[07:44] morning team [08:27] Morning Nafallo [08:28] good mroning desktopers [08:28] hey duflu, how is today? [08:28] Sorry my head is cloudy. Too many years with very little linear algebra makes for a slow duflu [08:28] Morning seb128 [08:28] See above [08:28] How are you seb128? [08:29] seb128: are we getting mutter 3.28? [08:29] I'm good, still tired though [08:29] tjaalton, https://bugs.launchpad.net/bugs/1751070 [08:29] Ubuntu bug 1751070 in budgie-desktop (Ubuntu) "mutter 3.28 transition" [Wishlist,In progress] [08:29] ok [08:30] * duflu crosses fingers [08:30] thing is that the diff to force 8bit rgb doesn't apply to 3.26 since it's missing stuff [08:31] I'm not even sure if it would matter there [08:31] since it doesn't have meta_egl_choose_config() [08:32] Morning desktopers [08:34] hey flexiondotorg [08:34] tjaalton, so that transition nows require us to update the shell stack to 3.27? [08:35] seb128: i'm saying it might not affect mutter 3.26 [08:35] ah [08:35] so back to investigate what's the issue that makes clutter autopkg red? [08:37] mutter disables some clutter tests that clutter itself doesn't [08:37] at least one of them was now failing [08:38] so back to recommending just skipping that one? [08:39] did anyone tested that the new stack doesn't result in a buggy session, with xorg and wayland? [08:39] I've been using it with xorg for weeks [08:40] and did you try to log into wayland session? [08:41] trying now [08:41] thx [08:42] seems to work [08:43] but this is intel, AIUI it would need a driver that uses gallium (radeon, swrast) [08:44] the discussions about that transition seem tedious [08:44] I would just like to know what's the way out [08:44] we need to unblock things, other work needs to land :/ [08:45] looks like moving i965_dri.so aside breaks logging in [08:45] thought it would use swrast then [08:46] guess it tries to respawn the session, and i'm locked out [08:46] can't access anything :) [08:47] :/ [08:50] duflu, we said no bluetooth meeting this week right? (iirc konrad is not there?) [08:51] seb128, sounds right [08:51] good, so I'm staying outside a bit longer :) [08:51] seb128, isn't it cold there? [08:53] duflu, only -7°C :) [08:57] k, I'm changing location, back in 5-10 min [08:57] morning all [08:59] morning willcooke :-) [08:59] 17:08:49 hehe [08:59] 17:08:50 and the release days :-D [08:59] 17:08:59 good luck [08:59] meh [08:59] not that button... [09:01] moin oin [09:11] hey willcooke Laney, how is u.k today? [09:19] hi seb128, some nice snow here over night [09:20] Quite nice this morning, the kids all get very excited when they open their curtains [09:20] :) [09:21] hey willcooke, seb128, Laney [09:22] seb128, hi, do you have a bug associated to the removal of ecryptfs? [09:22] in ubiquity [09:22] yo darkxst [09:22] jibel, hey, no, would that be useful/better? [09:22] hey darkxst [09:23] seb128, no, it was just to add it to the commit log for the sake of documentation. [09:24] willcooke, we must almost be due for some snow, but right now its just a smoky mess outside ;( fires down in the valley [09:24] jibel, the security team is due to write a post about that "this week" since january, we agreed that they are too busy and that it shouldn't block the commit [09:24] darkxst, :(( [09:25] hey seb128 darkxst [09:25] and willcooke jibel too [09:25] hoy Laney, snow? [09:25] poor showing [09:25] :( [09:26] hoping my fuschias are surviving [09:26] * Laney fleeced them up [09:27] seb128, isnt g-s-d using systemd user sessions? and a bunch of other services? really the only r-dep I can find that still uses X-GNOME-Autostart-Delay is probably update-notifier [09:28] ^(re gnome-session patches obviously) [09:28] darkxst, no, GNOME isn't using systemd user sessions [09:28] L_aney is working on that [09:28] see https://trello.com/c/9kiXF8rW/28-systemd-user-session-for-gnome-shell [09:30] tkamppeter, hey, did you see jbicha's question the other day about whether you plan to update system-config-printer in bionic? [09:30] tkamppeter, Debian has a newer version than us [09:31] Trevinho: https://launchpad.net/ubuntu/+source/qtbase-opensource-src/5.9.4+dfsg-0ubuntu3 contains your Qt patch [09:34] seb128, ok will have a look [09:36] tjaalton, so what's the conclusion about clutter/autopkg/transition? [09:38] jbicha, seb128, no, but I planned updating it today, to version 1.5.11 [09:40] tkamppeter, k, that version is already in Debian if that helps you [09:40] tkamppeter, also did you see my comment on https://trello.com/c/kMTto6kl/2-fix-printing-of-filled-pdf-forms-using-qpdf ? [09:41] seb128: xorg and totem work fine even with swrast, i can't test radeon but could add a workaround in /etc/drirc [09:41] needs mesa upload though [09:42] tjaalton, well I don't know if that's needed [09:42] what would you recommend doing? [09:42] accept what we have now [09:42] I'm not following you :/ do you have confidence the update is fine and should be forced in or reserves on whether it might screw users on some videocards/drivers? [09:43] Laney, ^ wdyt? [09:43] iff mutter 3.26 would be affected, then the workaround is to log in to x [09:45] or i'll force-disable the 10bit support in drirc [09:45] there is something wrong in our process if we can't test that case upfront without relying on screwing users and waiting them to report back [09:46] well, does anyone here have radeon? [09:46] no [09:46] (i bet) [09:47] I need to head out for a bit [09:47] if we don't even have a radeon machine in ubuntu engineering then something is indeed wrong and we need to get somebody to buy/expense one imho... [09:47] tjaalton, k, thanks for the update, let's see if Laney has an opinion on what we should do [09:48] I have at the office but it's -15C and won't be going there today [09:48] k [09:48] I'll just add the drirc workaround and upload [09:48] seb128, I have downloaded it. wiull merge it. [09:49] tjaalton, thx [09:49] I don't understand the problem [09:49] If the tests are now wrong, upload clutter to fix or drop them [09:49] tjaalton, thanks, sorry for being annoying about that, I just want things to get unblocked :/ [09:49] If they are right and there's a problem, it should be fixed [09:49] tjaalton, that's going to fix clutter's tests? [09:49] no [09:49] just gnome-shell [09:50] makes sure it doesn't break on wayland [09:50] so is anyone looking at the clutter test issue?. [09:50] I think tseliot was looking into the clutter/cogl thing specifically [09:50] that's blocking the transition still :/ [09:50] ah ok [09:50] tseliot, hey :) [09:51] i'd say screw clutter [09:51] we can't do that [09:51] totem works [09:51] as Laney said, either it's a real issue and we need to fix it [09:51] or it's not one but we understand what's going on and decide it's ok to skip [09:51] well, the mutter patch is for 3.27 [09:52] cogl is way older [09:52] but clutter tests aren't on a GNOME stack [09:52] so that update might still screw standalone clutter users [09:52] I'll upload this after lunch, sorry [09:52] no problem [09:52] enjoy lunch! [09:57] Laney, can we hint gtk to ignore the clutter autopkg so it gets in? [09:57] seb128, Laney: I'm working on it [09:58] or is it going to be blocked by that transition anyway? [09:58] I guess it it picked the proposed version that's because they need to go together? [09:58] tseliot, thanks [09:59] xserver, mesa, and the nvidias should go in together, yes [10:01] but gtk? [10:03] not sure why the test got the mesa stack from proposed [10:03] that indicates that it picked up a dep or something [10:11] I could add the hint though and then we'll see if it works or not [10:11] this isn't the fault of gtk [10:17] * Laney stabs bzr [10:17] "These branches have diverged" never fails to make me want to cry [10:18] :/ [10:19] snowy morning, desktoppers [10:20] ok, there, let's see [10:20] hey oSoMoN [10:20] been out making a snowman? [10:20] Laney, thx! [10:20] hey oSoMoN [10:24] Laney is watching me ⛄ [10:24] hey seb128 ! [10:25] anyone else having issues connecting to canonistack instances? [10:26] the one that I use for my IRC bouncer is off, and the web console is currently unavailable [10:27] they've been doing restarts for spectre [10:28] right, that might be it [10:28] could be [10:47] Laney: hey, how can I use local packages with autopkgtest ? [10:50] Laney: or, how do I test my new packages? [10:50] oh "Run tests from Debian .dsc source package" [10:55] oSoMoN: one for you maybe? https://www.reddit.com/r/libreoffice/comments/80uyx8/different_version_numbers_of_libreoffice_snap/ [10:55] you can give it a .dsc and that will be built and tested or an amd64+source.changes if you already built it [10:56] popey, that's weird, let me read this again and try to make sense of it [10:56] seb128, I reviewed your MP, it needs a small fix and it's all good. [10:57] tseliot: or if you fix in a different package you can give those debs on the commandline too [10:57] like if you wanted to add a new cogl in [10:57] and test clutter [10:57] Laney: how? [10:57] after the source package to test [10:58] see autopkgtest(1) - it calls those 'testbinary' [11:00] popey, that