[00:00] extended support thingy https://support.mozilla.org/en-US/kb/choosing-firefox-update-channel [00:00] the funny part: Disabling System Addon Updates [00:00] lolz [00:01] so, you *must* have a reason to use ESR [00:01] well, it's not trying to be drop-in compatible with all firefox extensions [00:14] No ideas?? [00:16] hi, is there a way of changing the icons' text color from white to black in Ubuntu Mate 18? Thanks... [00:19] Hi, using Lubuntu 16.04 LTS on an Asus Eee PC Flare series. I just installed the updates via the update centre today and now when I boot I can see the decrypt screen fine, but the login screen is all black. If I use the Fn keys to ak the machine hiberanate then upon pressing the 'on' button it resumes and screen is fine. I didn't have this issue before, but perhaps I should mention that the decrypt screen is always too dim unless I use the brightness keys. The [00:19] brightness keys have no effect on this new problem though. Any thoughts? Thanks! [00:22] uio, I think I have the same problem here. I switch TTY to work around it. [00:22] Try switching to another TTY, and then returning to TTY 7, if that's where the display manager appears. [00:23] If we're talking about the same problem, it will show you the login screen properly. [00:23] crimson_king, Is it since the update? [00:24] Hmmmm. Any ideas on what the cause is? [00:25] You're on Ubuntu 16. I run Ubuntu 18 and 19. Maybe the package that causes the problem was backported/updated to Ubuntu 16, and then you started having the problem. [00:25] Lubuntu* [00:26] crimson_king, I see. I was working fine until today when I applied the updates. Albeit somewhate late, as I had put it off. I'll give your workaround a go... I was having a similar issue in Debian and that's why I put Lubuntu on it, but Canonical has failed me too! [00:27] uio, if we're having the same problem, it is not Canonical's fault. This is reproducible everywhere, on Arch, Fedora, Debian... I don't know which component is causing it. I think I created a bug report about it, let me search it. [00:28] crimson_king: Are you tired of jokes about your court? :s [00:29] crimson_king, I think it's a different issue. On Debian when I had it before, tty didn't work. This one is a case where Canonical had something working… and then broke it. Alas. I'm looking into going to test out Debian 10 anyway, this was just for the last few days. [00:32] uio: may i suggest you take a look at your logs? this may bear a greater potential in solving the issues or coming up with a workaround at least. [00:35] hmm actually lubuntu 16.04 is !eol [00:35] at least the 'l' part in it [00:36] https://help.ubuntu.com/community/EOL#Ubuntu_16.04_Xenial_Xerus [00:45] i see you got replies in #lubuntu, too. [00:45] Sorry, thanks, but I have to go, I'll be back another time!a [00:46] yay! can boot the new amd vega on 5.* kernel when setting nomodeset in grub cmdline options [00:47] sure, because then you run in vesa mode [00:48] ikatamoonshots: ^^ nomodeset defeats kernel Mode Setting - not a long term solution. [00:48] better install latest kernel, oibaf or padoka mesa and newer amd firmware and boot without nomodeset. [00:49] maybe with amdgpu.dc=0 if it still doesn't work then. [00:49] does that affect performance ? [00:49] vesa? yes, very much. [00:49] also resolution [00:50] i applied the nomodeset to boot with the latest kernel in 18.04 which is 5.* , the older kernels boot fine without the grub option [00:50] which graphics card do you actually have? [00:50] the card is amd vega 20 "amd vii" [00:51] i want to run opencl applications [00:51] tomreyn, i found https://askubuntu.com/questions/1121782/radeon-vii-and-ubuntu-problems already [00:51] oh, hmm. well the mesa implementation is not that optimized, yet [00:51] * opencl that is [00:52] have seen the page already, chatted here not long ago about it, just installed a new ubuntu and set the nomodset to boot the kernel 5. [00:53] now need to figure if installing the headless amdpro is better than the full stack i guess [00:53] apparently only way to get opencl support on ubuntu? [00:53] thanks OerHeks. [00:54] ikatamoonshots: what do you mean by "headless amdpro"? [00:54] theres an option to install it headless , opencl only and keep the main driver installed - thats how i understood it [00:55] there are two ways to opencl. one if amdgpu-pro, you probably want to stay on 18.04 with the defualt or hwe kernel then, no padoka, no oibaf, downgrade everything. [00:55] and follow amds instructions and get support from there. [00:56] or you can try oibaf + padoka mesa and the kernel they recommend. [00:56] these are the stack variants https://amdgpu-install.readthedocs.io/en/latest/install-overview.html [00:58] i guess that's correct [00:58] the download repo has a slightly updated version [00:58] if you only care about computation then go the amdgpu-pro route [00:58] headless or full? [00:58] well if only computation matters then you certainly want headless [00:59] yeah, thats what the card is for [00:59] dont waste cycles on the gui [00:59] on it xD [01:04] ikatamoonshots: note this guide you'Re following is somewhat old, refers to outdated amdgpu-pro versions. maybe better follow their website. [01:04] oh i take this back, https://www.amd.com/en/support/kb/release-notes/rn-amdgpu-unified-navi-linux points to your guide [01:05] i follow the guide from the repo i downloadded, but round up its all the same. [01:05] they dont give any kernel instruction nor say on what kernel its build [01:06] installed amd pro headless and got a neat warning. [01:06] WARNING: nomodeset detected in kernel parameters, amdgpu requires KMS [01:07] say i disable nomodeset again and see if it boots in the 5 kernel - just have to check the grub and x11 optons [01:08] i don't think you need X11 for opencl [01:10] "Ubuntu 18.04.2" is listed as compatible. so i'd expect this refers to 18.04.2 with !HWE kernel. [01:10] default i hasnt booted in the 5 kernel, thats why i have set nomodeset in the first place [01:10] i'm not convinced AMD tested against the recently arrived 5.0 kernel in 18.04.2 (HWE) though. [01:11] they probably tested against 4.18 [01:11] from what i read the 4.15 kernel worked for others, but that was with amd linux 18.* , now its 1.* [01:12] either 4.15 or 4.18 ,yeah [01:12] 19.30 is the latest amdgpu-pro compatible to your hardware [01:12] wont take too long to figure that out [01:12] yeah, downloaded and isntalled the headless [01:12] actually 19.20, sorry [01:13] https://www.amd.com/en/support/kb/release-notes/rn-rad-lin-19-20-unified [01:13] anyways, that's as far as we can help here [01:14] already on it, last thing unclear is grub options for amd gpu - amdgpu.dc=1 ....... have to read up about those, yesterday when i was that far no app detected opencl devices [01:16] modinfo -p amdgpu [01:16] mom, need to reboot [01:25] so, after default install of AMDGPU-pro headless im now in the 4.15 kernel, both 4.15, 4.18 boot, 5.* doesent [01:25] modinfo -p amdgpu [01:25] https://pastebin.com/MWKgCvKt [01:32] $ sudo clinfo -a [01:32] Number of platforms 0 [01:32] ? === [b0b] is now known as b0b [01:58] hello community [01:59] :) === EliTheGamerFTW_ is now known as EliTheGamerFTW [03:03] I use an old psu that only has old connectors so I've connected an old hdd to the cdrom wires. is there something I need to do to make it possible for the live usb to install the bootloader? [03:04] it currently fails, it also cant find grub-install grub2-install boot-repair and wont install those [03:06] keep having to use gparted to make the disk usable again [03:10] nn nnnnnnnnnnnnnnnnn mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm [03:11] ooopps :( [04:47] . [05:21] mysql-server 5.7 ssems broken on 18.04? [05:21] dirtwash: details? [05:22] after upgrade just now dpkg reports failure: mysql_upgrade: [ERROR] 1146: Table 'mysql.gtid_executed' doesn't exist [05:22] was running fine before and earlier upgrades [05:22] well itstarts manually at least [05:24] hm possible remnants from an old upgrade, previous upgrades didnt trigger that though [05:25] dirtwash: it may well be, the post-install script had some rather complex scenarios to try to figure out, see the shell script at /var/lib/dpkg/info/mysql-server-5.7.postinst [05:35] i have some SBC orange pi zero's i am playing with. i loaded them up with ubuntu core and made a snap on my desktop system. the SBC's are using core18 and wont let me run snapcraft, the desktop is not responding to snapcraft --target-arch=armhf any ideas on the best way to get the snap on ubuntu core? [05:35] when the i try to snap install snapcraft from the core18 SBC i get "error: snap "snapcraft" requires classic confinement which is only available on classic systems" [05:38] Hi guys, is there a way to scan for all my hardware that may require propietary firmware *afterr* ubuntu is already installed? === karlthane_ is now known as karlthane [06:37] https://kernel.ubuntu.com/~kernel-ppa/mainline/v4.19.63/ [06:37] I'm failing to find shiftfs patch there ^ [06:38] Is there a separate repo/branch? [06:45] hey guys , how is it possible to share the user's default desktop in ubuntu using VNC? i am using TigerVNC, but everytime i connect to the server it creats new desktop session and i am missing all the icons and files added to desktop [06:45] i guess something has to be added to xstartup script but