[00:08] <ab2m> Hey folks.   First time.   Long time Linux user, new Multipass user.   Trying to run multipass on a Windows 10 Home host with Virtualbox, and need to use a USB-attached device, but can't seem to find a way to pass it through.   When I go into the Virtual Box GUI as System (https://multipass.run/docs/using-virtualbox-in-multipass-windows), I can't modify the settings (the GUI goes nuts) and USB
[00:08] <ab2m> is listed as "disabled" anyway.   Any ideas?
[00:11] <sarnold> ab2m: I suggest trying in #multipass, I wouldn't be surprised if there's more windows multipass users there than here
[00:11] <ab2m> Sarnold: Will do.  I'll hang out here too just in case.   :)
[00:12] <sarnold> cool cool
[00:40] <Hash> Hi
[00:40] <Hash> What's a good TTF font editor in linux ubuntu
[00:40] <Hash> I need to make a font's individual glyphs a bit more narrow they're too wide.
[00:42] <sarnold> Hash: maybe fontforge?
[00:43] <Hash> I have it
[00:44] <Hash> I'll try it
[00:44] <Hash> Looking to also edit HAccent vert to adjust line height also.
[00:44] <Hash> I found a python2 script that uses fonttools and adjusts all this for you on cli, but its python2, and I can't get it to work on ubuntu 20 which has python3
[00:46] <Hash> https://input.djr.com/build/?fontSelection=whole&a=0&g=0&i=0&l=0&zero=0&asterisk=0&braces=0&preset=default&line-height=0.9&accept=I+do&email= David makes fonts available, but he has also provided that script in his zip file.
[00:47] <Hash> https://termbin.com/tb7m This is the script that edits fonts on commandline
[00:47] <Hash> If I can get this working in ubuntu 20.04, I can try this out
[00:48] <sarnold> oh man, it looks old enough that it wasn't programnmed to run in both py2 and py3
[00:48] <Hash> :(
[00:50] <sarnold> yeah, I tried my hand at converting something from py2 to py3 yesterday and gave up after a few minutes, it needed someone more familiar with both languages :)
[00:52] <revmacian> Im using Konversation in Kubuntu. No matter what I try I cannot get #ubuntu channel to auto join on connect to Freenode. Is this a bug?
[00:52] <sarnold> what have you tried?
[00:53] <Hash> https://i.imgur.com/kIqx3C5.png
[00:53] <Hash> That's what the font looks like. It's a nice wide bold font, but it's too wide. I need to make it slightly compressed. Not edit actual font things, but just the width of the glyphs in it
[00:54] <Hash> For all of them, reduce it by a bit.
[00:54] <Hash> Trying to now figure out how to use this program
[00:55] <revmacian> sarnold: Were you talking to me? If so, here it is: https://imgur.com/a/VCV4n4H
[00:55] <Hash> I can change the width, but only left side. :/ I'll figure itout
[00:56] <sarnold> Hash: hah, that 'w' blows right through the bounding box..
[00:56] <Hash> Yup
[00:56] <Hash> I can make it less wide and center! Nice.
[00:56] <Hash> I think I can do this now. Tedious, bu doable
[00:56] <sarnold> revmacian: does the edit button next to 'default identify' lead to you some place to configure certificate authentication and perhaps a small delay before executing autojoins?
[00:57] <Hash> https://i.imgur.com/4R4kiKY.png It is for Diablo 2. So wide, but such a beautiful readable font in Diablo 2 OGM!
[00:57] <Hash> OMG
[00:57] <Hash> Just less narrow and we're in busienss
[00:57] <sarnold> so close
[00:59] <revmacian> sarnold: It leads to Auto-identify. You thinking that the identify is taking time and the auto-joins never get executed? Which other auto-identify types work on Freenode?
[00:59] <sarnold> revmacian: yes, this channel for example has +r on it whic requires you to be identified before you can join
[00:59] <sarnold> revmacian: most clients have a way to delay that for a bit to let services catch up
[01:00] <sarnold> I expected to see it in the dialog box you pasted :) but alas no :(
[01:00] <revmacian> sarnold: That's exactly what is happening. Now that I know this I can work towards fixing it. I don't think Konversation does that
[01:01] <revmacian> sarnold: So, you can't join here if you're not reg'd and id'd?
[01:01] <sarnold> right
[01:01] <revmacian> How are new users supposed to know that limitation?
[01:02] <sarnold> the ircd will send them an error message along hte lines of "you need to be registered and identified to join this channel"
[01:03] <revmacian> so someone is making it harder to ask a question than it has to be for newbies? That sounds like the opposite of user-friendly.
[01:03] <sarnold> yup]
[01:03] <sarnold> but the spam load is bad news otherwise
[01:03] <revmacian> Oh well, at least I know what's wrong now.
[01:03] <revmacian> Oh, right, let's make it easier on staff but more difficult for new users. Ok.
[01:04] <revmacian> I'll shutup now.
[01:04] <Hash> That may be advisable.
[01:17] <raecarruth> hello, can anyone tell me why my qbittorrent client looks so odd? https://imgur.com/JXpVysD.png
[01:22] <sarnold> raecarruth: wild guess, one of the libraries it uses was replaced while it was running
[01:22] <raecarruth> i just installed it on this system for the first time
[01:23] <sarnold> dang, I was way wrong..
[01:29] <maret> quick question , is it possible to search file contents by regex. I want to find all (text files), which contain certain string and I want to search by regex
[01:29] <sarnold> maret: grep is the usual tool for this, ripgrep is nice if you've got LOTS of files
[01:31] <maret> sarnold: make sense, how would I go about it ?
[01:32] <sarnold> maret: grep '^r' /etc/passwd   for example
[01:33] <maret> that would be search for using ^r regex in /etc/passwd ?
[01:33] <sarnold> yes
[04:35] <rodrigoty23> does anybody know to change cajas icon on the panel?
[04:35] <rodrigoty23> im seeing a blank@
[06:54] <JPSman> Ever since I did an install of KDE along with my gnome-flashback, the menu text color has been unreadable.  I'm also having a hard time GOOGLING this problem hence why I am here.  here is a picture of what Im talking about https://i.imgur.com/JrV8cr3.png
[06:54] <JPSman> the
[06:55] <JPSman> the "grayed out" text is the selectable one, and the darker one isn't
[06:55] <JPSman> 18.04 with gnome-flashback and metacity compiz
[07:11]  * JPSman watches a tumble weed blow through the chat room.
[07:16] <liuyonghao156> Hi there. I am a user of Ubuntu 20.04.2 LTS on AMD Ryzen platform. I find taskbar icons not showing for first logins every time I reboot, but after logging off and then back again, I have them displayed properly. Any ideas on that?
[07:36] <ThinkT510> JPSman: which apps are like that? all of them? your screenshot shows audacity which uses wxwidgets
[07:43] <JPSman> no, its not all of the apps, just some of them.  Thats a big clue though thank you :D
[07:44] <JPSman> hmm, whats another popular app that uses wxwidgets?
[07:46] <JPSman> actually, I am using XChat as my IRC client and it has the same problem too
[07:46] <JPSman> https://i.imgur.com/IyrQ8r2.png
[07:49] <ThinkT510> JPSman: you're better off using hexchat, xchat is no longer maintained
[07:49] <JPSman> ...... ok.  but does xchat also use wxwidgets?
[07:51] <ThinkT510> not sure, I can't tell
[07:52] <ThinkT510> audacity is the only popular program I'm aware of and use myself that uses wxwidgets
[08:26] <slidercrank> hi folks. I created a VPN connection in network-manager. The problem is the option "[ ] Automatically connect to VPN" is absent. It annoys me to turn it on manually every time
[08:37] <slidercrank> I've found the configuration file in /etc/NetworkManager/system-connections/<name>.nmconnection It has autoconnect=false. There is no way to change it via GUI. So you have to manually edit the file
[08:41] <kill-animals> Hey I need some advice
[08:42] <slidercrank> kill-animals, shoot
[08:42] <kill-animals> Is it possible to write an application that creates files that would only be writable by that application?
[08:42] <kill-animals> I am no chmod expert.
[08:43] <slidercrank> hm.
[08:44] <ThinkT510> kill-animals: root has access to everything to technically the answer would be no but it isn't really a ubuntu support question
[08:44] <kill-animals> slidercrank: Basically I am building an app, will be packaging it in a snap, and I want to prevent users from deleting files accidentally,
[08:45] <slidercrank> there is the concept of user. If you run an application as a specific user, all applications running as the same user could have access to the same file. The access on determined per user, not per application
[08:45] <kill-animals> slidercrank: Hmmmm.
[08:46] <kill-animals> Wait; can't you create write protected files? I know sometimes I can't even RM something that is write protected.
[08:46] <slidercrank> kill-animals, you can create a user to run your application. Like "myapp". Thus files created by your app will have the user access 'myapp'
[08:47] <kill-animals> slidercrank: Interesting.
[08:47] <slidercrank> if only one app runs as myapp user, only this app will have access to it
[08:47] <kill-animals> Ill take a look at that then.
