[00:17] <danielb> ash_worksi, there's a copy mode in every program if you run it in tmux ;)

[00:51] <gambl0re> is it possible to run apt update on live usb?
[00:58] <Bashing-om> gambl0re: Yes - but the change does not persist a re-boot.
[01:00] <Nate_Sturtz> when is 20.10 getting released?
[01:00] <Examknow> Nate_Sturtz: It is already released.
[01:01] <Nate_Sturtz> it is
[01:01] <Examknow> Yes.
[01:01] <Nate_Sturtz> ok, I did not know that, i am running a server should I upgrade?
[01:01] <Nate_Sturtz> or wait until 20.11?
[01:01] <Examknow> Nate_Sturtz: That's up to you
[01:02] <Nate_Sturtz> ok, thanks, where can I find the change log at?
[01:02] <rfm> Nate_Sturtz, next release will be 21.04, only two releases a year.  no 20.11.
[01:02] <Nate_Sturtz> ok, sorry
[01:04] <Nate_Sturtz> thank you
[01:05] <sonicwind> Nate_Sturtz, https://discourse.ubuntu.com/t/groovy-gorilla-release-notes/15533
[01:05] <Nate_Sturtz> ok I thouth that it was a lts
[01:06] <Nate_Sturtz> so I don't think I will update
[01:06] <Nate_Sturtz> thank you eveyone
[01:08] <gambl0re> this is what i get when trying to run apt update on live usb linux mint
[01:08] <gambl0re> https://pastebin.com/rj6aPBBM
[01:08] <Nate_Sturtz> may I ask why you are running apt update on a live usb
[01:08] <Nate_Sturtz> not on your system
[01:09] <gambl0re> cause i'm trying to install something
[01:09] <gambl0re> and it requires update
[01:09] <Nate_Sturtz> ok
[01:11] <Examknow> gambl0re: It seems you have a bad install disc
[01:11] <Bashing-om> !mint | gambl0re
[01:13] <gambl0re> !mint
[01:14] <Nate_Sturtz> 3
[01:14] <Examknow> Nate_Sturtz: Is there something we can help you with?
[01:14] <Nate_Sturtz> no wrong button
[01:15] <AppXprt> anyone familiar with Bacula?
[01:18] <Examknow> AppXprt: We can't help with that here. You might be better off asking in #bacula
[01:26] <Hash> I am trying to install https://flathub.org/apps/details/org.srb2.SRB2 but I get the following error in /.local/share/flatpak/repo): Permission denied
[01:26] <Hash> Any idea how to do this please
[01:27] <Examknow> Hash: How are you attempting to do this?
[01:27] <Hash> The command I use is: flatpak install org.srb2.SRB2.flatpakref
[01:27] <Examknow> Oh
[01:27] <Examknow> Hash: Try 'sudo flatpak install org.srb2.SRB2.flatpakref'
[01:27] <Hash> sudo shouldn't be writing to anything in my ~/.local
[01:28] <Examknow> What is the full error?
[01:28] <Hash> It should not require root perms
[01:28] <Examknow> We can't help you if you don't give us the full error
[01:28] <Hash> $  flatpak install org.srb2.SRB2.flatpakref
[01:28] <Hash> error: While opening repository /home/sert/.local/share/flatpak/repo: /home/sert/.local/share/flatpak/repo: opendir(/home/sert/.local/share/flatpak/repo): Permission denied
[01:29] <Examknow> Are you logged on as 'sert'?
[01:29] <Hash> Ohhh
[01:29] <Hash> Hang on
[01:29] <Examknow> If not, there's your problem right there
[01:30] <Hash> Everything inside ~/.local/share I own it, but flatpak is owned by root
[01:30] <Hash> This is very strange
[01:30] <Examknow> Yeah just run as sudo
[01:30] <Hash> Why?
[01:30] <Hash> It's my user directory
[01:30] <Examknow> Because why not?
[01:30] <Hash> why does root have anything to do with it?
[01:30] <Examknow> I don't know
[01:30] <Hash> What if I'm not in the sudo group
[01:30] <Hash> I can't have elevated perms.
[01:31] <Hash> I'm just a lowly user
[01:31] <Examknow> You or someone else must have used sudo when installing something
[01:31] <javi404> so on a new install of desktop version 20.4.1 gparted works fine
[01:32] <Examknow> The only reason it would be owned by root is if someone created it using sudo
[01:32] <javi404> on a patched 20.4.0->20.4.1, error.
[01:32] <javi404> uninstalled and reinstalled multiple times
[01:32] <Hash> So I changed the owner to me
[01:32] <Hash> then I did flatpak install again
[01:32] <Hash> then it asked me for a root password
[01:32] <Hash> I guess it's system that does i
[01:32] <Hash> inteeting
[01:32] <Hash> I dunno this flatpak thing at all
[01:33] <Hash> Installing now.
[01:33] <Examknow> Hash: Try asking in #flatpak
[01:33] <Hash> Ok thanks
[02:32] <YWH_1> I have windows 10 pro installed on my computer in legacy mode. Does anyone know if there is a way to install ubuntu 20.10 in BIOS legacy mode?
[02:33] <RingtailedFox> do you mean via WSL2?
[02:33] <RingtailedFox> or in a dual-boot configuration?
[02:33] <YWH_1> In a daul boot config
[02:33] <RingtailedFox> becuase.... BIOS booting is easy. just burn a CD or write to a USB stick, then  check bios to see what your boot order is... you might have to change which device boots first
[02:34] <YWH_1> OK when it does that I get a BRUBDOS0.4 screen and I can't load Ubuntu install
[02:34] <YWH_1> GRUBDOS
[02:36] <Rubato> hi
[02:38] <Examknow> Rubato: Hi. What can we help you with?
[02:40] <whoever> hi all usiing version 20.04 how am  I supose to set the default  sound card  with alsa if F6 dosin't work and asoundconf is missing
[02:40] <whoever> thank you in advance
[02:44] <Rubato> i have a problem with my brandnew laptop
[02:44] <Rubato> i cant boot on battery only
[02:44] <Rubato> then the splashscreen will show but after that i get a blackscreen
[02:44] <Rubato> and i dont get the login screen
[02:44] <Rubato> on power this happens also occasionly but not all the time
[02:44] <Rubato> with battery only this happens all the time
[02:46] <Examknow> Rubato: Have you tried reinstalling ubuntu?
[02:46] <Rubato> no and also it wont help
[02:46] <whoever> Rubato: look for laptop-utils
[02:47] <Rubato> i think its some kind of a splashscreen bug
[02:47] <Rubato> which is there for a long time
[02:47] <Rubato> but i forgot how to get rid of it
[02:47] <Bashing-om> Rubato: See: http://iam.tj/prototype/enhancements/Windows-acpi_osi.html - An ACPI issue ?
[02:47] <Rubato> no
[02:48] <Rubato> if this black screen occurs and i press the power button
[02:48] <Rubato> it will just do a tidy shutdown and the logo will appear while shutting down
[02:48] <Rubato> so it just ends up to a blackscreen but its still alive
[02:53] <Bashing-om> !nomodeset | Rubato
[03:08] <Rubato> im on ryzen 5
[03:08] <Rubato> ah the nomodeset yes that is related i think
[03:08] <Rubato> i think i did something with that in previous ubuntu versions aswell
[03:08] <Rubato> to fix this issue
[03:09] <Rubato> but how was that again
[03:28] <Rubato> ah its in grub
[03:48] <Rubato> it looks like this now
[03:48] <Rubato> GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
[03:49] <Rubato> so i do like this right
[03:49] <Rubato> GRUB_CMDLINE_LINUX_DEFAULT="quiet splash nomodeset"
[04:08] <DarkTrick> I guess new bug in ubuntu 20.10: When moving windows with ALT+mouse, the key-release event gets sent to the window. Can someone verify? Which package is this? "Xorg" does not exist.
[04:10] <saurabh009> Hi, does anyone know how to configure warmstart trap in snmp in Ubuntu ?
[04:11] <saurabh009> Or tell how can I get it enabled ? I can get the coldstart trap but not able to get the warmstart trap in the Ubuntu server
[05:12] <cortexman> https://news.ycombinator.com/item?id=24904582
[05:44] <strywgr> using lubuntu but whenever i plug in my headphones the pulseaudio speaker icons turns grey. It was fine till yesterday.
[06:04] <spandexVortex> hello?
[06:24] <freespirit-girl> anyone here familiar with virtual box?
[06:26] <DarkTrick> freespirit-girl, just ask your question
[06:26] <freespirit-girl> my display setting are messed up
[06:27] <tatertots> freespirit-girl: what is your HostOS/GuestOS?
[06:28] <freespirit-girl> windows to Ubuntu
[06:29] <tatertots> freespirit-girl: did you successfully install guest additions?
[06:30] <freespirit-girl> I think I did
[06:31] <freespirit-girl> how can i tell ?
[06:36] <freespirit-girl> shit! I should just try reinstalling the damn thing
[06:36] <freespirit-girl> thank you.
[08:03] <freespirit-girl> DarkTrick, I installed the guest tools using the command line while in Ubuntu and it did not work
[08:08] <ThinkT510> you mount the guest additions iso in the vm and install it in the guestos
[08:10] <freespirit-girl> so I install it like I installed the Ubuntu ISO?
[09:18] <chl_> where can I find crash logs?