looks like a bug in gnome-software that's displaying info from the candidate channel [11:01] oh wait, it's not a bug, you can click on the channel label in the details to switch channels [11:01] I'm not sure why candidate is selected by default though [11:07] Laney: ok, thanks [11:21] since the update of gnome-terminal this morning in bionic, symlinks in byobu are blinking. Is anyone seeing that? [11:21] my eyes are bleeding [11:23] yes [11:23] someone told me this happens in some other terminals a while back [11:23] popey, is there a bug for it? [11:24] dunno, I only just noticed when you mentioned it [11:24] okay, I'll report one [11:24] popey, thanks [11:24] its not just urls [11:24] symlinks do too [11:24] (here anyway) [11:24] jibel, great, thanks [11:24] yes, what I said. [11:24] oh, i read it as urls.. duh [11:25] and just in byobu, without it it's fine [11:25] dont think it's since this morning btw [11:25] I haven't updated for a few days [11:26] Laney, gtk migrated, thanks :) [11:30] np [11:32] jibel: I don't know what you're talking about but blinking text is a new feature 😉 https://git.gnome.org/browse/vte/commit/?id=38396ef8 [11:34] * popey blinks [11:41] * xnox pukes [11:42] Laney, tjaalton: I suspect the regression in virtualbox that's blocking X is more like the test looking for the vboxvideo.ko, which is not even mentioned in the sourceshttp://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#gdbm https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic/bionic/amd64/v/virtualbox/20180227_201549_c6f06@/log.gz [11:43] as all the modules build here [11:43] bah, we had that back in the 90s with the HTML blink element [11:43] tseliot: the maintainer knows, and said it should be badtested [11:44] Hmm Tilix has an option for Blink Mode, wonder if that'll switch it off, gnome-terminal doesn't seem to have such an option (at least in 17.10 :-) ) [11:44] tjaalton: badtested? [11:44] tseliot: ignored [11:45] tjaalton: ok, it makes sense, looking at the logs [11:45] seb128, is there anything blocking the promotion of libepubgen to main? [11:47] jibel, k, I pushed an update which should fix that and remove the autopilot corresponding file as well [11:47] tseliot: apw (oh hi) usually takes care of that one I think [11:52] oSoMoN, no, just too much to do, I did it now [11:52] seb128, thx [11:52] jibel, thank you for testing&reviewing [11:52] seb128, thanks! [11:52] oSoMoN, yw! [11:59] tseliot, wassup ? [12:00] apw: the tests for virtualbox are looking for a module (vboxvideo.ko) that is not there any more https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic/bionic/amd64/v/virtualbox/20180227_201549_c6f06@/log.gz [12:00] apw: and this is one of the things blocking xorg-server: http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#gdbm [12:01] vbox builds without problems here [12:03] tseliot, that log shows it failing to check the version correctly, but the thing _is_ there [12:03] tseliot, that said that is meant to be fixed ... [12:04] apw: oh, I suppose that comes from virtualbox-guest-dkms, I only tested virtualbox-dkms [12:05] tseliot, anyhow that can be ignored as it has passed otherwise [12:05] tseliot, but i did all the work to make this work, and yet it is not applied, wtf [12:06] heh [12:15] * duflu masters elliptical hysteresis [12:15] * duflu forgot to eat [12:33] tseliot, oh no it is there, just that that one driver is now officially in staging [12:33] oh [12:47] tseliot, and dkms is very very dumb when handling this [12:54] tseliot, hinted anyhow [13:14] seb128, should encrypt-home be dropped from user-setup udeb too? such the d-i does not ask those questions. [13:14] seb128, i know that we may use native-fisesystem encryption in the future, but until then, imho it's better to drop the code alltogether. [13:14] seb128, we have version control to resurrect things. [13:15] xnox, if you want to sure, I didn't want to have to chase kubuntu people to know what they wanted to do so I only removed the option from the installer we use in Ubuntu Desktop [13:16] but it probably makes sense to remove it from d-i is we support that one [13:16] is->if [13:16] seb128, my thinking was that the underlying tech is unsupportable, reliably, any more =/ [13:17] seb128, was this requested by security team, no? i wonder if i should double check with tyhicks [13:17] that might be the case, I'm not going to stop you if you want to go further [13:17] yeah, he said it was going to move to universe [13:17] not to be removed from the archive [13:17] so I guess flavors could still