i dont know how the gnome session works to create the initilization script to load the user's settings for each user [06:46] SaEeDIRHA: x11vnc [06:47] SaEeDIRHA: is it on an offline LAN, or are those machines connected to internet [06:47] lotuspsychje, its a LAN [06:47] SaEeDIRHA: there is also x2go [06:47] fling, i tried x11vnc but still the same problem [06:48] when i connect that way i get black screen with mouse cursor [06:48] SaEeDIRHA: x11vnc connects to your running session [06:48] SaEeDIRHA: it shares your screens over vnc [06:48] SaEeDIRHA: maybe you have multiple sessions running and x11vnc connects to a wrong one? [06:49] To one created by tigervnc server for example? :P [06:49] fling, i killed that session and tried with x11vnc [06:49] but the session wont be created unless i add the user to auto login to gdm3 config file [06:50] then it creates the display session [06:50] this is expected [06:50] SaEeDIRHA: x2go would be much convenient for your needs [06:50] fling, i will have a look at it now [06:50] thankx :) [06:59] Hello! Why does the man page for iptables not mention "dport" even though that is one of its parameters? [07:00] SaEeDIRHA: just be carefull vnc port doesnt show on the outside internet, or youl get hammered 24/7 [07:01] lotuspsychje, its on my local network server and i am the only user [07:01] but thanks [07:01] SaEeDIRHA: allrighty === juboxi is now known as jubo2 [08:23] Hi. I am having problems with graphical snap-apps. Can someone help me with this? https://paste.ubuntu.com/p/69hcPs6Cr8/ [08:33] hello :) [08:35] Well, i wanted know how to configure the "flick with pen" kind of scroll on wacom. It's default in windows but i use linux ( ubuntu 18.04.2 LTS). It would be nice to know that because it is a mess using that scroll bar all the time. much easier to navigate it with pen itself. [08:35] please help [08:44] TJ-: thanks again, waited for the resync to finish and now everything works! [08:47] napcae: took some time then?! [08:50] where does the Software Updater app get the description of each package that's shown in the list? [08:50] clarkk: the app-data package(s) [08:50] TJ-, How do I get that info from the command line? [08:52] clarkk: I think it's actually appstream. When you do "apt update" you'll sometimes see references to DEP-11 ... which is https://wiki.debian.org/DEP-11 [08:53] clarkk: e.g. for 18.04 bionic: http://archive.ubuntu.com/ubuntu/dists/bionic/main/dep11/ [08:54] TJ-, but what about 3rd party packages/repos? [08:54] TJ-, is the information in the .deb file? [08:54] If so, how do I query the package? [08:55] apt show package [08:55] clarkk: If you're on about snaps I believe they have their own appstream feed [08:55] clarkk: anything in the archives is looked for in the (relative path) I showed above [08:59] TJ-, I don't think I mean snaps. Just simple deb packages from 3rd party repos [09:00] where is the option to check SMART data of a hard drive? [09:01] hey when i do `lshw`, where does it get it's information from to print out? I'm trying to do C++ programme that checks for information like serial number and battery etc. It seems like a bad solution to execute the command and do string comparisons to get what i want. Is there a library of file i can open to get the information I'd want? [09:01] clarkk: the 3rd party would need to publish DEP-11 info as given in that spec then [09:02] eraserpencil1: read the source-code of lshw [09:02] gpio, TJ- Here's what I'm trying to get to the bottom of.... The skype repo/package works correctly, showing "Skype" and its icon in the "details of updates" list in the "Update Software" app. This is what is in my apt sources... deb [arch=amd64] https://repo.skype.com/deb stable main [09:02] gpio, TJ- However.... (one moment please) [09:02] clarkk: dig into the repo then, it's just an HTTP(S) server [09:03] oooh good idea [09:03] Hey, everyone I'm getting error while adding ppa, Here is the link to the terminal output https://paste.ubuntu.com/p/zwFhn5hXYc/ [09:03] For nordvpn, it shows the description instead of the name ("Protect your privacy online and access media content with no regional restrictions. Strong encryption and no-log policy with 3000+ servers in 60+ countries") and no icon. [09:03] clarkk: they don't have any DEP-11 there: https://repo.skype.com/deb/dists/stable/main/ [09:04] eraserpencil1, or use 'lshw -xml' or 'lshw -json' so the parsing becomes easy [09:04] gpio, TJ- It's very confusing, and I'm trying to get them to fix it [09:05] gpio, TJ- This is what is in my apt sources.... deb https://repo.nordvpn.com/deb/nordvpn/debian stable main [09:05] TJ-, it's nordvpn that I'm trying to get to the bottom of, not skype. Skype is