=== Class7_ is now known as Class7 [05:46] good morning [05:48] good morning [05:49] Salut didrocks [05:50] hey readlnh, salut jibel [06:00] Hi didrocks, readlnh, jibel [06:05] hey duflu [06:05] duflu: did you get anywhere on the grub delay issue ? [06:07] didrocks, foundations own it now and it seems to be a bionic regression-update so considered even more important [06:07] Whether they fix it this cycle, I don't know. I'm not pushing... [06:08] didrocks, oh, foundations In Progres 8h ago [06:09] duflu: gotcha :) [06:09] Turns out I was not subscribed because I wasn't the original reporter [06:11] well, as long as it's handled in some way… [06:12] new thunderbird has the fix for emoji in email titles to not take 20% of the screen \o/ [06:13] \o/ === pstolowski|afk is now known as pstolowski [07:31] good morning all :) [07:32] hey clobrano! Happy Friday [07:33] hey didrocks , comment ça va? :) [07:34] clobrano: bene bene ;) and you? [07:35] didrocks: :) I'm fine thanks [07:46] Trevinho, hi, why is bug 1794280 set to "won't fix" for gdm? [07:46] bug 1794280 in xorg-server (Ubuntu Cosmic) "gdm doesn't start on a fresh installation of Cosmic Desktop" [Critical,Confirmed] https://launchpad.net/bugs/1794280 [07:47] last comment from Will suggests there is something related to gdm [07:54] morning gang. I wonder what surprises await us today [07:55] No ISO today [07:57] Morning willcooke [07:57] * duflu looks [07:58] Hmm, yes. Although that doesn't bother me because it took all night to download yesterday's ISO [07:58] :) [07:59] hey willcooke [08:01] hello co-conspirators, what evil can we do today? [08:01] morning didrocks, happy friday [08:01] hi Laney [08:01] Laney, are you in London all next week? [08:02] not friday [08:02] but Monday onwards [08:02] beer fest innit [08:02] yes [08:02] oki [08:02] depending on what happens with this race condition, I might see you [08:02] on Monday [08:02] otherwise Weds [08:03] morning Laney [08:04] willcooke: oh, you'll be joining us as well? Sweet! [08:04] O_O [08:04] sil2100, someone's got to be in charge of getting pizza [08:06] willcooke, which race? Didn't yours get fix released last night? [08:06] duflu, I stopped being able to reproduce it yesterday [08:06] good morning desktopers [08:06] and after a lot of faffing about I gave up [08:06] but I'm trying again today [08:06] Sheet [08:06] hey seb128 [08:06] Morning seb128 [08:06] yo [08:07] Hi Laney [08:07] how is the desktop gang today? [08:07] chillin' [08:07] still full of pizza from last nite [08:07] little be of illin' [08:07] yummy :) [08:07] the pub has an arrangement with a pizza place [08:07] duflu, do you have access to a PPTP VPN? [08:07] you text them your order and you can have it in there [08:08] sweeeeeeeet [08:08] willcooke, I don't know, also what that is [08:08] :) [08:09] salut seb128 ! [08:10] willcooke, is your "disconnect from pptp-vpn screws up dns" reproducible or was it a one time issue? [08:10] seb128, yeah [08:10] Oh oh oh oh. Guess what... [08:10] it's baaaaaack [08:11] hehe [08:16] willcooke, sorry, was your "yeah" meaning it's reproducible or a one time? [08:16] gdm3 | 3.30.1-1ubuntu2 | cosmic-proposed | s390x [08:16] seb128: can you remove that pls [08:17] we forgot to add the gjs build-dep in ubuntu2 [08:17] oh, sorry seb128 - it's reproduceable on a VM and real hardwar - but I'm using the same VPN, so could be my condif [08:17] config [08:18] Laney, the other binaries as well or only gdm3? [08:20] willcooke, well, even if it's only some VPN configs, that remains a bug [08:21] seb128: all the arch specific binaries [08:21] did you figure out was is wrong exactly? [08:21] Laney, done [08:21] thx [08:23] didrocks, reading that gdm changelog, wasn't the idea/problem that wayland fails to work with the nvidia driver on some configs? upstream gave up on fixing the problem that we went with a workaround? [08:23] seb128, @ DNS bug. Yeah, fair. Restarting resolved fixes