[08:47] <slidercrank> I think it's the easiest solution
[08:48] <slidercrank> kill-animals, yes, you can open a file exclusively. So while your app is running, no one else will be open your files.
[08:48] <kill-animals> Also, if you were to store partially downloaded files, that are meant to be persistent between reboots (IE, I don't want to store it in /tmp/) is there an obvious place in ~/.xyz that I should put it in?
[08:49] <kill-animals> slidercrank: Interesting. I'll look into that.
[08:49] <slidercrank> so it was not clear from the beginning: do you want to limit access while your app is running or permanently
[08:49] <ThinkT510> wouldn't it matter where the files are stored though? anything under a user's /home will be editable by that user
[08:50] <kill-animals> slidercrank: Its a bit up in the air at the moment, as I strategize how I am going to deal with my issue.
[08:51] <kill-animals> ThinkT510: btw. I never know there was a T510 thinkpad
[08:52] <kill-animals> I thought it was T500, and T520.
[08:53] <ThinkT510> yes, Thinkpad T510 exists
[08:54] <kill-animals> ThinkT510: but no x210 exists. Thats what had me scratching my head
[09:34] <Hash> What's a good permanent url shorten service free?
[09:34] <Hash> I had a link to a latest master.zip but it's too long and I need to shorten it and put in topic
[09:43] <lotuspsychje> !ot | Hash
[10:09] <inerkick> Hi. I'm unable to connect to printer/scanner Canon MF241D. Kindly help how to use it
[10:32] <imrjd> how are you connecting?
[10:32] <imrjd> wired, or wifi?
[11:25] <Paddy_NI> Is there any screen annotation software available for linux that approximates the functionality one would have had in Compiz?
[11:26] <Paddy_NI> I have tried Gromit which is usable however it does seem a little bit flaky as to whether or not it wants to work at times.
[11:32] <lotuspsychje> Paddy_NI: ive seen some users before playing around with compiz on 18.04 and higher, but cant recall anymore what they did exactly
[11:39] <Paddy_NI> lotuspsychje: Yeah I did for a moment consider using compiz again but for the sake of the annotation feature it seems to be more hassle than it is worth.  I suspect I'll just have to deal with the quirks of gromit.
[11:40] <AlexP11223> What should I use to encrypt a disk? Not a system one, just a separate HDD
[11:41] <lotuspsychje> Paddy_NI: try also snap find annotate
[11:41] <Paddy_NI> lotuspsychje: been there done that
[11:42] <Paddy_NI> Normally the first thing I do before asking others is to search the repos, stack exchange, forums etc lotuspsychje
[11:43] <lotuspsychje> Paddy_NI: maybe someone in #ubuntu-discuss might know some alternates
[11:44] <Paddy_NI> Never heard of #ubuntu-discuss, is this new?
[11:44] <Paddy_NI> Is it official?
[11:44] <lotuspsychje> official, yes, new no
[11:45] <Paddy_NI> Cool
[11:45] <Paddy_NI> Delighted I was not told to use club ubuntu lol
[11:45] <Paddy_NI> X-D
[12:19] <TJ-> Investigating a very hard to reproduce issue on 20.04 with K3B - used for audio CD mastering regularly, we've had a recent session where the resulting CD's appeared good and the tracks and track lengths were correct, but every track was absolute silence.
[12:20] <slidercrank> TJ-, can I listen to these tracks in noisy environments to eliminate noise? :)
[12:20] <TJ-> slidercrank: no, we'd need to record inverse waveforms of your enviroment for that to cancel it out#
[12:50] <BluesKaj> Hi folks
[12:56] <ltl2> hi
[13:06] <john_rambo> This is the first time I am going to use a RSS reader. I want the app to play a sound when a new feed arrives. Which one should I install ?
[13:22] <slidercrank> I think it's better to use a browser extension to read RSS feeds. Because you'll have to click on links in RSS anyway.
[13:57] <zmagii> I think this is an Ubuntu question, but let me ask anyway. In Ubuntu there are some thing that require sudo / root that don't necessarily do on other distros? 1) Why is this? and 2) Can I allow some of these to run without having to invoke sudo everytime?
[13:58] <BluesKaj> !alias
[13:59] <zmagii> You mean in .bashrc? That's my current workaround, but I don't particularly like it since it calls root on a command that I may either forget what it does or could be changed.
[14:00] <BluesKaj> you could setup aliases for sudo commands in ~/.bash_aliases
[14:02] <zmagii> OK, so let's take Apt as an example. Does Ubuntu require sudo because installations are mostly in system directories and in other distros this isn't necessarily the case?
[14:05] <BluesKaj> most debian based distros use the same file system and structure ...dunno much about other linux OSs
[14:16] <rory> zmagii, yes apt installs to system directories, but the same is true for other distros package managers
[14:17] <rory> zmagii, it's possible to configure the /etc/sudoers file to not require a password for specific commands. do you have an example that prompted your question?
[14:20] <BluesKaj> oops, forgot about sudoers, good call, rory
[14:39] <silent-failure> From a technical perspective, how does Unity get programs to use the global menu bar?
[14:39] <silent-failure> Is there some hack it does with GTK?
[14:42] <silent-failure> Ah, looks like it has a patched version of GTK.
[14:53] <thatpythonboy> ssh works but locally what to do to get external ip or to make it work over globally
[14:54] <rory> thatpythonboy, get your IP by Googling "what is my IP" or using this web service: curl ifconfig.me
[14:55] <slidercrank> if you're behind NAT, it will be tricky
[14:55] <rory> thatpythonboy, in your router settings, port forwarding, set up port 22 to forward to the internal IP
[14:55] <silent-failure> You may also have to open ports on your router/modem.
[14:56] <PuttySandwich> Does anyone know how to reconnect a user to an encrypted home? I had to re-install my system and I have a user with an encrypted home. I had to do this once before a long time ago and I have a vague recollection of just having to symlink something. But my experiments with that have failed. Google has also been unhelpful.
[14:57] <thatpythonboy> any guides to do that
[14:59] <jeffro_working> is it normal for linux-generic to point to 5.4?
[14:59] <jeffro_working> !linux-generic
[15:00] <EriC^^> jeffro_working: which ubuntu version?
[15:00] <jeffro_working> 20.04.2
[15:00] <EriC^^> !info linux-generic focal
[15:00] <EriC^^> jeffro_working: yes
[15:01] <jeffro_working> okay, so if I want a newer kernel I need to manually install a linux-image-5.xx-generic package
[15:01] <EriC^^> maybe you could try the hwe
[15:01] <EriC^^> !hwe | jeffro_working
[15:01] <EriC^^> !mainline | or these
[15:05] <jeffro_working> I'm not interested in a mainline kernel.
[15:06] <jeffro_working> Thank you though. I was looking at adding a pcie intel card and just making sure the kernel I use will have the drivers, so that I don't have to mess with drivers on this machine.
[15:06] <jeffro_working> ax200 probably
[15:06] <EriC^^> i see
[15:10] <EriC^^> jeffro_working: perhaps try "locate ax200.ko" and see if anything pops up?
[15:12] <soulseeker-cloud> Can you get a refund for zephram alpha
[15:30] <rory> Isn't that a character from Star Trek?
[15:30] <rory> you should probably ask him.
[15:31]  * rory just made a trekkie quit in disgust
[15:41] <Gladwyn> Hey folks.  I just did-release-upgrade (heh) from 18.04 to 20.04 (server) and I'm having a problem with my mumble-server which I didn't have before.  It's not able to read my SSL cert (from /etc/ssl/certs).  However, when I just run murmurd (without options), it starts just fine.  I saw this (https://github.com/mumble-voip/mumble/issues/3297) but it really looks more like a systemd issue.  Any thoughts?
[15:45] <ioria> Gladwyn, https://bugs.launchpad.net/ubuntu/+source/openssl/+bug/1856428/comments/2
[15:46] <Sven_vB> hi! I'd like some sudo mount commands, and fstab entries, to fail if the mountpoint directory isn't empty. what's the easiest approach?
[15:46] <Sven_vB> in bionic
[15:47] <Gladwyn> Thanks, ioria, that's a good place to start!
[15:47] <Sven_vB> actually I was kinda surprised that it didn't fail with "mountpoint is not empty". Seems that's a fuse-specific default.
[15:47] <ioria> ok
[15:48] <longshanks> in the days of old there was a rule of thumb of 2xRAM is a good size for swap. Is there still a rule of thumb regarding swap and has it changed?
[15:48] <Sven_vB> longshanks, for the last decade I usually omitted any swap
[15:49] <Sven_vB> longshanks, you may need it though for stuff like suspend-to-disk
[15:49] <longshanks> Sven_vB: I wondered if i needed it. I've a box i use running Ubuntu that is used as a home theatre pc and it never seems to use the swap. was just thinking maybe i'm wasting the 16GB of disk
[15:50] <Sven_vB> nowadays lots of machines have so much actual RAM that they don't need a swap.
