've enabled the Force Composition Pipeline also because was having some problem with tearing while playing videos.
[13:33] <Texas> I don't use nvidia
[13:37] <Texas> Caim2f: Sorry, but I have to go do something very urgent for work but if you wish to stick around, perhaps someone else could help you or I can whenever I come back
[13:38] <Caim2f> I'm thinking of ditching these displays though. Maybe 27" 1080p is not ideal . Would look into a 2k display. I can't get into 4k monitors I found the text hard to read. :D
[13:38] <Texas> Sorry I couldn't be more helpful :(
[13:38] <Caim2f> Sure np thanks for the help :D  I'm learning something new every day :P
[13:38] <Texas> heh
[13:38] <Texas> I use mac hardware for everything lol
[13:38]  * Texas is an apple fanboy
[13:38] <Texas> I get a lot of shit for that on IRC
[13:40] <lotuspsychje> !ot | Texas Caim2f
[13:41] <Texas> lotuspsychje: Yes, I'm aware. That's as far as I was having the conversation go.
[13:42] <Tom01> Hello people.
[13:42] <Tom01> I have upgraded from Ubuntu 18 to 20.04. Now the shutdown is very slow. What could the reason?
[13:44] <Texas> Tom01: I don't really know any reason other than it's a newer os version and there may be more things to be done on shutdown.
[13:45] <oerheks> maybe you encountered a upgrade bug, the reason why upgrade path is not released yet.
[13:46] <lotuspsychje> Tom01: press F1 at shutdown to switch to text shutdown, to perhaps catch the culprit
[13:46] <Tom01> thanks I wil try
[13:48] <wyoung> hi Tom01
[14:12] <dbugger> Hi fellas
[14:12] <dbugger> Another magical day. I am using my PC without problems. I reboot. And now Ubuntu will only use one of my monitors.
[14:13] <dbugger> I have no idea where to start with... anyone has any suggestions?
[14:14] <Salatwurzel> dbugger, what does "xrandr" say? (terminal command)
[14:14] <Texas> dbugger: What are your display setting set to? Is the second monitor enabled?
[14:15] <dbugger> oh wait... apparently my computer has switched to noveau... without my permission o_O
[14:15] <dbugger> what the heck... how could that happen?
[14:16] <dbugger> Im gonna try resetting the propietary drivers and reboot. BRB
[14:18] <dbugger> Well, what do you know.. that fixed it
[14:18] <dbugger> No idea how in the hell the drivers changed to noveau by themselves...
[14:53] <MadLamb> I need to adjust my touchpad sensitivity settings, but its a bit hard with this xinput/synaptics variables. Is there a good gui tool that can help?
[15:31] <summonner> dbugger, which Ubuntu are you running?
[15:50] <j`ey> what's the best way to customise the MOTD w/o affecting updates, I was thinking `chmod -x`ing the files I didnt want
[15:59] <dreamscrypt> j`ey: I believe there is the /etc/motd and /etc/update-motd.d files.
[16:00] <j`ey> yeah, so I just chmod -x some of the files in /etc/update-motd.d, I still want them to be updated by the package manager, if theyre changed upstream
[16:03] <dreamscrypt> j`ey: You do still want motd output, correct?
[16:06] <j`ey> dreamscrypt: yeah, I'm just removing some parts of it
[16:06] <j`ey> like the news / documentation bits
[16:06] <dreamscrypt> j`ey: Oh, ok. :)
[16:07] <j`ey> just shows the system info now, much nicer than the big banner from before!
[16:07] <dreamscrypt> j`ey: I was going to mention that you could also disable the motd per user by creating a hidden .hushlogin within the users' directory.
[16:08] <dreamscrypt> j`ey: Sweet! :)
[16:09] <longshot> Is there a package for running an lxd image server? Or is that just lxd with network enabled but doesn't (necessarily) run containers itself?
[16:10] <nsaunders> how do I install mysql-connector-java for focal?  I can't find it in apt
[16:10] <longshot> I want to have a central server that I can publish private images to that all of my nodes can pull from.
[16:10] <nsaunders> mysql-connector-java
[16:10] <nsaunders> https://askubuntu.com/q/1276942/847449
[16:13] <longshot> nsaunders: Is different from libmysql-java?
[16:14] <nsaunders> longshot: I don't know
[16:14] <longshot> Have you tried that one yet?
[16:15] <nsaunders> it's also not in focal
[16:15] <oerheks> Bionic was the last one https://launchpad.net/ubuntu/+source/mysql-connector-java
[16:15] <nsaunders> I do see libmysqlclient21 but don't see anything in the documentation regarding jdbc
[16:16] <nsaunders> oerheks: yes.  that's very odd.  it must've been moved or renamed to something different?
[16:17] <nsaunders> regardless, I already have that installed.  I don't think it has JDBC, but I'm not sure.
[16:19] <longshot> Well there's always this: https://dev.mysql.com/downloads/connector/j/
[16:23] <leftyfb> longshot: you want #lxcontainers for your question
[16:24] <nsaunders> longshot: yes, thx
[16:24] <nsaunders> I'm intrigued.  what's lxcontainers about?  images?
[16:28] <JoeBk> I did a clean install of ubuntu this morning and found software center was not installed.  I think this is a serious bug.
[16:29] <JoeBk> I did "sudo apt install gnome-software" to fix.
[16:39] <pavlos> neteffect: btw, fsck runs in initramfs before the fisystem is mounted RW so there is no log of its activity.
[16:40] <pavlos> filesystem*
[16:47] <Kali_Yuga> Hello can someone tell me what kind of keyboard layout this is? https://www1.xup.in/exec/ximg.php?fid=15205854
[16:48] <Kali_Yuga> it's not US apparently
[16:48] <Kali_Yuga> even though it's very similar
[16:48] <Habbie> US International
[16:51] <JoeBk> I would say US becaise it has a "$" sign.
[16:51] <dreamscrypt> Kali_Yuga: I'm curious - is this for a fresh Ubuntu install on that particular laptop?
[16:52] <Habbie> JoeBk, many non-US layouts also have the $ there
[16:52] <pavlos> but it has a euro sign with alt-gr 5
[16:52] <dreamscrypt> ^
[16:52] <Habbie> pavlos, hence, US International
[16:52] <oerheks> international, with the € on 5
[16:53] <JoeBk> what happens when you set it to US?
[16:54] <Habbie> JoeBk, the euro sign might stop working :)
[16:54] <Habbie> but really , US International mostly means you enable 'dead keys' for turning 'e into é
[16:56] <Kali_Yuga> it was English (intl. with AltGr dead keys)
[16:56] <Kali_Yuga> thank you guys
[16:58] <dr`venom> Can someone help me setup the thumb button on the mx master logitech mouse on Ubuntu 20.04. Thank you.