just an example of one that works correctly [09:06] gpio, TJ-, it's nordvpn that I'm trying to get to the bottom of, not skype. Skype is just an example of one that works correctly [09:08] clarkk: again, no DEP-11 info https://repo.nordvpn.com/deb/nordvpn/debian/dists/stable/main/ [09:09] TJ-, skype works correctly, so then surely it's not related to DEP-11? [09:09] clarkk: could the info you see be coming from the debian package's debian/control Description fields, e.g. see the "Description:" field listed for each package in the Packages file at https://repo.skype.com/deb/dists/stable/main/binary-amd64/Packages [09:09] Hey, everyone I'm getting error while adding ppa, Here is the link to the terminal output https://paste.ubuntu.com/p/zwFhn5hXYc/ [09:09] Habbie: sounds good. the source code is abit beyond me [09:10] TJ-, How can nordvpn fix this? The name needs to go in that Software Updater app list, not the description [09:11] clarkk: I don't know. What info does Skype show that makes you think they are using DEP-11 ? I see no DEP-11 Appinfo files in the Skype repo [09:12] TJ-, I'd guess that it's nothing to do with DEP-11, because skype is showing correctly (just the word "Skype" with an icon next to it) in the Software Updater list [09:12] *the skype icon [09:13] clarkk: hmmm, that doesn't sound like correct to me! I thought "correct" meant the listing had additional description about it [09:14] clarkk: what is your definition of being correct? having an icon? [09:15] TJ-, ok, let me get you some screenshots [09:15] clarkk: thanks [09:21] gpio, TJ- See these screenshots. Skype is at the top of the list, and the nordvpn package is one above the one highlighted. Interestingly, using aptitude from the command line displays it all correctly (screenshots also included) https://www.dropbox.com/sh/954piyyd6u7hrlf/AAAXBb6CkWMtIft0BRj1IxRAa?dl=0 [09:22] TJ-, Do you see what I mean? [09:25] clarkk: aha, yes. With my 'programmers' hat on I'd guess that is caused by nordvpn not having some kind of "Name" field provided and therefore the "Description" field being used by Software Updater in its place [09:31] does Ubuntu take care of security patches for mainboards, like the Intel Management Engine patches? [09:33] 'dpkg-deb -I nordvpn_3.2.0-2_amd64.deb' shows the right info [09:34] My guess would be the DEP-11 file Software Updater uses [09:34] clarkk: do you have the Ubuntu "partner" component enabled? that has skype packages and info in it? Also, try "find /usr/share/app-install -name '*skype*' " -- see if there is any skype files there... if so we can find out which package added them [09:35] clarkk: and I checked the nordvpn control info and indeed they've got a badly formatted Description: field. there's supposed to be a short 1 line (less than 50 characters) summary followed by an empty line then the longer description [09:39] clarkk: but in your list you have also a standard package that shows "Developer documentation for Mesa" ... I searched for that and found it for the package mesa-common-dev. If you do "apt-cache show mesa-common-dev" you'll notice the Description: summary is what you see in your screenshot and is followed by a longer dsecription which isn't in your screenshot (which is the expanded info that aptitude shows) [09:39] clarkk: so the nordvpn issue is mostly badly formatted Description and no app-install data from appstream to add DEP-11 metadata and icons [09:41] Sven_vB: no, those would require BIOS/firmware upgrades from the manufacturer [09:45] TJ-, so you're saying that because ubuntu is shipped with skype in an optional repo, its icon and description come from there, but the nordvpn info comes from its deb file? [09:46] TJ-, " I checked the nordvpn control info and indeed they've got a badly formatted Description: field" where did you find the control info? [09:48] clarkk: in the Packages file from the nordvpn repo [09:48] clarkk: see https://repo.nordvpn.com/deb/nordvpn/debian/dists/stable/main/binary-amd64/ [09:48] clarkk: it has: Description: Protect your privacy online and access media content with no regional restrictions. Strong encryption and no-log policy with 3000+ servers in 60+ countries. [09:49] TJ-, where did you find the spec for that field? [09:50] clarkk: I know it because I'm a developer. Here's the spefication: https://www.debian.org/doc/debian-policy/ch-controlfields.html#description [09:52] TJ-, thanks! [09:52] TJ-, Thank you for all this help. I'll pass it on to nordvpn and hope they action it [09:52] clarkk: I'm surprised they don't publish DEP-11 metadata and a nice icon, it'd make them stand out :) [09:53] hey is there no way to install a package like postgresql-10 without autostarting the