use it if they wanted [13:18] but you are right, it's probably not a good thing to do [13:19] tseliot: https://paste.ubuntu.com/p/z44XsGY8NS/ I thought that might be useful to you [13:19] the remaining xorg-server failure [13:20] pasted it twice in there by mistake, go to line 100 [13:21] soup time, bbbbbbbbbbbbrb === ondra_ is now known as ondra [14:08] Laney: have you consider adding totem to the minimal-remove list? [14:09] and what about ubuntu-web-launchers?? [14:14] don't ask me, I'm not sure what the specification is, I just copied the list didrocks came up with [14:15] ok, I'll add it to the big stack I have for him :) [14:18] Laney: I'll have a look at it, thanks [14:20] jbicha: wi_llcooke could probably assist you [14:21] don't think you'll get away with ubuntu-web-launchers [14:21] nice try though :P [14:24] jibel, thanks for the reviews, sorry for the stupid mistake I didn't do any extra testing after the previous changes, should have I guess [14:24] jibel, going to fix that in a bit [14:26] seb128, np, I fixed them and continuing with the review [14:26] it would be controversial if u-web-launchers was included and it's more noticeable when the Dock & Activities Overview are more empty [14:26] jibel, great, thx [14:28] jbicha, totem seems like a good idea, but I dont want to cause more changes, so if it can be added to the list easily, wfm. u-w-l I think we'd need to keep really. But I will defer to didrocks who is back tomorrow [14:28] controversial to who? [14:28] I think it's as controversial as it's going to get anyway [14:29] there will be complaints and I expect reviewers will notice [14:30] as far as controversies goes, this one is kinda "minimal" though ;) [14:34] sorry I changed location and closed IRC, what's the suggestion? [14:36] jibel, if the rest is fine do you want to merge my branch and commit your fix after it, or me to include your changes? [14:37] seb128: I suggested that ubuntu-web-launchers be added to the minimal-remove list [14:37] ah [14:39] willcooke, don't make didrocks be unhappy on his day back by having to have it to make a call for things he doesn't care about/didn't make the requirement for :) [14:43] seb128, well, I said I didnt want to make more changes, so I've said no already, but if he wants to say yes, thats fine by me [14:43] k, wfm [14:43] :) [14:44] it's easy to tweak that list now, thanks to L_aney [14:50] seb128, I'll test on kubuntu just in case, then apply my changes and merge your proposal [14:51] jibel, k, you might hit the issue raised on #ubuntu-release with the kde installer though [14:52] seb128, oh, okay, I'll just approve and merge then [14:52] thx :) [14:52] and have a look at this kde crash [14:52] jibel, it looks weird, the diff has no mention of that secureboot label [14:53] seb128, the page layout changed, maybe the xml is broken a bit [14:53] i'll see [14:53] don't duplicate too much with tsimonq2 [14:56] jibel, blinking cursors is a new feature in gnome-terminal [14:57] you can turn that off in the preferences [14:58] =/ [14:59] i really hope we can ship a sensible default of non-blinking, for the sake of epileptic users. [14:59] (in your gnome-terminal profile to be exact) [15:00] amano, okay, but blinking symlinks is byobu is not nice. I suppose the default theme of byobu needs fixing [15:00] s/is/in/ [15:01] which reminds me I had to file a bug [15:03] seb128: should I go ahead and upload e-d-s 3.27 late today to avoid FFe paperwork? [15:03] I guess the reason for Egmond Kobleder was that if an app thinks that it has a good reason to blink (e. g. attention grabbing) , let it blink [15:03] jbicha, that's going to add up to the mesa/gnome-desktop transition if we do that right? [15:03] yes [15:03] so please don't [15:03] ok [15:03] tseliot and tjaalton are working on clearing those out [15:04] let them today at least [15:06] jibel: as a workaround, Terminal > Edit > Preferences > Allow blinking text [15:07] speaking of gnome-terminal: is there still a Unity CSD patch applied or does it still lack a symbol for the preferences in the title bar? [15:08] amano: ? [15:09] a preferences button next to minimise/maximise/close [15:10] there should be one, I guess [15:11] I still don't understand [15:11] jbicha, he seems to want an hamburger menu in a csd decoration? [15:11] upstream wishlist [15:13] ah ok, it is still not there upstream then. [15:13] jbicha, thanks, that's fine. I don't spend my day starring at symlinks and can bear it until the default is fixed :) [15:13] (yes, a hamburger menu ;) ) [15:14] amano: you mean