[16:59] <leibniz[m]> How to temporary comment-out an apt PPA? It's .list file resides in etc/apt/sources.list.d/target.list
[17:00] <tomreyn> if it's just temporarily, the easiest way is to rename the file to target.list.disabled
[17:01] <oerheks> or uncheck in softwaresettings?
[17:01] <tomreyn> you'll get warnings about this when running    apt update    but they're warnings, not errors, non-fatal
[17:01] <tomreyn> right, if it has a graphical desktop this would be even easier
[17:02] <pavlos> dr`venom: https://askubuntu.com/questions/973713/logitech-mx-master-thumb-button-with-ubuntu
[17:02] <tomreyn> leibniz[m]: software-properties-gtk --open-tab=1
[17:03] <Habbie> Kali_Yuga, the shape of the Enter means it is -not- English
[17:03] <Habbie> Kali_Yuga, but you won't notice until you try using the key above tab or one or two other keys
[17:04] <leibniz[m]> tomreyn: I changed it to .disabled. no warnings
[17:04] <leibniz[m]> `linux-image-5.4.0-48-generic` What is this?
[17:08] <tomreyn> leibniz[m]: the name of a kernel image package?
[17:09] <pavlos> leibniz[m]: apt show linux-image-5.4.0-48-generic
[17:11] <leibniz[m]> tomreyn: Is it the latest linux kernel?
[17:11] <leibniz[m]> Why after downloading it, apt asks again to autoremove it
[17:12] <Habbie> are you sure it is the same number it wants to autoremove?
[17:12]  * leibniz[m] sent a long message:  < https://matrix.org/_matrix/media/r0/download/matrix.org/mISjPHDeEjlGQRYMqAMVAlyZ/message.txt >
[17:12] <Habbie> leibniz[m], that message was too long for irc
[17:13] <Habbie> but, i clicked on the url for you, and it says -45-
[17:13] <Habbie> not -48-
[17:13] <leibniz[m]> Habbie: Sorry...ah thanks. so it asks to remove the older version
[17:13] <leibniz[m]> But still, I'm on Xubuntu 18.04...are linux kernels still provided for it?
[17:13] <nbusrone> hi , i am getting error on a DSL/PPPoE 'your current network has a .local domain which is not recommended and incompatible with the avahi"
[17:14] <Habbie> (to be fair, we also would not have liked it if your four lines still came into irc)
[17:14] <leibniz[m]> Shouldn't they provide only for 20.04?
[17:14] <Habbie> leibniz[m], 18.04 should get updates until 2023 i believe?
[17:14] <nbusrone> what does the message mean ?
[17:14] <Habbie> nbusrone, is something, that is not avahi, trying to use .local?
[17:14] <leibniz[m]> Habbie: Is that linux kernel updates?
[17:14] <Habbie> leibniz[m], also, yes
[17:15] <Habbie> leibniz[m], oh! it's 2021 for xubuntu - https://help.ubuntu.com/community/EOL
[17:15] <Habbie> leibniz[m], although i suppose kernels are not xubuntu-specific and would get updates until 2023 still
[17:16] <leibniz[m]> Habbie: So in terms of kernel, it doesn't matter which distro I use? Arch would also have 5.4.0-48?
[17:16] <Habbie> no, Arch would likely have something newer
[17:16] <Habbie> '-48' means ubuntu applied 48 updates on top of kernel.org's 5.4.0
[17:16] <Habbie> ubuntu, like debian, does not just import full new versions into an existing release
[17:16] <Habbie> they only import security fixes, other important fixes, and a few other exceptions
[17:16] <leibniz[m]> Habbie: What features are coming from kernel? driver support?
[17:17] <Habbie> leibniz[m], driver support, google for 'ubuntu hwe' i think, but i could be outdated on that
[17:18] <leibniz[m]> Habbie: I'm actually using i3wm. I should have used ubuntu-minimal or debian, right?
[17:18] <Habbie> leibniz[m], makes sense to me yes
[17:18] <Kireji> so I went through and took out all the entries in passwd and shadow that did not seem to be associated with any service or feature I could follow. all good. now on reboot, 5 users are consistently (repeatedly) added back into both the passwd and shadow file. I want to figure out where this is happening, and having trouble. how do I read the scripts/process for bootup on 20.04.01 LTS/Focal ?
[17:18] <Habbie> Kireji, what users?
[17:19] <Kireji> 1ec
[17:19] <Kireji> 1 sec
[17:19] <Kireji> lp,news,irc,list,gnats
[17:20] <Kireji> and, if they are part of packages still on my machine that I don't need, I want to remove those packages. I can't seem to find any indication gnats is on this machine, for example
[17:21] <Kireji> I did a sudo search through the entire /etc/ filetree, and found nothing there that would indicate a script that adds those users back into passwd and shadow
[17:22] <Kireji> I also want to read up on how the standard boot process is **re-writing my passwd file** --- this does not seem reasonable to me
[17:22] <jelly> Kireji, some groups and users are hardcoded in the OS
[17:22] <jelly> even if you never have a user member of group "admin" to read logs, the group will exist
[17:23] <Kireji> jelly: are there docs on this?  which users? what are they used for?
[17:23] <jelly> not sure about Ubuntu, but upstream lists them in Debian Policy
[17:23] <Kireji> jelly: not worried about groups - that I can fully understand, sharing and all. this is about users
[17:24] <Kireji> "they used for?
[17:24] <oerheks> Kireji, if those 'users' are program names and end with :/usr/sbin/nologin, don't worry
[17:24] <jelly> leave them be.  You'll be asked to recreate them on the next release upgrade.
[17:24] <Kireji> "upstream lists them in Debian Policy" googling
[17:24] <Kireji> oerheks: I'm not worried, I'm curious
[17:24] <Kireji> I want to understand boot
[17:25] <jelly> https://www.debian.org/doc/debian-policy/ch-opersys.html#users-and-groups
[17:26] <jelly> Kireji, basically, you shouldn't touch users and groups with uid or gid <= 99
[17:27] <jelly> Kireji, you'll find templates in /usr/share/base-passwd/
[17:27] <jelly> and docs in base-passwd package docs.
[17:27] <jelly> under /usr/share/doc/base-passwd
[17:28] <Kireji> "basically, you shouldn't touch" <== some Microsoft-level admin coddling, imo
[17:29] <nsaunders> where's the lib directory for BaseX which was installed from apt?  I'm trying to add a JDBC driver.