[09:19] <mgedmin> in /var/crash/
[09:19] <chl_> what to do if that folder is empty :D
[09:20] <mgedmin> be happy you had no crashes
[09:21] <chl_> where would gui freeze/crashes be located? My problem is that sometimes the gui crashes and I can only move the mouse
[09:21] <chl_> cant restart gui with M-F2 :/
[09:22] <chl_> Tried adding another keybind, but does not work
[09:22] <chl_> and I have no idea what caused it :/
[09:29] <mgedmin> that sounds like a freeze rather than a crash
[09:29] <mgedmin> usually this is caused by running out of memory, especially if you have no swap file/partition
[09:30] <mgedmin> but it could also be a driver bug of some kind
[09:30] <mgedmin> at least if the mouse cursor is moving, it's not totally frozen
[09:30] <chl_> might be driver related. will try another nvidia driver
[09:31] <chl_> thank you for your inputs
[09:44] <clarkk> on 18.04, what is the correct way to add a permanent ip route?
[09:48] <lotuspsychje> clarkk: are you on ubuntu server?
[09:48] <clarkk> lotuspsychje, I didn't realise there was a server version of ubuntu
[09:48] <clarkk> I'm just on the standard version, afaij=k
[09:49] <EriC^^> clarkk: it's basically the same minus the gui/desktop environment
[09:49] <clarkk> ah, ok, then I have the GUI environment
[09:50] <clarkk> but I'm happy to use the shell to set it up.  I don't completely trust network manager, after having some problems getting openvpn working reliably with it
[09:52] <clarkk> Is netplan the correct way in 18.04?  I added a configuration, but it removes my openvpn ip routes when I apply it
[09:53] <mgedmin> what renderer have you told netplan to use?
[10:02] <clarkk> mgedmin, by renderer, do you mean network manager, or gnome?
[10:02] <clarkk> btw, this is my configuration  https://gist.github.com/superclarkk/39563e8dee1cb3f88c55e1f3e3cda58b
[10:03] <mgedmin> okay, so far so good, and do the openvpn routes come back if you bring the vpn up using network manager?
[10:03] <mgedmin> (you can do that on the command line with nmcli con up "name of your openvpn connection")
[10:06] <clarkk> mgedmin, I'm hesitant to use network manager, because after a while it totally fails and I need to reboot my system.  It may be because I have an older version of the network manager ovpn package.  I open the openvpn tunnel from the shell, using openvpn myconfig.ovpn, and it's rock solid
[10:08] <clarkk> a workaround I could use is to have a script that adds the ip route and then runs the openvpn command.  I was just wondering if there is a commonly accepted "best" way
[10:09] <mgedmin> have you reported the network manager + openvpn bug?
[10:10] <mgedmin> iirc there are scripts that network manager runs after it brings up a connection, you could run the openvpn command-line client from those
[10:13] <mgedmin> the fun part is there are like four ways to configure networking on ubuntu (n-m, networkd, ifupdown, netplan), and which one is best depends on ... stuff
[10:20] <clarkk> mgedmin, that sounds _really_ fun ;)  Actually, thinking about it more, I think putting it in my script will be the easiest way.  Then all my VPS / VPN config is in one place
[10:21] <mgedmin> are these normal routes, or vpn routes?  because openvpn can set them up for you too
[10:21] <mgedmin> I mean, if the routes go over the vpn, I wouldn't add them to the system-wide network configuration
[10:21] <mgedmin> I would add them to the vpn configuration
[10:27] <clarkk> mgedmin, I just added them to the ovpn config, and that worked a treat.  Thank you :)
[10:42] <BluesKaj> Hi all
[10:48] <ufk> isn't 20.04.1 the latest ubuntu ? if i changed update from LTS to normal why do-release-upgrade is updating packages? it won't downgrade my system right ?
[10:49] <xtao> 20.10 is the latest. 20.04.1 is the latest LTS version
[10:49] <ufk> ahhhhh
[10:49] <ufk> right thanks :)
[10:53] <xBfrog> i have my camers set to ptp mode i can open the camera flies but when trying to delete any it says i don't have sufficient permissions. How do i change this?
[11:05] <javi404> how can i find why gparted works on a new 20.04.1 and not an upgraded 20.04.1 from 20.04
[11:10] <mgedmin> can you define 'not works'?
[11:11]  * mgedmin looks into his crystal ball and intones: wayland
[11:11] <javi404> mgedmin: issue with wayland
[11:11] <javi404> i was able to do some magic i found googling with xhosts
[11:12] <javi404> so then to reproduce, i did a fresh install in a virtualbox vm
[11:12] <javi404> works no problem.
[11:12] <javi404> but my physical workstation, which was a new build of 20.04 originally, still doesn't work.
[11:12] <javi404> it asks for password, and then nothing.
[11:12] <mgedmin> ubuntu doesn't use wayland by default
[11:12] <javi404> also, the icon doesn't sue.
[11:13] <javi404> mgedmin: i don't remember selecting it.
[11:13] <javi404> xrander would beg to differ?
[11:13] <mgedmin> you can check if your current session is wayland or xorg by doing env|grep WAYLAND
[11:14] <mgedmin> I remember that gparted used to plain not work in wayland sessions because Xwayland rejected connections from programs running as root
[11:14] <javi404> mgedmin: so turns out, on the virtualbox, it is using xorg
[11:14] <javi404> so that is why it works
[11:14] <mgedmin> I don't remember if xhost was a possible workaround
[11:14] <javi404> but i don't remember selecting wayland on my workstation
[11:14] <mgedmin> why not use gnome-disks instead of gparted?
[11:14] <javi404> how can i switch to x
[11:15] <mgedmin> you switch by choosing the session type during login, and it is then remembered
[11:15] <javi404> i don't need gparted anymore but would like it to work as to not need to reboot or learn gnome-disks
[11:15] <javi404> mgedmin: got it
[11:15] <javi404> ill take a look when im back home
[11:15] <mgedmin> enter your username, press enter, then in the bottom right corner there's a little cog icon that gives you a menu of session types
[11:35] <slimep> under what circumstances would a 20.04 installation be using NetworkManager to manage DNS instead of systemd-resovled? Is it always just the user changing it or could it be a remnant config from upgrading from an older release?
[11:36] <donofrio> what does it mean when nmon says "dm-1       0%    0.0    0.0|>disk busy not available " and how would I review this from dmesg or?
[11:37] <mgedmin> donofrio: is /etc/resolv.conf a symlink?  to where?
[11:37] <mgedmin> on a stock system it's a symlink to ../run/systemd/resolve/stub-resolv.conf
[11:38] <mgedmin> and afaiu network-manager tells systemd-resolved what DNS server to forward to
[11:38] <donofrio> but this is for disks not network?
[11:39] <slimep> mgedmin: yes, but I'm seeing a lot of reports from users where NM is actually manaing /etc/resolv.conf. And I'm wondering how did they end up there.
[11:39] <mgedmin> donofrio: sorry I meant that for slimep
[11:39] <donofrio> k
[11:44] <mgedmin> donofrio: your question is very interesting, and I don't know the answer
[11:44] <mgedmin> and I can't find it on google
[11:45] <mgedmin> and I can see how _some_ of the disks on my server show the same "disk busy not available", but I can't see why
[11:45] <mgedmin> loop7 (?!), all of my md-* devices, none of my dm-* devices, and sdc3
[11:47] <donofrio> mgedmin, has me intrested to, server doing just fine as far as I can tell but I still wonder...
[11:49] <mgedmin> donofrio: afaict nmon prints that when it sees 0 in the reads and writes columns in /proc/diskstats for that device
[11:49] <mgedmin> so it means "disk's busyness numbers are not available" rather than some error about the disk being busy and not available
[11:50] <mgedmin> it could just mean that nothing is currently reading/writing to that device
[11:50] <mgedmin> and yeah, the devices that show "busy not available" are all old partitions, not currently mounted
[11:50] <donofrio> mgedmin, ok I'll contiune on with my workflows....tnx
[11:51] <mgedmin> no, not all of them, md1 is mounted
[11:51] <mgedmin> maybe the kernel doesn't track these things for mdraid devices
[11:52] <mgedmin> interesting tool nmon, didn't know about it
[11:52] <donofrio> mgedmin, yah it's an AIX tool that finally made it to desktop
[12:31] <jakobbg> Hi. I have a Ubuntu in Azure running standard "Ubuntu Server 20.04 LTS" image. I'd like to upgrade to 20.10. Do I have to consider any other issues apart from the regular do-release-upgrade? Any problems which might arise to having begun using the Canonical image? Do I have to change the image instead of a manual upgrade? I assume not, but thought it might be wise to ask first :)
[12:35] <mhcerri> hi jakobbg. I don't think there's anything else to worry about. patviafore, any thought? ^
[12:37] <slimep> mgedmin: seems like upgrading from 18.04 leaves the installation in a state where /etc/resolv.conf is managed by resolvconf
[12:37] <mgedmin> well, consider that you're jumping off an LTS and will have to upgrade to 21.04, then 21.10, then to 22.04 within 3 months of each of those releases
[12:38] <mgedmin> slimep: was resolvconf once installed by default?
[12:40] <mgedmin> I had it on one machine, but removed it back in 2017, while running 17.04, due to some unfortunate interaction with systemd-resolved
[12:40] <slimep> mgedmin: don't think so.
[12:47] <patviafore> jakobbg: unfortunately, it's not a frequently tested workflow, the safer thing is to use a new image. That being said, it will be very dependent on which packages are used and your workflow. It is very likely that it would be fine but I cannot say that with 100% certainty. Is this for a production machine, a personal machine, etc?
