[00:11] <coryc257> LjL, are you certain your hardware doesn't support 64 bit? some embedded things ship with a 32bit kernel for reasons
[00:13] <B0RN> I am creating a user using this command, "useradd username -m -g users -G wheel,storage,power,network,video,audio,lp". Can someone explain to me why "group: users" is generally not recommended? Reference: https://wiki.archlinux.org/title/Users_and_groups#Unused_groups
[00:17] <coryc257> B0RN, my guess is to avoid potential issues if you run into a poorly written program that magically grants access to something without going through pam/polkit
[00:19] <tomreyn> B0RN: please don't cross post.
[00:19] <B0RN> sorry, my bad.
[00:21] <sarnold> B0RN: it used to be popular to have all users in a 'users' group or 'students' group or 'staff' group oor something like that
[00:21] <sarnold> B0RN: but it was dangerously easy for people to make files and directories with vastly more group access than they intended
[00:22] <B0RN> what should command should I use? I will only create one user, and I will never add another user.
[00:24] <LjL> coryc257, right now i'm on holiday so i can't even check, it was just a thought. but whatever it comes with, i'd have to use whatever coreelec ships with, i don't have a very wide choice of what Kodi distros i can install on it
[00:24] <coryc257> I'd just add the user and add it too groups as you find something is broken if you: useradd username -G wheel
[00:25] <coryc257> you will have a user with a default group of username and an additional group of wheel
[00:25] <coryc257> with wheel assuming your system is set up to sudo based upon wheel you can sudo add groups as you need them
[00:26] <URUGU4Y> HOLA
[00:26] <URUGU4Y> LLLALALA
[00:26] <tomreyn> URUGU4Y: welcome to ubuntu support
[00:27] <B0RN> coryc257, so this is useless, storage,power,network,video,audio,lp?
[00:27] <URUGU4Y> HOLA
[00:27] <URUGU4Y> HELLO
[00:28] <coryc257> B0RN, I'm not sure if you need them or not. that depends on your distro and how they have stuff set up
[00:28] <tomreyn> URUGU4Y: please don't type in ALL CAPS, ask a support question if you have one. keep things on as few lines as possible. you are currently muted.
[00:28] <coryc257> i think ubuntu defaults to: adm cdrom sudo dip plugdev lpadmin lxd sambashare
[00:30] <B0RN> i use kiss linux, coryc257.
[00:31] <tomreyn> B0RN: then you're not asking on the right channel
[00:31] <URUGU4Y> what?
[00:31] <B0RN> no, I mean use ubuntu and kiss both.
[00:31] <B0RN> I mainly use ubuntu
[00:32] <coryc257> what system are you trying to add the user to?
[00:32] <coryc257> ubuntu?
[00:32] <B0RN> ubuntu, yes.
[00:32] <oerheks> adduser ...
[00:33] <coryc257> B0RN, sudo useradd mrbill -m -G adm,cdrom,sudo,dip,plugdev,lpadmin,lxd,sambashare
[00:35] <B0RN> thanks, coryc257. last question, do I really need that cdrom? I don't have any CD Hardware
[00:36] <coryc257> B0RN, it may do usb mounting as well. I'm not sure
[00:38] <B0RN> thank you for your kindness, coryc257!
[01:08] <CanisElSaviore> Finally!
[01:09] <CanisElSaviore> The Freenode nonsense came as a surprise
[01:09] <sarnold> heh, yeah
[01:10] <CanisElSaviore> I just wanted to post an answer it took me way too long to find about dual booting, and linux not being able to use your NIC
[01:11] <CanisElSaviore> https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi#about_dual-boot_with_windows_and_fast-boot_enabled
[01:11] <CanisElSaviore> I didn't realize it was dual boot specific.
[01:12] <CanisElSaviore> you have to disable fast boot in windows to make a long story short
[01:12] <sarnold> windows fast boot is pretty frustrating
[01:14] <CanisElSaviore> I kept wondering how the linux guys broke it.  I felt kind of dumb after I read it.
[01:16] <sarnold> heh I'm surprised you were even able to get into linux, I thuoght the fastboot thing meant the startup was pretty much locked to windows
[01:19] <CanisElSaviore> Some folks on the linux side made it easier to deal with uefi, or whatever it is called.  The only issues I have been having was the inability of linux to use the NIC, and sometimes frequent Bluetooth disconnects.
[01:21] <pi0> on ubuntu 20.04, i have a little working to launch a program from cli, i need to add something to my .bashrc an export, if i want to launch the application from the menu
[01:21] <pi0> that does not work
[01:21] <pi0> is there a place to set a global export so i can launch my app directly from the menu
[01:24] <oerheks> make a proper starter. and add the program to yout %path%
[01:25] <oerheks> .bashrc is the place to be, for user level
[01:29] <pi0> ah ok
[01:35] <CanisElSaviore> Now if I can just figure out why my HDD is constantly being accessed on linux.  The cure for it on windows didn't work.  On windows it was FireFox and it's backups being rerouted to my Drive.  Found the fix.  The fix didn't work on linux
[01:39] <sarnold> CanisElSaviore: you might have luck with the biotop-bpfcc tool from the bpfcc-tools package
[01:39] <sarnold> CanisElSaviore: you can see an example of it on https://github.com/iovisor/bcc/blob/master/tools/biotop_example.txt
[01:39] <sarnold> CanisElSaviore: I'm pretty sure you have to disable secureboot to use it though :(
[01:39] <sarnold> there might be other tools that don't require disabling secureboot
[01:40] <CanisElSaviore> Hmm
[01:44] <tomreyn> CanisElSaviore: it's probably you irc client, configured to log to disk? other than "biotop", there's also the classic user space based "iotop", which should not depend on disabling secureboot. the results will probablly not be as exact as biotop's, but for answering the "what keeps waking my disk" question, it shoul dbe good enough.
[01:45] <sarnold> oh nice. I knew there was something else :)
[01:48] <tomreyn> and i didn't know about biotop, thanks for that!
[01:50] <CanisElSaviore> @tomreyn, It isn't my primary disk thogh.  That is an SSD.  It's one of my Hdd's.  I have been having this issue for years.  It might just be some idiosyncratic factory thing though. I have 2 that are suppose to be idential, but only one spins, and sounds as if it's being accessed constantly, while the other acts like it is supposed to.
[01:51] <tomreyn> maybe give them both a smartctl run
[01:51] <XV8> Are PPAs something the developer hosts or is it something that is ubuntu/canonical specific?
[01:51] <tomreyn> !smart | CanisElSaviore
[01:51] <CanisElSaviore> Using some FF workarounds I got it to stop spinning constantly.  But I can't see what is accessing the disk in linux.
[01:51] <CanisElSaviore> It's telling me that nothing is, though I can hear it.
[01:51] <sarnold> XV8: they're almost certainly canonical/ubuntu -- while there *are* other people running launchpad, it's very rare
[01:53] <CanisElSaviore> I'm still trying to sort my way through what biotop actually is.  There are a lot of words for a drunk guy that digs holes to absorb.
[01:53] <XV8> sarnold, There's a PPA I'm trying to add, but I'm getting GPG timeouts. Wondering if that's one of those rare cases.
[01:55] <CanisElSaviore> ok.  I remembered how to read from the beginning.
[01:55] <XV8> Oh, now it wants to work that I passed the debug flag... probably a hiccup
[02:30] <tianyu> h
[02:57] <d_rwin> my wifi interface name was changed to wlo1. how are they different from wlan0 ?
[02:58] <d_rwin> any help on ubuntu network configuration types here?
[03:20] <newdimension> How come these variables aren't being set? https://dpaste.org/Hg0w
[03:21] <newdimension> I'm on Ubuntu Ubuntu 20.04.2 LTS for reference
[03:24] <Ali-Karam> did you source file ?
[03:24] <newdimension> Ali-Karam, no I logged out and then logged in. The goal of this is for it to be run on login
[03:26] <Ali-Karam> these are the contents of .bashrc and .profile ?
[03:26] <newdimension> Yes
[03:26] <newdimension> My understanding is that .profile should be called on login so the variables should be set
[03:32] <Ali-Karam> try echo $PYENV_ROOT
[03:32] <Ali-Karam> no output?
[03:33] <Ali-Karam> I'm new to linux, just trying to help you if I can
[03:34] <newdimension> Ali-Karam: thanks for trying. I think this is going to be related to precedence of how the files load
[03:37] <shimbles> what does this do  sudo aptitude install '~T'
[03:38] <rud0lf> installs everything but "T", it's negation ;)
[03:39] <shimbles> i'm going to try it
[03:39] <Ali-Karam> newdimension, there is a note at the bottom of the command you entered. Did you do it?
[03:39] <Ali-Karam> Note: If you have ~/.bash_profile, make sure that it too executes the above-added commands, e.g. by copying them there or by source'ing ~/.profile
[03:40] <shimbles> yes, make sure to put sudo aptitude install '~T' in ~/.bashrc as well
[03:41] <Ali-Karam> that wasn't for you shimbles
[03:41] <shimbles> i'm sorry i am struggling with brain damage
[03:41] <shimbles> when i attempted to trigger your sense of humor, i noticed your brain was damaged
[03:42] <Ali-Karam> sorry for not getting that :/
[03:44] <shimbles> aptitude gave a Segmentation Fault with aptitude install '~T'
[05:56] <RsInc> How To Install And Configure Sendmail On Ubuntu
[06:01] <akshansh> hi i installed yandex.disk on my machine and now i uninstalled it but the option 'upload via yandex.disk' is still there when i right click any file. please help
[06:27] <RaimondRaj> hai all
[06:27] <RaimondRaj> how do i configure email on ubuntu
[06:28] <yvehan> You can use terminal mutt client
[06:28] <yvehan> Raimomdraj
[06:30] <RaimondRaj> can u guide me please
[06:30] <RaimondRaj> i see true youtube and some website
[06:30] <RaimondRaj> but still im lost
[06:30] <RaimondRaj> unable to do it
[06:31] <kushal> RaimondRaj, what exactly do you want to do?
[06:31] <RaimondRaj> acct i need to cofi for nickserv register email
[06:32] <kushal> sorry can not parse that line.
[06:33] <yvehan> You mean register your Nick name
[06:33] <yvehan> ...
[06:33] <yvehan> That's it oh my goodness
[06:33] <RaimondRaj> its like went user register new nickmail its will send email true email to send confirm nickname
[06:34] <yvehan> Just type /msg NickServ REGISTER YourPassword youremail
[06:35] <RaimondRaj> new nickname register will get email to confirm their nickname
[06:35] <RaimondRaj> some server dia setup email for confirmation
[06:35] <yvehan> In sever channel
[06:35] <RaimondRaj> oh god
[06:35] <yvehan> Are you a newbie
[06:36] <yvehan> Also here actually
[06:36] <RaimondRaj> yvehan : im running new irc server need to confi sendmail
[06:37] <yvehan> You are building a IRC server
[06:37] <yvehan> Right, I have no idea
[06:38] <RaimondRaj> yes
[06:38] <yvehan> Search with google
[06:38] <yvehan> I'm working now
[06:38] <yvehan> Workday
[06:38] <yvehan> Sorry
 "Could be your wine config is set..." <- Sorry, got pulled away and forgot I asked here. :) Would the matter with Proton? I thought proton used all its own internal stuff? That being said, I've got everything installed as far as I can tell?