[17:29] <matsaman> nsaunders: could be dpkg -L packagename | grep -i lib would tell you
[17:29] <Kireji> it's an open source os for a reason. we can understand what it does
[17:29] <jelly> Kireji, the implied "without good reason" is implied
[17:30] <Kireji> exactly
[17:30] <Kireji> which is *why( I want to read the script to see the reasons for them being added in
[17:30] <Kireji> I have good reasons for removing them
[17:30] <jelly> read the docs, many of those are legacy but not all
[17:30] <jelly> some are used by packages that are optional
[17:30] <ElNomReal> The reasons for the different users and groups?
[17:31] <Kireji> gnats is not on this server
[17:31] <jelly> Kireji, care to explain some of those reasons?
[17:31] <Kireji> lp is a print daemon,  and nothing will ever print on this machine
[17:31] <thetoblin> is anyone familiar with Conda, Numba and Cuda? I have installed numba and condatoolkit through conda, but when I try to use cuda in a python script I get the error: libNVVM cannot be found
[17:32] <jelly> Kireji, you don't need those.  Do you actually gain anything by removing them, except for adding extra changes compared to a default install?
[17:32] <matsaman> I'm not sure it's the daemon, but sure, to do with printing
[17:32] <thetoblin> I have looked at every forum post I could find on the error, such as this one https://stackoverflow.com/questions/48385686/cuda-library-nvvm-not-found but to no avail
[17:32] <Kireji> jelly: basic security. if I don't understand why X needs to be on the machine, then X is not on the machine
[17:33] <jelly> Kireji, ask yourself whether removing a user that does nothing, has no extra privileges, and cannot log in, adds to "security" and in what way
[17:33] <Kireji> for other interpretations, see aparigraha
[17:34] <jelly> Kireji, then ask yourself whether changing OS settings from defaults and most walked paths adds or removes from "securit", and in what way
[17:34] <nsaunders> matsaman: thx
[17:34] <Kireji> jelly: I'm not looking for a debate on how to keep a machine secure, I've been doing that for 20 years. I'm interested in finding where in the boot process in 20.04 the passwd file is edited each time the machine boots
[17:34] <oerheks> " where in the boot process in 20.04 the passwd file is edited each time the machine boots" .. never?
[17:34] <Kireji> oerheks: nope
[17:34] <jelly> Kireji, ah, I don't have a 20.04 to verify that happens at all
[17:35] <Kireji> oerheks: ***bingo*** chicken winner dinner
[17:35] <oerheks> give proof of that?
[17:35] <jelly> there's still no upgrade path from 18.04 to 20.04
[17:35] <Kireji> it's happening on my machine. /etc/paswd and /etc/shadow have 5 lines added each boot if those users are not there
[17:36] <jelly> Kireji, which files get changed on boot of your machine and which users are added?
[17:36] <ElNomReal> Are they daemons? That might make sense.
[17:36] <Kireji> jelly: uh ??!? -- I just did a 18.04 to 20.04 upgrade last week, direct. 0 issues
[17:36] <jelly> do-release-upgrade did not work last week for me yet
[17:36] <Habbie> perhaps jelly means that do-release-upgrade still does not offer it?
[17:36] <Kireji> the users re-added on each boot are "lp,news,irc,list,gnats"
[17:37] <Kireji> Habbie: it does
[17:37] <Kireji> it did for me at least, last week
[17:37] <jelly> Kireji, sounds like something runs an equivalent of dpkg-reconfigure base-passwd
[17:38] <jelly> Checking for a new Ubuntu release
[17:38] <jelly> There is no development version of an LTS available.
[17:38] <Habbie> development version?
[17:38] <jelly> nope, still not there.  I won't do -d
[17:38] <compdoc> sounds right
[17:38] <oerheks> indeed, jelly, but the fix might come soon.
[17:38] <jelly> Habbie, the output is weird like that. :-)
[17:39] <jelly> Habbie, it correctly assumes non-LTS releases are purely for development
[17:39] <Kireji> jelly: curious. so you have 1 more data point, I did a do-relea-upg with -d from 18.04 to 20.04 and had no issues
[17:39] <jelly> yeah, I'm not doing that on a machine I do daily work on
[17:40] <jelly> did 16.04 -> 18.04 earlier this year, it broke boot
[17:42] <jelly> Kireji, I'd hunt for base-passwd related stuff in service definitions, esp. frivolous running of update-passwd command
[17:42] <Kireji> jelly: thank you
[17:42] <jelly> this ought to only be done during the release upgrade, not afterwards
[17:43] <jelly> but perhaps your do-release-upgrade -d was not completely without lingering issues
[17:43] <Kireji> lol, perhaps
[17:43] <jelly> (more precisely, during upgrade of that base-passwd package)
[17:43] <Kireji> 20.04 has a few notable differences
[17:44]  * jelly liked 16.04 Unity with menus on top as is sad to see them abandoned
[17:45] <jelly> switch to wayland was really painless, bar that boot issue (gdm was to fault)
[17:51] <pavlos> Kireji: https://www.gnu.org/software/gnats/
[18:20] <lordcirth> I want to reinstall the default config for syslog-ng; but when I do "apt purge syslog-ng" it doesn't remove config; and even if I rm it, "apt install syslog-ng" doesn't place any config. What's the correct way to install the config?
[18:23] <Texas> lordcirth: Have you ran 'apt-get remove syslog-ng' and 'apt-get autoremove syslog-ng'?
[18:23] <lordcirth> Texas, I used purge
[18:24] <lordcirth> I'll try autoremove - is the config in a dependency?
[18:26] <Texas> lordcirth: Try 'apt-get autoremove --purge'
[18:26] <Texas> That purges the files for all the dependencies as well
[18:29] <lordcirth> purge --autoremove worked
[18:30] <Texas> lordcirth: I think either one works
[18:30] <Texas> But I'm glad it worked out
[18:31] <tonsofpcs> so apparently Sept 1 - Sept 2 overnight my system tried updating and got hung, I've killed all the apt/dpkg processes and am trying dpkg --configure -a and it has been sitting at this for the past 6 hours (which I presume is what it was stuck on the past 21 days):  update-initramfs: Generating /boot/initrd.img-5.4.0-45-generic   -- is there a way to cancel that new kernel package - either roll back or
[18:32] <tonsofpcs> skip the build?
[18:35] <Texas> tonsofpcs: So you are able to boot without any problems?
[18:36] <tonsofpcs> Texas: system has been running.  I'd rather not try to find out.
[18:37] <tonsofpcs> (unless there's a way to test that the kernel is valid without turning it off, of course)
[18:37] <Texas> Oh you haven't shit it down?
[18:37] <Texas> *shut
[18:37] <tonsofpcs> nope.