[13:07] <jakobbg> patviafore: mainly a home computer, but having to set up a new machine means a lot of configuration on the zfs datasets +++
[13:07] <javi404> mgedmin: thanks, ill check it out.
[13:29] <patviafore> jakobg: understood. I'm looking through any changes that might affect an upgrade. Do you know what serial you are running? (cat /etc/cloud/build.info)
[13:39] <patviafore> jakobbg: ^^
[13:48] <jakobbg> patviafore: build_name: server serial: 20200610
[13:50] <patviafore> jakobbg: I don't see anything that stands out as an issue on an upgrade. Upgrading an Azure instance in your case should be similar risk to upgrading a home computer.
[13:53] <jakobbg> awesome. sounds like i'll just create a snapshot and see what happens :p
[13:53] <jakobbg> thank you patviafore
[13:53] <nate_KF0BXU> 2
[14:05] <zamba> i'm trying to deploy the ova from ubuntu.. but what is the default username/password for this ova?
[14:05] <zamba> i attempted to set the password through the gui on the esxi host, but it doesn't seem like they are being honored
[14:07] <EriC^^> zamba: not sure what ova is but usually in the live usb the login is "ubuntu" with an empty password
[14:43] <nbusrone> Anyone using nemo as Desktop ? how do I adjust icon size fix whenever I create an icon shortcut or new document ?
[14:45] <Maik> nbusrone: nemo isn't a desktop, it's a file manager
[14:46] <nbusrone> Maik : I replace Nautilus Desktop to Nemo.So nemo icon appear
[14:49] <nbusrone> but the size is standard , how do I edit it to a smaller size ?
[14:50] <nbusrone> where does nemo icon size config location ?
[14:50] <lotuspsychje> nbusrone: try dconf-editor perhaps
[14:52] <javi404> mgedmin: just an FYI, on 20.10 fresh install, Wayland is default.
[14:52] <javi404> i see i can change it at login where you suggested, bottom right corner, but just an FYI
[14:54] <nuala> thinkcentre 720q: 20.04 won't boot up until i add nomodeset. I am not sure this is expected behaviour or some uefi-settings I dont understand is wonky again (Intel SuperSecure ChipsetProtocol Guard or something?) Anyone has experience to share? :S
[14:55] <javi404> mgedmin: I take my previous statement back
[14:55] <javi404> haha
[14:56] <javi404> there is some logic that made my physical workstation default to wayland
[14:56] <javi404> and virtual xorg
[14:56] <javi404> oh well easy fix
[15:06] <nbusrone> lotuspsychje : there is no setting for nemo icon size at /org/nemo/desktop.Possible to create a customize config on dconfig ?
[15:24] <madeti> Can anyone help me to change the default log directory of snaps on ubuntu?
[15:24] <madeti> by default they log in /var/log and lets say I don't have enough space there and I want to log in /home/madeti/logs
[15:24] <madeti> how do I do that?
[15:25] <mgedmin> madeti: symlink or bind mount
[15:26] <madeti> @mgedmin i used a symlink (`ln -s`) to point /var/log to /home/madeti/logs
[15:26] <madeti> syslog and all other services are logging fine
[15:26] <madeti> but snap throws a permission error
[15:27] <madeti> apparmor="DENIED" operation="mount" info="failed flags match" error=-13 profile="/snap/core/10185/usr/lib/snapd/snap-confine" name="/tmp/snap.rootfs_pRGXOp/var/log/" pid=10892 comm="snap-confine" srcname="/home/provakil/logs/" flags="rw, rbind"
[15:27] <Examknow> madeti: Did you run it as root?
[15:27] <mgedmin> hm, snaps play games with mount namespaces... and also the sandboxing might not allow access to /home/
[15:27] <madeti> what do I need to run as root? @Examknow
[15:28] <mgedmin> where do snaps log anyway?
[15:28] <madeti> @mgedmin yeah. I think sandboxing may have something to do with it. But where do I give it permission to log in /home/
[15:28] <nbusrone> may I ask , how do I install ppa update ? example  3.28.0-1ubuntu1.5	updates (universe)	2020-10-06 ?
[15:28] <madeti> I think they log in /var/log only
[15:28] <Examknow> sudo would fix that issue
[15:29] <madeti> @Examknow what command do I run as sudo?
[15:29] <mgedmin> madeti: the best place to find experts in snaps is probably https://forum.snapcraft.io/
[15:30] <Examknow> madeti: The command that returned the error
[15:31] <madeti> The error I pasted was from syslog. I am running `sudo service snap.rocketchat-server.rocketchat-server restart`, which generated the log
[15:31] <madeti> @Examknow ^^
[15:32] <Examknow> Strange...
[15:57] <rocksteady> hi all i would like to learn about how to make key combinations do certain things i am not sure if i should be looking up keybinds or emacs or something else (example if i want to make pressing composite key open a different menu
[15:58] <whoever> hi all
[16:00] <whoever> under accessability I have mouse location on. so thaT when you hit ctrl it should pulse an orb around the pointer but nothing is happening  and when I try ~$ gsettings set org.gnome.settings-daemon.peripherals.mouse locate-pointer true
[16:00] <whoever> I get no such 'locate-pointer ' can someone assist
[16:02] <mgedmin> hmm this seems to've regressed in 20.10, I remember it working in earlier releases
[16:02] <mgedmin> whoever: are you using wayland the default xorg session?
[16:03] <mgedmin> lol for lt_LT that setting is mistranslated as "locate printer" :D
[16:04] <whoever> megoix: most likly
[16:05] <hofer> works for me under locate pointer
[16:06] <whoever> hofer: i am on 20.01
[16:06] <whoever> *20.04
[16:06] <hofer> me too
[16:06] <rocksteady> whois mgedmin
[16:06]  * mgedmin 
[16:06] <rocksteady> exit
[16:07] <hofer> universal access
[16:08] <mgedmin> whoever: I think the setting moved to org.gnome.desktop.interface locate-pointer
[16:09] <mgedmin> and these days it's gnome-shell itself that's supposed to be doing it
[16:11] <mgedmin> https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/1875056
[16:15] <whoever> manopola: thank you that worked , but it there a place to change the colour intensity and the button( it only worcs with left ctrl but  i want to use it with right)
[16:15] <mgedmin> heyyyyyyy _left_ ctrl?  since when does it use left ctrl?
[16:16] <manopola> whoever, ?
[16:23] <Bill73> ssh problem on my ubuntu 20.04  whenever I ssh to a remote machine I get the security warning with a fingerprint and yes/no/[fingerprint] before I get the password prompt. now it was working but subsequent reinstalls and I had a bad key in my known_hosts, prompter to remove it so I did   Now I get the failed to add hosts for any remote system I attempt to ssh to from this setup. suggested fix?
[16:26] <pavlos> Bill73: are the permissions for .ssh/known_hosts 600 ?
[16:26] <pavlos> Bill73: and 700 for .ssh/
[16:26] <Bill73> pavlos: not sure
[16:27] <Bill73> I'm a novice so it'll take me a minute to find out how to check
[16:28] <pavlos> Bill73: ls -al | grep .ssh should show drwx------
[16:29] <pavlos> Bill73: ls -al .ssh | grep known should show -rw-------
[16:32] <Bill73> this is the dir     drwx------ 2 bill bill 4096 Oct 26 16:44 .ssh/
[16:32] <Bill73> this is known_hosts      -rw------- 1 root root 1422 Oct 26 16:42 known_hosts
[16:33] <Bill73> so they are good on permissions...
[16:34] <pavlos> Bill73: ahhh ... see the permissions on the known_hosts, not good. It should be owned by bill, that's why you cannot write in there
[16:34] <mgedmin> Bill73: your known_hosts file is owned by root
[16:34] <dionysus69> is it worth trying to fully encrypt hd with existing system on it ?
[16:34] <dionysus69> and is there a utility that makes it simpler ?
[16:34] <mgedmin> the permissions per se are right, ownership is not -- do a sudo chmod bill:bill ~/.ssh/known_hosts and try not to use 'sudo ssh ...'
[16:34] <Bill73> pavlos: ok
[16:35] <mgedmin> dionysus69: this is a question that interests me a lot too!
[16:35] <pavlos> Bill73: to fix, sudo chown bill:bill .ssh/known_hosts
[16:35] <dionysus69> yea I remember in past ubuntu came with an option to fully encrypt disk at install time, but now its missing
[16:36] <Bill73> pavlos: chmod: invalid mode: ‘bill:bill’   Also I dont remember doing a sudo ssh unless it was for the removal of the bad key sith ssh-keygen  arg
[16:36] <mgedmin> dionysus69: I think it's possible if you use ext4's native encryption feature (fscrypt)
[16:36] <Bill73> should that have been chown?
[16:37] <dionysus69> mgedmin: will look it up thanks
[16:37] <mgedmin> Bill73: sorry I typed chmod instead of chown aaaaugh why do I always make this mistake
[16:37] <Bill73> pavlos: lol i see it
[16:37] <mgedmin> dionysus69: I'm a bit scared to try tbh
[16:37] <dionysus69> yea it's kind of scary you are right
[16:38] <Bill73> pavlos: thx  all is working again.
[16:38] <dionysus69> mgedmin: especially these kind of warnings on stackoverflow :D :D The API to fscrypt might change in the future, so make sure to back up your important files if you attempt to upgrade your system. (This feature is not widely used on Desktop. Use at your own risk.)
[16:39] <Bill73> pavlos: scrolled back through term  there it is... sudo ssh-keygen -R 192.168.1.95 that messed it up lol
[16:47] <Perdurabo> Hello, IRC test - do you see this ?