[08:13] <confluence> Is there a specific channel here for Debian and Ubuntu package developers/maintainers? I specifically have a question about apt's conflict resolution mechanism.
[08:15] <guiverc> confluence, the package itself may provide clues as to which room is appropriate; ie. which team is responsibile for the unspecified package?
[08:16] <confluence> I *am* the team responsible for the package. I'm looking for information for package maintainers.
[08:16] <confluence> I'll try in #ubuntu-packaging; it's small, but I'll see if someone respondes. :)
[08:18] <guiverc> yep that would be the most appropriate
[08:32] <alkisg> guiverc: for packages in universe etc, there used to be #ubuntu-motu, not sure if it's active anymore
[08:33] <alkisg> While if it's a personal package not in universe, then yeah #ubuntu-packaging sounds appropriate
[08:48] <confluence> Yeah, it's a custom package in a PPA.
[09:18] <alzgh> Hello, I'm on an Ubuntu focal. I'm wondering why snap mounts a device for every app I install? Devices names "loop0", "loop1"...
[09:36] <Liblx> Hello terminal-lovers, the buffer is limited, so I can only scroll up a defined number of lines. Can I change this number/buffer, so I can scroll more "up" in the terminal?
[09:41] <alzgh> if you are using an emulator like Konsole, there is a setting for it in the profile settings
[09:41] <Liblx> alzgh: not for the terminal?
[09:42] <Liblx> alzgh: yes, it's really weird. one of the reasons many uninstall snap.
[09:43] <alzgh> terminal like a login terminal or is it an emulator that you run from a desktop environment?
[10:03] <Psy-Q> the likelihood of getting iptables 1.6.2 packaged for 18.04 is not very high judging by this, right? https://bugs.launchpad.net/ubuntu/+source/iptables/+bug/1805543
[10:05] <Psy-Q> it causes connection resets on any kubernetes setup with 18.04 nodes and e.g. Calico as CNI since it can't make use of --random-fully for MASQUERADE (Calico can detect that it needs to do this, but only iptables >=1.6.2 exposes the option)
[10:12] <bob2> Hello everyone. I can't install the "gi module" with python3. It is driving me nuts. Any help greatly appreciated.
[10:13] <zaggynl> hey bob2, what have you tried and what (error) messages are you getting?
[10:14] <bob2> zaggynl, I've tried "pip install gi" and I've tried "apt-get install python3 gi". I even tried "conda install gi".
[10:14] <bob2> python3-gi*
[10:15] <zaggynl> I forget if ubuntu 20.04 defaults to python3 or python2.7, does it work with pip3 install gi?
[10:16] <zaggynl> assuming you're on 20.04
[10:16] <matsaman> bob2: talking about pygobject?
[10:19] <bob2> zaggynl, Ubuntu 20 comes with python2.7 installed which is no longer supported. "pip2.7 install gi" works fine. But "pip3 install gi" fails.
[10:20] <matsaman> what's it say when it fails?
[10:21] <bob2> ERROR: Could not find a version that satisfies the requirement gi (from versions: none)
[10:21] <bob2> ERROR: No matching distribution found for gi
[10:21] <matsaman> 'cause there's no such package in pypi
[10:21] <matsaman> try pip install pygobject
[10:22] <bob2> ERROR: Could not find a version that satisfies the requirement pyobject (from versions: --1.1.1, -1.1, -1.1.4, 1.0, 1.1.2, 1.1.3)
[10:22] <bob2> ERROR: No matching distribution found for pyobject
[10:24] <matsaman> bob2: what command is that output from?
[10:24] <bob2> oops typo
[10:24] <bob2> pyggobject and pycairo are both installed
[10:28] <bob2> This command fails: python3 -m pip install -U PyGObject
[10:28] <bob2> ERROR: Could not build wheels for PyGObject which use PEP 517 and cannot be installed directly
[10:31] <matsaman> (you don't have to worry about case with pip, 'pygobject' will suffice)
[10:32] <matsaman> bob2: does it say to upgrade pip?
[10:33] <bob2> matsaman, for pip2.7 yes.
[10:33] <matsaman> what you should probably do is python -m venv ~/my-venv; source ~/my-venv/bin/activate; pip install -U pip; pip install gobject
[10:37] <matsaman> bob2: or you could try pip install --no-binary :all: pygobject
[10:37] <bob2> That seems to have worked but I have a new error message now.
[10:38] <bob2> gi.require_version("Gtk","3.0")......  ValueError: Namespace Gtk not available
[10:41] <matsaman> https://stackoverflow.com/questions/56823857/valueerror-namespace-gtk-not-available
[10:44] <bob2> matsaman, yes I tried everything on that page. It also seems to have gone back to the original "gi error" for some inexplicable reason.
[10:44] <bob2> I don't have a good understanding of python environments
[10:44] <bob2> I am in (v-env)(base). I think I want to go back to (v-env)
[10:45] <matsaman> oh okay
[10:45] <matsaman> you probably need to run 'deactivate' a couple times
[10:45] <matsaman> till your prompt is 100% normal
[10:45] <matsaman> and then source path/to/venv/bin/activate _once_ for the right path
[10:46] <matsaman> if you're not sure which is right, then just start from scratch again: python3 -m venv new/venv/path; source new/venv/path/bin/activate
[10:46] <matsaman> #python can probably help with whatever you're doing, too
[10:47] <Guest64> Hi
[10:47] <matsaman> hi
[10:49] <Guest64> I have a laptop, bought in 2013-14 with 8 gb ram. I have been looking at different DEs but not able to settle on one. The problem is that the icons and fonts of all the DEs are very big. Can anyone suggest any DE
[10:51] <bob2> Thanks matsaman.
[10:52] <matsaman> Guest64: most of the popular DEs, you can configure icon size
[10:52] <matsaman> and font size
[10:55] <Guest64> The launcher for example covers half of the screen and the windows cover the whole area, the minimise, maximise buttons are big
[10:55] <Guest64> i mean fo desktop i understand
[10:55] <Guest64> *for
[10:56] <Guest64> but for laptop its really annoying
[11:31] <Ali-Karam> how do I configure my cpu so that it's set for maximum performance?
[11:43] <cbreak> that should be automatic.
[11:44] <cbreak> unless you mean overclocking?
[12:00] <Liblx> alzgh: no, I mean: 'gnome-terminal'
[12:01] <Liblx> How can I see more lines when I scroll up in gnome-terminal?
[12:02] <alzgh> Liblx Preferences->Settings->Scrolling tab->Limit Scrollback to...
[12:04] <Liblx> alzgh: cool! thanks. And oh, it's already 10,000 ... hm. :D
[12:07] <Liblx> alzgh: do you also know how to change the sound of the 'terminal bell'?
[12:32] <ledeni> Liblx, Settings--Sounds--Alerts--Sounds and if you prefers change it look for /usr/share/sounds/gnome/default/alerts/
[12:48] <lotuspsychje> ioria: seems like that guy from yesterday ran into mesa on that bug #1939447
[12:48] <ioria> let me see lotuspsychje
[12:50] <alzgh> hey, Whenever I pause/play on vlc, if the pause is longer than a few seconds, I get a silence of a few seconds in the audio when I click play again. I looked this up and saw some other folks have this too, but couldn't find a definite solution
[12:50] <alzgh> your talk of vlc on Ubuntu reminded me of this
[12:50] <alzgh> I'm on focal
[12:53] <ioria> lotuspsychje, i see; i have already read somewhere about vdpau problems with vlc ...
[12:54] <lotuspsychje> ok, tnx ioria lets keep an eye on those
[12:54] <ioria> sure
[12:58] <lotuspsychje> alzgh: other mediaplayers dont have that lag?
[12:59] <Guest9659> Is there a way to down load the source code of "linux-image-extra-4.4.0-66-generic" package?
[12:59] <alzgh> lotuspsychje haven't checked any others. Is there any you would recommend checking?
[12:59] <lotuspsychje> alzgh: smplayer would be a good test
[13:00] <alzgh> thanks, will check and come back
[13:06] <coconut> Guest9659, you can with apt source, but you have to comment out deb-src lines first in your sources.list config.
[13:08] <alzgh> lotuspsychje checked with smplayer and mpv player, they don't have the few seconds of silence like VLC.
[13:10] <lotuspsychje> alzgh: something like bug #1856036 ?
[13:12] <coconut> Guest9659, assuming version 4.4.0-66 is still part of your repo
[13:16] <alzgh> hey lotuspsychje sorry, I lost connection. Can you post the issue # again?
[13:17] <lotuspsychje> alzgh: can you check if you have the symptons of bug #1856036
[13:17] <alzgh> yes, will do
[13:21] <Guest9659> coconut: sudo apt-get source linux-image-extra-4.4.0-66-generic
[13:21] <Guest9659> ?
[13:21] <Guest9659> coconut: is the above command correct?
[13:21] <leftyfb> Guest9659: what version of ubuntu are you running?
[13:22] <Guest9659> leftyfb: Ubuntu 16.04.4 LTS
[13:22] <leftyfb> 4.4
[13:22] <leftyfb> !eol | Guest9659
[13:23] <leftyfb> Guest9659: your efforts should be put into upgrading the release, not looking for the kernel source
[13:23] <leftyfb> !esm | Guest9659 if you'd like support with EOL releases, you can get ESM support through Canonical
[13:26] <lotuspsychje> maybe if Guest9659 would clarify his purpose, volunteers could help think along?
[13:27] <BluesKaj> Hi all
[13:30] <alzgh> lotuspsychje it's more like this https://code.videolan.org/videolan/vlc/-/issues/25964 .
[13:31] <lotuspsychje> alzgh: if they claim to be fixed in vlc 4, you could perhaps test the vlc snap dev channel
[13:31] <lotuspsychje> alzgh: snap info vlc
[13:33] <alzgh> will check it, lotuspsychje . thanks
[14:14]  * wonko[m]  < https://libera.ems.host/_matrix/media/r0/download/libera.chat/04fdc00197d52cd7d8f8836913618f3fc1b77043/message.txt >
[14:19] <wonko[m]> weird, I can get there with a browser
[14:19] <wonko[m]> but pastebinit gets a 502
[14:34] <Liblx> ledeni: thanks for the sound alert infos. perfect.
[14:55] <coconut> wonko[m], my pastebinit does not work either... have you tried the <command> | nc termbin.com 9999 #alternative
[14:56] <coconut> ?
[15:37] <nick7465> Hello! I am running 20.04.2 and am having a bit of difficulty with the nvidia drivers. I tried from both the software updater GUI and terminal to install/update the latest driver from nvidia (470) but am receiving "unmet dependencies" and "you have held broken packages" errors.
[15:38] <nick7465> I ran dpkg --get selections | grep hold and it doesn't show anything so I'm sure it's only the unmet dependencies
[15:38] <nick7465> How can I figure out how to fix this?
[15:39] <nick7465> the dependencies listed are all nvidia-based, so I'm confused as to why it doesn't just pull them as part of the install
[15:53] <coconut> nick7465, show the error you got so volunteers can try and help you
[15:54] <nick7465> sudo apt install nvidia-driver-470
[15:54] <nick7465> Reading package lists... Done
[15:54] <nick7465> Building dependency tree
[15:54] <nick7465> Reading state information... Done
[15:54] <nick7465> Some packages could not be installed. This may mean that you have
[15:54] <nick7465> requested an impossible situation or if you are using the unstable
[15:54] <coconut> !paste | nick7465
[15:55] <nick7465> Oops! Sorry about that, thank you. Let me try the other paste method.