it, so I've tried logged it as a systemd bug ;) [08:24] willcooke, sounds like a good bug to get fixed but not important enough to be a release targetted one, right? I still wonder what goes wrong, you said the DNS config was correct ... maybe resolved getting confusing or something? [08:27] I agree, not a rls blocker [08:30] seb128: it's an issue on wayland gdm starting a wayland xorg user session [08:30] sorry [08:30] "wayland xorg user session" [08:30] wdym? [08:30] starting a wayland user session [08:31] ah, you use wayland for your user session? [08:31] it's not an issue, apparently, if we start a xorg user session [08:31] no [08:31] ah, you mean nvidia [08:31] right, the reason why they blacklisted it [08:31] they disabled it because glx doesn't work properly even if the session itself will [08:31] so it means users on nvidia binary driver that pick wayland session are going to get a non working login? [08:31] as they have today on bionic [08:32] we are speaking about cosmic no? [08:32] right [08:32] * seb128 is getting confused [08:32] what has bionic to do with it now? [08:32] but we are talking about regressino vs non regression [08:32] I was rather trying to figure out fix vs workaround [08:32] the fix isn't going to come on time [08:32] so, the question is: [08:32] I'm not arguing that the workaround/going back to the old state is wrong [08:33] k [08:33] well no question then [08:33] if the fix is not possible [08:33] right, it's the less evil path [08:33] did we give up on being able to fix that? [08:33] and as told in the changelog and patch, this solution was acked by upstream [08:33] I don't think we will be able to, on our side, we rely on upstream [08:33] do they know what the issue is and it's hard to fix? [08:33] or just ENOCLUE atm? [08:33] I did some tests with initial ideas [08:33] but it creates even more issues [08:33] like flicker boots [08:33] or they are too busy to work on it? [08:33] sessionc rashing [08:33] right [08:34] you should join #gnome-shell :p [08:34] willcooke, there is a new iso, published 30 min ago [08:34] but basically, it will be a lot of poking from what I understood, and non trivial changes [08:34] and I can still reproduce the gdm bug on hw [08:35] which is why I decided, after checking with them, to go to the safest path [08:35] meaning "no regression if you compare to bionic" [08:35] didrocks, I joined #gdm, wrong pick :) [08:35] as the issue was then GNOME Shell related, we continued there yesterday [08:35] and did some tests on a branch, but yeah, it even created more issues than it solved [08:35] didrocks, yeah, good choice, I jsut did see any discussing on #ubuntu-desktop about the solution yesterday so the upload came a bit as a surprise [08:35] it makes sense now [08:36] didrocks, thanks for poking and pushing/working with upstream :) [08:36] didn't see* [08:36] yw! :) [08:36] at least, I wanted their blessing on the workaround, that's why I mentioned it in the changelog as well [08:36] right [08:36] I guess they will have more feedbacks when fedora users start to upgrade [08:37] as in few days, we already have 2 people, with different cards, on a pre-released ubuntu version [08:37] I guess maybe that will help getting that prioritized [08:37] btw, we know why GNOME Shell exits [08:38] (invalid drawing context due to nvidia driver) [08:38] the Shell behavior is just "restart" in that case, without any threshold [08:38] right [08:38] we need systemd user session to be able to do better things [08:38] and maybe a xfailsafe mode again [08:38] futur wishlish items [08:39] yeah, that might help [08:39] tearing down Xorg though is a complicated story apparently [08:39] jibel, thx [08:39] ah, also a side effect for people fallbacked to Xorg is that Xorg + GNOME Shell process will stay up [08:40] so the memory gain on having gdm GNOME Shell stopped once logged in doesn't apply in that case [08:41] * seb128 still believes that we