[16:47] <ThinkT510> indeed we do
[16:47] <Perdurabo> haha ! Superb !
[16:47] <ThinkT510> you can also /join #test for that sort of thing
[16:48] <Perdurabo> Thank you
[16:48] <Maik> Perdurabo: if you have a support question just ask, all other non related talk takes place in #ubuntu-offtopic or #ubuntu-discuss :)
[16:49] <Perdurabo> mhm
[16:49] <Perdurabo> i am new to irc
[16:49] <Perdurabo> great thx
[16:49] <Maik> no problem, just pointing you the way
[16:51] <mgedmin> I would like nautilus to start in a certain directory instead of my home when I click the icon in the dock; how can I achieve that?
[16:54] <Maik> Perdurabo: no unwanted DM's please. There isn't any conversation going on here so you won't see much. It's a support channel which is at times not that busy.
[16:55] <Maik> with support channel i mean Ubuntu support channel
[16:57] <madeti> @mgedmin bind mount worked like a charm
[16:57] <madeti> thanks for your help
[16:57] <mgedmin> oh cool, and here I was beginning to think snaps would have problems with bind mounts too
[17:13] <wasutton3> what repo can I add so I can download/install libthrift-0.11.0 like its shown here? https://launchpad.net/ubuntu/focal/+source/thrift
[17:13] <vasilakisfil> hi I am trying to run sudo loginctl enable-linger $USER on ubuntu 19.10 but I get Could not enable linger: No such file or directory, how can I solve it ?
[17:14] <Maik> vasilakisfil: 19.10 is EOL and thus not supported anymore
[17:15] <vasilakisfil> yeah I get it but I was wondering if you know a clever fix that could solve this
[17:16] <Maik> we don't support 19.10 anymore, either install 20.04 LTS or the just released 20.10
[17:18] <wasutton3> https://github.com/gnuradio/gnuradio/issues/3673#issuecomment-716364065
[17:18] <vasilakisfil> ok I am upgrading now to 20.04, in theory problem should be solved right ?
[17:21] <lotuspsychje> wasutton3: just install gnuradio from the ubuntu repos?
[17:22] <wasutton3> lotuspsychje, that is built on libthrift 0.13.0 which breaks my flowgraph
[17:23] <lotuspsychje> wasutton3: break on gnuradio from the repos or from the git?
[17:23] <wasutton3> lotuspsychje, https://github.com/gnuradio/gnuradio/issues/3673#issuecomment-716364065
[17:23] <lotuspsychje> !info gnuradio bionic
[17:24] <wasutton3> lotuspsychje, this is on focal, not bionic
[17:24] <lotuspsychje> !info gnuradio focal
[17:25] <wasutton3> their ppa is 3.8.2
[17:26] <lotuspsychje> wasutton3: if its about the git version, file a bug there please
[17:26] <lotuspsychje> wasutton3: we can only support ubuntu and its packages from the official repos
[17:26] <wasutton3> lotuspsychje, its not. its an issue with using libthrift 0.13.0
[17:26] <wasutton3> if 0.11.0 could be built for focal, the problem would go away.
[17:28] <lotuspsychje> wasutton3: did you test this on the focal version 3.8.1?
[17:28] <wasutton3> yes.
[17:29] <wasutton3> the advice given was to use the PPA which supposedly had 0.11.0 built in, but they had that for a previous version (bionic iirc)
[17:29] <lotuspsychje> wasutton3: can you see if its bug #1742448 please
[17:30] <wasutton3> lotuspsychje, no, that bug is for bionic and doesn't match the error.
[17:30] <lotuspsychje> wasutton3: but it also breaks your flowgraph right?
[17:31] <wasutton3> let me revert all my changes. FWIW i'm on 20.04. so all errors will be related to that LTS version
[17:31] <wasutton3> 18.04 had different incompatibility issues with ZMQ
[17:31] <lotuspsychje> wasutton3: if you can reproduce this on 20.04 with all packages from the repos, we can file a !bug
[17:32] <wasutton3> in the process of reverting,
[17:32] <lotuspsychje> cool
[17:32] <Descriptioned> Good evening, i recently reinstalled ubuntu, i was have qbittorent in dark theme, i dont even remember who i did it, but now i cant change theme, i downloaded theme and i choose custom ui theme but nothing..
[17:36] <wasutton3> lotuspsychje, ok, so i've removed the gnuradio ppa, reinstalled all the local packages, and created a super basic flowgraph.
[17:36] <wasutton3> constant source -> throttle -> null sink
[17:36] <wasutton3> with a floating CtrlPort Performance Monitor
[17:36] <wasutton3> with it disabled, the flowgraph executes as expected
[17:37] <wasutton3> with it enabled, I get `python3: ../gnuradio-runtime/lib/controlport/rpcmanager.cc:38: static rpcserver_booter_base* rpcmanager::get(): Assertion `booter_registered || aggregator_registered' failed.`
[17:39] <lotuspsychje> wasutton3: i installed it here, tell me what to do to make it break?
[17:39] <wasutton3> open gnuradio-companion
[17:39] <lotuspsychje> yes
[17:40] <wasutton3> create this flowgraph https://i.imgur.com/sLerFHL.png
[17:41] <lotuspsychje> wasutton3: how do i do that
[17:41] <wasutton3> are you staring at a blank screen?
[17:41] <lotuspsychje> yes
[17:41] <wasutton3> there should be a create button in the top left.
[17:42] <wasutton3> alt text reads "Create new flow graph"
[17:42] <lotuspsychje> yes
[17:43] <arch-nemesis> Is there a listing somewhere of boot options on ubuntu 20.04 server live CD. I've got it setup to pxe boot and auto-provision servers with cloud-config and this works great. I want another pxe boot option that will load the ISO and just provide a troubleshooting environment. So far the best I can do is boot the installer and hit ctrl-z to get out of subiquiti.
[17:43] <wasutton3> on the far right you should see Core, Audio, Boolean Operaters, etc
[17:44] <arch-nemesis> Hoping there is a boot option for "justalivecd" or something. Otherwise I'll just have to use a different ISO
[17:44] <lotuspsychje> wasutton3: i see, then what
[17:44] <wasutton3> from Core/Control Port, grab CtrlPort Performance Monitor and drag it to the flow graph
[17:45] <lotuspsychje> wasutton3: done
[17:45] <wasutton3> repeat for Core/Misc/Null Sink
[17:45] <wasutton3> Core/Misc/Throttle
[17:45] <wasutton3> and Core/Waveform Generators/Constant Source
[17:46] <lotuspsychje> wasutton3: done, how to connect them?
[17:49] <wasutton3> click the blue box on the constant source, and drag it to the left blue box on the throttle
[17:50] <wasutton3> repeat for throttle -> null sink
[17:51] <lotuspsychje> wasutton3: done
[17:52] <lotuspsychje> arch-nemesis: you might wanna check the #ubuntu-server channel
[17:52] <wasutton3> go to Run->Execute
[17:53] <lotuspsychje> wasutton3: its greyed out
[17:54] <wasutton3> fill in the title on the options box. Double click it to enter a value in the fields.
[17:56] <lotuspsychje> wasutton3: title:test still greyed out run
[17:57] <mgedmin> arch-nemesis: there's a shell waiting for you on tty2, iirc
[17:57] <wasutton3> lotuspsychje, can you send me a screenshot?
[17:58] <lotuspsychje> wasutton3: https://imgur.com/a/Knau7nR
[17:58] <wasutton3> lotuspsychje, try changing generate options to No gui
[17:59] <whoever> mgedmin: ok, am here , apparently today it does
[17:59] <lotuspsychje> wasutton3: no dice
[18:00] <whoever> Maik: just got out of class sorry for the silence
[18:00] <wasutton3> lotuspsychje, you might need more packages... try installing these
[18:00] <wasutton3> `sudo apt install git cmake g++ libboost-all-dev libgmp-dev swig python3-numpy python3-mako python3-sphinx python3-lxml doxygen libfftw3-dev libsdl1.2-dev libgsl-dev libqwt-qt5-dev libqt5opengl5-dev python3-pyqt5 liblog4cpp5-dev libzmq3-dev python3-yaml python3-click python3-click-plugins python3-zmq python3-scipy thrift-compiler python3-twisted libtool openssl libssl-dev libthrift-dev flex libevent-dev bison
[18:00] <wasutton3> liborc-0.4-dev libgnuradio-qtgui libgnuradio-channels`
[18:01] <lotuspsychje> wasutton3: it installed all the depends (912mb) on install
[18:01] <wasutton3> not in my experience
[18:01] <wasutton3> i bet there are a few on that list its missing.
[18:02] <wasutton3> probably libgnuradio-qtgui
[18:02] <wasutton3> if i had to guess
[18:05] <lotuspsychje> wasutton3: thats installed, does ID need to be filled in or something?
[18:06] <wasutton3> lotuspsychje, yes it does.
[18:08] <lotuspsychje> wasutton3: ok it wants to execute but wants xterm executable
[18:10] <wasutton3> lotuspsychje, maybe switch back to qt-gui rather than no-gui
[18:11] <lotuspsychje> wasutton3: ok works now; >> Done (return code -6)
[18:11] <wasutton3> look at the python log at the bottom of the screen. you should still see the python3 error
[18:12] <lotuspsychje> wasutton3: https://dpaste.com/H8LYF4WYX
[18:13] <wasutton3> lotuspsychje, huzzah! the bug is reproducable!