[18:38] <tonsofpcs> it literally still had a dpkg instance running for the update from Sep 02 earlier today.
[18:38] <bugLe> exit
[18:39] <Texas> tonsofpcs: What is the output of 'sudo killall dpkg'?
[18:41] <Texas> tonsofpcs: Hello?
[18:42] <tonsofpcs> nothing there
[18:42] <tonsofpcs> this is what I get from ps aux after the killall:
[18:42] <tonsofpcs> root      555746  0.0  0.0   2608   548 ?        S    Sep02   0:00 /bin/sh /var/lib/dpkg/info/linux-image-5.4.0-45-generic.postinst triggered linux-update-5.4.0-45-generic
[18:42] <tonsofpcs> root     1591982  0.0  0.0   2608   548 pts/11   S    15:27   0:00 /bin/sh /var/lib/dpkg/info/linux-image-5.4.0-45-generic.postinst triggered linux-update-5.4.0-45-generic
[18:43] <Texas> Hmm
[18:44] <Texas> Let's try and kill the pids
[18:44] <Texas> 'sudo kill 555746' and 'sudo kill 1591982'?
[18:44] <Texas> and then try the ps -auc
[18:44] <Texas> *aux
[18:44] <tonsofpcs> gone from ps :)
[18:44] <Texas> yey
[18:44] <Texas> Is that everything?
[18:45] <tonsofpcs> that's everything that greps with dpkg, yeah
[18:46] <Texas> tonsofpcs: I meant for the update issue
[18:46] <tonsofpcs> ~$ sudo apt upgrade
[18:46] <tonsofpcs> E: dpkg was interrupted, you must manually run 'sudo dpkg --configure -a' to correct the problem.
[18:46] <Texas> Yeah run that
[18:46] <tonsofpcs> so then I run that and we get back to where we started...
[18:46] <tonsofpcs> update-initramfs: Generating /boot/initrd.img-5.4.0-45-generic
[18:47] <Texas> tonsofpcs: Try sudo update-initramfs -k 5.4.0-45-generic -d
[18:47] <Texas> That will remove the kernel
[18:48] <tonsofpcs> ok, but then dpkg still wants to regenerate it.  Should I let it try and see if it fails again or should I try to remove the package?
[18:48] <Texas> And then run sudo apt-get install -f
[18:48] <ovrh> Hello everyone, I just rebooted into Ubuntu 20.04 and my second monitor isn't being detected anymore. I'm not sure how to go about debugging this :/
[18:49] <Texas> ovrh: Did your graphics driver get changed somehow?
[18:49] <tonsofpcs> ok, so now that's doing it...
[18:49] <Texas> ovrh: I'm pretty sure this happened to someone else today
[18:50] <Texas> tonsofpcs: It's working now?
[18:50] <ovrh> Texas, Unless something updated automatically, then no. I was using it a few hours ago, and everything worked fine. Didn't install or remove anything then. Then I rebooted into Windows to play some games, and just rebooted back into Ubuntu to find the external monitor isn't working anymore
[18:51] <Texas> ovrh: The other person today said that theirs got changed without their knowledge
[18:51] <tonsofpcs> no, it's doing the same thing.  generating initrd.img-5.4.0-45-generic .  apt also points out that 5.4.0-40 packages are no longer in use
[18:51] <ovrh> Texas, What got changed?
[18:51] <Texas> Hang on
[18:52] <Texas> I'm looking at my logs from earlier
[18:52] <Texas> tonsofpcs: Did you run sudo apt-get install -f?
[18:52] <Texas> And then reboot
[18:54] <tonsofpcs> Texas: I ran apt-get install -f, that's what it is doing.
[18:54] <Texas> ovrh: What is the output of lspci?
[18:54] <tonsofpcs> https://pastebin.com/b9FHmQmU
[18:55] <ovrh> Texas, this: https://paste.ubuntu.com/p/t3zBvMQ4k5/
[18:57] <tonsofpcs> watching /boot, here's what I see.  It's like it can't build or something - https://pastebin.com/MQ4EKBP1  (yes, there's plenty of free space in /boot)
[18:59] <tonsofpcs> looks like some old build attempts are hanging out there too, going to try killing them and give it a last go: https://pastebin.com/7ScJr71r
[19:02] <tonsofpcs> whelp, looks like that lets me run apt and now 5.4.0-48 is available so let's see if that will build
[19:04] <Texas> tonsofpcs: Try and reboot without doing the apt-get command then
[19:05] <Texas> But do take note that you may need a live ubuntu cd to proceed
[19:05] <tonsofpcs> there's no valid kernel set to boot, I'd have to manually choose one.  Trying to fix it first.  It's really not happy though so I may need to.
[19:10] <Texas> I'd be curious to know why it got stuck in the first place
[19:11] <Texas> ovrh: Are you still here?
[19:11] <tomreyn> disk falling apart and/or inconcistent file system. or no inodes left, but that should have printed an erro on stdout.
[19:11] <ovrh> Texas, yeah, trying to read about this online, but nothing so far
[19:13] <Texas> yeah
[19:15] <tomreyn> dmesg | grep -v ' IN=' | tail | nc termbin.com 9999
[19:19] <ovrh> rebooting
[19:20] <Texas> ovrh: What is the output of lspci?
[19:20] <Texas> (type 'lspci' into your terminal)
[19:27] <Texas> ovrh: Welcome back
[19:27] <Texas> Did it work?
[19:27] <ovrh> Nope
[19:27] <ovrh> I tried removing a blacklist-framebuffer.com.conf from /etc/modprobe.d/ as someone suggested somewhere in a nvidia forum, but nothing
[19:28] <Texas> ovrh: What did you try to change?
[19:28] <Texas> I did not tell you to do anything
[19:28] <Texas> Oh
 ovrh: What is the output of lspci?
[19:28] <Texas> You still haven't done that
[19:28] <ovrh> Did find out though that trying to run `nvidia-smi` (never done that before) doesn't seem to work
[19:28] <ovrh> Texas, I have, it's a bit above
[19:28] <ovrh> Texas, here it is again:https://paste.ubuntu.com/p/7QHdjhjHZz/
[19:29]  * Texas looks
[19:29] <Texas> Hmm
[19:29] <Texas> Yeah it's not the issue I thought
[19:30] <tonsofpcs> Texas: no idea but it seems to be progressing on building -42-generic now (oddly not -48....)
[19:30] <ovrh> I'll have to check if the journalctl says anything. I'm gonna grab a bit for now though. Thanks for your help, Texas
[19:30] <tonsofpcs> Texas: I also see a few mailing list and forum posts about this kernel failing to build for others leaving them unable to boot
[19:31] <Texas> ovrh: No problem. If you need anymore help, you may ask in this channel or you can PM me.