service? is --no-triggers safe to run? because there might be other triggers like building manpages or whatever. [09:54] I think the package maintainers screwed up here because they put pgbench in postgresql-10 package on bionic. so to run that tool I need to install the service. [09:54] on an app server. [09:57] and even worse is this pg_wrapper thing. which means you find /usr/bin/pgbench in postgresql-client-common but it's useless without the real pgbench in postgresql-10. [10:01] stemid: it used to be using invoke-rc.d but I'm not sure if that works with systemd, see https://serverfault.com/questions/567474/how-can-i-install-packages-without-starting-their-associated-services [10:10] it's just awfully hacky. in my case it was easier to install it and then disable it. at least disabling is easier now with systemd. === lemsip is now known as emilsp [10:30] hi, I want disable ipv6 then I have added the following lines http://paste.ubuntu.com/p/MggDWhq8yT/ in /etc/sysctl.conf. If I run it with 'sysctl -p' it works but after reboot looks again enabled. [10:31] sysctl looks running by systemd with systemd-sysctl service.. but it doesn't work [10:31] i recall reading that systemd-sysctl does not read /etc/sysctl.conf [10:32] MMYStic: check there aren't other options over-riding it: "grep -rn ipv6 /etc/sysctl.{conf,d/*}" [10:32] ah, on my (debian) system, /etc/sysctl.d/sysctl.conf is a symlink to /etc/sysctl.conf to avoid that problem [10:33] eh, /etc/sysctl/99-sysctl.conf actually [10:33] MMYStic: if you read "man systemd-sysctl" you'll note it only reads the files in sysctl.d/ diretories (see "man 5 sysctl.d") [10:37] Hi [10:38] Hlo [10:40] TJ-: yeah but under sysctl.d there is link to /etc/sysctl.conf, the following is the output of grep -rn ipv6 /etc/sysctl.d/* https://paste.ubuntu.com/p/RKNfD36xn2/ [10:41] Habbie: exactly the same in ubuntu [10:44] MMYStic: which network manager is being used? maybe it over-rides that later [10:45] MMYStic: can you apply them manually with "sudo /lib/systemd/systemd-sysctl --prefix net.ipv6.conf" ? [10:48] TJ-: ok just a momento I have to reboot to check it [10:48] Howdy folks [11:13] TJ-: well, if I run /lib/systemd/systemd-sysctl --prefix net.ipv6.conf it works. The following is my systemctl status systemd-sysctl https://paste.ubuntu.com/p/jyw8GXRX82/ [11:14] MMYStic: so it is possible something else is resetting it after systemd-sysctl initially sets it disabled [11:15] ok I try to grep /etc/* for any ipv6 conf [11:15] MMYStic: which network manager is in use? systemd-networkd (server) or NetworkManager (desktop) ? I'd suspect either of those may set it [11:19] TJ-: was done longer already but I just replied now to test some more stuff before confirming [11:27] TJ-: looks network manager but I am not sure because I have lubuntu, if I click on the icon to configure the network it open a term with a cui. On ps I have /usr/sbin/NetworkManager --no-daemon [11:28] MMYStic: yes, NetworkManager is the service (package: network-manager) [11:29] MMYStic: I'm not sure if NM will re-enable IPv6 or not; I looked in my local logs where I *do* use IPv6 and see no obvious mention it is enabling it but then again, if it is already enabled, why would it redo that? [11:29] MMYStic: you could check the log with "journalctl -b 0 -u NetworkManager" and search for "ipv6" "v6" and so on strings [11:38] TJ-: look this https://paste.ubuntu.com/p/vdx8qnwCQc/ [11:50] TJ-: ok I have resolved, and yeah it was the NM. I have disabled ipv6 conf on NM. [11:50] thx [11:58] hi [11:59] which emulator can you suggest me to install a simple android APP ? [12:02] sacarde: I use https://www.android-x86.org/ with good results [12:03] is it in repo? [12:04] sacarde: as far as I am aware, no. [12:06] Hi. I am having problems with graphical snap-apps. Can someone help me with this? https://paste.ubuntu.com/p/69hcPs6Cr8/ [12:07] Aavar: How are you trying to launch the application, are you on the local machine? [12:09] pragmaticenigma, hmm... I first tried without sudo and then with sudo from terminal. I also tried from the menu and it would not start. I am on the local machine, yes. [12:09] Aavar: Are you currently in a graphical environment such as Gnome-Shell or KDE? [12:09] Aavar, and are you using wayland (or just logged out from it) ? [12:09] mouses, but android-x86_64-8.1-r2.iso is a complete SO? not only an emulator [12:10] ssiegl: yes [12:10] sorry, sacarde ^^ [12:10] ah.. ok [12:10] pragmaticenigma, ioria: I am running Unity in X11 [12:10] Aavar: what version of Ubuntu are you running? [12:10] I am looking for only an emulator ... like anbox [12:11] pragmaticenigma, . [12:11] pragmaticenigma, 19.04 [12:11] sacarde: Oh, sorry - I