[18:13] <lotuspsychje> wasutton3: ok please; ubuntu-bug gnuradio and ill affect it for you
[18:23] <wasutton3> lotuspsychje, https://bugs.launchpad.net/ubuntu/+source/gnuradio/+bug/1901746
[18:24] <lotuspsychje> wasutton3: thank you for filing the bug, can you apport-collect 1901746 please?
[18:26] <wasutton3> lotuspsychje, sent
[18:26] <lotuspsychje> wasutton3: affected & confirmed for you, tnx
[18:27] <wasutton3> lotuspsychje, awesome. thank you.
[18:29] <lotuspsychje> wasutton3: and added a comment
[18:36] <jayjo_> I'm looking into packaging a python cli program into a ubuntu/debian package to understand how the system works a bit more. Does creating a debian package make it available for ubuntu automatically? Is is that the tools are the same but the repositories are different?
[18:37] <|subz3r0|> jayjo_: can you be a little more precise what you want to achieve? maybe for the others its clear, but i dont get your point
[18:38] <|subz3r0|> since ubuntu is debian based, yes you should be able to install a *.deb file via dpkg i the depenecies fit. correct naming scheme etc...
[18:39] <sarnold> jayjo_: if your package is accepted into debian, it'll eventually make its way into ubuntu, yes
[18:40] <Guess59061> Sorry if this is a stupid question, but I added a new disk to my VM (a fixed-size disk in VirtualBox), and I used the Disks utility program to format it, wherein I chose ext4 as the disk's filesystem. Now, I edited /etc/fstab, etc. /dev/sdc is the name of this disk. Everything works. What I want to know is why there's nothing in the first 512B (the
[18:40] <Guess59061> first sector) of this disk (it's empty, all nulls), because I want to see the partition table. My primary disk, /dev/sda, has a standard MBR in its first 512B. Is this not needed when there's a single "partition" of the disk? How does the OS know where to find the superblock, etc.?
[18:41] <lordcirth> Guess59061, sounds like it's formatted GPT
[18:41] <sarnold> Guess59061: it may have been formatted with gpt, instead
[18:41] <|subz3r0|> jayjo_: also its not "the same". because every distro follows their own versioning. example: RHEL 6 uses ssh server v5.4. but that does not mean they are not implementing security fixes from more recent versions. so after all versions dont tell you more in comparison to other distro
[18:42] <Guess59061> sarnold lordcirth But aren't they backwards compatible and contain MBR anyway?
[18:42] <Guess59061> I mean *it, the GPT.
[18:42] <lordcirth> Guess59061, GPT leaves the 512b MBR space unused in case GPT-unware software tries to overwrite the MBR.
[18:42] <sarnold> Guess59061: I think I always had the impression an old-style partition table was often included as a convenience, but I don't think that's strictly necessary
[18:43] <lordcirth> Or, sometimes it writes a "protective MBR" so that software won't do that
[18:44] <Guess59061> Alright, thank you very much.
[18:51] <ash_worksi> whats the deal with "wayland"? Do people use that on top of ubuntu? Are there any benefits to it over X?
[18:51] <mgedmin> Guess59061: you can find out by doing sudo fdisk -l /dev/sdc, look for "Disklabel type: gpt"
[18:51]  * ash_worksi just heard of wayland
[18:52] <lordcirth> ash_worksi, it is an option on ubuntu, and is planned to eventually replace X
[18:53] <mgedmin> some things work better, some things don't work (yet)
[18:53] <Maik> ash_worksi: Wayland will replace X over time. Fedora uses it as default since a few releases, Ubuntu might give it another shot using it as default in one of the future releases.
[18:53] <mgedmin> Ubuntu gave it a try in one of the lts+1 releases, then switched back to xorg by default in the next lts because wayland wasn't ready
[18:53] <mgedmin> e.g. currently you cannot use chromium on wayland in 20.10 (but a SRU is in progress to fix that)
[18:54] <mgedmin> and currently you can't share your desktop over VNC/RDP if you use wayland (but hopefully that'll be fixed in 21.04)
[18:54] <mgedmin> and I think nvidia has special compatibility problems like nvidia usually does
[18:55] <Maik> ash_worksi: to some it still feels like a Alpha/Beta after 11 years of development
[18:55] <mgedmin> things that work better: HiDPI, fractional scaling, smoother window resizing, you can take screenshots (or lock the screen) while a popup menu is open
[18:56] <sarnold> does copy and paste work?
[18:56] <mgedmin> mostly
[18:56] <sarnold> I heard that was up to each 'window manager' kind of thing, and wondered if clients would be able to use the same protocol if they were using sway or using  .. weston?
[18:58] <mgedmin> sway and weston are wayland compositors (like gnome-shell), not clients
[18:58] <mgedmin> and there's a standard protocol for the clipboard on wayland
[18:58] <mgedmin> I'm less sure about middle-click paste
[18:59] <mgedmin> it works in gnome, but I don't remember if it uses the same protocol, or some gnome-specific extension
[18:59] <sarnold> well yeah but I heard one wayland compositor thing supported copy and paste and most didn't .. and didn't know which ones were which
[18:59] <sarnold> how about shift+insert?
[19:00] <Guess59061> mgedmin Hm, I don't see any information there. parted -l tells me it's of type "loop", which means raw and unpartitioned.
[19:00] <mgedmin> Guess59061: I'm starting to think you formatted your entire disk as ext4 instead of creating a partition table in gnome-disks
[19:01] <Guess59061> Yep, that's what I did.
[19:01] <mgedmin> sudo file -s /dev/hdc can confirm (by saying "Linux rev 1.0 ext4 filesystem data")
[19:02] <Guess59061> Yeah.
[19:03] <lordcirth> TIL about file -s
[19:06] <ash_worksi> wow
[19:06] <ash_worksi> wow @ Maik , lordcirth
[19:07] <ash_worksi> I don't really grok what wayland is, tbh; like, what differentiates wayland from dwm?
[19:07] <mgedmin> I have a feeling that mkfs'ing an entire /dev/sdc is a bad idea, but I canno't defend that feeling this late at night
[19:07] <mgedmin> ash_worksi: wayland is a protocol, dwm is a window manager
[19:07] <Maik> ash_worksi: https://wayland.freedesktop.org/
[19:08] <Guess59061> mgedmin Where is information about /dev/sdc's file system stored?
[19:08] <mgedmin> there are recorded conference talks on youtube explaining wayland nicely
[19:08] <mgedmin> Guess59061: on /dev/sdc itself
[19:08] <ash_worksi> oh? is sway a protocol?
[19:09] <ash_worksi> mgedmin: I watched a ytb vid and it was just a guy saying, "yeah.... I'm impressed" a lot
[19:09] <Maik>  since it's not support related it might be better to take the wayland conversation over to #ubuntu-discuss
[19:09] <ash_worksi> not saying they're aren't better videos, but yeah
[19:09] <mgedmin> ash_worksi: sway is a compositor, I think
[19:10] <ash_worksi> so is sway akin to dwm ?
[19:10] <mgedmin> and yeah, good call on the topic-ness of this conversation
[19:10] <ash_worksi> Maik: it's (at least for me) a more abstract conversation, but I'll move if you want
[19:11] <Maik> yep, better to do so i guess. This channel is after all for support with Ubuntu :)
[19:28] <Junior> hello
[19:29] <Junior> on a ubuntu 18.04 i have performed upgrade on packages and did not recognised netplan.io so i have removed it. of course that i have no eth0 and no connection to the internet but i am able to conenct to the console, how can i fix this as i don't have ifup/ifdown. Thanks
[19:30] <sarnold> Junior: the .deb for it may still be in /var/cache/apt/archives -- you can install it with dpkg -i
[19:31] <Junior> @sarnothing in var/cache/apt
[19:31] <Junior> cd ..
[19:34] <sarnold> Junior: dang, if you had the old pckage around that would have saved a lot of trouble.
[19:34] <mgedmin> Junior: what's your network configuration?  standard dhcp?
[19:34] <sarnold> Junior: okay, you can use the 'ip a' and 'ip r' commands to add an IP address and check routing tables, so you can get off the machine
[19:36] <mgedmin> in times long gone you could run `dhclient eth0` and you'd be up and talking to the network
[19:36] <Junior> no dhcp :)
[19:37] <mgedmin> (and then the dhclient you ran would start fighting with the dhclient spawned by the OS and fun times would happen)
[19:37] <Junior> cannot find device eth0
[19:37] <mgedmin> ip link will show you what it's called
[19:38] <Junior> ens3 ...
[19:39] <mgedmin> (is iproute in 18.04 recent enough to support all the colorful goodness of ip -c -br link ?)
[19:40] <mgedmin> bringing the network up is something like `ip link set ens3 up; ip addr add X.X.X.X/NN dev ens3; ip r add default via Y.Y.Y.Y; echo nameserver 8.8.8.8 >> /etc/resolv.conf`
[19:40] <mgedmin> I assume you know the right address/netmask/gateway?  if not your netplan config should still exist in /etc/netplan
[19:40] <mgedmin> assuming you apt removed the package, not purged it
[19:41] <Junior> yes, the config was there, one second mgedmin
[19:42] <Junior> thank worked, installed netplan
[19:42] <Junior> reboot should be enough or should i do something else before that?