[15:50] <longshanks> Sven_vB: so yeah sure on a laptop it might still be useful for suspend. But the home theatre pc is a desktop and never suspends
[15:51] <Gladwyn> I've managed to lock up a laptop with 16 GB ram and next to no swap (whatever was default, I think just below one gig) by having a bajillion (like really 500+) browser tabs open plus Teams and Slack.  OOM-killer would _eventually_ kick in but the laptop was unresponsive for up to 10 minutes.
[15:51] <longshanks> Sven_vB: yeah that's the conclusion I've (recently) come to too..
[15:51] <Gladwyn> Like "I'd hard reboot now if I didn't know I have open files" unresponsive.
[15:51] <Sven_vB> Gladwyn, ulimit can help
[15:51] <Sven_vB> I launch all my "hungry" programs ulimit-ed
[15:52] <Sven_vB> also it's a good idea to enable on-demand OOM kill via magic sysrq
[15:52] <Gladwyn> Nice!  I mean, I'm admittedly an idiot for having so many browser tabs open (they all were important _at some time!_) but I'd love my OS to handle that sort of shenanigans rather than be swarmed by it =P
[15:53] <Sven_vB> (and practice it in advance. :D yes it may take a while until results become visible)
[15:55] <Sven_vB> yeah on my netbook sometimes ulimit wasn't an option, VLC really needed that much RAM. and when I forgot a movie was paused, and returned after a few hours, I had to request OOM. Which worked reliably. After a few seconds, VLC was killed and the mouse was able to move again.
[15:56] <Gladwyn> Cool, I'll look into that, thanks for a great tip, Sven_vB!  I knew about SysRq from the olden days but I had completely forgotten about it, and didn't realise how perfect it would be for just this kind of thing.  Much obliged!
[16:01] <longshanks> What is OOM?
[16:01] <longshanks> one of the magic sysreqs ?
[16:01] <Gladwyn> longshanks: "Out of Mana" -- OOM killer is a kernel sub-routine that tries to find the biggest memory hog and kill it off, if your system is all out of RAM.
[16:02] <longshanks> Gladwyn: neat. I'll have to go read more. Cheers :)
[16:02] <Gladwyn> And yes, as I just learned from Sven_vB, you can actually trigger it manually, if you know you're close to the precipice
[16:14] <longshanks> is there a (relatively) easy way to install kernel 5.10 in Ubuntu 20.10 ? Am running into this bug https://github.com/linux-on-mac/mbpfan/issues/221
[16:17] <rory> To be clear longshanks, OOM means "out of memory" in this context. Your system is manaless.
[16:17] <longshanks> rory: thanks :)
[16:20] <yates> is there a better channel for asking about mimetype associations?
[16:21] <yates> xdg-mime query default application/pdf
[16:21] <nunya> how to make nsmbd and nmbd on demand rather than started at boot? I think I could improve my boot time by 30 seconds or more if these services were opened only when I want to send a file to my phone, aother laptop, etc.
[16:21] <yates> ..returns /usr/share/applications/okularApplication_pdf.desktop
[16:22] <yates> yet xdg-open x.pdf uses evince
[16:26] <nunya> lotuspsychje: Last time I was here you recommended  install preload stacer haveged and tweak systemd services and startup items. If anything my boot time got slower! Is there a revert to a time before like Restore Point in Windows?
[16:33] <longshanks> nunya: you can do time machine like snapshots I just don't know if that's enabled by default
[16:38] <leftyfb> nunya: no. Just undo everything you did if you feel they are the cause
[16:40] <longshanks> nunya: also checking your shell history to see what you did sometimes helps jog the memory
[16:43] <nunya> longshanks: I wish I could remember what I did. I've uninstalled and reinstalled so many things trying to get a faster boot time so was hoping to return my system to what it was a couple of days ago.
[16:44] <nunya> leftyfb: I wish I could remember what I did. I've uninstalled and reinstalled so many things trying to get a faster boot time so was hoping to return my system to what it was a couple of days ago.
[16:44] <longshanks> nunya: there is a systemd-profiler have you looked at that? produces some swanky graph for you to see where it's taking time
[16:48] <nunya> longshanks:systemd-profiler in terminal does nothing, sudo Apt-get install gives Unable to locate package systemd-profiler
[16:52] <leftyfb> ptah: please do not PM
[16:53] <ptah> leftyfb are you the same guy in efnet #linuxhelp?
[16:53] <leftyfb> ptah: that is offtopic here. What can we help you with?
[16:55] <longshanks> nunya: the systemd profiling is called systemd-analyze I believe
[16:57] <longshanks> nunya: any program which helps is going to require you at least look at the man-pages/docs to see how to use it.
[16:58] <RonWhoCaresOffic> Is someone able to walk me through a boot repair for Ubuntu Studio 20.04 ?  The reason this has come along is to try and setup a dual boot Windows 10 and Ubuntu on this computer.
[16:59] <leftyfb> RonWhoCaresOffic: https://help.ubuntu.com/community/RecoveringUbuntuAfterInstallingWindows
[17:02] <nunya> longshanks: 1systemd-analyze blame  shows 1min 35.001s man-db.service, 51.336s snapd.service Those 2 services if I could disable at boot time and make them on demand services would save 2:20 on boot time. Have googled but having hard time finding or understanding what to do.
[17:02] <nunya> systemd-analyze blame
[17:22] <leftyfb> ptah: this is a support channel. Do you have a support question?
[17:28] <RonWhoCaresOffic> leftyfb:I have received an error with a command Boot Repair gave me.  Would you look at this pastebin? https://pastebin.com/vxkiXTUg
[17:30] <leftyfb> RonWhoCaresOffic: neither of those commands were in the article I sent you and have nothing to do with re-installing grub after you installed Windows
[17:31] <Kraus> Morning! I have been sent by the folk in ##networking to ask about possible bugs in the DHCP client.
[17:31] <RonWhoCaresOffic> ptah: I am not on the computer that has the issue
[17:32] <RonWhoCaresOffic> ptah: The CTCP version request you just sent is irrelevent
[17:32] <leftyfb> RonWhoCaresOffic: ignore them, they are not here to help
[17:32] <RonWhoCaresOffic> leftyfb: I did exactly as you stated.  "boot-repair" has given me a list of ~5 commands to run in Terminal to repair this
[17:33] <leftyfb> RonWhoCaresOffic: right, and none of those commands involve apt/dpkg
[17:34] <leftyfb> RonWhoCaresOffic: you can also try https://www.howtogeek.com/114884/how-to-repair-grub2-when-ubuntu-wont-boot/
[17:35] <diofantoz> hi, i'm looking to "cap" my ubuntu system for mobile internet usage
[17:35] <Kraus> I'm having an issue where suddenly, overnight, something is occuring to my  Ubuntu Server (20.04). In the morning I discover I can't resolve its DHCP. All my fstab entries are nullified and upon ssh'ing to the hostname, Temporary failure in name resolution. DNS is being handled by my Linksys router. Just bought it a couple months ago. This happened with both the current and previous router I owned. The server's IP address is
[17:35] <Kraus> reserved in DHCP (192.168.1.16)
[17:35] <diofantoz> is there anything more than restricting updates i should think of?
[17:36] <Kraus> Rebooting generally solves the issue, but I'm curious why this happening in the first place. Someone in ##networking suggested there might be bugs in Ubuntu's DHCP client.
[17:36] <leftyfb> diofantoz: nope. Not much else runs by default that will use any significant amount of data
[17:36] <Kraus> (doubtful but they wanted me to ask about this here)
[17:36] <diofantoz> leftyfb: cool
[17:36] <leftyfb> Kraus: I would look at logs to see what happens with networking around the time of the issue
[17:36] <Kraus> https://pastebin.com/WDhN5Y2F
[17:37] <Kraus> The lost carrier, gainer carrier, is when I reboot the router.
[17:37] <Kraus> But you can see, the last router restart was today, and it was able to get its IP back.
[17:37] <Kraus> So that doesn't seem to be the issue...
[17:39] <leftyfb> Kraus: your pastebin is invalid.
[17:40] <Kraus> Oy, one sec, it must have expired.
[17:40] <Kraus> https://pastebin.com/kJXvnSfw
[17:40] <leftyfb> what time on here did the issue happen?
[17:41] <Kraus> Overnight whilst I slept
[17:41] <Kraus> And I went to sleep around 2 am, so there wasn't much time. :)
[17:42] <leftyfb> Kraus: you need to be more specific than that. Which of the times in the pastebin you posted reflect the issue? All of them? Also, it looks like this isn't happening often enough to point to any sort of DHCP client issue
[17:42] <leftyfb> Kraus: otherwise it would be happening pretty consistently
[17:42] <Kraus> The post does not reflect the issue. That's the log for my DHCP.
[17:43] <leftyfb> not once or twice every 4-5 days
[17:43] <leftyfb> Kraus: ok, please share the networking logs from the server reflecting the time(s) when the issue is happening
[17:43] <Kraus> Isn't that what I did? $ journalctl -u systemd-networkd
[17:44] <leftyfb> Kraus: ok, then if nothing is showing up in the logs, and the only symptom you have is not being able to resolve DNS for the hostname of the machine, it sounds like the issue is not with the machine. Can you ping the ip at the time that the hostname isn't being resolved?