should have stayed with lightdm [08:41] but I lost that argument :( [09:02] Trevinho, here's 11megs of logs: https://drive.google.com/file/d/1tTpZYe_cQs14qdRNJUhGNH3hdHam2xum/view?usp=sharing [09:04] jibel: what hw are you using for that? inspiron 11 too? [09:19] His [09:20] Hi* [09:20] hi Trevinho [09:20] morning Trevinho [09:21] jibel: mhmh.. In the contextes I was testing was all more related to lower level issues. Like FB more than gdm itself [09:22] Morning willcooke and didrocks [09:22] good morning Trevinho [09:22] ... and seb128 :-) [09:24] :) [09:24] jamesh, hey, do you think you could nudge alex to comment on that portal fix? would be better to get the fix uploaded today to cosmic if possible [09:24] it's getting late [09:25] seb128: I'll ping him about it. If we don't get a response, we can merge the patch as is [09:26] jamesh, thx === cpaelzer_ is now known as cpaelzer [10:06] Laney, inspiron 15 [10:06] nod [10:06] rls note: do not use dell inspirons (except this one laney has which is fine) [10:06] :) [10:27] seb128: Alex has merged my fix as is: https://github.com/flatpak/xdg-desktop-portal/pull/236. He also suggested a couple of document-portal crash fixes from 1.0.3 we might want to include if we backport [10:27] flatpak issue (Pull request) 236 in xdg-desktop-portal "desktop-portal: don't attempt to export NULL GDBusInterfaceSkeletons" [Closed] [10:38] jamesh, ah, nice, do you want/can you handle that backport today still? If not I can do it if you want [10:39] jamesh, for the other fixes unsure how important they are, we are supposed to focus on/land only release targetted fix now [10:39] jamesh, I would probably delay the other fixes to a SRU, unless they are trivial/make sense to backport as well [10:40] (the segfault is probably worth trying to get fixed on the iso since it impacts the live session, apport triggering there makes for a slower systemd/prompt annoyance) [10:54] k, moving location, going to eat to a place next week and work during lunch today, bbiab === ylli_ is now known as ylel9 [11:30] Trevinho, andyrock, you have a stack of cards "in review" in the 18.10 board, could you update them as fit and move them to the next cycle board if you don't plan to land those changes in cosmic now [11:30] some probably make more sense for d now [11:31] seb128: kk I'll take a look before eow [11:32] andyrock, thx [11:33] andyrock, I somewhat lost track, but what's the stack of the osk fixes that Carlos was supposed to review/help landing in 3.30? [11:33] did they end up being fixed or still ongoing work? [11:33] seb128: I asked him to review it several times, every time he tells me that he's going to take a look [11:34] k [11:34] we could SRU it after C is released [11:35] OSK is broken in several ways in Xorg [11:35] and in Bionic as well [11:36] right [11:36] it's too late for the release now and not critical to the ISO [11:36] this is what happens when upstream has no test infrastructure [11:36] I'm going to nag him nicely about the reviewing, can you give me the url for those waiting for him? [11:36] yeah :/ [11:37] https://gitlab.gnome.org/GNOME/mutter/merge_requests/162 [11:37] GNOME issue (Merge request) 162 in mutter "clutter/x11: Implement keycode remap to keysyms on virtual key devices" [1. Bug, 5. Input, Opened] [11:37] it looks like I need to rebase the changes [11:40] k [11:41] seb128: hi. I'm not going to be able to get to the x-d-p backport today. I've left the info about the patches on the bug report: https://bugs.launchpad.net/ubuntu/+source/xdg-desktop-portal/+bug/1691649/comments/10 [11:41] Ubuntu bug 1691649 in xdg-desktop-portal (Ubuntu Cosmic) "xdg-desktop-portal crashed with SIGSEGV" [High,Confirmed] [11:41] jamesh, k, no worry, thx! And enjoy the weekend :) [11:41] you too! [11:42] thx [11:43] Laney, ^ from a r-t member perspective do you have an opinion on whether including the other fixes listed in that comment? [11:44] the rule for me is: after final freeze I prefer not to upload stuff unless it's important for the ISO, if it can be SRUed then assume that's the way to go [11:46] people can argue on borderline cases if they want [11:46] Laney, right, my point earlier was that having a service segfaulting/triggering apport on the live session (which is what the main portal bug/fix is about) add IO pressure on the live session which is already slow on that side so worth having on the ISO [11:47] unsure how much you are agree with that [11:47] so assuming we want that fix on the ISO and do an upload, do you prefer having only one fix or take the opportunity to bundle the other ones? [11:50] seb128: rebased [11:50] andyrock, thx [11:50] seb128: I prefer keeping changes minimal, a crash fix that impacts the live session sounds good to me [11:50] Laney, great, thx! [12:22] shrug, just managed to make a cup of coffee fall over and drown my laptop keyboard :/ [12:22] well, my laptop rather [12:23] I reversed it and cleaned up a bit, seems to be fine so far, I should perhaps still stop to get the keyboard out and clean up [12:34] seb128, time for a new laptop [12:34] hehe [12:53] willcooke, do you see the boot logo with the red and white dots after a fresh installation with latest iso? [12:53] it's all purple for a while then it flashes and display the ubuntu logo with white dots then gdm, no progress anymore [12:56] ^ this is on hw [12:57] I'm going to do a test install in a bit and can comment on that then [12:57] changing location first, brb [13:01] jibel, I dont want to "break" my Inspiron if I can help it, let me try in a vm [13:11] seb128: ok [13:14] Trevinho, thx [13:16] and I installed the wrong version [13:16] I've been keeping quite a collection of ISOs just in case [13:26] jibel, comfirmed on a vbox vm [13:26] thx, i'll report a bug [13:26] bios -> blank purple screen -> text message about sda being clean -> gdm [13:26] no dot animation either here after updating my inspiron, but I didn't see plymouth/dots for long before so I don't think it's very conclusive on a few boots [13:28] Trevinho, I'm thinking that I will wipe my Insiron and reinstall [13:28] Before I do, anything you want me to do with it in it's current state? [13:29] seb128, andyrock - g-i-s; I used to be able to quit if from the global menu, but now I can't. Is that intended? [13:29] willcooke: mh, we've quite some logging, but hopefully the fails will happen again. I guess you might do a copy of the whole disk to be sure, but that shouldn't change much if it was fresh [13:30] willcooke: nope [13:30] Trevinho, I've got the ISO if I need to reinstall [13:30] I'll wipe it [13:30] k [13:30] willcooke: let me check [13:31] andyrock, I think we spoke about /something/ to do with that menu before, but I cant remember what, but I have been using that menu for a while and I think it stopped working a couple of days ago. I might be wrong though [13:31] willcooke: yeah to not show the wizard on the next setup [13:31] ahh, thats it [13:31] willcooke: I can reproduce, let me give a quick look [13:33] thanks andyrock [13:33] willcooke: is there already a bug? [13:34] andyrock, no, I'll create one [13:34] kk thx otherwise I can do it if you're busy [13:34] I'm re-installing the inspiron atm, I try later [13:34] but if andyrock can confirm it should be enough [13:34] andyrock, just testing, s;fine I will do it [13:37] andyrock, pour vu https://bugs.launchpad.net/ubuntu/+source/gnome-initial-setup/+bug/1797565 [13:37] Ubuntu bug 1797565 in gnome-initial-setup (Ubuntu) "Can't exit via global menu "Quit" option" [Undecided,New] [13:37] good, the langpack export correctly included the fixed gnome-shell translations [13:42] nice! [13:47] * didrocks reloaded with the new langpacks and confirm [13:48] jibel, same on real hardware [13:48] oh [13:49] ha [13:49] yes, on this one I did get the logo for a second [13:49] which I didnt on the VM [13:49] but yeah, same [14:08] willcooke, thx for confirming [14:20] willcooke: upstream is affected by the same issue [14:26] andyrock, still need fixing :) [14:38] jibel: willcooke: make gdm sleep for 5 seconds and see what happens then [14:39] Laney, in the preexec in the unit? [14:40] ExecStartPre=/bin/sleep 5 or something [14:42] That looks like it fixed it [14:42] retryinh [14:42] g [14:43] "fixed" [14:43] hegh [14:43] (ship it!) [14:43] then, we can SRU 20% boot time improvements by changing it to 4 :p [14:43] my next statement is that what you are seeing is not a bug, we are just better at deactivating plymouth now and your boot is fast enough that it's not needed [14:44] neat! [14:44] the boot visual experience is not really nice but it's nothing new [14:44] didrocks, cunning [14:44] like my inspiron spends way more time showing grub/a purple empty screen than anything else [14:45] yeah, the 10s timeout… [14:45] plymouth is kicking in too late to be useful and that part of the boot is quick enough that it seems pointless [14:45] didrocks, you seem quite bothered by then, not the first time you mention it today :) [14:46] seb128: the first one was before you start working :p [14:46] cyphermox is assigned to that grub issue, we might have it fixed before cosmic is out [14:46] didrocks, I saw it in the IRC log yes :) [14:46] seb128: but I was just reacting to you mentioning it in that case [14:46] which is why I said it's not the first time you mention it today :p [14:46] yeah, my comment was not specific to that grub timeout though [14:47] I hope there will be a fix, but TBH, before it was mentioned on the channel, I just found boot time to be a little bit longish [14:47] ah, I thought so [14:47] I saw some movement on the bug earlier today I think [14:47] the grub/a purple empty screen isn't that issue for you? [14:47] no different thing [14:48] ah k, I associated it with that one [14:48] but the thing that jibel and I are talking about is what Laney described ^ [14:48] and the bug I saw some movement on was the grub one [14:49] Has anyone found n-m to be a bit flaky still? [14:49] for wifi [14:49] didn't notice anything, but I don't use much of my bandwidth (emails/uploads/IRC) [14:49] not me, but I'm not doing much out of connecting to my local AP [14:50] "flacky" as disconnecting? [14:50] it's got an IP address, but no traffic can pass [14:50] but I'm also at the edge of wifi coverage here I think [14:50] I'll do some more testing from other places [14:51] man, I really dig journalctl now I'm getting used to it [14:53] I'm definitely uploading the timeout fix today, waiting a bit to see if an unrelated FFe will be approved or rejected to know if I include that change as well [14:53] cyphermox, great, thx! did you have a chance to test the vpn update again to confirm it really doesn't fix your issue? [14:53] shrug, Robert [14:53] thanks cyphermox [14:55] willcooke: wifi> please 'iw dev wlp3s0 station dump' [14:56] https://bugs.launchpad.net/ubuntu/+source/gnome-software/+bug/1797543 [14:56] Ubuntu bug 1797543 in gnome-software (Ubuntu) "Gnome-software - unable to update software - Button Restart & Update, restarts no updates applied" [High,Confirmed] [14:56] :( [14:57] gnome-software displays "updates are available" on a fresh install (after refreshing the index through update-manager) and display the "restart&update" control [14:57] cyphermox, lemme see if I can recreate it [14:58] willcooke: even now, it's telling [14:58] cyphermox, I already rebooted [14:59] if signal is a high negative number, and you have tx failed, etc. it essentially says you're at the edge of signal and stuff sometimes can't pass [15:00] cyphermox, ah cool, thanks! I'll see what I find out [15:01] cyphermox, -73 [15:01] avg -76 [15:01] yeah, that's pretty low; but not terrible [15:02] lets see what packet loss is like [15:02] anyway, when you have issues, this can tell you a lot [15:03] say the signal is really bad (since that