[19:44] <sarnold> you may need to re-write your netplan configs, too, depending if you purged the package or just removed the package
[19:44] <Junior> just removed the package
[19:45] <Junior> before removing netplan, i have visited https://netplan.io/ and it looked like a new relic if you know about it
[19:45] <Junior> and i knew that i have installed new relic some time ago
[19:45] <sarnold> Junior: hahaha those logos are pretty similar :)
[19:46] <Junior> https://newrelic.com/
[19:46] <Junior> i have used a paid plan on newrelic for aprox 6 months so when i saw the logo i knew what i had to do
[19:46] <Junior> as new relic installs a daemon on server if i recall well
[19:47] <mgedmin> haha
[19:47] <Junior> am i the first one that does this ?
[19:47] <Junior> :))
[19:47] <mgedmin> you should probably do a netplan try?
[19:48] <mgedmin> I don't know; I haven't actually used netplan ever
[19:48] <Junior> me neither, used to /etc/network/interfaces :)
[19:48] <mgedmin> (well, ubuntu installed it for me, but I haven't touched the config)
[19:49] <mgedmin> I haven't ever brought a network up using ip a also; the last bad time I had with networking was when we still had ifconfig
[19:50] <Junior> mgedmin this is the first time for me in 20 years also to bring up using ip a
[19:50] <Junior> the ip link set ens3 up was the "trick"
[19:50]  * mgedmin cheated and looked it up in the man page
[19:50] <Junior> that's not cheating:)
[19:50] <mgedmin> (but I'd seen it before and recognized what it was when I saw it)
[19:50] <Junior> so thank you all for your time
[19:51] <Junior> new relic... fmm
[19:58] <settermjd> Hey all. For some reason, the first time I cold boot ubuntu, the grub boot menu doesn’t show on the screen and the laptop doesn’t boot. However, if I leave it for about 60 - 90 seconds, hold the power button for four seconds to turn it off, and then start it up again, the boot menu appears and it boots normally. Any suggestions as to how to debug this?
[20:17] <hansh_> why is ubuntu 20.04 fetching updates over http by default?
[20:17] <jpds> hansh_: It always has
[20:18] <hansh_> maybe that made sense in 2004, but why haven't it been changed to httpS?
[20:18] <jpds> hansh_: Because everything is PGP signed
[20:18] <hansh_> (for one, httpS would protect your privacy, your ISP will no longer be able to spy on what packages you've installed)
[20:19] <hansh_> s/ISP/MITM attacker
[20:19] <jpds> hansh_: You could always use Debian, they offer APT updates over Tor
[20:21] <jpds> hansh_: And you could always find a mirror that offers HTTPS access
[20:24] <sarnold> hansh_: because there's a lot of country-level mirrors that provide a fair chunk of bandwidth, and those are run by volunteers who perhaps shouldn't have certs for ubuntu
[20:24] <sarnold> hansh_: and terminating tls isn't free, the hardware that canonical runs the archives on isn't up to the task for that many connections
[20:25] <sarnold> hansh_: you're right that once or twice a year someone comes in here with unexplainable hashsum mismatches due to terrible proxies run by their ISP or their VPN, and telling them to switch to https kinda sucks
[20:25] <sarnold> hansh_: hopefully we'll be able to offer tls services some day
[20:27] <hansh_> i see, thanks
[20:31] <mgedmin> hmm, those country mirrors already have subdomains under ubuntu.com, so they could get a domain-validated cert if they wanted to
[20:31] <mgedmin> ubuntu.com isn't publishing a CAA record
[20:37] <jayjo_> re: packaging... do debian packages have to be accepted into the distribution itself, either in backports, testing, or edge, and ubuntu packages will have the packages from those debian packages in addition to repositories on launchpad.net, and then you can also install *personal* repos from launchpad.net? And roughly the build procedures for them all are the same? Is that right?
[21:03] <pagios> hello, how can i disable the local dns and just keep reading dns config from /etc/resolv.conf ?
[21:04] <Examknow> pagios: What do you mean by 'local dns'?
[21:04] <sarnold> probably systemd-resolved
[21:04] <pagios> i need to point it to 8.8.8.8
[21:05] <pagios> yea i need to disable that and use an external dns like in the old days
[21:06] <Examknow> Try using 'sudo systemctl disable systemd-resolved' and then 'sudo systemctl stop systemd-resolved'
[21:09] <pagios> Examknow:  i did but that doesnt do the job, i cant resolve any hosts then
[21:11] <Examknow> pagios: Do you have 'dns=default' in '/etc/NetworkManager/NetworkManager.conf'?
[21:11] <Examknow> (goes in main section)
[21:12] <pagios> Examknow: i dont have that file /etc/NetworkManager/NetworkManager.conf
[21:12] <Examknow> That's strange
[21:12] <pagios> -> /etc/NetworkManager/ doesnt exist as a dir
[21:12] <Examknow> I do
[21:12] <Examknow> What version are you using?
[21:13] <pagios> Description:    Ubuntu 20.04 LTS
[21:13] <Examknow> So am I. Server or desktop?
[21:14] <pagios> server
[21:14] <Examknow> oh
[21:14] <Examknow> pagios: What does 'sudo systemctl status NetworkManager' return?
[21:15] <pagios> Unit NetworkManager.service could not be found.
[21:15] <Examknow> Ah so you don't have NetworkManager at all
[21:15] <pagios> i have systemd-resolved i gusess
[21:16] <sarnold> the server install doesn't install networkmanager
[21:16] <Bashing-om> pagios: Server generally uses systemd-networkd, rather then the desktop network-manager.
[21:16] <pagios> ok  i need to make that use an external dns in /etc/resolv.conf
[21:16] <pagios> like the old days
[21:17] <Examknow> ok i'll link you to instructions to install netmanager
[21:17] <Examknow> pagios: https://help.ubuntu.com/community/NetworkManager#Installation and then https://help.ubuntu.com/community/NetworkManager#Start
[21:17] <pagios> why do i have to install it
[21:18] <Examknow> actually hang on
[21:18] <Examknow> you may not need to
[21:20] <Examknow> pagios: Yeah see this article https://www.tecmint.com/set-permanent-dns-nameservers-in-ubuntu-debian/
[21:20] <sarnold> pagios: please be sure to file a bug against systemd-resolved before you unconfigure it, to describe what usecase it can't support
[21:20] <Examknow> And go ahead and revert the changes we made to systemd-resolved
[21:20] <pi0> within ubuntu is there a way to see my video card chipset or model number
[21:21] <pi0> i did a dual boot mac
[21:21] <pagios> i just removed /etc/resolv.conf and replaced it with a static file , it does the job
[21:21] <pagios> it was using a symlink
[21:21] <Examknow> pi0: Settings -> About
[21:21] <sarnold> pi0: lshw -C video  or lspci  (with more -v if desired)
[21:23] <pi0> nice!
[21:23] <Bashing-om> pi0: ^^ Alternately Try : sudo lshw -C display ; lspci -k | grep -iEA5 'vga|3d .
[21:23] <pi0> thank you
[21:23] <pi0> Intel® Iris(R) Pro Graphics P5200 (HSW GT3)
[21:23] <pi0> reason for the dual boot is to support opengl
[21:23] <pi0> i think mac is moving away from it
[21:23] <pi0> and i would like to create a godot engine game
[21:24] <pi0> does opengl work out of the box with ubuntu or do i need to tweak it a bit
[21:24] <sarnold> pi0: good question, I did some opengl through java back around 2000 or so... :)
[21:25] <sarnold> pi0: try 'glxgears' and see if you get 60 fps on it or not
[21:25] <Bashing-om> pio See the result from terminal command: glxinfo | grep OpenGL
[21:25] <sarnold> pi0: there's also 'glxinfo' which dumps gobs of information, I'm not sure how much of it is useful, but the first few dozen lines are probably most uesful
[21:28] <pi0> looks like i need to install mesa-tools
[21:28] <pi0> mesa-tools are for opengl?
[21:28] <pi0> or just to view the configuration of the system
[21:29] <sarnold> I don't have a mesa-tools package on my system; I can't find anyting in the archive that looks like that..
[21:29] <sarnold> aha, typo :)
[21:29] <sarnold> $ dpkg -S `which glxgears`
[21:29] <sarnold> mesa-utils: /usr/bin/glxgears
[21:29] <sarnold> it's mesa-utils, not mesa-tools
[21:31] <pi0> you are correct
[21:31] <pi0> installed it
[21:33] <pi0> glxgears 304 frames in 5.0 seconds = 60.699 FPS
[21:33] <pi0> 300 frames in 5.0 seconds = 59.994 FPS
[21:33] <pi0> decent? i have no idea on video cards
[21:35] <sarnold> iirc glxgears caps at 60fps regardless of hardware, because people were going silly comparing 500fps results with each other, hehe
[21:36] <Bashing-om> pi0: Old dual core Athlon system here: " 303 frames in 5.0 seconds = 60.538 FPS ; 300 frames in 5.0 seconds = 59.946 FPS".
[21:36] <sarnold> I don't know intel iris well, but I assume it's probably no real competitor for amd or nvidia cards, but the drivers might be less annoying :)
[21:38] <legreffier> it would be irrelevant to take 100 non-textured polygons as a benchmark.
[21:38] <pi0> that is true
[21:39] <sarnold> once upon a time it was very relevant :)
[21:39] <sarnold> hehe
[21:39] <legreffier> it was useful 20 years ago when non-accelerated rendering would give you 1FPS, and >100fps with a gpu.
[21:40] <legreffier> your <100$ smartphone is more powerful by any 2000's gaming rig by a long shot.
[21:41] <sarnold> but I had a 1600x1200 monitor in those days!