kinda misunderstood [12:11] sacarde: For software recommendations please use the #ubuntu-offtopic. This channel focuses on support issues for Official Ubuntu flavors and the software provided through its official software distribution platform. [12:11] * mouses finds more coffee [12:12] ok ok [12:12] Aavar: Unity is not supported in 19.04... try launching the application from Gnome-Shell or another support Desktop Environmnet [12:13] I created usb startup. I works on other computers but on a specific dell xps 15, it says Failed to open \EFI\Boot\mnx64.efi. failed to load image. Not found. Failed to start MokManager: Not Found. Somehting has gone seriously wrong. Import+mok_state*( failed." I re created the startup and same thing. What can I do now? [12:14] intelo: Did you create the USB drive with UEFI support? [12:15] pragmaticenigma, OK, I will try this later. I am running remote (teamviewer) so I dont want to risk not being able to reconnect :) [12:15] pragmaticenigma how to do that? i used kubuntu startup disk creater [12:16] pragmaticenigma, Do you believe this is the cause, or is it just something to check along the way? I dont get why it should not be working as it is the latest release and unity is in the repo and upgraded. [12:17] Aavar: Unity as a desktop is no longer provided by Ubuntu as of 17.xx series. Depending on how you installed your desktop, there may be libraries that were not included during installation or are missing. [12:17] !uefi | intelo [12:17] intelo: UEFI is a specification that defines a software interface between an operating system and platform firmware. It is meant as a replacement for the BIOS. For information on how to set up and install Ubuntu and its derivatives on UEFI machines please read https://help.ubuntu.com/community/UEFI [12:18] Aavar, might sounds silly, but try to 'export DISPLAY=:0' [12:18] hi anyone using acronis backup? It looks like there is no gui, is that possible? [12:19] intelo: mmx64.efi is part of the 'shim' package to handle Secure Boot. If that is missing thne it is possible the ISO doesn't have EFI support. What is the .iso file you created the USB image from? [12:23] pragmaticenigma, I am not being difficult here, but I have to ask :) It is included in the repo. so it is provided by ubuntu? [12:23] kek_: This channel only support official Ubuntu desktop flavors and software provided through its official software distribution channels. For all other software recommendations and help, please use #ubuntu-offtopic. Thanks [12:23] pragmaticenigma, thanks [12:24] !info unity [12:24] unity (source: unity): Interface designed for efficiency of space and interaction.. In component universe, is optional. Version 7.5.0+18.04.20190304-0ubuntu1 (bionic), package size 1755 kB, installed size 6812 kB [12:24] Aavar: It is in the Universe repo, which means it is community maintained and not closely monitored by Canonical [12:24] ioria, thanks, but that gave the same result :) [12:24] pragmaticenigma, aha, I understand :) [12:25] Aavar, does the other snaps have the same problem ? [12:26] ioria, I am not sure. LEt me check ioria. [12:26] ioria, yes. [12:28] Aavar: trace it: "strace -o /tmp/snapstore.strace -f -e trace=file snap-store" then "pastebinit /tmp/snapstore.strace" [12:28] Aavar, xhost +local: and launch again [12:30] ioria, no change :( [12:30] Aavar, see what TJ- suggested [12:31] TJ-, "need to run as root or suid" [12:32] maybe refresh the snap core ... [12:32] TJ-, with sudo: https://paste.ubuntu.com/p/pC3WVh2Fxw/ [12:32] Aavar: huh!? that's not normal [12:33] pragmaticenigma I cannot change the file names in /boot/efi as its a usb disk. read only [12:33] Aavar: do the non-sudo one again and then "pastebinit /tmp/snapstore.strace" [12:33] Aavar: strace should work unless you're in some kind of restricted environment [12:35] TJ-, he's on teamviewer [12:35] I thing I have to test this later when I am home. [12:38] ioria: right but I would not imagine it is going to prevent strace running [12:39] yep [12:40] ioria: do we have any technical info on how, exactly, TV handles things? presumably it hooks in to before the greeter so it can show the log-in greeter dialogs? [12:40] TJ-, nope [12:41] TJ-, ioria I am now running xubuntu-session. same result running snap-store. I am trying the strace thing next [12:41] (still via tv) [12:42] Aavar, can you run without problems gparted or any other root gui applications ? [12:42] I cannot find the string "need to run as root or suid" in /usr/bin/strace or see a shared-object it links in that might report that [12:43] TJ-, ioria, https://paste.ubuntu.com/p/WSM2BFdwPf/ [12:43] same result i'm afraid. [12:43] Is it an idea to remove all snaps and