[15:56] <nick7465> https://paste.ubuntu.com/p/zr8XjsjjbQ/
[15:57] <nick7465> In the GUI it shows that I'm using a proprietary driver now (460) but in the installed applications it also shows the Xserver app which I believe is the open source Nouveau driver
[15:58] <nick7465> Performance certainly feels like the Nouveau driver
[15:59] <nick7465> "ubottu" haha that's a clever bot name
[16:02] <ioria> nick7465, can you paste dpkg -l | grep nvidia   ?
[16:04] <nick7465> https://paste.ubuntu.com/p/PHgggn9pTf/
[16:05] <ioria> nick7465,  lspci -nnk | pastebinit
[16:06] <nick7465> What does this command do? It's asking me to install pastebinit
[16:06] <ioria> nick7465, no worries ; paste '  lspci -nnk '
[16:07] <nick7465> https://paste.ubuntu.com/p/pHzZnhy5f4/
[16:09] <ioria> nick7465, as you can see, you're using nvidia module and you have multiple versions  installed ... and you want to install  also 470 ?
[16:11] <nick7465> Hmm, I am guessing the multiple versions is potentially the issue... I'm mainly trying to troubleshoot why I get such poor performance compared to using Windows. I can run games 144fps on Windows but the same game (made for Linux version of course) I will get 20fps.
[16:11] <nick7465> I thought upgrading to 470 might help
[16:11] <nick7465> But I'm kind of just throwing darts at a wall here
[16:12] <ioria> nick7465, if you ask me, purge the previous version(s), before installing a new one
[16:12] <Walex> nick7465: my experiene with Steam games (not the latest though) is that they run pretty well. 20 FPS seems to show that it is running without GPU accel.
[16:13] <nick7465> How can I do that?
[16:13] <Walex> nick7465: it is not easy to install NVIDIA drivers "right", in many cases.
[16:13] <nick7465> Yes I agree, Walex. I for sure am having an issue I really don't think it's the game itself
[16:13] <Walex> nick7465: the first thing to check is whether you do have acceleration.
[16:13] <ioria> nick7465, you have installed 440 450 460
[16:14] <nick7465> Oh my. How can I purge all of them, Ioria?
[16:14] <ioria> nick7465, does it work your card with nouveau ?
[16:15] <nick7465> Walex, I am very new to Linux still. I am not sure how to do that
[16:15] <Walex> nick7465: that
[16:15] <Walex> nick7465: that's why we are here and can tell you.
[16:15] <nick7465> Ioria, I have no idea.
[16:15] <Walex> nick7465:  which package manager do you usually use for installation?
[16:15] <nick7465> I feel really dumb, I'm sorry. But thank you guys very much for helping.
[16:15] <ioria> nick7465, what's the nvidia version actually in  use ?
[16:16] <nick7465> I usually use apt via terminal. I think I have one snap package installed but can't remember.
[16:16] <nick7465> According to the software updater app, it says the 460 is in use.
[16:17] <ioria> nick7465, nvidia-smi what it says ?
[16:18] <nick7465> https://paste.ubuntu.com/p/3XfPmbFQ24/
[16:18] <varaindemian> running nested virtualization (ubuntu host -> ubuntu guest -> another ubuntu guest). Any idea why everthing is so slow? Also, Virtualbox
[16:19] <ioria> nick7465, sudo apt nvidia-driver-440  && sudo apt autopurge
[16:19] <ioria> nick7465, sudo apt purge nvidia-driver-440  && sudo apt autopurge
[16:21] <nick7465> Done. Should I also do the same for 450, Ioria?
[16:21] <Walex> nick7465: your latest paste shows that various processes are using the NVIDIA libraries...
[16:22] <Walex> nick7465: 'nvidia-smi' seems to show that your installation is mostly right.
[16:22] <nick7465> Walex, that makes sense. I run dual monitors though my secondary monitor's max resolution is 1024x768 according to the software/OS but the monitor's actual capability is far greater (1080 monitor)
[16:22] <ioria> nick7465, yes
[16:23] <nick7465> Oh hm, it says 450 is not installed so it wasn't removed. Logical, but I thought we saw it was. Interesting.
[16:23] <nick7465> Maybe it was purged when I did the 440
[16:24] <ioria> nick7465, again dpkg -l | grep nvidia please
[16:24] <Walex> nick7465: you also have CUDA installed and then it is possible to do some quick speed tests.
[16:24] <amcsi> hey
[16:25] <ioria> nick7465, if you have cuda that's probably the culprit
[16:25] <nick7465> https://paste.ubuntu.com/p/cDv25pcrZQ/
[16:25] <nick7465> I don't know if I have cuda or not, why would that be problematic? Don't nvidia GPUs have cuda cores regardless?
[16:25] <Walex> ioria: CUDA does not make OpenGL become slower.
[16:26] <jpmh> I'm managing to get very confused.  Does fail2ban block ipv6 miscreants as well as i?v4?
[16:26] <ioria> Walex, but can mess up, with the package manager
[16:26] <nick7465> dpkg -l | grep nvidia looks a lot cleaner now, but nvidia-smi no longer works
[16:26] <Walex> also a little known secret: the most reliable NVIDIA drivers are those in the CUDA repository, because they have to be :-)
[16:26] <nick7465> Should I purge 460 and install 470?
[16:27] <ioria> nick7465,  no it's wrong i don't see 460
[16:27] <nick7465> Uh oh
[16:27] <nick7465> In that paste it mentions 460?
[16:27] <nick7465> Ah okay in software-updater it shows I'm using the nouveau driver now
[16:28] <ioria> nick7465,  the pkg is calld nvidia-driver-460
[16:28] <Walex> nick7465: not necessariy, you have not rebooted.
[16:28] <nick7465> oh I see there's still two 440s listed in that last paste
[16:28] <ioria> nick7465,  dpkg -l | grep -i cuda
[16:28] <Walex> nick7465: existing processes will continue to use the NVIDIA libraries.
[16:28] <nick7465> oh my this is confusing! haha. Should I reboot?
[16:28] <ioria> nope
[16:28] <ioria> nick7465,  dpkg -l | grep -i cuda
[16:29] <Walex> nick7465: not yet reboot
[16:29] <nick7465> That command returns nothing
[16:29] <nick7465> no error, just nothing
[16:29] <ioria> ok
[16:29] <amcsi> I seemed to have broken my ubuntu install by starting my dual-booted windows and shutting down before it started to boot. While trying to set up grub again on my ubuntu (which is on a different SSD than my windows), I seemed to have trouble setting up GRUB+EFI for my ubuntu. Based on searching around on the internet, I think the source of the problem may be that there's no EFI partition on my ubuntu SSD. When listing the partitions, I saw
[16:29] <amcsi> that there were only the / partition, the /home partition, and swap. Assuming nothing outright removed any other partition, and that there needs to be an efi partition to install an efi bootloader, then might the only explanation be that my EFI partition was on my Windows drive all along? Because I do see an EFI System partition there. Also I believe I installed Windows first
[16:30] <ioria> nick7465,  sudo apt purge inux-modules-nvidia-440-5.4.0-42-generic inux-modules-nvidia-440-5.4.0-47-generic
[16:30] <Walex> amcsi: there must be only one EFI partition.
[16:30] <ioria> nick7465,  sudo apt purge inux-modules-nvidia-440-5.4.0-42-generic inux-modules-nvidia-440-5.4.0-47-generic
[16:31] <ioria> nick7465,  sorry,   sudo apt purge linux-modules-nvidia-440-5.4.0-42-generic linux-modules-nvidia-440-5.4.0-47-generic
[16:31] <Walex> amcsi: unless you do really weird things like switching the "primary" drive in the BIOS before booting.
[16:31] <nick7465> hehe figured that one, no problem! I ran it with the missing l's and it removed both
[16:31] <amcsi> Walex, oh, I mounted /sdb2 (the windows drive's EFI partition), and in there I see an EFI/ubuntu
[16:31] <ioria> nick7465,  sudo apt purge libnvidia-compute-460
[16:32] <Walex> amcsi: amcsi: the way Ubuntu boots with EFI is slightly weird, so don't look into that.
[16:32] <amcsi> so that means the EFI partition was on the drive meant for windows all along most likely, right?
[16:32] <amcsi> ok
[16:32] <nick7465> ioria done!
[16:32] <ioria> nick7465,  sudo apt autopurge
[16:32] <nick7465> ran, nothing changed
[16:33] <ioria> nick7465, uname -r  (you can paste here)
[16:33] <nick7465> 5.11.0-25-generic
[16:33] <Walex> amcsi: how much do you know about system administration? Can you boot a standalone "liveCD"/"liveUSB" Linux without further details?
[16:33] <ioria> nick7465, sudo apt update && sudo apt full-upgrade
[16:33] <amcsi> Walex, yes
[16:34] <nick7465> ran, no changes (I ran this about an hour ago)
[16:34] <amcsi> Walex, so I would like to appreciate a little help in the following:
[16:34] <Walex> amcsi: do you know how to then 'chroot' to your '/' partition and run commands in it?
[16:34] <ioria> nick7465, dpkg -l | grep nvidia
[16:34] <amcsi> Walex, yeah, mostly
[16:34] <nick7465> didn't return anything ioria
[16:34] <ioria> ok
[16:34] <amcsi> but wait a second please
[16:35] <ioria> nick7465, run this 'apt -s install nvidia-driver-470 --no-install-recommends'
[16:35] <Walex> amcsi: ok, give us the further details.
[16:35] <nick7465> Ran very quickly, seemed successful?
[16:35] <ioria> can we see ?
[16:36] <amcsi> so initially I tried to run grubinstall --boot-directory=/mnt/boot /dev/sda, were /mnt is where my ubuntu root is mounted to, and /dev/sda being the device obviously. But I got: warning this GPT partition label contains no BIOS Boot partition yada yada
[16:36] <nick7465> Oh "this is only a simulation" haha! no wonder it ran quick. Yes let me paste ioria
[16:36] <amcsi> but now I realized that the EFI partition was on the other device
[16:36] <amcsi> in /dev/sdb2
[16:36] <nick7465> ioria https://paste.ubuntu.com/p/f6q39Mybs8/
[16:36] <amcsi> now I mounted /dev/sdb2 to /mnteffi
[16:36] <amcsi> I mean /mntefi
[16:37] <ioria> nick7465, now this :  'apt -s install nvidia-driver-470'
[16:37] <amcsi> Walex, so might you know how I would need to continue from here to finish the grub reinstall?
[16:37] <amcsi> I'll need to somehow let the grub installer know about where my efi is
[16:38] <nick7465> ioria https://paste.ubuntu.com/p/cgvrNNrwsZ/
[16:38] <Walex> amcsi: that is "all wrong" unfortunately.
[16:38] <ioria> nick7465, looks ok now
[16:38] <Walex> amcsi: your EFI partition must be on the prima boot device.
[16:39] <nick7465> I agree, Ioria. Should I run one of those two simulated commands for real now?