[21:42] <sarnold> (iirc I played quake at 512x384 @ ~28 fps. and liked it.)
[21:42] <pi0> only issue now is scaling
[21:42] <pi0> most of my other apps icons are tiny, example gimp inkscape
[21:42] <pi0> unless there is a zoom tool for ubuntu
[21:42] <pi0> lol
[21:43] <pagios> when i run ethtool -s ethx autoneg off speed 1000 duplex full will this be preserved after a restart?
[21:45] <genii> Nope
[21:47] <pagios> how can i make it survive a restart
[21:48] <genii> The old way was with a pre-up command specified in  /etc/network/interfaces but with netplan now, I'm not sure
[21:49] <pagios> @reboot cron?
[21:49] <genii> If commands in the file /etc/rc.local  afile are still homoured you could put the command in there
[21:49]  * genii curses his typos
[21:50] <sarnold> pagios: https://netplan.io/faq/#use-pre-up%2C-post-up%2C-etc.-hook-scripts
[21:50] <genii> sarnold: Ah, nice
[21:51] <pagios> i have something called smartspeed that i need to disable too
[21:51] <sarnold> genii: yeah, I'm pleasantly surprised, iirc once upon a time the answer was as you suggested :)
[21:54] <clarkk1> On Ubuntu 18.04, I'm connecting to an "anonymous" VPN provider. When a tunnel is established, via a terminal, systemd-resolved has the new DNS configured, but also the existing one for the LAN.  It continues to query both during the connection, meaning that I get a DNS leak.  Is there any way with systemd-resolved to add and remove a DNS from the command line? Also, can it be done when the when an interface comes up or down?
[21:55] <gordonjcp> hello
[21:56] <gordonjcp> installing Ubuntu 20.10, at the end of the install it pops up a dialogue box "Executing 'grub-install /dev/sdb' failed / This is a fatal error."
[21:57] <gordonjcp> that's not a very user-friendly error message because it doesn't tell you what's wrong
[21:57] <gordonjcp> running grub-install /dev/sdb in a terminal just pops up a very generic error message "grub-install: error: cannot find EFI directory"
[21:57] <gordonjcp> that leads to lots of very old posts referring to other distros, but nothing terribly specific
[21:58] <sarnold> gordonjcp: that sounds a bit like you've booted using efi but didn't create an efi partition on the drive
[21:58] <gordonjcp> sarnold: I simply followed the prompts in the installer, without choosing anything other than the defaults
[21:58] <gordonjcp> sarnold: so, the installer is buggy
[21:59] <sarnold> gordonjcp: perhaps; did the installer give you a bug number?
[21:59] <gordonjcp> sarnold: no
[22:00] <gordonjcp> apport offers to send a bug report for ubiquity
[22:00] <sarnold> gordonjcp: what's your launchpad id?
[22:00] <gordonjcp> sarnold: no idea
[22:00] <gordonjcp> maybe gordonjcp, I vaguely recall playing with launchpad when it first came ot
[22:00] <gordonjcp> *out
[22:00] <bumblefuzz> when I leave my computer to sleep for a given period of time, when I come back the audio controller is disabled
[22:01] <bumblefuzz> can someone help me debug how this is happening?
[22:01] <bumblefuzz> moreover, I don't even see the option in sound preferences to renable the device
[22:01] <sarnold> gordonjcp: hmm, i wonder what happened to your bug report :( there's nothing newer than 2016 here https://bugs.launchpad.net/~gordonjcp
[22:01] <bumblefuzz> I don't know enough about these from the cli
[22:02] <TJ-> bumblefuzz: is that suspend to RAM? Could be an ACPI issue. See https://iam.tj/prototype/enhancements/Windows-acpi_osi.html
[22:02] <gordonjcp> sarnold: I didn't put a launchpad ID in or anything, just told apport to go ahead and report a bug
[22:04] <TJ-> gordonjcp: when you say "...running grub-install /dev/sdb in a terminal..." do you mean whilst using the installer's "Try Ubuntu" session and opening a terminal and shell after the installer failed?
[22:04] <gordonjcp> TJ-: yes
[22:05] <bumblefuzz> TJ-, where in the logs would I look to see how the audio controller became disabled in the first place
[22:05] <bumblefuzz> ?
[22:05] <TJ-> gordonjcp: OK, thought so. A bit of background. In UEFI mode grub looks for /boot/efi/EFI/ being a mount-point for the EFI system partition in order to write its core image there. This won't be present in the 'ive' environment but would be in the installed /target/ - which would need to be accessed via 'chroot' and may need mounting correctly if the installer itself has terminated already
[22:06] <gordonjcp> TJ-: makes sense
[22:06] <TJ-> bumblefuzz: "journalctl -k -n 250" would show the latest 250 lines of the kernel log
[22:07] <TJ-> gordonjcp: so, I'd first check there is actually an EFI-SP on /dev/sdb with "lsblk -f /dev/sdb"
[22:07] <gordonjcp> TJ-: so why doesn't the "proper" installer work then
[22:07] <crash0verride> Hi, I'm trying to resolve a small but annoying issue I'm having on 20.04 where the mouse cursor is tiny (1/4 size) for certain applications. Most Ubuntu default apps like Files, Settings, Software Updater are affected, but other applications like Firefox and VS Code are fine, same for hovering over an empty desktop. The issue appears to be Wayland specific as when I'm not running Wayland it works as expected. I'm running a
[22:07] <crash0verride> laptop plus a second monitor and the issue is only on the Laptop as it's 4k and running at 200% Scaling, the external monitor running at 100% scaling is fine, no issue with the cursor.
[22:07] <gordonjcp> TJ-: there's a vfat partition of some sort
[22:08] <bumblefuzz> TJ-, it doesn't show anything about the audio controller
[22:08] <TJ-> gordonjcp: if so, then I'd check if the new install is still mounted, or needs to be remounted with "mount | grep target" --- it should be mounted to /target/
[22:08] <gordonjcp> TJ-: I'll continue with this in the morning, it's too late to work on it now
[22:08] <gordonjcp> TJ-: tbh I'll likely switch back to 20.04
[22:08] <TJ-> gordonjcp: you'd expect to see something like "├─nvme0n1p1                vfat        EFI-SP"
[22:08] <gordonjcp> TJ-: the new install is mounted on /target/ definitely
[22:09] <gordonjcp> TJ-: says vfat FAT32 DBFF-26CF
[22:09] <TJ-> gordonjcp: if still mounted then you can do "for n in proc sys dev etc/resolv.conf; do sudo mount --rbind /$n /target/$n; done"
[22:09] <gordonjcp> yeah, ftn
[22:09] <gordonjcp> no way
[22:09] <TJ-> gordonjcp: then you're ready to fix/test it: start with "sudo chroot /target mount -a"
[22:09] <gordonjcp> I'm not talking users through that
[22:09] <roothorick> Is there a way to give CAP_SYS_NICE to *all* processes under a certain UID, current and future?
[22:09] <crash0verride> The issue seems to be that when running Ubuntu on Wayland, the Cursor is running at 100% scaling in these apps and everything else is at 200%, it's a weird one and I'm not sure where to even start looking to resolve it
[22:10] <gordonjcp> TJ-: this has simply taught me not to stray from the path of LTS
[22:10] <TJ-> gordonjcp: this is to discover the issue, not as a workaround. If the issue can be identified it whould be possible to solve it
[22:10] <gordonjcp> TJ-: haven't really got time right now, maybe in a month or so
[22:10] <TJ-> gordonjcp: it's less than 3 minutes to do and grab logs
[22:10] <gordonjcp> TJ-: I'll copy the scrollback for then
[22:10] <gordonjcp> TJ-: yes, I don't have time right now
[22:11] <TJ-> gordonjcp: "sudo chroot /target grub-install -v /dev/sdb |& tee /tmp/grub-install.log" and then create a bug report and attach /tmp/grub-install.log
[22:12] <bumblefuzz> how do I reenable an audio controller?
[22:13] <gordonjcp> TJ-: aha
[22:13] <gordonjcp> TJ-: mount shows that /boot/efi is mounting /dev/sda2 which is the install media
[22:14] <gordonjcp> TJ-: so, the installer is broken I guess
[22:15] <TJ-> gordonjcp: if it is doing that in the chroot then that does look like a regression
[22:15] <clarkk1> On Ubuntu 18.04, I'm connecting to an "anonymous" VPN provider. When a tunnel is established, via a terminal, systemd-resolved has the new DNS configured, but also the existing one for the LAN.  It continues to query both during the connection, meaning that I get a DNS leak.  Is there any way with systemd-resolved to add and remove a DNS from the command line? Also, can it be done when the when an interface comes up or down?
[22:16] <gordonjcp> TJ-: I'm surprised, it's just an iso, dd'ed to the USB stick, passes all its file system checks on startup
[22:16] <gordonjcp> TJ-: following default values for the prompt
[22:16] <gordonjcp> TJ-: both the USB stick and SSD wer zeroed out before starting the install, to at least "simulate" installing on brand new box fresh hardware
[22:17] <TJ-> clarkk1: systemd-resolved switches between available resolvers when the current resolver doesn't respond promptly. Newer versions have "dns" options in systemd-resolve/resolvectl to alter DNS settings
[22:18] <TJ-> gordonjcp: it sounds on the face of it as if the first EFI-SP found is being used BUT I suspect this could be user error - the installer *should* look for the EFI-SP on the device that is set as the Bootloader device in the partitioning stage, and that may well default to /dev/sda and can be missed by users
[22:19] <clarkk1> TJ-, I'm afraid I'm stuck on 18.04 at the moment.  What options are available to me?