[17:45] <Kraus> Yeah, I'm actually SSHed into it via the IP address right now.
[17:45] <Kraus> It's only the hostname that's not resolving. If I ssh [hostname] it says "Temporary failure in name resolution"
[17:46] <leftyfb> Kraus: ok, then the issue is not with the ubuntu server at all but with your DHCP/DNS server (router)
[17:47] <Kraus> Both my current and previous ones? That seems odd... only for that one ubuntu machine? Why not my desktop as well if that's the case?
[17:48] <nunya> Made /etc/systemd/system.conf.bak, Edited as sudo /etc/systemd/system.conf, edited lines DefaultTimeoutStartSec=90s to DefaultTimeoutStartSec=10s, DefaultTimeoutStopSec=90s to DefaultTimeoutStopSec=10s Now have 2:40 second from boot to icons on Desktop Better than 3:15 before. I also removed package brltty since I am not blind using a braille terminal. Having one less daemon at boot may have helped.
[17:48] <leftyfb> Kraus: no clue. You can test by running:   host <hostname-of-server-> <ip-of-router>
[17:50] <nunya> leftyb: longshanks: Made /etc/systemd/system.conf.bak, Edited as sudo /etc/systemd/system.conf, edited lines DefaultTimeoutStartSec=90s to DefaultTimeoutStartSec=10s, DefaultTimeoutStopSec=90s to DefaultTimeoutStopSec=10s Now have 2:40 second from boot to icons on Desktop Better than 3:15 before. I also removed package brltty since I am not blind using a braille terminal. Having one less daemon at boot may have helped
[17:51] <Kraus> leftyfb: Seems to be working.. it's giving the name, the address (192.168.1.1#53) alias, then it gives the name of the host and says it has an address of 192.168.1.16
[17:51] <leftyfb> Kraus: what OS are you running on your client?
[17:52] <Kraus> My client?
[17:52] <leftyfb> Kraus: yes, the machine you are trying to resolve from
[17:52] <tomreyn> diofantoz: if you're using network manager, tag the connection as metered.
[17:52] <Kraus> Oh! Ubuntu Studio 20.04
[17:53] <leftyfb> Kraus: ok, pastebin: nmcli device show|grep DNS
[17:53] <Kraus> https://pastebin.com/xUXL5bSq
[17:54] <leftyfb> Kraus: is 2600:6c4e:117f:d878:ea9f:80ff:fe79:b217 the IPv6 of your router?
[17:55] <leftyfb> Kraus: also: cat /etc/resolv.conf
[17:55] <Kraus> leftyfb: I have nooooo idea.
[17:55] <Kraus> I'm not familiar with how to work with IPv6 addresses
[17:57] <Kraus> Both my server and desktop have this in the resolv.conf nameserver 127.0.0.53 \ options edns0 trust-ad
[17:58] <leftyfb> Kraus: My initial thought is to try to disable ipv6 on your Ubuntu studio machine. My thought is the ipv6 address (possibly from your router) isn't resolving properly. Either way, it seems to me that the issue is with your studio machine having issues resolving local addresses and not actually a problem with the server.
[18:00] <Kraus> I'm curious to see where that v6 address is going to.
[18:01] <Kraus> I'm able to ping it easily.
[18:01] <leftyfb> Kraus: try to pull it up in a browser
[18:02] <Kraus> traceroute6 on the other hand... it's having some issues...
[18:02] <Kraus> leftyfb: How do I do that? I copy/paste it to the address bar and it thinks I'm doing a web search
[18:03] <leftyfb> Kraus: http://2600:6c4e:117f:d878:ea9f:80ff:fe79:b217
[18:03] <Kraus> Nope, still taking me to a web search...
[18:03] <leftyfb> Kraus: you could also look through the settings on your router. That might tell you if the address originated from there
[18:06] <Kraus> leftyfb: I translated it to IPv4 and it's 254.121.178.23
[18:06] <leftyfb> Kraus: that ipv6 nameserver is provided by your router. You should be able to delete it so you only get the ip of your router for DNS
[18:07] <Kraus> Wait, sorry, I'm confused..
[18:08] <leftyfb> Kraus: ipv6 addresses don't natively "translate" to an ipv4 address. That's not how they work.
[18:08] <Kraus> https://ipworld.info/ipv6-to-ipv4/ ?
[18:09] <Kraus> Anyways, what I'm confused about is.. why would I need to disable anything to make it stop not-working?
[18:11] <Kraus> Would it not be logical that I would have to disable IPv6 on all other machines too? That makes no sense. I also cannot ssh to my desktop from my server via hostname for the same failure to resolve.
[18:12] <leftyfb> Kraus: it seems you are intermittently not able to resolve local addresses. You have an ipv4 nameserver and an ipv6 being assigned to your DHCP clients on your network. If you are not actively using ipv6, disabling that on your router and/or your clients is a good place to start. You should only ever have 1 address being assigned for local DNS anyway ... unless you're running an ipv6 network locally as well. In which case, you would more than
[18:12] <leftyfb> likely know enough to sort all of this out on your own to begin with.
[18:12] <Kraus> leftyfb: Resolving local addresses is fine. It's resolving local hostnames that's not working.
[18:13] <leftyfb> Kraus: does your server also have multiple nameservers assigned? systemd-resolve --status|grep -i "dns servers"
[18:13] <rfm> leftyfb, Kraus: kind of late, but the way to use an IPv6 numeric address in a url is square brackets, like http://[2600:6c4e:117f:d878:ea9f:80ff:fe79:b217]
[18:13] <leftyfb> Kraus: I mean local hostnames, not addresses. I know the issue
[18:15] <Kraus> rfm: Yeah, that's pointing to 192.168.1.1
[18:15] <leftyfb> Kraus: disable ipv6 on either your router or client. Reboot client, test and then we go from there
[18:15] <Kraus> But what causes this?
[18:16] <Kraus> I hate treating a patient without understanding the cause.
[18:16] <cncr04s> ipv6 is used by at least 30% of websites, I wouldnt advise turning it off
[18:16] <leftyfb> Kraus: problem #1: multiple nameservers. problem #2, trying to use an ipv6 nameserver to resolve the ipv4 address of hostnames
[18:17] <Kraus> Multiple nameservers?
[18:17] <leftyfb> cncr04s: that's not how this works. They will still be able to get to all websites
[18:17] <Kraus> leftyfb: You may call me he/him/your-majesty ^_^
[18:17] <leftyfb> Kraus: pardon?
[18:17] <Kraus> You used 'they'. I was just letting you know what I am :)
[18:18] <Kraus> Sorry, can you show me where it was established I have multiple nameservers?
[18:19] <Kraus> (leftyfb
[18:19] <Kraus> )
[18:19] <leftyfb> Kraus: https://pastebin.com/xUXL5bSq
[18:20] <Kraus> leftyfb: Isn't that the default output for a machine that has IPv6 enabled?
[18:21] <leftyfb> Kraus: let me know when you disable ipv6 on either the client or the router and we'll continue troubleshooting
[18:21]  * Kraus blinks.
[18:21] <Kraus> I just want to understand why we need to do that first? Why would anything need to be altered from standard configuration before it will work?
[18:22] <Kraus> Because that doesn't make sense why it would be fine for months... then suddenly start having issues...
[18:22] <leftyfb> Kraus: because something isn't working, so we test a hypothesis before moving onto coming up with new ones
[18:23] <leftyfb> Kraus: because "intermittent"
[18:23] <Kraus> Ahhh! so this is just temporarily disabling it.
[18:23] <leftyfb> Kraus: to test, yes. Though if it solves the issue, I'd be hard pressed to be convinced of a good reason to re-enable it
[18:25] <Kraus> Gotcha! I thought we were just ducttaping the problem (which some are prone to do) instead of finding the cause
[18:29] <Kraus> leftyfb: Alright, looks like I disabled it, but the v6 address is still showing after $ nmcli device show|grep DNS
[18:29] <leftyfb> Kraus: to be honest, if disabling ipv6 fixes the problem, then I'm pretty happy with that. Also, you said this is intermittent. How often? The only way to test is to wait X amount of time you feel it should have happened
[18:29] <leftyfb> Kraus: reconnect to the network and/or reboot. Where did you disable ipv6?
[18:30] <Kraus> leftyfb: Every few weeks? ish? Hard to tell. I haven't been keeping a log.
[18:30] <Kraus> I followed these instructions https://itsfoss.com/disable-ipv6-ubuntu-linux/
[18:30] <Kraus> Says it's only a temporary disable.
[18:30] <leftyfb> Kraus: ok, then you'll have to wait a few weeks or longer before knowing if what you did fixed the issue. Nothing to do in the meantime
[18:31] <Kraus> There has to be a more definite way...
[18:31] <leftyfb> Kraus: ok, lets get it verified and disabled properly so it'll stay disabled across reboots.
[18:31] <Kraus> Could you explain how to do it properly?