the framework all togheter and start from scratch (if that is posible)? [12:44] Aavar: that doesn't look like what I asked for.. is there no /tmp/snapstore.strace file? [12:46] TJ-, my bad. There is yes. let me try adain. [12:46] again [12:46] TJ-, https://paste.ubuntu.com/p/msKJFkXsRf/ [12:48] ioria, if they are not snaps. Yes [12:49] Aavar, snap issue then [12:49] ioria, I think so. [12:49] since the upgrade from 4.18 to 5.0 on ubuntu 18.04, some of my dkms drivers (intel drivers, in particular), throw "exec format error", any ideas how to fix this? [12:50] an example dkms.conf: https://paste.ubuntu.com/p/Cr2yGHDkGq/ [12:50] and btw, ioria TJ-: I am pretty sure snaps with gui worked earlier this week, but I am not sure what I have done to break this. [12:51] (last wek obvoiusly) [12:51] week [12:54] After new update 18.04 boot is stuck on "Started GNOME display manager" . I tried switching to LightDm but that doesn't work either, just blank screen with blinking cursor. [12:54] https://www.picpasteplus.com/view.php?i=o4a4q4qjapx4i4f4hjo424xp.jpg [12:55] Aavar, i'd try to roll back the version if already updated (or install a beta or edge version) ; awk, sy [12:56] Aavar: the command that is failing appears to be execve("/snap/core/7396/usr/lib/snapd/snap-confine", ["/snap/core/7396/usr/lib/snapd/sn"..., "--base", "core18", "snap.snap-store.snap-store", "/usr/lib/snapd/snap-exec", "snap-store"], 0xc42014d8c0 /* 83 vars */ Aavar: and that is the process that applies confinement (security) to the snap application itself, and that is failing. Shortly after that message you can see the system is checking if suid debugging is enabled (presumably to optionally tell you more about why it failed). [12:57] Aavar: so you might gain more info by finding out what needs adding to /etc/suid-debug. The message: access("/etc/suid-debug", F_OK) = -1 ENOENT (No such file or directory) [12:59] Aavar: aha, seems you just need to create that file, see https://renenyffenegger.ch/notes/Linux/fhs/etc/suid-debug [12:59] Gotango: this link says "image does not exist" for me [12:59] Gotango: which graphics card do you have there? lspci -knn | grep -A3 VGA | nc termbin.com 9999 [13:00] Aavar: so try "sudo touch /etc/suid-debug" and the re-run the strace command and then pastebin any output you see in the terminal and then "pastebinit /tmp/snapstore.strace" again [13:05] @tomreyn https://termbin.com/8n7h [13:05] so apparently dkms is a completely useless tool, as it only works if you manually remove the module from dkms, then re-add it, which isn't really any different from just manually recompiling the module on your own, gg [13:07] bviktor: did the kernel upgrades not trigger DKMS builds, or did those builds fail? [13:07] neither [13:07] it's "built" [13:07] it just won't load [13:07] modprobe: ERROR: could not insert 'igb': Exec format error [13:08] TJ-, I am not following, but I have touched the file and rerun the strace-command: https://paste.ubuntu.com/p/4vc3kJhN2b/ [13:08] @tomreyn https://paste.pics/93c77d961105a46be7d2a52e653f453d [13:08] Aavar: did you see any 'debug' messages in the terminal as well? [13:09] bviktor: was/is "modinfo igb" able to report it or also report problems? [13:09] modinfo works fine. [13:09] Gotango: So Intel HD Graphics 5500 [8086:1616] (rev 09). you can check in /var/log/apt/history.log which packages were upgraded recently. There are a couple reports about issues with the !HWE kernel in Ubuntu 18.04.2 which recently got upgraded to Linux 5.0 (was 4.18). you should post your full system log and have someone review it for error messages: journalctl -b | nc termbin.com 9999 [13:09] TJ-, "need to run as root or suid" i the terminal when I ran the strace-command [13:09] https://paste.ubuntu.com/p/33XGBKfBF3/ [13:10] vermagic: 4.18.0-25-generic SMP mod_unload [13:10] wonderful, isn't it [13:10] Aavar: so nothing more despite our adding /etc/suid-debug? In which case I'm going to guess the debug info may be in the kernel log. Can you "pastebinit <( journalctl -b 0 -n 250 )" [13:10] but then again, why is there an igb.ko with a newer timestamp in /var/lib/dkms/igb/5.3.5.36/5.0.0-23-generic/x86_64/module [13:11] Gotango: alternatively, but then only based on my rough guess that it could be related to the kernel update, you could install a bios upgrade or downgrade to the general availability kernel (4.15) which still gets security updates, too: sudo apt install --purge linux-generic linux-generic-hwe-18.04- linux-generic-hwe-18.04-edge- xserver-xorg xserver-xorg-hwe-18.04- [13:11] is dkms somehow using the running kernel to build? but then why is the path showing the newer version? i just don't get it [13:11] Gotango: i cuold review