[22:19] <gordonjcp> TJ-: there's no option to set where the bootloader installs
[22:19] <gordonjcp> TJ-: as a dumb user I shouldn't have to know that kind of thing
[22:19] <gordonjcp> TJ-: I should be able to say "next next next done"
[22:20] <clarkk1> TJ-, they may switch between resolvers when the current one doesn't respond promptly, but I'm not noticing any problems if I block communication with the local resolver
[22:22] <TJ-> gordonjcp: there is an option, I was just trying to find a screenshot of it, but the one I'm thinking about is in the partitioning "Something Else" option ... I've not run the installer in automatic mode recently
[22:23] <crash0verride> Hi, quick easy one. What framework does the default Ubuntu applications use GTK, QT etc? By default applications I mean Files, Settings, Software Updater etc
[22:24] <tomreyn> Gnome application UI are based on the 'graphical toolkit' (GTK)
[22:25] <TJ-> clarkk1: you mean you alter /etc/resolv.conf ?
[22:26] <clarkk1> TJ-, sorry, I forgot to say, it's using systemd-resolve
[22:27] <TJ-> clarkk1: the entry in /etc/resolv.conf (which should be a symlink to /run/systemd/resolve/stub-resolv.conf ) contains the IPv4 address of the systemd-resolved daemon which does local name resolution via the upstream DNS servers, reported by "systemd-resolve --status" I think on 18.04 - you'd need to check the options with "man systemd-resolve"
[22:27] <tomreyn> (i think it was actually called 'gnome toolkit' at some point)
[22:30] <gordonjcp> tomreyn: in the olden days, yes, it was called gnome toolkit
[22:30] <crash0verride> tomreyn - Thanks for that, seems like a simple one but with such generic names eg Files it was proving hard to find via Google. Is there any difference in graphical toolkit the default apps I mentioned use vs the default Calculator app in Ubuntu or Firefox uses?
[22:31] <crash0verride> The issue I'm trying to track down only affects some default Ubuntu apps but not others, so I'm just trying to identify what's different
[22:32] <gordonjcp> TJ-: can confirm that a "next next next done" approach works in 20.04
[22:32] <crash0verride> Or even better, is there a way to identify which toolkit an application is using via command line, logs etc?
[22:34] <tomreyn> crash0verride: you can inspect package dependencies, this will usually reveal it.
[22:34] <tomreyn> what's the issue you're trying to track down?
[22:34] <sarnold> crash0verride: try this: for p in `pidof firefox` ; do cat /proc/$p/maps ; done | awk '$2 ~ /r-xp/ {print $6;}' | sort -u
[22:35] <crash0verride> tomreyn - Hi, I'm trying to resolve a small but annoying issue I'm having on 20.04 where the mouse cursor is tiny (1/4 size) for certain applications. Most Ubuntu default apps like Files, Settings, Software Updater are affected, but other applications like Firefox and VS Code are fine, same for hovering over an empty desktop. The issue appears to be Wayland specific as when I'm not running Wayland it works as expected. I'm
[22:35] <crash0verride> running a laptop plus a second monitor and the issue is only on the Laptop as it's 4k and running at 200% Scaling, the external monitor running at 100% scaling is fine, no issue with the cursor.
[22:37] <TJ-> gordonjcp: that's a pain then; could be a regression in ubiquity, or possibly the underlying debian-installer/partman component
[22:37] <tomreyn> i see.so not too big a problem, but still worth a bug report if there's none, yet (but there probably is)
[22:38] <crash0verride> tomreyn - Seems to be the cursor isn't respecting the scaling settings in certain applications, but others it's fine.
[22:38] <clarkk1> thanks for your help TJ- - I'll go through this in the morning
[22:39] <gordonjcp> TJ-: if I get a chance in the next few weeks I'll try and narrow it down
[22:39] <tomreyn> i agree this is what it sounds like
[22:40] <TJ-> gordonjcp: now we know we can likely reproduce it in VMs
[22:40] <crash0verride> tomreyn - Not a big problem, just annoying and looks pretty shite. I'm mainly trying to be able to narrow down so I can describe it accurately in a bug report.
[22:40] <TJ-> gordonjcp: the installer writes copious logs in the live environment and puts some into the installed /var/log/... somewhere so we should be able to capture info
[22:41] <tomreyn> crash0verride: good luck. and keep asking if you need more details, but i'll need to head to bed now.
[22:41] <crash0verride> tomreyn - It's just a bit strange it occurs in Files, Settings etc but is totally fine in Calculator
[22:41] <TJ-> gordonjcp: I'll do a quick test run tomorrow morning when I get into the office
[22:41] <tomreyn> crash0verride: not really, calculator is a snap
[22:41] <TJ-> crash0verride: could the difference be in 'snap' applications vs native ?
[22:43] <TJ-> crash0verride: Firefox, although native (it is still a .deb isn't it?) ships with its own libraries if I recall correctly, which might explain it being fine
[22:44] <crash0verride> TJ - Yep, I think Firefox was a deb, same for VS Code which also works fine
[22:46] <crash0verride> I'm not seeing Calculator in the list of Snaps (ie snap list), sure it's a snap and not native?
[22:46] <crash0verride> Sorry I take that back, gnome-calculator :facepalm:
[22:48] <crash0verride> TJ - Confirmed all of the snap applications I've tried all work as expected
[22:51] <crash0verride> TJ - Who'd be the best to report the issue to and/or resolve it, Ubuntu, Gnome or Wayland?
[22:55] <TJ-> crash0verride: Ubuntu bug to begin with and then identify any upstream bugs and link them to it
[23:02] <Sven_vB> what's a good image viewer for renaming or deleting lots of images while viewing them? Shutter at least has a rename function, but it always pops up centered inside its window, hiding the image. I could make an xdotool script to always move the rename window, but maybe there are better tools?
[23:03] <sarnold> Sven_vB: qiv can do deletes super-conveniently
[23:04] <Sven_vB> thanks! I totally forgot that one
[23:04] <Sven_vB> maybe it can rename, too
[23:05] <sarnold> Sven_vB: ooh it looks like it can -- you'd need to supply a qiv-command program, but then you could use ^ to rename
[23:05] <sarnold> that's not super-convenient
[23:12] <Sven_vB> or maybe a program that easily lets me add some text annotation that I can later parse to mass-rename them?
[23:15] <TJ-> Sven_vB: maybe some tool targeted at EXIF data might also do it?
[23:16] <Sven_vB> TJ-, thanks for the idea!
[23:17] <crash0verride> sarnold - Thanks for the "for p in 'pidof xyz'" snippit earlier, I've just been trying that on applications I'm having trouble with vs ones that work fine
[23:18] <crash0verride> The issue seems to only affect apps using ... usr/lib/x86_64-linux-gnu/gtk-3.0/3.0.0/immodules/im-wayland.so
[23:18] <Sven_vB> well I guess the easiest approach for now is to use "save as" and then match duplicates by checksum.
[23:19] <sarnold> crash0verride: sweet! that sounds promising :D
[23:19] <Sven_vB> (or press "end" key before typing the new name and then matching by filename prefix. :D)
[23:21] <sarnold> Sven_vB: fdupes or jdupes or dozens of other similar tools can do identical-by-checksum things
[23:21] <Sven_vB> sarnold, in my case md5sum is good enough. :)
[23:22] <sarnold> Sven_vB: ah good :)
[23:22] <sarnold> when I used one of thse tools, I had hundreds of dupes, and didn't want to deal with all the cases of symlinks and so on by hand
[23:39] <Newb101> Hi, is virtualbox in the ubuntu software store on 20.04 I cant find it?
[23:39]  * Examknow looks
[23:40] <Examknow> Newb101: It is for me
[23:40] <Newb101> weird
[23:40] <Newb101> i did minimal install would that make a difference
[23:40] <Newb101> ?
[23:41] <Examknow> Shouldn't
[23:41] <Examknow> If you need to download it, you can do so from https://www.virtualbox.org/wiki/Linux_Downloads
[23:41] <sarnold> Examknow: where'd it come from? I don't see it on my focal system, nor do I see it in the snap store
[23:41] <Examknow> ubuntu-focal-updates-multiverse
[23:41] <Newb101> is that an option I need to enable
[23:43] <sarnold> Examknow: aha!
[23:43] <Newb101> how do i get to that
[23:43] <Newb101> source
[23:44] <sarnold> Newb101: try: sudo apt-add-repository multiverse
[23:44] <Examknow> yeah
[23:44] <Newb101> 'multiverse' distribution component is already enabled for all sources.
[23:45] <Newb101> ?
[23:46] <sarnold> Newb101: hmm, funny. how about: apt policy virtualbox ?
[23:48] <Newb101> sudo apt install virtualbox virtualbox-ext-pack
[23:48] <Newb101> worked for me
[23:48] <sarnold> cool cool
[23:52] <signofzeta> sorry to interrupt, but can you run virtualbox and KVM/libvirt on the same system?  i know my old Windows PC would blue-screen with both Hyper-V and virtualbox.
[23:53] <sarnold> signofzeta: you can't use kvm-accelerated qemu on the same machine at the same time as virtualbox, no; one or the other will fail to load its kernel modules and won't run
[23:53] <signofzeta> thanks, sarnold!
[23:53] <sarnold> signofzeta: you may have some luck unloading modules and loading the other modules *HOWEVER* unloading kernel modules is not a generally safe operation, so you may have trouble with this