[16:39] <Walex> amcsi: the EFI is actually a MS.DOS like operating system
[16:39] <ioria> nick7465, try to install it :  sudo apt install nvidia-driver-470'
[16:39] <Walex> amcsi: that gets booted by the BIOS built into the motherboard.
[16:39] <ioria> nick7465, try to install it :  sudo apt install nvidia-driver-470
[16:39] <amcsi> Walex, it couldn't be all wrong, because this is how it's been working...
[16:39] <Walex> amcsi: and its "root" partition must be on the primary boot device.
[16:40] <Walex> amcsi: then perhaps you have a BIOS that can boot the EFI OS from a secondary boot device, but that is unlikely.
[16:40] <amcsi> I installed Windows first on the windows drive which created an EFI partition there. Then I installed ubuntu on the other drive, and it likely reused the same EFI partition as windows, no?
[16:40] <amcsi> on the other device
[16:40] <nick7465> It would then install Grub wouldn't it?
[16:41] <Walex> amcsi: what is then more likely is that the BIOS and Linux list devices on a different order
[16:41] <Walex> amcsi: so 'sdb' is actually at the BIOS level the primary device (0x80 in MS-DOS terms).
[16:41] <nick7465> Ioria, it seems to have worked fine! Installed about 788mb. No errors. Do I need to change any configuration or settings or whatnot to use it? Or reboot etc?
[16:42] <Walex> amcsi: in that case it is a bit of a mess to get all the details right.
[16:42] <Walex> amcsi: does MS-Windows boot right?
[16:42] <ioria> nick7465, you know... reboot (if you got a blank screen you need to boot with nomodeset)... i suggest to make grub show at startup
[16:42] <amcsi> Walex, yes, windows does boot. And btw remember, I said that on the windows drive's efi partition I saw ubuntu
[16:43] <Walex> amcsi: again things are a bit strange there.
[16:43] <Walex> amcsi: does GRUB2 start at all?
[16:43] <Walex> amcsi: or rather:
[16:43] <amcsi> Walex, no. If I start the Linux partition, I don'T even get there...
[16:44] <amcsi> Walex, I get this: https://www.dropbox.com/s/dormqhnjkmj547a/2021-08-11%2016.43.55.jpg?dl=0
[16:44] <Walex> amcsi: when MS-Windows boots, does it show first the GRUB menu or just the MS-Windows boot menu?
[16:44] <ioria> nick7465, to make grub show you can use this config in /etc/defgault/grub : https://paste.ubuntu.com/p/FvFkXhf3Qz/    then run sudo update-grub
[16:44] <nick7465> ioria, grub shows at startup since I dual boot ubuntu + windows. Do I need to do anything else?
[16:44] <ioria> nope, good
[16:45] <nick7465> Okay perfect
[16:45] <amcsi> Walex, if I choose the windows partition as the boot device, then it just boots straight to windows
[16:45] <Walex> amcsi: in what kind of menu do you choose MS-Windows?
[16:45] <amcsi> if I choose the ubuntu partition, then I get that failure in the dropbox link, and no grub
[16:45] <nick7465> I hope to return soon with good news! I think I might need to relog here as different nick since I'm not registered but we'll see. I'll ping you either way Ioria
[16:45] <ioria> ok gl
[16:45] <amcsi> Walex, I simply choose the boot device manually in my BIOS to get windows to run
[16:45] <nick7465> Thank you so so much for your help Ioria and Walex. I will return shortly!
[16:46] <Walex> amcsi: chnaging the boot device often creates a lot of problems unless you really know the details how how BIOS, EFI, GRUB2 work.
[16:47] <Walex> amcsi: the crucial details from the Ubuntu point of view is that when 'grub-install' runs '/boot' and '/boot/efi' must have the right filestems/partitions mounted.
[16:48] <amcsi> Walex, this is just the simple Boot Order stuff. Like how if your live usb does not run, you'd tell your bios to prioritize your usb devicwe
[16:48] <Walex> amcsi: if you know all the details, it can be made to work, but obviously it does not work.
[16:49] <amcsi> Walex, one quick question: does grub-install check to see if the circumstances are valid before setting up grub? so if I mounted the wrong things to the wrong folders, will it warn me?
[16:49] <Walex> amcsi: no.
[16:51] <Walex> ioria: in the case of "nick7465" cleaning up his packages is a good thing, but first I would have checked *why* it was slow... It could something simple like i386/amd64 differences.
[16:51] <ioria> Walex, yeah
[16:52] <Walex> ioria: the Steam installer does not check that all i386 dependencies needed by Steam games are installed, and many i386 Steam games are still i386, so it could have been that.
[16:52] <amcsi> Walex, so for starters are you saying I have to create a new partition acting as EFI on the same drive as my ubuntu?
[16:52] <Walex> ioria: or as you probably know quite a few other things.
[16:52] <ioria> Walex,  what worried me was the apt errors
[16:52] <Walex> amcsi: I would really avoid doing that...
[16:52] <Walex> ioria: oh yes, I can understand that.
[16:53] <amcsi> Walex, why? besides, I'm planning on installing a new physical drive in place of the windows drive, so I think it would be beneficial to have the ubuntu efi partition on the same device as my ubuntu
[16:54] <Walex> BTW as I was mentioning the most reliable NVIDIA drivers are those in the CUDA repo: http://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64
[16:54] <nick4765> ioria I can't even remember if this is the same nick I used before, but I'm back. So far it doesn't seem any different from before. My second monitor is still stuck on a low resolution, and attempting to play the game still presents 20-30 fps. Another game is asking "launch" or "launch using OpenGL" I am guessing choose the OpenGL?
[16:55] <Walex> amcsi: because unless you are completely sure that you know all the details of BIOS, EFI, GRUB2 configuration and booting, having multiple EFI partitions creates problems and ambiguities.
[16:55] <ioria> nick7465, again nvidia-smi please
[16:55] <TJ-> any way to persuade {user,group}add that an ephemeral user/group doesn't exist so it/they can be added to /etc/{passwd,group} ? I've tried 'tricking' with --prefix=/ but they still uses NSS and therefore 'find' the ephemeral user/group
[16:56] <nick4765> https://paste.ubuntu.com/p/xYD4NH74Wb/
[16:56] <amcsi> Walex, going on a little tangent, if I wanted to replace my windows drive (that has the efi partition), should I be prepared to have to reinstall ubuntu, cause it's better to not mess around with efis?
[16:57] <ioria> nick7465, now 470 is correctly in use and is the latest
[16:57] <Walex> amcsi: you never need to reinstall Ubuntu because it can always be fixed from a liveCD/liveUSB.
[16:57] <amcsi> Walex, ok... could I do the same right now with my current situation?
[16:57] <Walex> amcsi: the details may be a bit complex, but it is not like MS-Windows yet where the details are unknowable.
[16:58] <amcsi> I don't know if I mentioned, but the live usb is already running as we speak
[16:58] <amcsi> Walex
[16:58] <noarb-> I tried to upgrade to hirsuite and now I've got a broken system. It just said "failed; please contact your system administrator" during the upgrade. I restarted, and I can operate on the machine, but all the apt packages seem to be messed up. I tried sudo apt --fix-broken install and got this: https://bpa.st/3BLA
[16:58] <nick4765> ioria - sweet!! That's good news. I'm not sure what my problems are then
[16:58] <Walex> amcsi: the best and simplest thing is; a single EFI partition on the primary boot drive, which boots the MS-Windows bootloader or the GRUB2 bootloader.
[16:58] <ioria> nick7465, maybe Walex has some suggestions
[16:59] <nick4765> ioria, thank you so much for your help so far you've been amazing
[16:59] <amcsi> Walex, fun fact, the ubuntu drive has been the primary drive
[16:59] <nick4765> in the future if I wanted to update the driver, what would be the best way Ioria?
[16:59] <amcsi> so I don't even understand how that works when the EFI partition is on the windows drive
[17:00] <Walex> amcsi: please don't be offended when I say that I don't trust you to *know* which drive is the primary boot drive unless you know why that is aclled the '0x80' drive.
[17:01] <noarb-> I've tried these commands: https://askubuntu.com/a/346795
[17:01] <Walex> amcsi: some BIOSes will boot EFI and mount the first EFI partition they find on any drive, some don't
[17:01] <amcsi> Walex, I'm absolutely not offended, you are super kind
[17:01] <Walex> nick4765: which game is it?
[17:02] <amcsi> Walex, so you're saying the primary boot drive has nothing to do with the boot device order in my BIOS?
[17:02] <nick4765> Walex the one I was testing in comparison was Tropico 6, but I just installed Valheim to test that one. It's asking to launch or launch with OpenGL not sure which to do. Hate to try one and then not have that option return in the future
[17:02] <Walex> amcsi: in theory the BIOS assigns MS-DOS drive number 0x80 to the first drive listed in the boot device order.
[17:03] <Walex> nick4765: it has to be OpenGL.
[17:03] <amcsi> Walex, even if it's the live cd, right?
[17:03] <nick4765> Okay awesome, let me try Valheim and see. Though I am still having the issues with my second monitor reso being low
[17:03] <Walex> nick4765: but there is a complication: some games run under MS-Windows emulation ("proton")
[17:03] <Walex> nick4765: try Team Fortess 2 as a baseline, it is free.
[17:04] <nick4765> Yeah, I enabled Proton. should I not have?
[17:04] <nick4765> Okay I will try TF2
[17:04] <Walex> nick4765: Proton usually works, but it can add a layer of extra complexity.
[17:04] <Walex> nick4765: so to solve problems it is best layer by layer.
[17:05] <Walex> nick4765: "ioria" has got you to clean up the package install layer, now we need to check the other layers.
[17:05] <Walex> nick4765: while you wat for the download to finish, please install 'pastebinit'
[17:06] <Walex> nick4765: and then 'pastebinit /var/log/Xorg.0.log' and 'nvidia-smi | pastebinit' and 'xdpyinfo | pastebinit' and 'xrandr | pastebinit'
[17:07] <nick4765> Awesome! I just tried Valheim using OpenGL and I get significantly better performance in that game, 60-80 fps. This is still down from 144 over in Windows but way better than the 20-30 seen in Tropico. I am downloading tf2 now and will do the pastebin things
[17:08] <Walex> amcsi: when you boot a liveCD or liveUSB the primary driver gets set to the CD/USB, but also the liveCD(liveUSB things contains code that scans most of the partitions on the system to figure out how to continue booting.
[17:08] <Walex> amcsi: again, unless you really know what you are doing, never switching boot drive and using a single EFI partition is a lot easier.
[17:09] <amcsi> Walex, alright... so how do you suggest that I proceed?
[17:10] <nick4765> Walex https://paste.ubuntu.com/p/DRsmMnD7g3/ https://paste.ubuntu.com/p/VJxYh9QbF5/ https://paste.ubuntu.com/p/8k9bXyZX24/ https://paste.ubuntu.com/p/fXkqTW6jh7/
[17:10] <amcsi> you saw my dropbox image, right? I'm wondering if my EFI is screwed up and if it could simply be fixed somehow
[17:11] <Walex> amcsi: the simplest configuration is this: the primary boot drive is fixed, and it is the MS-Windows drive with the only EFI partition in the system. Given that, ubuntu can be anywhere, because it is a lot less fussy then MS-Windows as to booting locations.