your journal but not right now, need to leave for ~ 1 hour now. [13:12] TJ-, https://paste.ubuntu.com/p/DWbGZfXPWg/ [13:12] bviktor: earlier I wwas going to point out that the dkms.conf looks wrong, specifically the line "BUILT_MODULE_LOCATION=src//" ... I presume the module built whilst 4.18.0-25-generic was in use? [13:13] TJ-, and I got this in terminal: https://paste.ubuntu.com/p/7yx4F9jbBS/ [13:13] why is that wrong? [13:13] yup, 4.18 was running [13:13] @tomreyn ok thanks [13:13] Aavar: I think I know what you've done! I suspect you've changed ownership of files in your user's $HOME [13:14] Aavar: check that with "sudo find $HOME -not -user $USER -ls" [13:16] bviktor: vermagic should match the kernel version it is intended for [13:16] that i already know. [13:16] how does that make BUILT_MODULE_LOCATION wrong [13:17] I'm trying to pin phpunit so I can get the version from Disco (I'm on Bionic). apt-cache shows it pinned, but it's not upgrading [13:17] bviktor: it doesn't but the "src//" is an obvious typo... one such always makes me suspect others. Where's the repo that you got igb from? I'll build it here [13:18] it's not a typo, it's the result of some ansible parameterization. [13:18] what makes you think 4.18 tolerated src// but 5.0 suddenly won't ? [13:18] if the path was incorrect the build would fail in the first place [13:18] TJ-, there is quite a lot of files owned by root. whould I chown them? [13:18] should [13:18] https://sourceforge.net/projects/e1000/files/igb%20stable/5.3.5.36/ [13:19] i have /mnt/seagate mountpoint for an external hdd, but transmission (debian-transmission) cannot read or write onto that .. how can i change that? what rights i need to set? [13:19] Aavar: yes, do this: "sudo find $HOME -not -user $USER -exec chown $USER:$USER {} \; " [13:19] i mean, using src// instead of src/ shouldn't make dkms use a random previous kernel version for the build, correct? [13:19] intelo: If you followed the imaging instructions to making the USB drive, you can modify it. It is very, very rare to have a read only USB drive. [13:19] Aavar: I suspect then you'll be OK [13:20] bviktor: no it won't my point is such an obvious typo often points to sloppiness elsewhere [13:20] apt-cache policy output: https://termbin.com/y1ec [13:20] i already showed my dkms.conf [13:20] i think the sloppiness is on dkms' part btw [13:21] i mean all i did was apt dist-upgrade [13:21] and the module's been working fine before [13:21] and dkms has created the path for the new kernel version, the newer igb.ko is there. [13:21] TJ yes. its 64 bit kubuntu 18 [13:21] there are no hardcoded versions in dkms.conf either [13:22] TJ-, still the same. I am leaving work now, but I think I will resume later. THank you for your helt so far :) [13:24] Output from `apt list --upgradeable -a`: https://termbin.com/z5c5 shows that phpunit can be upgraded, but `apt dist-upgrade` doesn't upgrade it [13:25] I've followed this guide: https://help.ubuntu.com/community/PinningHowto [13:25] TJ-, https://paste.ubuntu.com/p/zrzjBnYD2D/ [13:26] How can I fix this? [13:26] i can't see any other explanation than dkms being completely stupid [13:26] "DKMS make.log for igb-5.3.5.36 for kernel 5.0.0-23-generic (x86_64)" ... "vermagic: 4.18.0-25-generic SMP mod_unload" [13:27] can ubuntu please stop trying to use Snap? [13:30] make[1]: Entering directory '/usr/src/linux-headers-4.18.0-25-generic' [13:30] why, just why [13:32] is $kernelver the offending line? is that a variable for the _current_ kernel? but then 1) why is it documented like that in the help pages 2) what should be used instead? kernel_source_dir? [13:34] why would you EVER use the current, running kernel for building instead of the target kernel? i just don't understand the use for such a variable... === NickG365_ is now known as NickG365 [13:45] I want to install firefox developer edition from ppa:ubuntu-mozilla-daily/firefox-aurora but the normal firefox has the same package name as well so it installs the normal firefox [13:45] MrSirPotato google the method to specify to use a specific repo. [13:46] MrSirPotato: For software provided by PPAs you will have to refer to the developer for help. Also, have you looked over the PPA to see if there are instructions on how to use it? [13:48] turns out, $kernel_source_dir is also the running kernel. that's just hilarious [13:48] now on 5.0, if i build the module for 4.18, it uses the 5.0 headers [13:48] bviktor: Please refrain from idle chat here [13:48] idle? i'm trying to solve the problem [13:49] but you're right, you're useless, so i'll just figure it out on my own and keep the info to myself too, why would anyone want to know how to solve dkms crap [13:49]