[18:31] <leftyfb> Kraus: the only way to test a fix for a problem is to replicate the problem. If the problem is intermittent and only happens every few weeks, you have to replicate that scenario
[18:32] <Kraus> And that's the thing, is I'm not sure what event is triggering it. So far all we've done is temporarily disable IPv6 on a machine that isn't being affected.
[18:33] <leftyfb> Kraus: it is 100% affected
[18:34] <Kraus> Why I would want to permanently disable something on an unrelated machine to fix the problem occuring on the source machine... makes no sense. Would I not logically have to disable it on all machines other than the problem machine?
[18:34] <leftyfb> Kraus: the issue is said machine not being able to resolve a local hostname on it's own but can when you specify the nameserver manually. That means the issue is not with the hostname you're trying to resolve but the machine trying to do the resolving
[18:35] <Kraus> (which machine is 'said' machine)?
[18:35] <Kraus> Oh waitwaitwait, maybe you misunderstand.
[18:35] <leftyfb> Kraus: source(problem) machine = all DHCP clients who are receiving multiple nameservers
[18:36] <Kraus> I am not specifying a nameserver. I stated that I can ssh to the affected machine(server) when specifying its IP address.
[18:36] <leftyfb> Kraus: you did when you ran the host command I gave you earlier
[18:36] <leftyfb> Kraus: here, lets test.... is your studio machine still having the issue resolving the hostname of the server?
[18:39] <Kraus> Whoaaahhh... wait, what you just said made me think about something... I just tried pinging the hostname from my phone using LANDroid and it's able to find it...
[18:39] <Kraus> Veerrrryyy interesting....
[18:39] <Kraus> So it's not the server itself...
[18:39] <leftyfb> Kraus: so that shows you the problem is not with the server
[18:39] <Kraus> Right!
[18:39] <Kraus> Then why does rebooting the server usually solve the issue?
[18:39] <leftyfb> it doesn't
[18:40] <Kraus> It has in the past...
[18:40] <leftyfb> Kraus: coincidence
[18:40] <Kraus> Also! Remember what I mentioned about suddenly my sound not working on my ubuntu-studio machine? Coincidence? I think NOT.
[18:40] <Kraus> Both these issues began after re-logging into my studio machine.
[18:41] <leftyfb> Kraus: sounds not working is completely unrelated to anything we've discussed thus gar
[18:41] <leftyfb> far*
[18:41] <Kraus> I have a feeling the problem may be related though... the odds are too great.
[18:41] <leftyfb> Kraus: negative
[18:42] <leftyfb> Kraus: would you like to continue with the setup for troubleshooting your DNS issue?
[18:42] <Kraus> I thought you said I had to wait for weeks?
[18:43] <leftyfb> Kraus: we want to verify ipv6 is disabled and then make sure it will take across reboots and reconnections
[18:44] <Kraus> According to these that I'm reading, I have to add some lines to my sysctl.conf file.
[18:45] <leftyfb> Kraus: yup, and we need to test
[18:45] <leftyfb> Kraus: actually, with later versions of ubuntu, I think that doesn't 100% work
[18:46] <leftyfb> Kraus: https://linuxconfig.org/how-to-disable-ipv6-address-on-ubuntu-20-04-lts-focal-fossa   look at step 1 and 2
[18:46] <Kraus> Yeah, I'm seeing different versions of these instructions...
[18:47] <Kraus> *sighs* Yknow, I'm really.. really considering just redoing my machine. I can't even upgrade to 20.10 because the studio devs switched to cinnamon or something.
[18:47] <leftyfb> Kraus: that will 100% not fix the issue
[18:47] <Kraus> True but why configure lines on a config file that will soon be wiped out?
[18:48] <Kraus> I'm not doing anything more important right now, I should just get working on this system redo.
[18:49] <Kraus> Err, not Cinnamon.. Plasma..
[18:50] <leftyfb> Kraus: the issue is(my hypothesis) with your router giving out an ipv4 and ipv6 nameservers and you machine(s) having issues trying to resolve an ipv4 address from the hostname using the ipv6 address it might intermittently try to use. If we disable ipv6 on the client, then no ipv6 nameserver will be assigned to it. If we disable ipv6 on the router(my recommendation), then it will not assign ipv6 addresses anywhere on your local network
[18:51] <Kraus> leftyfb: Shouldn't I just disable IPv6 on the router then?
[18:51] <leftyfb> Kraus: that is preferred, yes. I have recommended that multiple times above
[18:51] <Kraus> You told me to disable IPv6 on my studio machine. O_o
[18:52] <Kraus> Anyways, let me disable it. One moment.
[18:54] <leftyfb> Kraus: you will need to reconnect to your network or possible reboot machines in order for them to update their network assignments from dhcp
[18:55] <nikolam> where is exactly /tmp folder for snap application? I started 'namebench' and it is supposed to save some report at /tmp , but where it is actually saved, when it is an snap application??
[18:57] <nikolam> I have just found it under "/tmp/snap.namebench-snap/tmp" but it is only visible to sudoer/root account and not for user that is using the snap app??
[19:00] <Kraus> Test
[19:00] <Kraus> Wow, were you able to see anything I just typed? I was still able to type even though I got booted from the channel.
[19:01] <nikolam> Kraus, there is kick, it's not the ban. One can rejoin and learn lesson :P
[19:01] <Kraus> Nono, what I mean is I lost connection to Freenode all together. ANyways here's what I said:
[19:02] <leftyfb> Kraus: https://pastebin.ubuntu.com/p/KFpzYYFFBf/
[19:02] <Kraus> [10:57:16 am] <Kraus> Okay.. let me see if I can describe precisely all what just happened right now.
[19:02] <Kraus> [10:58:59 am] <Kraus> Okay, disabled IPv6 on the router, '$ nmcli device show|grep DNS' still showed two nameservers, so I rebooted my client machine(studio), still showing two nameservers, AND just now, I got disconnected from IRC (I'm using Quassel Core), which means my server lost connection, regained connection... so I'm wondering if disabling IPv6 caused something to happen on the server.
[19:02] <Kraus> Anyways, still cannot resolve the hostname
[19:03] <leftyfb> Kraus: the router could certainly had to restart some services in order to make the change
[19:03] <Kraus> Right.
[19:03] <leftyfb> Kraus: you are still getting the ipv6 assigned?
[19:03] <Kraus> leftyfb: Yes, still showing two nameservers.
[19:03] <leftyfb> Kraus: are you 100% sure you disabled the router assigning an ipv6 nameserver via DHCP?
[19:04] <leftyfb> Kraus: your router might have separate configs for local ipv6 and public
[19:04] <leftyfb> Kraus: you might have only disabled ipv6 for the internet side, not locally
[19:05] <Kraus> Yes, logged into it > Connectivity > Internet Settings tab > Type of Internet Connection | Edit > Clicked Edit, Disabled IPv6, and it currently says "IPv6 - Automatic: Disabled"
[19:05] <Kraus> AHhh
[19:06] <Kraus> I suppose there's no purpose in disabling the internet side of things is there?
[19:06] <leftyfb> right
[19:07] <Kraus> And under the Local Network tab, there isn't anything mentioning enabling/disabling IPv6.
[19:07] <leftyfb> Kraus: ok, then we might have to do this only on the client-side
[19:08] <Kraus> One sec, searching for a solution onlnie
[19:10] <Kraus> (What's a 6rd tunnel?)
[19:11] <leftyfb> it routes ipv6 traffic through ipv6 tunnels. Nothing related to what we're doing
[19:11] <leftyfb> sorry, ipv4 traffic through ipv6 tunnels
[19:19] <Kraus> Insane.. unrelated rant.. they published all langauges of my router's manual together as one massive PDF.. and "ipv6" isn't mentioned once in all of its 1258 pages.
[19:24] <quantum> Is there a way to boot to another device at the cli?
[19:25] <quantum> G**gle is baffled.
[19:28] <leftyfb> quantum: you mean booting another OS? Your BIOS/EFI handles that
[19:29] <leftyfb> quantum: or grub if you have multiple OS's listen in GRUB, you can modify the default item in grub then reboot
[19:29] <leftyfb> listen/listed
[19:29] <quantum> This is an Arm board, and I have no access to its bios.  The mechanism written about doesn't work.
[19:30] <leftyfb> quantum: ok, then you need to seek support from the devices manufacturer
[19:30] <leftyfb> quantum: or maye uboot or whatever loaded that board uses
[19:31] <quantum> uboot, but I can't make it go to the USB flash drive.
[19:32] <leftyfb> quantum: maybe try #ubuntu-arm
[19:32] <quantum> k
[19:35] <lmat> How do I make X not start at boot? I looked online, but there is conflicting information and recommendations and more importantly, none of them make sense :-)
[19:35] <lmat> Seems like I could disable a systemd unit that starts my window manager?
[19:36] <lmat> Oh yeah, this is an XY problem: I'm wanting ubuntu to start up with only my program running X.