[19:31] <Texas> I'm always connected.
[19:33] <tomreyn> tonsofpcs: did you ever look at dmesg?
[19:34] <tomreyn> (or journalctl -kf)
[19:36] <tonsofpcs> tomreyn: nothing related afaict in dmesg (not even around the same time[s]).  checking journalctl now
[19:36] <tonsofpcs> oh, that's the same.  yeah
[19:37] <tonsofpcs> /var/log/dmesg: 2020-09-02 06:15:36 trigproc linux-image-5.4.0-45-generic:amd64 5.4.0-45.49 <none>
[19:37] <tonsofpcs> 2020-09-02 06:15:36 status half-configured linux-image-5.4.0-45-generic:amd64 5.4.0-45.49
[19:37] <tonsofpcs> it actually goes through that quite a bit that morning...
[19:39] <tomreyn> tonsofpcs: mount | grep '/boot '
[19:39] <tonsofpcs> /dev/sda2 on /boot type ext4 (rw,relatime)
[19:40] <tonsofpcs> /dev/sda2       976M  322M  588M  36% /boot
[19:41] <tomreyn> sudo dumpe2fs -h /dev/sda2 2>&1 | nc termbin.com 9999
[19:43] <tonsofpcs> https://pastebin.com/ntiDCjB1
[19:46] <tomreyn> "Last write time:          Fri May 22 12:46:31 2020"
[19:46] <tomreyn> what is the     date    ?
[19:46] <tonsofpcs> looks like it might be hanging on update-initramfs: Generating /boot/initrd.img-5.4.0-42-generic too...
[19:46] <tonsofpcs> date is proper.
[19:46] <tonsofpcs> Tue 22 Sep 2020 07:46:45 PM UTC
[19:47] <tomreyn> you have the file system configured to be never checked for errors
[19:47] <tonsofpcs> and there's definitely files in there with dates of July 3 and Jul 22 and Sep 22...
[19:48] <tomreyn> unmount it, do a fsck -f,
[19:48] <tomreyn> remount it
[19:49] <tonsofpcs> ok
[19:52] <tonsofpcs> it still doesn't seem to be doing much.  Nothing is touching the CPU or RAM.
[19:52] <tonsofpcs> ( update-initramfs: Generating /boot/initrd.img-5.4.0-42-generic )
[19:52] <tomreyn> you should cancel this before trying to unmount
[19:52] <tonsofpcs> I did
[19:53] <tonsofpcs> killed all the initramfs and dpkg processes, lsof'd, killed everything touching boot, unmounted, fsck -f'd, remounted, ran apt install
[19:54] <tomreyn> sorry i mean e2fsck -f
[19:55] <tonsofpcs> fsck automatically selected e2fsck, does it not pass -f through?  https://pastebin.com/eEqMVuU8
[19:55] <tomreyn> yes i think it does
[19:56] <tomreyn> it did run a check, so it did
[19:56] <tomreyn> now all i can think of is that this kernel you are running is buggy, or it's a VM and the problem is on the host
[19:57] <tomreyn> cat /proc/version
[19:57] <tonsofpcs> Linux version 5.4.0-31-generic (buildd@lgw01-amd64-059) (gcc version 9.3.0 (Ubuntu 9.3.0-10ubuntu2)) #35-Ubuntu SMP Thu May 7 20:20:34 UTC 2020
[19:57] <tonsofpcs> and it is a VM
[19:57] <tomreyn> do you have access to the host?
[19:58] <tonsofpcs> not any more than to pop into a terminal for the guest.  (ESX)
[19:59] <tomreyn> does     journalctl -b -p3    have anything related?
[20:00] <tonsofpcs> systemd[1]: Failed to start Daily apt download activities. starting the evening of Sep 02 (Sep 02 morning was when the auto-update started and hung out until today)
[20:00] <tonsofpcs> interesting about an hour ago  systemd[1]: Failed to start Firmware update daemon.
[20:01] <tomreyn> i still suspect it's a storage problem on the host. but that's just a gut feeling. you could maybe work around it by trashing /boot and creating a new one.
[20:01] <tomreyn> or you could just reboot and pick one of the working initrds
[20:02] <tonsofpcs> it's able to move files around on /boot and it's making System.map and vmlinuz for each version, it just isn't isn't building the initrd.img
[20:02] <tonsofpcs> which is strange.
[20:03] <tomreyn> you could run    update-initramfs -k 5.4.0-42-generic -c -v
[20:04] <tomreyn> if this is an 18.04 system it should really be at 5.4.0-47-generic by now, though
[20:04] <tonsofpcs> -42 was the sep 02 update, -48 is what it wants to install now
[20:04] <tonsofpcs> and it's 20.04.1
[20:05] <tonsofpcs> err, sorry, -45 was the sep 02 update
[20:05] <tomreyn> oh ok
[20:06] <tomreyn> but what it fails to generate is initrd.img-5.4.0-42-generic, right?
[20:06] <tomreyn> so you may want to manually generate that
[20:06] <tonsofpcs> 45, 48, and 42 (despite 42 already existing)
[20:06] <tonsofpcs> running update-initramfs -k 5.4.0-42-generic -c -v
[20:06] <tonsofpcs> to see what happens
[20:07] <tonsofpcs> looks like it is hanging on/after: Adding module /lib/modules/5.4.0-42-generic/kernel/drivers/md/raid10.ko
[20:10] <tonsofpcs> hmmm https://askubuntu.com/questions/1255546/update-initramfs-stuck
[20:10] <tomreyn> hmm and /var has also not run full?
[20:11] <tomreyn> removing mdadm may be an option, if just temporarily
[20:11] <tomreyn> afer all you wont have raid in the vm, i assume?
[20:12] <tonsofpcs> /var isn't its own partition, / has 154 G free
[20:12] <tonsofpcs> right
[20:12] <tomreyn> and spare inodes on / as well?
[20:12] <tonsofpcs> forcing removing mdadm looks like it did it.
[20:12] <tomreyn> hmm, ok
[20:13] <tonsofpcs> 12.7 million Ifree, yeah.
[20:14] <tomreyn> i think i'd want to check my apt sources to ensure i'm not using some critical packages with newer versions from a third party repository, check apt-forktracer output, run debsums
[20:14] <tonsofpcs> and I have kernels now.
[20:14] <tomreyn> and certainly move to the current kernel sooner ratehr than later.
[20:15] <tonsofpcs> yeah, running apt upgrade now.