[17:12] <Walex> amcsi: the boot then proceeds like this: the BIOS boots EFI, EFI mounts th EFI partition, scans it and finds the MS-Windows bootloader and the GRUB2 bootloader, and then gives you a menu. When you select GRUB2 it mounts '/boot' as its root partition and there it goes.
[17:13] <Walex> amcsi: it is very difficult to screw up EFI itself. the kernel is in the BIOS "RO" storage, and the EFI root partition only contains the EFI shell or a set of bootloaders.
[17:14] <Walex> nick4765: having a look...
[17:14] <Walex> nick4765: BTW the critical bit to look for is 'glx'
[17:15] <nick4765> walex, I just queued into a TF2 game and the performance was perfect. 250-280 fps and smooth as butter.
[17:15] <nick4765> I know it's an old game, but still.
[17:15] <nick4765> I think that's a good sign
[17:16] <Walex> nick4765: that means that your install has full OpenGL acceleration. To be really sure check with 'top' while the game is running that the CPU time is low.
[17:17] <Walex> But 250-280 FPS at full 2K resolution is probably unattainable with software OpenGL alone.
[17:19] <Walex> so far the pastes show everything is dine, e.g.  NV-GLX is paert of of the X11 extensions available, etc.
[17:19] <amcsi> by the way I have run boot-repair: https://paste.ubuntu.com/p/S5dGyHj69B/
[17:20] <amcsi> I don't have high hopes, but I'll restart and see if it fixed it
[17:21] <amcsi> woah
[17:21] <amcsi> it fixed it!
[17:21] <Walex> amcsi: the boot repair log is a bit weird but it seems good.
[17:21] <amcsi> I'm so happy right now. Thanks for all the help!
[17:21] <Walex> amcsi: note that it mounted '/dev/sdb2' as '/mnt/boot/efi'.
[17:21] <nick4765> Walex, running top while playing shows roughly 200-230% cpu usage (how is that even possible?) while it is not in focus, it's around 20-30%
[17:22] <amcsi> Walex, that's what I said was on the windows drive
[17:22] <Walex> amcsi: you need to make sure that '/dev/sdb2' is mounted as '/boot/efi' in '/etc/fstab'
[17:23] <Walex> nick4765: each CPU counts as 100% in "Irix" mode of 'top'.
[17:23] <nick4765> Oh you said Time not % usage - the time is 3:47 to 3:48 while not in focus and while playing it goes up to 4:25
[17:23] <nick4765> Ah gotcha, good to know!
[17:23] <Walex> nick4765: % is good
[17:23] <Walex> nick4765: it is a bit high for TF2 on a latest system like that.
[17:24] <nick4765> Meanwhile the "time" on the steam process is 12:06 ! weird
[17:24] <TheFork> Recently encountered an issue: "fork: retry: Resource temporarily unavailable" ends up erroring out a ton of my processes.  Looked at the fix, they mentioned it could be a memory, ulimit or kernel limitation.  So I end up checking memory (53% usage only), upping ulimit for processes to their maximum, upping files opened to their maximum, putting
[17:24] <TheFork> pid limit to the highest and doing the same within the kernel.  Memory usage is probably not an issue since I got about 50 gb left.  So after doing all that, can't think of what else would cause the problem.  It's an AWS instance so might have to do with virtualization issue.  So far have 2 computers with this issue.  Wondering if someone else has
[17:24] <TheFork> encountered this problem.
[17:24] <Walex> nick4765: is it running at full 3584x1440?
[17:24] <nick4765> It's running at my native which is 2560 x 1440
[17:25] <Walex> TheFork: check 'dmesg' it should have more details, but you should rather try #Ubuntu-server
[17:25] <TheFork> Ah thanks!
[17:26] <Walex> nick4765: I mean 2560x1440
[17:28] <nick4765> Yeah it's running at that. What <should> the % and time be? (Also, what's CPU time? As in the timing like in RAM?)
[17:28] <Walex> nick4765: there is something rather weird in your configuration: the NVIDIA driver only reports 8 DisplayPort outputs, but 'xrandr' shows a los res display as 'HDMI-0'
[17:29] <sakura> hello, I have a script to compress, encrypt and upload my files to google drive, I want to run this script as a cronjob weekly, for that, I need to set the --passphrase flag to gpg so I can encrypty my files without my interaction, reading gpg's manual, they explicitly say that the --passphrase flag is unsecure, since the value can be logged and generally unsafe on a multi-user, the same for the
[17:29] <sakura> --passphrase-file, which reads from a file. My question is, how can I safely provide my password to gpg, without user interaction, on a multi-user system?
[17:29] <Walex> nick4765: since 'xrandr' report 6 DisplayPort and 1 HDMI, perhaps your system converts one DP port to HDMI.
[17:30] <amcsi> Walex, /dev/sdb2 has EFI/ubuntu and EFI/boot as well. Still /dev/sdb2 should be mounted as /boot/efi as-is?
[17:30] <nick4765> That is weird, I agree. I also definitely don't have 8 physical ports! judging from a quick glance, it shows 3 DP ports, 1 usb-c port, and 1 HDMI port Walex. My secondary display is hooked via HDMI
[17:30] <Walex> nick4765: nick4765: those are the chipset ports, which may or may not be connected to external sockets.
[17:30] <nick4765> Ahhh, gotcha.
[17:31] <Walex> amcsi: yes, as said. Don't worry about the subdirectories.
[17:31] <nick4765> The secondary display is in HDMI, and the monitor physically can go up to 1080 native but that isn't even an option in settings. The highest I can go is 1024 x 768 which is what it's at.
[17:32] <Walex> amcsi: I have written some of the details: http://www.sabi.co.uk/blog/20-two.html?200810#200810
[17:32] <Walex> nick4765: for the secondary try to connect with DP if you can, the reports by the NVIDIA driver are a bit weird.
[17:33] <nick4765> Walex The monitor itself doesn't have a DP port, sadly
[17:33] <Walex> nick4765: I am looking for more tests, sorry I am bit slow
[17:33] <TheFork> sakura this might help: https://stackoverflow.com/questions/9460140/gpg-encrypt-file-without-keyboard-interaction
[17:33] <nick4765> walex, don't even worry! You're doing amazing work here, thank you so much for your time. I know you're busy!
[17:34] <Walex> nick4765: one of the usual tests is the 'atalantis' demo: 'apt install xscreensaver-gl-extra'
[17:34] <TheFork> usually those tools have options, but just buried in the documentation.  You can also probably use open-ssh aes encryption and store the key in a vault.  Good example being something like AWS Secret Manager service.  That kind of limits you to how safe the service login is.
[17:37] <Apachez> how do I troubleshoot why /usr/bin/gnome-shell takes so much cpu?
[17:37] <Walex> nick4765: nick4765: then run and resize the window to take most of the screen: /usr/lib/xscreensaver/atlantis -fps -count 50
[17:39] <TheFork> sakura, another option that I didn't really consider, but AWS/GCP/Azure should have blob storage with an encryption option where you store the keys within the cloud service itself.  It's maybe like .02 per gb plus nominal fee for transfers + key usage.
[17:40] <Walex> nick4765: on a really weak GPU on my laptop I get 37FPS and 6% load. The important thing is that the load should not exceed single digits.
[17:40] <nick4765> So first run apt install xscreensaver-gl-extra to install it, then /usr/lib/xscreensaver/atlantis -fps -count 50  to run?
[17:41] <Walex> nick4765: yes
[17:42] <nick4765> Walex I am seeing 38.x fps and load ranging from 3.4 to 4%
[17:43] <nick4765> Weird that it can only do that low of FPS
[17:43] <Walex> nick4765: it is good, I wonder why the FPS is low here too.
[17:44] <sakura> TheFork thanks for the answers, I just didnt want to expose my passphrase on the terminal, which can be logged by the admin, just thought a little more about it and decided to create a new passwordless key just to encrypt these files, but I'm open for better ideas.
[17:44] <TheFork> I mean that's probably a pretty good option overall
[17:45] <Walex> nick4765: I just started Steram and it looks like Valheim is native Linux
[17:46] <nick4765> Walex so in theory I could launch it without OpenGL? Or I could launch it without proton? I think I have proton set to only be used if it NEEDS to be used, so I would think in the case of Valheim it wouldn't use proton?
[17:46] <Walex> nick4765: the only libraries that are native Linux are Vulkan and OpenGL
[17:46] <nick4765> Ahh, gotcha
[17:46] <Walex> nick4765: using Proton "usually" works.
[17:47] <Walex> nick4765: you may not have Vulkan installed, so it is better to default to OpenGL.
[17:48] <nick4765> Fair enough
[17:48] <nick4765> Can you think of why my display isn't showing 1080 capability?
[17:48] <nick4765> the secondary, I mean
[17:48] <Walex> nick4765: so far all tests succeed as to speed.
[17:48] <TheFork> So long as server is well patched etc.  If it's like major corporation, then you'd typically do key vaults and key checkouts, but think for most use cases, especially personal use, it's probably overkill.  Yeah, definitely don't put the password plain text on a cronjob.  That's definitely not koshure.  All you need is someone to read off the file
[17:48] <TheFork> or check history and boom.  It's all compromised.
[17:49] <nick4765> Yeah and TF2 and Valehim both work really well and smooth so I'm not AS concerned now. I think it's mostly straightened out. I think Tropico6 just has issues period despite supposedly being native linux.
[17:50] <Walex> nick4765: for the external display I can think of several reasons, one of them is that it is the wrong cable.
[17:50] <nick4765> Walex Even if it works fine on Windows?
[17:51] <Walex> nick4765: in the Xorg server log file the only mention of HDMI is for input devices, so it is weird indeed.
[17:51] <Walex> nick4765: have you tried using the NVIDIA control panel app?
[17:52] <Walex> nick4765: 'nvidia-xconfig'
[17:52] <nick4765> Walex the only one I have is the x server settings, is this correct?
[17:53] <nick4765> I can screenshot you what I see, sec
[17:53] <nick4765> !paste
[17:53] <nick4765> imgur, got it.
[17:54] <nick4765> Walex https://imgur.com/a/D40XVhi
[17:54] <nick4765> Now when I click "auto" to change reso on the second display I see resolutions up to 1600x900 which is higher than before. Still not up to native though.
[17:56] <nick4765> Walex: Interesting when I navigate to the two screens under the "GPU 0 -" tab, there's one for HDMI-0 and DP-2; DP-2 shows the "signal" to be DisplayPort while "HDMI-0" shows the signal to be "TMDS"
[17:56] <Walex> nick4765: try putting th second monitor *under* the main monitor, not to thew right.
[17:56] <Walex> nick4765: that's weird, usually TMDS is the builtin one
[17:57] <Walex> ahhh no, LVDS.
[17:58] <nick4765> LVDS?
[17:58] <nick4765> Putting the secondary underneath my main works fine, stays the same reso though
[17:59] <Walex> nick4765: have you tried manually selecting a higher res for the secondary?
[17:59] <nick4765> Although it does make me a bit dizzy
[18:00] <nick4765> Yes, the option doesn't even exist for higher. And there's no manual option to enter in my own values
[18:00] <Walex> nick4765: some drivers have a 4096 pixel height and width limitation for the virtual display
[18:00] <Walex> nick4765: it could be a bad cable but it would be strange.