[19:36] <lmat> My plan is to turn off the desktop manager, and adjust /usr/bin/startx only to run x& /path/to/my_prog;
[19:36] <leftyfb> lmat: https://askubuntu.com/a/79682/1151311
[19:37] <leftyfb> lmat: look at "For systems that use systemd"
[19:37] <lmat> leftyfb: I can skip the GRUB stuff, right? (what does grub have to do with anything?)
[19:37] <leftyfb> right
[19:38] <lmat> leftyfb: What systemd target is ordinarily run?
[19:38] <oerheks> graphical.target
[19:38] <lmat> oerheks: Ahh, okay.
[19:39] <lmat> And then how should I make it so that all that runs is X& my_prog; ?  I was thinking of something like    echo "x&my_prog;" | sudo tee /usr/bin/startx;   but I'm guessing there's some other more elegant mechanism?
[19:39] <lmat> of course s/x/X/
[19:41] <oerheks> weird approach, why asking for a more elegant mechanism than just the regular desktop?
[19:41] <Kraus> Okay. Here's what Linksys' rep said, which I believe, translated from their rough English, means the router isn't capable of disabling IPv6 locally.
[19:41] <Kraus> Mikaelh Joshua A: I believe you have also Ipv6 internet connection. However, let me set your expectation, Damien. Our router doesn't have an option to disable the Ipv6 or Ipv4 local connection on the router, because the IP address the client devices will depend from the connection type provided by the router. What we can only do on our end is to disable the Ipv6 internet connection type of the router, so that those affected devices will
[19:41] <Kraus> automatically be connected on the Ipv4 connection.
[19:41] <silent-failure2> putting `exec myprog` in your ~/.xinitrc (be sure to remove/comment out all the other stuff), then running startx should work.
[19:42] <lmat> oerheks: I have a special application and I want it to look like an embedded display. It's not one, but think of an arcade computer.
[19:42] <lmat> oerheks: I don't want multi-tasking. I just want my one client.
[19:42] <silent-failure2> If it's Qt or SDL based, it can even run without X at all.
[19:42] <EriC^^> +1 on ~/.xinitrc
[19:43] <lmat> silent-failure2: Right, that's why I said   X& my_app; that should start the X server, no?
[19:43] <lmat> silent-failure2: Ahh, thanks!
[19:44] <silent-failure2> I'm not sure. X has some complex initialization crap it has to do in order to start properly.
[19:44] <leftyfb> Kraus: ok, then just disable it on your client
[19:45] <lmat> silent-failure2: "Ahh, thanks!" was regarding .xinitrc .
[19:46] <lmat> multi-user target doesn't even run getty, right?
[19:47] <silent-failure2> I think it does. getty (or agetty) is normally what does that text based login prompt.
[19:47] <lmat> silent-failure2: Right, I don't see a prompt on the screen. it just says something about /dev/sda being okay.
[19:48] <lmat> silent-failure2: But how do I make it automatically run startx? Should I be creating a systemd unit for that?
[19:48] <silent-failure2> Try pressing Ctrl+Alt+one of the F1-F12 keys.
[19:49] <Kraus> leftyfb: Before I do that, I'm going to just reboot the server first to get the issue cleared up, because I need to redo my desktop machine anyways, so I need to have access to it in the short term.
[19:49] <silent-failure2> There are multiple ways to do that. A systemd unit could work.
[19:50] <lmat> silent-failure2: Okay. I think I understand that way best, so maybe I'll try that :-D
[19:50] <EriC^^> lmat: you could add "if [[ -z $DISPLAY ]]; then startx; fi" to your ~/.profile
[19:51] <lmat> EriC^^: Then it would only run when I log in, right? I'm wanting more automatic.
[19:51] <silent-failure2> EriC^^, yeah, provided it actually logs you in at boot.
[19:52] <EriC^^> lmat: ah i see
[19:52] <Kraus> Be back in a bit.
[19:52] <silent-failure2> And if you have to log into a TTY to fix something, it will run when you don't want it to.
[19:53] <silent-failure2> I imagine systemd has some way to run a unit as non-root.
[19:53] <lmat> So I can create a new unit file that runs startx as my user and add it to multi-user wants. But is that okay? All the multi-user wants start up at the same time, right? Maybe I should create a new target AFTER multi-user to run my startx? Or maybe I should remove all the wants of graphical target and put my one unit file in there?
[19:53] <lmat> silent-failure2: Of course. User=<not root> :-)
[19:55] <silent-failure2> I've done some things like that on Raspberry Pi to run a single application, but I use void linux on it, which doesn't have systemd.
[19:56] <lmat> silent-failure2: ahh
[20:05] <Kraus> Test
[20:05] <Kraus> leftyfb:
[20:05] <Kraus> I just remembered what DOES fix it.
[20:05] <lmat> I get https://paste.rs/EY6 now. I created a unit that runs startx as the user, and this is the error part of the output.
[20:06] <lmat> (I added set -x; to /usr/bin/startx so that I could see what was running.)
[20:06] <Kraus> leftyfb: It's not a reboot of either machine.. yknow what fixes it? I enter a specified static DNS entry of 190.168.1.1 ... This fixes it for a while.. until it happens again, and when it happens again I undo that and revert the DNS entry to 0.0.0.0
[20:07] <leftyfb> Kraus: where are you editing that?
[20:07] <Sven_vB> asking again in hopes other people are online now: In bionic, how can I make non-fuse mount fail if the mountpoint directory isn't empty?
[20:08] <Kraus> leftyfb: In the Linksys Smart Wifi control panel, under Connectivity > Local Network > DHCP Server > Static DNS 1: ___.___.___.___
[20:08] <leftyfb> Kraus: wait, what is that normally set to?
[20:09] <leftyfb> Kraus: you should only ever have 192.168.1.1 in there. Nothing else.
[20:09] <Kraus> It's normally set to 0.0.0.0, because, by default, that's how the router comes.
[20:09] <leftyfb> ah, that could be it then
[20:10] <leftyfb> Kraus: put in the router's ip and make sure all other dns entries are empty. Not 0.0.0.0 or anything else. Empty
[20:10] <Kraus> I would not expect a consumer product to expect 100% of its users manually edit a static DNS entry. Given the technical illiteracy of the vast majority of home internet users, I could never imagine a brandname company expecting them to even know how to access their configuration.
[20:10] <lmat> Sven_vB: I've done that many times, but not in bionic.   if mountpoint /path/to/mount >& /dev/null; then echo "it's already mounted!"; fi;
[20:13] <Kraus> leftyfb: Yes, that seems reasonable.. until once again the problem occurs even when the field is set to 192.168.1.1 ... it /will/ happen again, because through experience I can confirm it's happened to me. When it happens again, I just have to revert it again to 0.0.0.0 ... then THAT works for a while.. then when it reoccurs, I set it back to 192.168.1.1... Conversely I can also go from 0.0.0.0 to 192.168.1.1, then hit apply, then
[20:13] <Kraus> immediately return it to 0.0.0.0 (or vice versa) and it will still fix it... So in other words, what I'm suggesting is: the action of changing the static DNS seems to jog something temporarily in the router to where my machines suddenly see each other by hostname.
[20:14] <Kraus> The strange thing is, back in 18.04 and 18.10, to the best of my recollection, this didn't happen. These problems began to occur once I upgraded my machines. It's very strange, and I can't put my finger on it.
[20:14] <leftyfb> Kraus: that is 100% a router issue
[20:14] <Kraus> A router issue, meaning incompatible with Ubuntu?
[20:15] <Kraus> This has happened both with my current router model, and the previous router I owned for 5 years.... but ONLY once I switched to ubuntu.
[20:15] <Kraus> Both were Linksys routers.
[20:16] <Kraus> I seem to recall someone posing the possibility that Windows doesn't get affected by hostname settings on Routers due to... something or other... like.. it's too dumb to even recognize there's a problem.
[20:18] <Kraus> Again, all this seems to be speculation, opinion..
[20:19] <Kraus> I'd rather find a concrete cause for it instead of speculating.
[20:20] <Sven_vB> lmat, in my case that won't help. I want to guard against a directory becoming invisible
[20:21] <kill-animals> can you make a shortcut that follows a file around if you move it?
[20:24] <leftyfb> Kraus: in your GUI network settings on the desktop, under the ipv4 tab, under DNS, click off the "Automatic" button and put "192.168.1.1" for your DNS. Then in the ipv6 tab, under ipv6 method, click disabled. Reboot and verify your DNS server using nmcli from before
[20:24] <silent-failure2> Hard links are the only thing I know of that can do that.
[20:25] <leftyfb> kill-animals: no. A hard link will break and leave behind a snapshot copy if the file is moved
[20:25] <kill-animals> leftyfb: hmmm I might be fine with that.
[20:26] <leftyfb> kill-animals: then use hard links
[20:26] <leftyfb> kill-animals: mind you, if you move a file an modify the file in the new location, the link will still point to the original, unmodified version
[20:27] <kill-animals> leftyfb: I am building a library that will deduplicate repeat downloads, and so I want to store its hash and path in a database, and when I download something, it simply checks for it before going through with the actual downlaod
[20:28] <Kraus> leftyfb: Note, I'm on XFCE, not Gnome, so I might have a different GUI than that.