[20:18] <Kon> A Focal system ran into problems after upgrading Nvidia to 450.66. I tried rolling back to 440.100 which is listed as an available alternative, but trying apt install nvidia-driver-450=440.100-0ubuntu0.20.04.1 doesn't work
[20:18] <tonsofpcs> it is not happy.  trying a reboot since the kernels are all there... :gulp:
[20:18] <Kon> And to make matters worse, nvidia-driver-440 now just points to nvidia-driver-450
[20:20] <tonsofpcs> system shutdown waiting for a lot of blkid and mdadm lol
[20:20] <tonsofpcs> and it came back \o/
[20:21] <tonsofpcs> 5.4.0-48-generic
[20:21] <tonsofpcs> running apt upgrade again...
[20:22] <Bashing-om> !info nvidia-driver-450
[20:23] <kwhat4> Hi, I am currently using armhf (dpkg --add-architecture armhf and some other stuff) to cross compile for arm and it seems to work well.  My question is this toolchain suitable for aarch64?
[20:24] <Kon> !info nvidia-driver-440
[20:24] <Kon> And yet in focal, not focal-updates, 440.100 should exist
[20:27] <Bashing-om> Kon: Does exist. what shows for you ' sudo ubuntu-drivers list ' ?
[20:29] <Kon> Bashing-om: https://paste.ubuntu.com/p/8kVb52sCDQ/
[20:31] <Kon> What's the difference with the -server packages?
[20:33] <tonsofpcs> thanks tomreyn and Texas it's all happy now :)
[20:35] <kwhat4> maybe I should be asking in #ubuntu-arm or #ubuntu-toolchain?
[20:35] <Bashing-om> Kon: Strange - I have an older nvidia card, and the list command returns 5 options for me.
[20:36] <Kon> Bashing-om: Is it possible to have disabled the "focal" repository by accident and only have "focal-updates" ?
[20:36] <Kon> Or better question, can I force a specific package to ignore focal-updates
[20:38] <Bashing-om> Kon: What shows - lspci -k | grep -EA2 'VGA|3D' - ? repos are in /etc/apt/sources.list file . One can have a look and see that the respective repos are indeed available to the system.
[20:40] <Kon> Bashing-om: https://paste.ubuntu.com/p/GwTxw6bGf3/
[20:41] <Bashing-om> Kon: checking ^.
[20:42] <Kon> It's an RTX 2060 Super
[20:45] <Bashing-om> Kon: Nvidia recommends the 450 driver for this card : https://www.nvidia.com/Download/driverResults.aspx/163238/en-us .
[20:46] <Kon> Bashing-om: Okay, and there's a problem with it. The X-session has to be manually invoked each system start. There are regressions in video drivers all the time
[20:46] <Kon> Nvidia themselves have a one-click Rollback button in their Windows drivers
[20:47] <Kon> Ubuntu has traditionally separated Nvidia branches to allow rollback
[20:47] <Bashing-om> !info nvidia-driver-440 focal
[20:47] <Kon> But now, it seems that behavior has changed, and everything is a transtition metapackage to the latest bleeding edge release
[20:48] <Kon> Even installing nvidia-driver-430 points to 450.66
[20:48] <Bashing-om> Kon: ^ is the "restricted" repo enabled in your sources ?
[20:48] <Kon> Bashing-om: deb http://us.archive.ubuntu.com/ubuntu/ focal main restricted
[20:50] <Bashing-om> Kon: Hummm -- what shows then ' apt list nvidia-driver-440 ' ?
[20:54] <Kon> Bashing-om:  apt list -a nvidia-driver-440 https://paste.ubuntu.com/p/pCykBtBz6H/
[20:56] <Bashing-om> Kon: We;;. you know now that the 440 version driver is there, want to try and purge the 450 driver and install the 440 version ?
[20:56] <Bashing-om> well*
[21:01] <Kon> Bashing-om: If that would solve the issue, why didn't apt install 440=440.100 work?
[21:05] <Bashing-om> Kon: Well. perhaps others can know the better - but I suspect that the commands should have been as ' sudo apt --purge autoremove nvidia* ' *IF* OEM is not a factor. then to install ' sudo apt update ; sudo apt upgrade ; sudo apt install nvidia-driver-4440 '.
[21:07] <Kon> Bashing-om: !info nvidia-driver-440
[21:07] <Kon> If that would work you would see it straight up points to 450.66
[21:08] <Kon> This seems like a packaging issue to me
[21:08] <Bashing-om> !info nvidia-driver-440
[21:10] <Bashing-om> Kon: Ouch ^^ Looks like the re-direction is on purpose. Above my skill set now to know .
[21:11] <oerheks> try the driver ppa..
[21:14] <Kon> oerheks: Tried that and for some reason the repo appeared empty
[21:16] <Bashing-om> Kon: This one: https://launchpad.net/~graphics-drivers/+archive/ubuntu/ppa ?
[21:16] <Kon> Yes Bashing-om
[21:18] <tomreyn> it's not empty, but the package version is the same
[21:19] <Bashing-om> Kon: "
[21:19] <Bashing-om> 1 → 75 of 82 results"
[21:20] <retran> And I've never fouled up my install using PPA provided video drivers!
[21:20] <retran> tee-hee
[21:20] <oerheks> retran, how helpfull .. but then again you might not own a RTX 2060 Super
[21:22] <coreyhuinker> can someone suggest a strategy for getting this issue to get more attention: https://askubuntu.com/questions/1160348/openvpn-via-network-manager-using-2fa-overwrites-saved-password
[21:23] <ovrh> Hey everyone. I asked this earlier as well, and I've been searching online as well without hope, so I'm trying again: after rebooting into Ubuntu (20.04) from Windows, it doesn't detect my second monitor anymore. Was working fine before going into windows, and I confirmed the monitor works for sure (since it does work on Windows). I tried removing the blacklist-framebuffer.conf file as I read online, but that didn't work. Someone also suggested checking
[21:23] <ovrh> that there's no "nomodeset" in the grub commandline option, and I confirmed there isn't.
[21:23] <coreyhuinker> basically, if you use openvpn on ubuntu, and you have a 2fa challenge, the UI doesn't distinguish between the two values, and attempts to store each in the same spot - thus preventing the password from ever being saved
[21:23] <ovrh> Only thing I found that's weird so far is that `nvidia-smi` doesn't seem to work. But I never tried it before so I'm not sure it ever did.
[21:24] <Kon> tomreyn: So I just tested on a Bionic system as well. Adding the PPA, I see it does add lots of packages from the Launchpad source, but they are all the ones unsupported by Ubuntu. For 440, 450, etc I look at my available versions and they all come from Bionic
[21:24] <Kon> The Focal guy is not actually me but a friend who asked me for help. So this behavior is consistent for both of us
[21:24] <ticketron> Hey folks, I'm trying to use the O_EXEC flag and my GCC is telling me it doesn't exist, at least not in <fnctl.h>. Is it somewhere else?