[18:01] <nick4765> I changed it to 1600x900 and it is working fine at that resolution
[18:01] <nick4765> Yeah, I'm not ruling out the cable thing for sure. Just because it works on Windows doesn't mean it's NOT bad persay.
[18:02] <Walex> nick4765: it is possible to manually add a 1920x1080 mode and force that, but it is not simple.
[18:02] <nick4765> Ahh
[18:02] <TJ-> nick4765: what does this report? "pastebinit <( hexdump -C /sys/class/drm/card0-HDMI-0/edid )"
[18:02] <Walex> I'd like to disappear for while, I will be back in around 1 hour
[18:02] <nick4765> Oh under advanced, I could potentially edit the "viewportin" and "viewportout" values
[18:03] <nick4765> No problem! Have a great one, I should probably go do stuff myself. I really appreciate all of your help Walex, seriously. Thank you.
[18:03] <nick4765> TJ-: says I'm trying to send an empty document, so I'm guessing nothing in that.
[18:04] <TJ-> nick4765: great, that confirms there is no EDID obtained from the display, which is a problem
[18:04] <nick4765> That is good news!
[18:04] <TJ-> nick4765: your Xorg log file suggested that but this confirms it; so what happens is Xorg and the GPU driver offer a set of 'standard' resolutions
[18:05] <TJ-> nick4765: it points to the source of your problem - no EDID
[18:05] <nick4765> Which I think is some type of marking that let's the driver interpret what monitor is present? So it has no idea what the monitor is, so it can't suggest a good reso?
[18:05] <TJ-> nick4765: try it on DP-2 I expect that'll have data: "hexdump -C /sys/class/drm/card0-DP-2/edid"
[18:06] <nick4765> it says no such file or directory
[18:06] <TJ-> EDID is a packet of info sent from the display to the GPU describing make, model, and modes it can support
[18:06] <nick4765> Ahhh, gotcha.
[18:06] <dob1> firefox on ubuntu live cannot reproduce twitch streams. It's my problem related to my test or a known one?
[18:07] <TJ-> nick4765: check what outputs are listed, but DP-2 was shown in the xrandr output. "ls -l /sys/class/drm/
[18:08] <nick4765> Can I copy that info here or does that need to be pastebinned?
[18:09] <nick4765> not sure what rule of thumb is for length
[18:09] <nick4765> it just shows "card0" and "renderD128"
[18:09] <nick4765> as the two lines
[18:11] <TJ-> wow!
[18:12] <nick4765> (is that a good wow or a bad wow?)
[18:13] <TJ-> an unexpected 'wow'
[18:13] <nick4765> Fair enough. Haha!
[18:13] <nick4765> Above these two lines it says "Total 0"
[18:13] <nick4765> What does this information mean?
[18:14] <TJ-> that's fine. ... usually the kernel driver will create an entry there for each output of each card. I haven't touched nvidia in a long time so it may do something different. Let's first check if it lists the outputs only under the card node. "pastebinit <( ls -l /sys/class/drm/card0/) "
[18:18] <nick4765> Ironically enough I was AMD for a long time with GPU until the 5xxx series and just gave up because of shitty drivers. Went Nvidia for the first time since the Riva TNT2 Pro I had back in '99 with this 2080 Super and it's been stellar on Windows. Then I decide I really want to move to Linux as my main because of Windows repeated blatant privacy
[18:18] <nick4765> violations and I'm back full circle with "oh, AMD works way better on Linux because their drivers are open source"
[18:18] <nick4765> Makes me chuckle
[18:18] <nick4765> https://paste.ubuntu.com/p/Fd5srXxrR2/
[18:22] <nick4765> If it matters, Windows also doesn't see the EDID of my secondary monitor. It just shows as "generic pnp"
[18:23] <TJ-> There's something screwey about the GPU I feel
[18:24] <TJ-> nick4765: long shot but ... is it possible it's not in the 'correct' slot of the motherboard? sometimes the slot can make a difference for some weird reason only known to the motherboard firmware
[18:25] <TJ-> nick4765: I have a headless server that needed a GPU in during install and so I got a really low-spec PCIe fanless adapter, nothing special, but the firmware setup complained about the slot I put it in and I had to move it :)
[18:25] <nick4765> Interesting! I hadn't considered this. I could look up the mobo manual and see which x16 slot is preferred. I did actually know that from my own headless server I run! but hadn't considered it for this... haha
[18:26] <nick4765> Could we just force a custom reso and see what happens? I am looking up the mobo info now
[18:27] <amcsi> Walex, thanks!
[18:27] <nick4765> I want to add a GPU to my server as well but can't find a decent quadro that's low-profile. My server is a 4u but the rear is split, half backplane with drives half mobo/etc
[18:29] <nick4765> So in the way only Asus can, the manual is largely unhelpful but it does identify the slots as PCIE 3.0/2.0 x16_1 slot and PCIE 3.0/2.0 x16_2 slot; currently my GPU is installed in the 1 slot
[18:32] <Walex> nick4765: quadros are terrible value.
[18:32] <Apachez> r 993.938 boot -H -uD
[18:32] <nick4765> Hey TJ I need to step out a moment, I should be back in about an hour or so. Sorry. I really appreciate your time and help, I will follow up with you soon if you don't mind. Cheers!
[18:32] <nick4765> Walex I agree but I think they're the best at transcoding
[18:33] <nick4765> I will be back in a little bit! thanks again everyone
[18:36] <Apachez> seems like appindicators in 21.04 is broken somehow... makes gnome-shell utilize 100% of a single core after about 1 day of uptime... once disabled appindicators extension the gnome-shell stops hogging cpu
[18:38] <oerheks> depends what extention you installed ..
[18:48] <cornfeedhobo> hello. does anyone know if there is a place to search for repos with newer/updated packages?
[18:48] <cornfeedhobo> e.g. github's hub package is *way* out of date
[18:49] <cornfeedhobo> hopefully something like OBS/COPR
[18:49] <leftyfb> cornfeedhobo: can you be more specific on what it is you're trying to accomplish? Is there a particular package in mind you are looking to find a newer version of?
[18:49] <oerheks> newer.. we surely backport bugfixes
[18:50] <cornfeedhobo> leftyfb: in this case it's the example i gave. the hub package is 2.7.0, but current release is 2.14.2
[18:51] <cornfeedhobo> okay, *way* is an exaggeration, buuuuut, the fact that the api subcommand is missing is frustrating :p
[18:51] <leftyfb> cornfeedhobo: the snap version is at 2.14.2
[18:51] <leftyfb> cornfeedhobo: sudo snap install hub
[18:51] <cornfeedhobo> they recommend to not use the snap package specifically
[18:52] <leftyfb> cornfeedhobo: who is "they" and why do they recommend not to use it?
[18:52] <cornfeedhobo> as well, i'm vehemently against snap. it's a fundamentally flawed "solution"
[18:52] <leftyfb> ok, good luck
[18:52] <leftyfb> !latest | cornfeedhobo
[18:52] <oerheks> then install the flatpak?
[18:53] <cornfeedhobo> na. this packaging of dependencies is garbage. the core package just needs a bump
[18:53] <cornfeedhobo> or, like the many other distros, i'd love a community maintained version
[18:53] <cornfeedhobo> !ppa
[18:53] <cornfeedhobo> this is the only thing i asked for ^
[18:53] <oerheks> ha, it is maintained by the 'Debian Go Packaging Team' https://launchpad.net/ubuntu/+source/hub/2.14.2~ds1-1
[18:54] <cornfeedhobo> lol
[18:54] <oerheks> contact them for an SRU?
[18:54] <oerheks> !sru
[18:54] <cornfeedhobo> will do. thanks!
[18:54] <oerheks> I guess the answer will be ; use snaps
[18:55] <cornfeedhobo> leftyfb: not to be rude, but this is the second day you've bothered me with pedantia and non-solutions. please, i'm not new to irc or linux. I ask the specific question i'm looking for 90% of the time.
[18:55] <cornfeedhobo> oerheks: that's just a non-answer for packagers being lazy, imho
[18:56] <oerheks> same for chromium browser, find a team for every distro to maintain it.. almost impossible.
[18:56] <oerheks> so, snaps universal packaging is the answer, yes
[18:57] <leftyfb> cornfeedhobo: I asked you what version of ubuntu you were running yesterday. That is very important in most cases when supporting someone
[18:57] <cornfeedhobo> oerheks: centos, fedora, opensuse... all of them seem to get it done, with integration testing too
[18:57] <leftyfb> cornfeedhobo: ubuntu is not a rolling release
[18:57] <cornfeedhobo> neither is LEAP
[18:58] <cornfeedhobo> but again, it gets it done
[18:58] <leftyfb> cornfeedhobo: then use that
[18:58] <cornfeedhobo> if only i could
[19:01] <cornfeedhobo> lol. irony of irony, i found the ubuntu package i want on the opensuse open build service.
[19:04] <superboot> I've got an old install I'm migrating of Ubuntu 14.04 server. When I run /etc/init.d/mysql stop, it says it shuts down, and the logs confirm it too, but it starts right back up. What is the correct way to shut it down and have it stay down?
[19:06] <leftyfb> superboot: you probably have some service like monit watching over services to restart them if they are stopped
[19:22] <nick4765> Hello Walex and TJ- ! I am back :)
[19:22] <superboot> leftyfb: Turns out it was Upstart.
[19:23] <cornfeedhobo> haha
[19:24] <tomreyn> superboot: make sure you get off the EOL release soon
[19:30] <ChinaisSOY> How do I RANDOMIZE my ETHERNET MAC address on my Raspberry PI Debian Buster Lite?
[19:30] <leftyfb> ChinaisSOY: try #raspberrypi
[19:31] <ChinaisSOY> I already asked there
[19:31] <ChinaisSOY> Its a general linux question
[19:31] <leftyfb> ChinaisSOY: then try #linux
[19:31] <ChinaisSOY> Does anybody here know the answer?
[19:31] <ChinaisSOY> its essentialy just a debian system
[19:31] <leftyfb> ChinaisSOY: this is an Ubuntu support channel. You are not running Ubuntu. Try #raspberrypi, #debian or #linux
[19:32] <ChinaisSOY> Im using Lubuntu here
[19:32] <alkisg> ChinaisSOY: ok then go to network-manager menus and select"cloned mac address"
[19:33] <alkisg> But if you're on raspbian, there's no network-manager there...
[19:33] <ChinaisSOY> Im using Raspbian Debian Buster Lite
[19:34] <ChinaisSOY> I dont have a GUI/desktop environment, just TERMINAL
[19:35] <leftyfb> ChinaisSOY: people have responded to you in both #raspberrypi and #linux. Good luck
[19:36] <ChinaisSOY> Great, still can get more answers here
[19:36] <nick4765> TJ- if you have a moment I'd love to continue where we left off earlier
[19:36] <leftyfb> ChinaisSOY: no
[19:36] <ChinaisSOY> Yes, fuck off lefty and go do something else
[19:36] <nick4765> Wow. lol
[19:36] <ChinaisSOY> Theres 930 people here who can possibly answer my question
[19:36] <leftyfb> ChinaisSOY: you are welcome to /part now. Good luck
[19:36] <ChinaisSOY> fuck off
[19:36] <ChinaisSOY> leave me alone
[19:36] <onelegend> ChinaisSOY: lol
[19:36] <ChinaisSOY> Im awaiting answers
[19:37] <onelegend> you're an answer
[19:37] <nick4765> For someone who wants something you sure aren't going about this the right way Sir or Madam.