[20:28] <leftyfb> Kraus: you might. Look for something similar
[20:29] <Kraus> Yep, "Method: Manual" .. "DNS Server", and I put my 192.168.1.1 address in there.
[20:29] <Kraus> There IS a table with Address Netmask and Gateway entries which I can "add" something...
[20:31] <Kraus> Anyways, guess what. It now only shows my IPv4 DNS server in nmcli
[20:34] <leftyfb> Kraus: ok, now you wait until you can reproduce the issue. Until you decide it's not coming back
[20:35] <Kraus> leftyfb: How can I reproduce it? It happens on its own at unknown times...
[20:35] <leftyfb> Kraus: you wait and keep testing
[20:35] <Kraus> Well.. I mean it either happens or it doesn't!
[20:35] <leftyfb> right
[20:36] <leftyfb> Kraus: does your server run a GUI?
[20:37] <leftyfb> Kraus: If not, https://askubuntu.com/a/1130475/1151311  follow that for your netplan network settings. Of course plug in your network interface name and ip address for the nameserver
[20:38] <leftyfb> Kraus: leave out the renderer: part
[20:38] <Kraus> leftyfb: Nah, it's ubuntu server, headless machine, cli only.
[20:39] <Kraus> leftyfb: Thanks for all your help with this. :)
[20:39] <Hash> Hello, in Ubuntu 20.04, I try to compile this: https://github.com/Revan600/d2tweaks and I can satisfy all the dependencies with -dev packages, but I cannot find out what this means:   set_target_properties Can not find target to add properties to: fmt::fmt-header-only - Which package might contain this?
[20:40] <leftyfb> Kraus: I think yours should look like https://pastebin.ubuntu.com/p/CPbVm93qgS/
[20:41] <leftyfb> Hash: you'll need to contact the developer of that application for support
[20:41] <Kraus> Yeah right now it's very minimal. It's a file called 50-cloud-init.yaml and consists of:
[20:41] <Kraus> network:
[20:41] <Kraus>     ethernets:
[20:41] <Kraus>         enp5s0:
[20:41] <Kraus>             dhcp4: true
[20:41] <Kraus>     version: 2
[20:41] <leftyfb> Kraus: pastebin
[20:42] <leftyfb> Kraus: wait a minute or 2 for the quiet on you to be removed
[20:42] <silent-failure2> Is it really worth using pastebin for just 5 lines?
[20:42] <Kraus> Oh shoot, it thought I was spamming? XD
[20:43] <Hash> It is not.
[20:43] <Hash> It's just, IRC isn't like discord, so people joining, parting, quitting, modes being set, etc. etc. can destroy your paste, nevermind flooding the channel.
[20:44] <silent-failure2> If only IRC supported line breaks within messages.
[20:44] <Kraus> Alright, I copypasta'ed your recommended config lines.. do I need to restart network services?
[20:45] <leftyfb> Kraus: sudo netplan apply
[20:45] <leftyfb> Kraus: warning, you better have a way to access the console locally should something go wrong
[20:45] <Kraus> Yeah I do.
[20:46] <Kraus> INCONSISTENT INDENTATION. Wow, I love things that are sticklers for details.
[20:47] <silent-failure2> You'd love programming in python. Indentation is part of the syntax.
[20:47] <leftyfb> Kraus: http://www.yamllint.com/
[20:47] <Kraus> Hhehehe :)
[20:48] <leftyfb> Kraus: it's yaml, need to check the syntax/whitespace
[20:48] <Kraus> I really, really have always wanted to learn to code... it's just never... materalized.
[20:51] <Kraus> Testing 1 2... am I still here?
[20:51] <silent-failure2> yes
[20:51] <sarnold> pong
[20:52] <Kraus> Alright, applied the netplan! And I'm still here...
[20:52] <Kraus> Yknow, funny thing is, when I do 'nmcli device show|grep DNS' on the server itself, nothing shows up. No nameservers listed.
[20:53] <leftyfb> Karasu: nmcli is for network manager, which you're not running on the server ...
[20:53] <Kraus> Ahh.. doing only 'nmcli device', it shows that both my NIC device and lo are "unmanaged"
[20:53] <leftyfb> Kraus: systemd-resolve --status|grep -i "dns servers"
[20:54] <Kraus> How do you guys memorize all these commands? Jeesh!
[20:54] <Kraus> DNS Servers: 192.168.1.1
[20:54] <leftyfb> Kraus: you should be good then
[20:55]  * Kraus collapses.
[20:55] <Kraus> See!? EASY!
[20:55] <Kraus> Totally beginner friendly. XD
[20:58] <leftyfb> Kraus: basically, we've taken the router of the equation of assigning DNS servers to your 2 machines. You have to think about this being a router issue. I swore off those typical consumer level routers years ago due to issues like this and I'll never go back
[20:59] <Kraus> Yeah, it's not my plan to stay on a consumer router for much longer. Money being the primary setback, I eventually want to be on something rackmount, with a switch perhaps, and 10gig LAN.
[21:00] <leftyfb> Kraus: I recommend Ubiquiti. But that's offtopic here
[21:00] <Kraus> I'll look into it :)
[21:04] <Kraus> Now to solve the ever-persistent problem of Numlock not being active on startup. ;)
[21:05] <rory> that was a BIOS setting on my old PC
[21:05] <Kraus> Oh it's set.
[21:05] <Kraus> There's something about Ubuntu that just doesn't like having it active by default.
[21:07] <Elliria> This is old, but might help with the NumLock key: https://askubuntu.com/questions/155679/how-to-enable-numlock-at-boot-time-for-login-screen
[21:07] <Kraus> And I found a slew of options-- yes, one moment, let me see if that's one of them
[21:07] <Kraus> Ahhh yes, 8 years ago. Apparently something changed after Ubuntu 16 or 17.
[21:08] <Elliria> Actually, it seems to have recent information down at the bottom. It's worth a look.
[21:08] <Kraus> Looks like it's gnome-specific
[21:08] <Elliria> Ah, sorry about that.
[21:09] <Kraus> It's fine :) I ran into this thread months ago.
[21:09] <Kraus> There's a "Restore num lock state on startup" that might be worth trying. Not exactly the same but still would work.
[21:10] <Elliria> There may yet be a simple solution to it. You might want to go poking around in all the options. I don't use Ubuntu (currently use Ubuntu MATE) and am switching to Kubuntu next. Otherwise, I'd poke around.
[21:10] <yates> i have "apt install manpages" installed but i still don't have manpages for the c lib stuff like "man strftime" - how to get?
[21:10] <Kraus> Ahh, no, the Restore function only activates AFTER you've logged in.
[21:11] <Kraus> It's fine, Elliria :) I'm going to be reinstalling things soon anyways so I'll worry about it then.
[21:11] <yates> nm
[21:28] <tomreyn> yates: manpages-dev
[21:47] <xnat834[m]> Does anyone have a problem while trying to connect to a hotspot with a Windows 10 client?
[21:47] <xnat834[m]> It's Kubuntu in my case, but maybe someone had the problem with Ubuntu
[21:48] <xnat834[m]> My Windows 10 client can connect to my regular router, the hotspot of my Android 10 phone, etc.
[21:49] <xnat834[m]> Only to the hotspot on my notebook with Kubuntu it doesn't work and the Windows 10 client also want me to input a PIN for some weird reason
[21:50] <silent-failure2> The PIN seems like the router is using WPS.
[21:51] <xnat834[m]> The router would be my Kubuntu notebook hotspot...
[21:51] <xnat834[m]> I definitely didn't set WPS there and my Android 10 phone also doesn't ask for that when I connect there
[21:51] <xnat834[m]> So weird
[22:07] <xnat834[m]> Ok, if anyone has problems with the integrated hotspot functionality: https://unix.stackexchange.com/questions/626619/ubuntu-20-04-lts-wifi-hotspot-not-working
[22:07] <xnat834[m]> This worked for me
[22:08] <xnat834[m]> Creating a hotspot via "create_ap" from this repository works flawlessly for me
[22:19] <jas02> Hi, I am having trouble with Ubuntu 18.04 Server. I am testing LXD on the server, but with manually created bridge br0 (managed by Netplan). But for some reason there is all the time pop-up other bridge, called virbr0 with virbr0-nic attached. I uninstall the LXD snap, but the bridge is still present.
[22:20] <jas02> I tried to remove bridge using brctl, I found that the bridge is managed by systemd and I deleted /lib/systemd/network/99-default.link, but the systemd(probably) is still starting bridge froom some unknown place.
[22:20] <jas02> How can I get rid of this bridge for good? I have no idea where is still bridge hooked.
[22:25] <OrenTanay> My work requires me to use a windows10 machine, so in the past I've used Oracle VM VirtualBox to have access to Ubuntu, and Virtualbox has a simple feature to make snapshots that have saved me several times when I've messed up my Ubuntu install trying to make changes. I now want to try WSL2 with ubuntu 20.04, but I would really like some sort of snapshot feature. Does anyone have a recommentation
[22:25] <OrenTanay> to accomplish something equivelent? Thanks.