[21:25] <ticketron> fcntl sorry
[21:27] <tomreyn> Kon: i didn't follow the complete conversation, and don't know what the root problem is. i'm also not convinced i could be of any more help than Bashing-om is.
[21:28] <Kon> Well, the problem is a packaging one I am fairly confident. I'm not sure where to report the issue
[21:28] <Kon> But I really feel there's no value in Ubuntu packaing separate branches if they are all just going to redirect to bleeding edge
[21:29] <tomreyn> Kon: if there's a packaging issue, then identify the affected binary packages, then their source package(s), and report  bugs against them.
[21:29] <Kon> At least 440 is still technically hosted on the focal-security and bionic-security repos, but because it's all marked as transition packages, they are not accessible to the user
[21:32] <pavlos> ovrh: I had a similar issue this morning, it lost the second monitor, I did updates, rebooted, and it is back to normal. ubuntu 20.04 and nvidia 450.66 ... and nvidia-smi works.
[21:34] <ovrh> pavlos, I've just opened the updater little thingy and it's telling me "Not all updates can be installed". Never seen that before, but it doesn't look very good
[21:35] <ovrh> Have to say that I never installed the nvidia drivers manually, just checked the third-party box when I installed Ubuntu
[21:36] <oerheks> sudo apt update && sudo apt full-upgrade # could fix that
[21:38] <pavlos> ovrh: if it complains about broken packages, try "sudo apt install -f"
[21:38] <ovrh> Eh, it gave me a button saying "Partial upgrade" and I clicked it o.o
[21:41] <oerheks> maybe a reboot is needed after this.
[21:41] <oerheks> then try again
[21:43] <ovrh> yeah it just finished, asking me to reboot. brb
[21:48] <frad> if you work with youtube-dl, have they discontinued updates via wget and pip is the default update method now?
[21:48] <j`ey> anyone managed to run the netboot mini.iso on arm64 with qemu?
[21:49] <ovrh> Welp, Partial upgrade button worked, monitor is back on
[21:49] <ovrh> The question now is: why?
[21:50] <ovrh> Did the other part of the upgrade that this partial upgrade just completed mess everything up then?
[21:50] <instantp10neer> An Ubuntu 18.04 system will not allow checking the partition (via 'Disks').  It displays, "Error unmounting filesystem" "Error unmounting /dev/sda1: target is busy (udisks-error-quark, 14)" - Is checking the live system not an option?
[21:51] <oerheks> frad  get it from github? https://github.com/ytdl-org/youtube-dl
[21:52] <oerheks> their site gives pip as option too, indeed
[21:52] <tomreyn> instantp10neer: what do you mean by "check the partition", are you trying to run a fil esystem check on a mounted file system?
[21:53] <instantp10neer> The built-in utility 'Disks' offers 'Check Filesystem...
[21:53] <tomreyn> it seems to be trying to unmount the file system before running fsck against it. but it cannot unmount it because files are loaded / in use from there.
[21:53] <instantp10neer> Yes, live.  I assumed there exists a read-only check.  The same menu offers, "Repair filesystem..." , the same utility with read/write?
[21:54] <tomreyn> instantp10neer: why are you trying to run a file system check against a live iso?
[21:54] <instantp10neer> It isn't an ISO, it is a VM-installed operating system.
[21:55] <tomreyn> ok, i misinterpreted "Is checking the live system not an option" then
[21:55] <tomreyn> some file systems can be checked while they're mounted, other can't be
[21:56] <tomreyn> gnome gisks seems to assume it must unmount file systems before they can be checked.
[21:56] <tomreyn> but this one can't be unmounted (see above), so the check fails.
[21:56] <tomreyn> *gnome disks
[21:57] <pavlos> ovrh: glad you got the monitor back. No idea why the hiccup ...
[21:57] <oerheks> if a VM gives such issues, you might have 2 issues, check the host first
[21:58] <ovrh> pavlos, Thank you for the tip :) I think I'm gonna find a way to just disable all automatic upgrades forever
[21:58] <instantp10neer> The host is fine.  I ran into trouble export/importing and I want to ensure integrity.  I would have liked to fsck from recovery.  That gives an error.
[21:58] <instantp10neer> fsck from util-linux 2.31.1
[21:58] <instantp10neer> e2fsck: Cannot continue, aborting.
[21:58] <instantp10neer>  /lib/recovery-mode/recovery-menu: line 80: /etc/default/rcS: No such file or directory
[21:59] <pavlos> instantp10neer: sudo touch /forcefsck and reboot the VM. It will fsck upon reboot
[21:59] <MazinKaesar> hi guys succesfully intalled ubunto on my DELL M4800 solving all wifi issues :D
[21:59] <instantp10neer> pavlos thank you
[22:00] <pavlos> instantp10neer: there are no logs, the only evidence is that file will no longer exist in /
[22:00] <oerheks> !yay | MazinKaesar
[22:00] <instantp10neer> sorry?
[22:02] <tomreyn> pavlos: i think this applies: https://www.linuxuprising.com/2019/05/how-to-force-fsck-filesystem.html
[22:05] <giaco> hello! Why do I get "unable to locate package" for this one? https://packages.ubuntu.com/focal/ns3
[22:06] <giaco> I am running ubuntu 20.04 and apt update has been executed correctly
[22:06] <tomreyn> instantp10neer: if the /forcefsck file will still be present after reboot, you should read use !kernelparm      fsck.mode=force fsck.repair=yes     instead
[22:06] <giaco> but there could be a disabled repo as I'm coming from 18.04
[22:07] <tomreyn> giaco: this package is in universe. is universe enabled on your system?
[22:07] <tomreyn> giaco: also, which architecture is this?
[22:08] <giaco> running x64 laptop here
[22:08] <tomreyn> this package isn't available for this architecture
[22:08] <daveinlv> Got a quick yes/no question regarding Ubuntu 20.04.. Does the old "GRUB_CMDLINE_LINUX="net.ifnames=0 biosdevname=0"" in /etc/default/grub still work on an UEFI based system?
[22:08] <giaco> tomreyn: oh, wow, you're right
[22:08] <giaco> well thats surprising
[22:09] <daveinlv> I HATE those goofy devnames that you withOUT doing this trick
[22:10] <oerheks> daveinlv, my best advise, get used to it. and yes, it should work, rolling back.
[22:10] <daveinlv> Thanks.. Am kinda new to UEFI and all my old legacy systems got the above treatment...
[22:11] <daveinlv> Thanks oerheks!!
[22:17] <instantp10neer> tomreyn where would the file be present?