[19:37] <nick4765> Don't bite the hand that feeds you.
[19:38] <ChinaisSOY> Leftyfb isnt feeding me
[19:38] <ChinaisSOY> so Im not biting the hand that feeds me
[19:38] <leftyfb> ChinaisSOY: that's enough. You are being helped elsewhere.
[19:40] <ChinaisSOY> Fuck off
[19:40] <ChinaisSOY> leave me alone
[19:40] <leftyfb> !op | ChinaisSOY
[19:40] <ChinaisSOY> Im talking to the other 937 people here
[19:40] <ChinaisSOY> fuck the hell off you fucking trolls
[19:40] <ChinaisSOY> let me talk to the other people here
[19:40] <matsaman> ChinaisSOY: what's up?
[19:41] <ChinaisSOY> How do I reliably change my Ethernet MAC address in Raspberry Pi debian buster lite?
[19:41] <onelegend> you use macchange
[19:41] <ChinaisSOY> I want 100% spoofed mac
[19:41] <cornfeedhobo> ChinaisSOY: fwiw, debian and ubuntu differ quite a lot in out of the box configuration. what's the network manager in debian buster?
[19:41] <ChinaisSOY> does macchanger even work
[19:41] <RedHead> hey, I use irssi, is there a better irc client for ubuntu? :)
[19:41] <ChinaisSOY> has anybody tested it
[19:41] <matsaman> ChinaisSOY: yes, you can confirm with ifconfig or ip or whatever
[19:41] <matsaman> ChinaisSOY: you need to confirm it's changed _before_ you do whatever it is you're doing that you want it to be changed for
[19:41] <coconut> RedHead, hexchat is popular
[19:42] <RedHead> thanks coconut!
[19:42] <cornfeedhobo> RedHead: konversation if you are a kde fan
[19:42] <leftyfb> RedHead: feel free to search the software store for other IRC clients, try them and decide which one you like best. Or ask in #ubuntu-offtopic for opinions on software
[19:42] <coconut> or weechat if you like terminal client
[19:43] <ChinaisSOY> How do i script this
[19:43] <leftyfb> matsaman: ChinaisSOY is actively being helped in #raspberrypi, where they should be asking for help. You should go there and chime in. Otherwise someone is just wasting their time
[19:43] <cornfeedhobo> ChinaisSOY: matsaman has the right idea. essentially you get what ever network manager you are using to invoke macchanger before connecting, or after disconnecting and on boot (so it's just always random)
[19:43] <ChinaisSOY> I need it to be changed first thing
[19:43] <ChinaisSOY> I dont want the Router to see my origional ISP ever
[19:43] <cornfeedhobo> you will need to figure that out on your own
[19:43] <ChinaisSOY> MAC*
[19:43] <ChinaisSOY> I need to setup randomized MAC upon boot
[19:44] <matsaman> ChinaisSOY: that part might actually be specific to your distro
[19:44] <cornfeedhobo> ^
[19:45] <cornfeedhobo> you have the pieces you need. but the rest is debian specific
[19:45] <ChinaisSOY> Can I permanantly change my built-in MAC
[19:45] <cornfeedhobo> no
[19:45] <ChinaisSOY> Why not
[19:45] <ChinaisSOY> I just want to edit the file that says 'This is the permanant mac'
[19:45] <RedHead> it burn to the physical network adapter
[19:46] <JonathanD> That "file" is a piece of hardware.
[19:46] <cornfeedhobo> maybe in udev? *maybe*
[19:46] <RedHead> your mac address
[19:46] <JonathanD> You can edit it by replacing the network adapter.
[19:46] <cornfeedhobo> but yeah, what they are saying. the only thing you can do is configure it after boot. the network adapter mac is hardcoded
[19:46] <ChinaisSOY> Can I wrap the hardcoded mac in a randomized mac cover?
[19:47] <ChinaisSOY> cover identity
[19:47] <ChinaisSOY> so all network connections have to travel through the randimzed mac network interfacE?
[19:47] <ChinaisSOY> Preventing leaks 100%
[19:47] <cornfeedhobo> that is a more complex question. You should look into distributions like Tails and Qubes for that level of abstraction
[19:48] <ChinaisSOY> Yes but I need that anonymity on my Standard OS
[19:48] <JonathanD> Beyond that, your nic is active before boot.
[19:48] <ChinaisSOY> I just need a randomized Ethernet MAC
[19:48] <[twisti]> can i somehow see what version a package is in 21.04 without having 21.04 installed ?
[19:48] <JonathanD> Possibly even when the machine is powered off.
[19:48] <ChinaisSOY> yeah Im aware Jon
[19:48] <ChinaisSOY> Thats why I need permanant spoofing
[19:48] <leftyfb> ChinaisSOY: your "standard OS" in not supported here. Go to #raspberrypi for support
[19:48] <matsaman> ChinaisSOY: you can boot with networking mostly disabled, and then run a script to randomize the mac, and then run a script to ensure the mac was randomized and only then enable networking
[19:48] <cornfeedhobo> you should switch to a privacy focused os that does this out of the box, it sounds like. otherwise, you need to be willing to spend more time picking apart your setup and fixing it
[19:48] <JonathanD> You would need something hardware level, then.
[19:49] <JonathanD> If I was that worried about the end device leaking, I might put something on the lan in between them.
[19:49] <matsaman> but this isn't something you'd want to do with a computer at your home or place of business
[19:49] <matsaman> if you care about changing your mac, you should be using someplace else's network also
[19:49] <leftyfb> I love the whole purpose behind this though. They say there router is "backdoored". Like randomizing your MAC is going to do anything
[19:50] <JonathanD> I missed that detail :)
[19:50] <ChinaisSOY> Such as what?
[19:50] <leftyfb> JonathanD: it's because they didn't post it here
[19:50] <JonathanD> Ah.
[19:50] <ChinaisSOY> Jonathan? What kind of Network condom can I use
[19:50] <ChinaisSOY> To shield my PIs raw MAC
[19:50] <leftyfb> JonathanD: and this is why we don't cross-post
[19:50] <ChinaisSOY> I could use a WIFi to Ethernet device
[19:50] <ChinaisSOY> Then the wifi to ethernet devices MAC would be seen by the router, instead of my Raspberry PIs
[19:51] <JonathanD> I don't know of one. You'd have to implement something. Like anything that does NAT in between would cover it.
[19:51] <cornfeedhobo> ChinaisSOY: you don't need to hide your mac unless you are connecting to a router you don't control.
[19:51] <leftyfb> or ask for support for an OS in a completely different OS's support channel
[19:51] <ChinaisSOY> But i feel that we should be able to permanantly randomize MAC more easily and reliably
[19:51] <ChinaisSOY> Right, the router Im connecting to, I dont control, its a backdoored ISp router
[19:51] <cornfeedhobo> .... then just put a router in front of it
[19:51] <TJ-> nick4765: sorry, been working on something else. Did you try a slot-swap of the GPU?
[19:51] <nick4765> ^
[19:51] <ChinaisSOY> I dont have one
[19:51] <JonathanD> ChinaisSOY: you'd want to talk to the pi development/manufacturing team if you want permament random. That would have to be a hardware solution.
[19:51] <cornfeedhobo> go buy one
[19:52] <ChinaisSOY> I live under a train station
[19:52] <JonathanD> Welp, I'm out.
[19:52] <cornfeedhobo> okay, i don't care. routers cost $35.
[19:52] <cornfeedhobo> do you see my name?!?
[19:52] <ChinaisSOY> Im stealing power from a lamppost and an ehternet cable from the local coffe shops wall
[19:52] <cornfeedhobo> yeah, i get you man
[19:52] <nick4765> TJ-  all good! No I didn't try that, I looked in the manual and it shows that the one I have is slot 1, the only two listed are slots 1 and 2.
[19:52] <ChinaisSOY> :]
[19:53] <TJ-> nick4765: mine was the same; Asus Prime X370-Pro
[19:53] <nick4765> Oh no way! Yeah, good mobo so far. Has been solid.
[19:54] <nick4765> The mobo manual is trash though, doesn't go into which busses are which for the PCI, etc.
[19:54] <TJ-> nick4765: It was only in the firmware's setup, for GPU, where it shows a pretty diagram of the slots that there was a little note recommending changing the slot I had it in... like you I had it in slot 1
[19:54] <TJ-> I think in the firmware setup's Advanced mode, it was accessed via the last-but-one tab
[19:55] <[twisti]> i hope its okay to post this again, it seems like the question was lost in that bizarre troll story about living under a bridge; can i somehow see what version a package is in 21.04 without having 21.04 installed ?
[19:55] <nick4765> Wow. Well, I mean, hey what have I got to lose? Can always put it back. Let me try that. I'll be back in a little bit. So your slot 1 is "top" and slot 2 is "down" from there, correct?
[19:55] <nick4765> if you have it oriented in case setup
[19:55] <TJ-> [twisti]: yes, using the rmadison tool
[19:55] <TJ-> nick4765: i think so; let me look before you go!
[19:56] <nick4765> Twisti I think you can just run lsb_release –a
[19:56] <[twisti]> nick4765: and then ?
[19:56] <nick4765> Also yeah, that was incredibly bizarre story. Haha.
[19:56] <TJ-> nick4765: not for checking a package. Slot 3 is middle of the bunch. slot 1 nearest CPU
[19:57] <nick4765> Oh whoops! Sorry, I was thinking release version .My mistake
[19:57] <TJ-> [twisti]: which package did you want to know about?
[19:57] <[twisti]> podman
[19:57] <TJ-> !info podman hirsute | [twisti]
[19:57] <nick4765> I only have two slots on mine, TJ-. Well, x16 slots at least. I think the bottom-most full length slot is an x8. But either way yeah I think slot 2 is the middle one
[19:58] <[twisti]> thanks
[19:58] <TJ-> [twisti]: do you use podman with crun rather than runc?
[19:58] <nick4765> I will be back in a bit! (I really need to start using an IRC client...)
[19:59] <[twisti]> i dont know what that is or means. im not using podman at all at the moment, havent even decided what distro im going with, thats why i asked
[19:59] <[twisti]> i dont want to set up a system only to realize it comes with an ancient podman version
[19:59] <TJ-> [twisti]: ahhh... crun is a nice, simple, C-language container runtime, rather than golang. Advantages are it can natively interface with the kernel whereas golanf has to pull all sorts of weird tricks
[20:00] <TJ-> [twisti]: runc and others are usually golang
[20:00] <[twisti]> im setting this whole thing up to learn podman, and i dont know anything about runc, crunc, golang or any of the other things you said yet, looking forward to learning about it though :D
[20:01] <TJ-> [twisti]: it's a steep learning curve :)
[20:10] <cornfeedhobo> ^ truth
[20:11] <TJ-> Also, a lot of the worst of devops and NIH in the OCI world
[20:11] <TJ-> takes many months to realise that it's abstractions for abstractions sake, and unnecessary
[20:20] <nick4765> TJ- No difference, notably.