headerbars (aka client-side decorations)? [15:14] yes, headerbars ;) [15:15] that's a rough spot with the Terminal maintainer, there is a bit of progress there but it may take a while, see GNOME bug 756798 [15:15] Gnome bug 756798 in general "Gnome-Terminal Header Bar (Client-Side Decorations)" [Enhancement,New] http://bugzilla.gnome.org/show_bug.cgi?id=756798 [15:16] maybe the kind of person who enjoys dealing with low-level terminal emulation coding isn't as quick to adopt new UI stuff :) [15:16] also, terminal is more complex than most GNOME apps [15:17] thanks. I always thought that was hidden in Unity . But I like Egmond's work with 2.27/28. The new preferences window looks very clean 😍 [15:18] it always felt a bit weird to have separate preferences and profile preferences [15:33] s/2.27/3.27 [16:22] Laney, there is no minimal install on kubuntu with your patch. Is it expecteD? [16:23] OTOH it does not crash [16:24] jibel: yeah they don't have it [16:24] it's per flavour === amano_ is now known as amano [16:25] okay. Everything else is going well [16:30] great [16:30] thanks for being a ubi reviewer [16:46] Laney, seb128: I've verified that the clutter test failures were indeed due to the 10bit configs, verified by adding workarounds to drirc for each failing test binary [16:46] now hoping tseliot has a working cogl soon :) [16:48] tjaalton: unfortunately my patch didn't really work. I'm getting the exact same failures [16:49] ok, bah [16:52] tjaalton: well, the autopkgtest command ignored my local packages, that's why [16:53] oh, no it didn't [16:53] I misread [16:58] I'll try something quick.. [17:09] no dice [17:11] actually [17:11] it looks like clutter itself is to blame after all, the tests I mean [17:11] would be interesting to see how cogl tests would fare [17:12] tjaalton, hum, I've difficulty following, we are basically at the same point as yesterday then? [17:12] yes [17:12] clutter is failing and nobody has a clue why/how to fix it [17:15] I'll check if cogl tests (which aren't built) fail [17:15] tjaalton, that puts us in a tricky position :/ [17:15] unsure what to do from there [17:15] ff is tomorrow and thats blocking other work to land [17:16] we could delete that stack from proposed since it seems not ready [17:16] i know [17:16] or pile other things [17:16] or delay other lands, screw schedule, need extra ffe, etc [17:16] none of options is great :/ [17:17] tjaalton, do you need help from somebody and if so what kind of somebody? like what bits are not understandable in the test failure? [17:18] I'll just revert stuff from mesa then [17:19] tjaalton: would that allow libglvnd and nvidia to work? [17:19] unrelated [17:20] ok, good [17:24] kenvandine hey hey [17:25] kenvandine do you know if GsAuth is used just for snapd and ubuntu-reviews g-s plugins? [17:26] we might need to modify it in order to get g-s integrated with g-o-a [17:29] seb128, tseliot: I'll just flip the default in mesa and be done with it for now.. [17:30] no need to revert anything [17:30] andyrock, i really don't know [17:30] andyrock, robert_ancell would probably know [17:30] thanks I'll ping him tonight [17:31] andyrock: It is, yes, but I think there were plans to support donations which might use this [17:31] You should go to #gnome-software and discuss your changes there with hughsie I think [17:31] kk [17:38] mesa uploaded [17:38] tjaalton: yes, we can figure this out later. Thanks [17:40] right, I'll start by checking the cogl tests first [17:42] tjaalton, thx! [18:09] calling it EOD, night all [18:10] same here, it's been a productive day, see ya all tomorrow [18:22] o/ too [19:48] jbicha: is vanilla-gnome still something? because your tasksel changes from last cycle do not work (and aren't supposed to, as-is) [20:04] cyphermox: it's a thing, what's not working? [20:19] vanilla-gnome reminds me of the old days of stracciatella-session to get a "mostly GNOME upstream" (vanilla) session [20:19] jaunty i think [20:20] https://wiki.ubuntu.com/DesktopTeam/Specs/Jaunty/StracciatellaSession [20:20] wow, that was created the very day I joined canonical :) [20:21] nice :) [20:37] jbicha, debconf on gtk3 is landing it seems :) [20:37] shame that all the tests are getting screwed by that font conflict :/ [20:37] update_excuses is a stack of red [20:38] I hope somebody has a magic retry-all script [21:44] jbicha: hey hey [21:44] building the deb of gnome-online-account I get several warnings treated as errors [21:46] I just do "debuild -b" [21:54] I use sbuild [21:54] you could pastebin the build log?