[22:18] <tomreyn> instantp10neer: where you created it, if you did.
[22:19] <instantp10neer> I didn't understand I was creating a file.  I don't know 'force'
[22:20] <tomreyn> instantp10neer: if you ran what pavlos suggested - "sudo touch /forcefsck and reboot the VM. It will fsck upon reboot" - you created a file named "forcefsck" on the uppermost path, "/"
[22:20] <instantp10neer> I see.  Thank you.
[22:20] <tomreyn> after reboot it will eithe rbe gone, and your system carried out a file system check, or it will still be present, and then you system didn't carry out a file system check.
[22:21] <tomreyn> i assume the latter is going to happen, as discussed at the article i had linked
[22:21] <instantp10neer> It did appear to flash 'clean' past.
[22:25] <instantp10neer> The file is not present.
[22:27] <tomreyn> ok, so i was wrong. :-)
[22:29] <pavlos> instantp10neer: per the article, "sudo tune2fs -l /dev/sda1 | grep checked"
[22:31] <instantp10neer> How long should fsck have taken on an nvme ssd?  It was less than 10 seconds.  The drive has 30gb stored.
[22:33] <instantp10neer> Anyone know if an encrypted VBox VM will be encrypted after it is exported/imported?
[22:37] <pavlos> tomreyn: I made a /forcefsck, rebooted, sudo tune2fs -l /dev/sda1 | grep checked reports yesterday (Last checked:             Mon Sep 21 20:20:04 2020)
[22:38] <tomreyn> pavlos: so you're saying /forcefsck did not trigger an fsck?
[22:39] <pavlos> tomreyn: correct, and forcefsck file is gone from /
[22:39] <tomreyn> so the worst of two worlds ;-)
[22:39] <pavlos> maybe I need to try the /etc/default/grub method
[22:40] <instantp10neer> OK, it didn't check.  I will try tune2fs
[22:40] <tomreyn> you just need to use    fsck.mode=force    on the grub prompt once
[22:43] <pavlos> tried the grub method, it did fsck ... Last checked:             Tue Sep 22 15:41:55 2020
[22:45] <tomreyn> thanks, pavlos
[22:45] <pavlos> learned something ... thanks tomreyn
[23:07] <Chuckfu> can someone help with the export function for bashrc
[23:09] <Texas> Chuckfu: Depends on what you're trying to do :)
[23:10] <Chuckfu> I am following some install instructions and this is what they tell me to insert in bashrc https://pastebin.com/MBfry5ft
[23:10]  * Texas looks
[23:10] <Chuckfu> and it error out
[23:10] <Texas> What's the error?
[23:11] <Chuckfu> let me put it back in and run it
[23:12] <Chuckfu> ok rebooting system take just a moment
[23:12] <Texas> Ok.
[23:14] <Chuckfu> when I sudo -i I get -bash: export: `=/home/minds/minds': not a valid identifier
[23:14] <Chuckfu> I get what they are trying to do, but I don't know the proper syntax
[23:15] <pavlos> the first line, omit the $
[23:15] <pavlos> export MINDSROOT=/home/minds/minds
[23:16] <Chuckfu> leave the $ infront of the rest
[23:16] <Texas> Yeah
[23:17] <Chuckfu> ok is there a command to do to re-read teh bashrc without rebooting
[23:17] <pavlos> source .bashrc
[23:18] <Texas> Chuckfu: You rarely ever need to reboot to do anything
[23:19] <Chuckfu> Wow awsome it start to install minds, hope it works with such a simple thing to fix it
[23:19] <Chuckfu> yeah not a complete newbie but not an expert either
[23:20] <Chuckfu> really appreciate the help on that,  I have been trying to install it for like ever
[23:20] <Texas> Chuckfu: Not a problem
[23:20] <Texas> You're never too good to learn
[23:20] <Chuckfu> lol now I have a problem for the minds people to fix
[23:25] <hggdh> Chuckfu: the syntax is wrong for the export
[23:26] <hggdh> Chuckfu: it is: export MINDSROOT=/home/minds/minds    # NO '$' in front of MINDSROOT
[23:26] <hggdh> ah, already done :-) should really read the backlog
[23:27] <Kon> Bashing-om: Don't know if you're still around but we gave it another try and found a workaround for the packages being marked as transition
[23:27] <Chuckfu> thanks for trying anyway
[23:28] <Bashing-om> Kon: Oh ? I am interested . How so ?
[23:28] <Kon> Bashing-om: So instead of trying package=version, we did package/focal-security for the metapackage and every single dependency it complained about
[23:30] <Bashing-om> Kon: Sounds like a real pain. What the package manager is supposed to be doing for us :(
[23:32] <Kon> It's because it was comparing those two versions against focal and focal-update, but not focal-security, I believe
[23:33] <Kon> Because technically speaking focal-security is not a separate source in the list files
[23:33] <Kon> But it does exist to the Ubuntu servers
[23:35] <Bashing-om> Kon: Noted - thanks for that feed back.
[23:37] <Kon> Bashing-om: But again, that's the only reason focal-security had the files we needed. The core issue here is all Nvidia metapackages regardless of version number are currently redirected to 450
[23:37] <Kon> Which is very dangerous if the latest driver update ever has a regression that isn't caught by Ubuntu testers
[23:38] <Kon> Or indeed, if some users need or prefer the Nvidia "long-lived" branch
[23:39] <Bashing-om> Kon: Agreed - our testers can not catch everything that might be; I do like haing control - FOSS !
[23:39] <Bashing-om> haveing*
[23:40] <Kon> I see why Nvidia updates were only supported by PPA for such a long time :)
[23:43] <Bashing-om> Kon: Old now but: https://lists.ubuntu.com/archives/ubuntu-desktop/2015-August/004693.html where Nvidia and other vendors are working with our developers to provide tested drivers.
[23:45] <Kon> Bashing-om: Ubuntu already addressed this in 19.10 by starting to push latest stable Nvidia updates. Which I and most users support. Other Ubuntu-based spinoff distros do it as well
[23:46] <Kon> However, I think instead of ONLY supporting latest release, Ubuntu should take a lesson from the old PPA and support both official Nvidia branches. Nvidia always has a short-lived and long-lived branch.
[23:46] <Kon> Prior to 19.10, Ubuntu only supported the long-lived branch and provided updates during point releases
[23:46] <Kon> And now it's the polar opposite, forcing users onto the latest short-lived update
[23:49] <Kon> Maybe I sound like Goldilocks, but troubleshooting "tested updates" is a perk of proprietary drivers :)
[23:50] <Kon> All my hardware runs on free drivers now, so I hadn't experienced this situation until someone else came to me about it