[20:20] <nick4765> Although now my taskbar in Windows is different but I can deal with that later haha
[20:22] <TJ-> nick4765: darn- would have been an easy solution
[20:23] <TJ-> nick4765: it is booted into Ubuntu now? if so, show us " pastebinit <(lspci -nn; journalctl -k) "
[20:25] <nick4765> No doubt, and yes I'm in Ubuntu at the moment.
[20:26] <nick4765> https://paste.ubuntu.com/p/t8fKk3RYG2/
[20:31] <TJ-> nick4765: might not be related but: !kernel: pci 0000:0a:00.0: 16.000 Gb/s available PCIe bandwidth, limited by 2.5 GT/s PCIe x8 link at 0000:00:03.2 (capable of 126.016 Gb/s with 8.0 GT/s PCIe x16 link)"
[20:32] <nick4765> Whoa, that's kind of weird... so that's saying that one of my x16 slots is being limited to x8 speeds?
[20:33] <nick4765> I wonder if it said the same before I moved slots
[20:34] <nick4765> In the Nvidia X Server settings under GPU 0 it also says PCIe Generation Gen 3, x16 but then link speed 8.0 GT/s
[20:35] <nick4765> I didn't notice if it said the same before
[20:35] <nick4765> I am 93.5% sure the physical port is set to x16 link speed in BIOS but I can check that
[20:35] <nick4765> Either way, it shouldn't saturate an x8 slot I don't think? Not sure on that
[20:39] <Simmer> TJ- This is nick4765, I grabbed Hex chat.
[20:39] <TJ-> there's nothing else in the log file and I've read every line
[20:39] <TJ-> Simmer: ^^^
[20:39] <Simmer> So freaking weird. Well, let me restart and poke into BIOS to see if the link speed changed at all there. I shall return! I really appreciate all your help, truly. Thank you so much for your time.
[20:40] <TJ-> Simmer: show us "pastebinit <( sudo strings /sys/firmware/acpi/tables/DSDT | grep -i windows | sort )"
[20:41] <Simmer> Shows empty document
[20:41] <Simmer> oh wait, one sec
[20:42] <Simmer> it asks for my pw but doesn't wait and says unable to read, I/O error. Then says you're trying to send empty document.
[20:42] <Simmer> So I'm not sure if it's truly empty or it just can't access
[20:43] <Simmer> I'm guessing it's interpreting the 'sudo strings' part as root access?
[20:43] <Simmer> er, superuser
[20:44] <Simmer> I shall return, let me poke into BIOS real quick.
[21:08] <Simmer> Hi TJ sorry about the delay there. I am back, though now on my laptop with IRC so I don't have to keep leaving and coming back.
[21:09] <TJ-> Simmer: show us "pastebinit <( sudo strings /sys/firmware/acpi/tables/DSDT | grep -i windows | sort )" from the problem machine. the sudo command shouled cause you to be asked for your password
[21:11] <Simmer> And now my laptop freaked out and I lost what you just said. I'm sorry. I'm pretty sure at this point I'm going to set fire to everything and never touch a computer again.
[21:11] <TJ-> Simmer: show us "pastebinit <( sudo strings /sys/firmware/acpi/tables/DSDT | grep -i windows | sort )" from the problem machine. the sudo command shouled cause you to be asked for your password
[21:12] <TJ-> Simmer: I know the feeling; technology takes one look at me and fails in unique ways, just to test me!
[21:12] <Simmer> In BIOS I noticed two things. First, under Tool > GPU Post it states to use slot 1 if using a single GPU so I went ahead and swapped back. Second, it shows that slot as x8 and there isn't an option to switch to x16 so I'm looking into that on the hardware side of things. Might need to update BIOS
[21:13] <Simmer> Yes, no doubt. I'm still in BIOS on my main PC so it will be a few for that command. I will paste as soon as possible.
[21:16] <TJ-> Simmer: that's OK :) the output of tha command lists the Windows versions hard-coded into the motherboard, and we need to identify the 'latest' and ensure it is being faked by Linux to ensure all ACPI functions are available. Those can affect discovery of outputs on GPUs.
[21:17] <TJ-> Simmer: another thing, if this is related to ACPI, is to boot from cold power off, suspend to RAM, resume, and check if the outputs have appeared
[21:20] <AsenMx> Hello, I know this is not the right channel, but i will ask anyway. Anyone familiar with gitlab CICD deployment to AWS ECS ? I
[21:20] <Simmer> Interesting that Windows is hard-coded into the motherboard. That seems..... anti-competitive...
[21:20] <TJ-> Simmer: different functionality is enabled depending on what OS reports as running
[21:21] <TJ-> Linux has fake being the 'latest' version of Windows to ensure ACPI is fully enabled
[21:21] <TJ-> so we check which Windows versions are encoded in the ACPI's DSDT
[21:22] <Simmer> Ahh
[21:23] <TJ-> I've had a system that lost its outputs after a suspend/resume for example, unless the acpi_osi was set correctly
[21:24] <Simmer> sudo: unable to read password: Input/output error
[21:24] <Simmer> You are trying to send an empty document, exiting.
[21:24] <Simmer> That's what I get on that command
[21:24] <TJ-> Hmmm, are you doing that over ssh?
[21:24] <Simmer> No, direct
[21:24] <TJ-> hmmm, that is VERY weird. If it were over ssh, I'd say you need "ssh -t" to create a pseudo-tty
[21:25] <Simmer> Can I try forcing native resolution to the secondary just to see what happens?
[21:25] <TJ-> it suggests devpts isn't available
[21:25] <Simmer> What's devpts?
[21:25] <TJ-> pseudo-tty devices
[21:26] <Simmer> Oh
[21:26] <TJ-> check it's there "mount | grep devpts" should show it mounted at /dev/pts
[21:26] <Simmer> I'm almost tempted to just buy a new motherboard... haha
[21:27] <TJ-> Simmer: one thing I'd try here, is to remove the nvidia driver and let the nouveau driver run the GPU, and see if it discovers the outputs and can get the modes
[21:27] <Simmer> devpts on /dev/pts type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000)
[21:27] <tomreyn> this is an Asus PRIME X370-PRO, BIOS 5220 09/12/2019 with matisse cpu. matisse cpus were sold starting july 7 2019.
[21:27] <TJ-> I know it's not easy to remove cleanly though
[21:28] <TJ-> Simmer: install the latest firmware - it was released last month
[21:28] <Simmer> Yeah, it was the first BIOS released that supported 3rd gen Ryzen
[21:29] <TJ-> I updated the firmware on mine a couple days ago, so I know :)
[21:29] <Simmer> I had a first gen ryzen in this mobo first, then upgraded when they released BIOS support for the 3rd gen
[21:29] <Simmer> am running a 3700x now
[21:29] <TJ-> Mine's used for running thousands of containers
[21:29] <Simmer> And yeah, not going to lie it's slightly discouraging that some things take so many steps to do like cleanly removing driver, etc.
[21:30] <Simmer> I run a supermicro board for that TJ! Haha but wow thousands? and I thought my docker server was big... I used to run 40 containers until I cleaned it up a bit. Do you run kubernetics or whatever it's called?
[21:31] <TJ-> Simmer: no, we dumped k8s because its a total load of devops NIH  %^$$"
[21:31] <Simmer> Ah, I see! Fair enough
[21:31] <TJ-> Been developing our own alternative using standard distro packages, network protocols, etc.
[21:32] <Simmer> Sweet! That's awesome
[21:33] <Simmer> It looks like there's a BIOS that was released two days ago! wow! Not sure I should go that new..... it's version 5606 which one did you update to?
[21:33] <Simmer> The previous was last year
[21:34] <alzgh> Doing `man apt` I get under "SEE ALSO" `apt-get(8), apt-cache(8), sources.list(5), apt.conf(5), apt-config(8),`. But `man sources.list` doesn't give me anything. Do I need to install something?
[21:34] <TJ-> Simmer: 5606
[21:34] <Simmer> Ah okay I will use that one. It shows release date of August 9th
[21:35] <TJ-> Simmer: that's the one, but the CAPsule file inside is dated end of July
[21:35] <Simmer> Ahhhh gotcha fair enough
[21:36] <TJ-> extract the .CAP file from the ZIP, but it on a FAT-formatted USB, stick it in a port, and use the firmware's Eazy-Update function to select the USB device and the file
[21:36] <tomreyn> alzgh: sources.list(5) is part of the APT manual (at least in version 1.6.14)
[21:36] <TJ-> Simmer: if this fixes it, saves a lot of messing about
[21:36] <Simmer> Will do! There's also an EXE in the zip called BIOSRenamer ... why would you need to rename it?
[21:37] <Simmer> Yeah no doubt
[21:37] <TJ-> Simmer: ignore it :)
[21:37] <Simmer> I did see on some changelogs for previous versions "added fixes for certain linux distros"
[21:37] <TJ-> Simmer: but by all means rename the long-name to something shorter using standard tools
[21:37] <Simmer> which I think I have one older
[21:37] <tomreyn> alzgh: also on focal and impish it's part of apt: http://manpages.ubuntu.com/manpages/focal/en/man5/sources.list.5.html
[21:37] <TJ-> Simmer: I think I renamed it to 5606.cap
[21:38] <Simmer> Fair. I would just use windows explorer's rename method if I did
[21:38] <alzgh> tomreyn O, man. I was typing "source" instead of "sources". Thanks.
[21:39] <tomreyn> oops :)
[21:41] <Simmer> My current BIOS verison is 5220 - updating now
[21:43] <kotze3000> gf
[21:53] <Simmer> BIOS update complete, went smooth. Interestingly enough the option I was looking at prior that showed only x8 was renamed to "PCIEx16_2 Bifurcation" where auto is x8 or you can split it to do a hardware raid or other. Also guessing it ONLY affects slow 2 which I am not using anymore, back to slot 1. In GPU post it says nvidia GPU running at native
[21:53] <Simmer> x16 so it all seems okay and I think I was just confused on the previous name before it was updated to reflect the bifurcation ability.
[21:53] <Simmer> Makes way more sense now, I use bifurcation on my server board.
[21:54] <Simmer> Re-enabled XMP and rebooting now to get back into Ubuntu
[22:02] <Simmer> Hmm, well BIOS says it's fine running max but the nvidia control panel still shows 4 lanes @ 5.4gbps link. Not sure why.
[22:20] <Simmer> TJ You've been amazing and I so very much appreciate all your time and help today. I'm going to call it for the day. I don't think I'm going to get too much farther on this issue with my current hardware sadly. I think that second monitor is just too old. Thanks agian man, really appreciate it.
[22:27] <TJ-> Simmer: good luck with it :)
[22:27] <TJ-> Simmer: if you feel up to it, don't forget the nouveau experiment
[22:29] <hoppity> Is it more secure to have two login screens in a web app? (one for regular users and one for admins) or is this fake security?
[22:31] <coryc257> hoppity, if it's a second login and not just same same username/password it's more secure. especially if you make the admin section have no images/javascript except when absolutely required so you can have a web browser with all the features turned off
[22:32] <leftyfb> hoppity: try #security. Your question is not an ubuntu support question
[22:33] <hoppity> thanks guys
[22:36] <Simmer> Good call TJ! Sorry my thing disconnected. Alrighty, cheers! I will follow up with ya later on. Thanks again