[22:28] <tomreyn> !wsl | OrenTanay
[22:29] <tomreyn> this is probably more of a ##windows / hyperv question, though, if you're talking about snapshotting outside of the guest.
[22:32] <OrenTanay> Yes, it was the ability to create snapshots I was looking for, but thank you.
[22:33] <OrenTanay> i'm trying the ubuntu-on-windows group now.
[22:33] <HackerII> I now know why your new world agenda is relevant'
[22:34] <HackerII> im changing distros.. you have been taken ver... crash my system.. idiots
[22:34] <tomreyn> OrenTanay: this can be from libvirt / qemu
[22:34] <tomreyn> jas02: : this can be from libvirt / qemu
[22:35] <tomreyn> HackerII: do you have an ubuntu support question?
[22:35] <HackerII> i can do nothing on ubuntu as compared as the previous versions
[22:35] <HackerII> yes i do tomreyn
[22:35] <tomreyn> HackerII: do you have an ubuntu support question?
[22:35] <tomreyn> let's hear it then
[22:35] <HackerII> why did your distro give in
[22:36] <tomreyn> that's not a support question, that's trying to start a discussion, which you could do in #ubuntu-discuss
[22:36]  * HackerII waits
[22:36] <HackerII> tomreyn,  why did your distro give in
[22:36] <HackerII> discuss it
[22:37] <oerheks> for discussions, there is a dedicated channel, support only please, thanks.
[22:37] <HackerII> oerheks,  stop proselyting
[22:38] <HackerII> i have 100 vpn.. dont ruse me as well as the general public as to why you sold out
[22:38] <tomreyn> HackerII: please see /topic for the guidelines applying to this channel.
[22:39] <HackerII> your os resembles windoze
[22:39] <HackerII> i signed onto ubuntu back in6.o6.. what  happened
[22:40] <oerheks> HackerII, please stop the random noise
[22:40] <HackerII> nise.. gimme a break
[22:40] <HackerII> is that all you git
[22:40] <HackerII> got
[22:41] <HackerII> i belong to many forums.. dont do me dirty
[22:43] <HackerII> no worries.. my forums will make a note of all your doings.. remember.. you dupped the people.  and I will make it known..... FAST
[22:43] <HackerII> bye
[22:44] <jas02> tomreyn: Many thanks, that was it!
[22:45] <tomreyn> you're welcome
[23:22] <ptah> is it possible to upgrade firefox on Linux kernel 2.6.24 Ubuntu 8.04.2?
[23:22] <hansh> i want to multiboot windows and ubuntu, should i install windows first or ubuntu first?
[23:23] <ptah> thats 2008 system on mini dell laptop
[23:24] <ptah> it use to play youtube vids and facebook worked, but now it doesnt work
[23:24] <Bashing-om> hansh: ubuntu 1st - as Windows will over-write the boot code.
[23:24] <tomreyn> ptah: theoretically, yes, but it's a very bad idea. much better would be to do a fresh installation of something that's not reached EOL.
[23:25] <sarnold> ptah: probably not -- newer firefox binary packages will have requirements on newer libraries, and those newer libraries will have requirements for newer kernels. it'd be a lot easier to upgrade to something much newer
[23:25] <ptah> i see
[23:25] <sarnold> ptah: you could build a newer firefox on that machine, but I strongly doubt that machine is powerful enough to run the build in any reasonable time
[23:26] <ptah> hiow come old windows from like 2008 will still work?
[23:26] <ptah> how*
[23:26] <Bashing-om> !dual boot | hansh
[23:26] <tomreyn> ptah: note also that current ubuntu releases may not work well on such old hardware, if it doesn't match the minimum requirements.
[23:26] <ptah> linux is suppose to be more costumzable
[23:27] <hansh> thanks
[23:27] <ptah> ok
[23:28] <ptah> thank you
[23:28] <tomreyn> ptah: IF this system lacks the resources needed: there are other linux distributions which are customized to work with limited system resources. ask in ##linux
[23:28] <hansh> does Lubuntu still make 32bit builds?
[23:28] <ptah> then ubuntu is the problem not another distro?
[23:29] <hansh> ptah, is it a 32bit or 64bit cpu ?
[23:29] <ptah> hansh, how do i check that?
[23:29] <tomreyn> hansh: there's #lubuntu for lubuntu specific questions
[23:30] <ptah> i forgot lol
[23:30] <sarnold> cat /proc/cpuinfo or lscpu
[23:31] <sarnold> uname -a
[23:31] <tomreyn> grep -o -w 'lm' /proc/cpuinfo
[23:31] <ptah> Linux ptah 2.6.24-29-lpia #1 SMP Thu May 26 09:13:33 UTC 2011 i686 GNU/Linux
[23:31] <sarnold> 32 bit
[23:32] <ptah> 64
[23:32] <tomreyn> this is a 32-bit kernel. is it a 32-bit cpu?
[23:32] <ptah> clflush 64?
[23:32] <sarnold> tomreyn: what's the lm flag mean?
[23:33] <Bashing-om> sarnold: It's a flag for "long mode", meaning it's 64 bit capable :D
[23:34] <tomreyn> thanks Bashing-om, i didn't have that answer handy
[23:36] <ptah> its 64
[23:36] <ptah> will it work?
[23:36] <ptah> clflush size 64
[23:39] <sarnold> Bashing-om: aha :) thanks
[23:40] <sarnold> ptah: that's definitely encouraging; I think the last LTS release with a 32 bit x86 installer was 16.04, which ends standard support in a few weeks
[23:40] <sarnold> ptah: it might be worth trying a 20.04 installer; if it works, it'll give new firefox, and you'll get security updates again
[23:41] <ptah> sarnold, yeah i notice certain websites give a security error
[23:41] <ptah> :)
[23:42] <ptah> that means i have to upgrade the entire kernel?
[23:42] <euxneks> for some reason a directory in a samba share is not showing files, but showing directories on the clients, but other directories are showing files. As far as I can tell there is no selinux - where can I start to troubleshoot this?
[23:44] <ptah> sarnold, so apt-get upgrade firefox wont do anything right?
[23:44] <ptah> :)
[23:44] <tomreyn> ptah: as i mentioned earlier, you'll want to do a fresh install.
[23:44] <euxneks> I am, however, able to view the files using ls if I open a terminal in the mounted samba share
[23:45] <ptah> i see, ok
[23:45] <tomreyn> ptah: first of all, i'd check whether the system provides enough resources, though.
[23:46] <ptah> tomreyn, probably have to use a regular laptop
[23:46] <ptah> this is just a free gift mini laptop
[23:46] <sarnold> ptah: yeah that apt-get update won't work :)
[23:47] <ptah> ok :)
[23:50] <soulseeker-cloud> No it's tainted
[23:50] <soulseeker-cloud> See my previous post
[23:50] <soulseeker-cloud> If you can salvage the cluster
[23:51] <soulseeker-cloud> Use it for warp core
[23:51] <ptah> sarnold, do people still use redhat ?
[23:51] <sarnold> ptah: yes
[23:51] <tomreyn> soulseeker-cloud: it's not clear what you're referring to. do you have an ubuntu support question?
[23:51] <ptah> sarnold, in companies, my old boss switched to ubuntu
[23:51] <soulseeker-cloud> Warp core trajectory
[23:51] <soulseeker-cloud> Zephram alpha calculator
[23:51] <soulseeker-cloud> Ptah 'people'?
[23:51] <sarnold> soulseeker-cloud: please stop
[23:52] <tomreyn> soulseeker-cloud: let's please stay on topic, ubuntu support.
[23:52] <raffaem2> my ubuntu doesn't switch focus to applications when it should. For example, I double click a file in Dolphin, and VLC media player gets opened but in the background. I have to manually swithc to it. I click a magnet link on the web and Transmission gets opened in the bakground
[23:52] <soulseeker-cloud> Yes tom refund project zephram alpha
[23:53] <soulseeker-cloud> Serious talk quit being simple weed sprouts
[23:53] <soulseeker-cloud> Thorny weeds
[23:53] <sarnold> ptah: there's lots of reasons why one or another distro may make more sense for any given user; while ubuntu is in one sense a competitor to fedora and rhel, in another sense we're callaborators and colleagues
[23:53] <Sonderblade> anyone know how to get rid of the "agetty" processes? /sbin/agetty -o -p -- \u --keep-baud 115200,38400,9600 ttyS0 vt220
[23:54] <soulseeker-cloud> Yes tom refund project zephram alpha
[23:54] <soulseeker-cloud> Use all ubuntu's got the whole distributed computing network
[23:55] <ptah> soulseeker-cloud, lol
[23:56] <soulseeker-cloud> Ptah does it matter?
[23:56] <euxneks> looks like samba has some stored attributes somewhere
[23:59] <sarnold> Sonderblade: hopefully helpful http://0pointer.de/blog/projects/serial-console.html -- I found the link from systemd-getty-generator(8)