varies over time, weather, etc.) and what bitrate is (it'll tell you what speed is negotiated) [15:06] nice [15:06] thanks [15:11] Trevinho, k, reinstalled and can recreate the issue again. [15:11] good and bad :) [15:12] andyrock, if I head to the office on Monday, are you free to come in as well and see if you can help Trevinho with working out what's going on? [15:12] yup! [15:12] super! [15:13] oki, I'll get the first off peak train, so should be in the office by around 1030 [15:13] kk I'm half an hour away from the office by metro so it should not be a problem [15:14] nice, I'll Telegram you when I'm nearly there [15:14] Did you get your pass sorted out already? [15:14] Shall I put your name on the visitors list? [15:15] I know that my pass is waiting for me [15:15] but I had no time to get there and pick it [15:15] what the F [15:15] I ran bzr commit in a git repository [15:16] it didn't tell me to sod off but actually tried to do it [15:16] crashed of course [15:16] my beautiful message commit message, lost :( [15:17] you don't have some bzr-git plugin or something similar? [15:18] probably [15:18] andyrock, I'll add you anyway, no harm [15:18] I never tried to actually use bzr commands in a git repo though [15:18] I used those things to convert and push/pull from one to another [15:18] but you were always in the native one [15:18] madness [15:19] what a world we live in today [15:20] willcooke: seb128: g-i-s fails to close because of this https://gitlab.gnome.org/GNOME/gnome-shell/commit/87a645aa2b96d5db4da4c2ec80bf5b1c6e8cc7d3 [15:21] the commit is not wrong but introduced the bug in g-i-s [15:21] because it doesnt have a close button? [15:22] I guess [15:25] hrm. Not sure what to do there. [15:25] I'm not sure if closable and "don't have the close button" are the same thing [15:26] willcooke: I can try to understand if closable and "don't have the close button" are the same thing [15:26] willcooke: otherwise we can add a custom quit action [15:26] on g-i-s, that should work [15:26] why's this a bug? [15:27] Well, I think being able to quit was a nice feature, and that's been taken away [15:27] andyrock, dont worry about it for now, I don't think it's important enough to look at before release [15:28] also, gnome-shell shows a quit that does nothing [15:28] that part I can agree with (although that won't happen with 3.22 after the app menus are gone) [15:29] but appmenu are going away so.. [15:29] the other part, debatable, there are arguments for having it not have a quit action [15:29] yeah [15:29] We're using it in a manner where you can quit and still have a working machine [15:29] but that's non standard [15:30] even so, Ubuntu might want users to have to go through the g-i-s process for its own reasons [15:31] I would suggest to just hide the quit menuitem [15:31] it's three pages and you can skip them [15:31] yeah, sounds fine to me [15:32] the one from the dock still works by the way [15:32] ha, interesting [15:33] andyrock, ok, you're on the list [15:33] thx [15:34] Laney: https://github.com/micheleg/dash-to-dock/blob/ubuntu-dock/appIcons.js#L685 [15:35] the dock knows how to kill stuff 😂 [15:35] 🔫 [15:52] didrocks, ah, Laney send a report/patch the "language-selector prompt with an non empty list after install" issue we discussed the other day :) [15:52] https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1797579 [15:52] Ubuntu bug 1797579 in ubiquity (Ubuntu) "ubiquity removes packages for non-en_US english locales" [Undecided,New] [15:53] seb128: we discussed/debugged it a little bit together today [15:53] that's probably only part of what you said [15:53] it's even more complex than this [15:53] here is a subpart of the issue [15:53] shrug, somehow I'm missing all the debug discussions this week [15:53] the vpn thing one [15:53] the gdm one [15:53] that one [15:54] :( [15:54] but I checked and it was the case in finale iso for 16.04(.5 at least) [15:54]