=== duflu_ is now known as duflu [05:25] Good morning [05:25] hi pitti [05:25] hey darkxst [05:32] Laney, any chance you could do a saucy build of trusty webkitgtk on your special ppa? ppa builders still dislike webkit ;( === jhernand1z is now known as jhernandez [09:06] morning [09:07] hey Laney [09:08] darkxst: hmm [09:08] Laney, please! [09:09] they all build for too many architectures [09:09] also I'd rather not have to do it for every release [09:09] did you try disabling debug symbols etc? [09:10] Laney, its only to fix Bug 1163886 [09:10] Launchpad bug 1163886 in webkit (Ubuntu) "software-center crashed with signal 5 with WebKit 2.0+" [Critical,Confirmed] https://launchpad.net/bugs/1163886 [09:10] oh man we should SRU that [09:10] it has been SRU'ed [09:11] but we have a new webkit on the PPA's [09:11] are you already ahead of saucy then? [09:11] ._. [09:11] yes [09:13] hey Laney, good morning [09:13] hey pitti [09:13] how are you? [09:14] Laney: quite fine thanks! last week before holidays, and still so much stuff to do .. :) [09:14] Laney: can merge your new dbusmock template today when you can clean it up (or tell me to do it myself if you don't have time) [09:14] I'll do it, sure, just catching up on weekend stuff atm [09:15] all the things will still be there after holidays ;-) [09:15] yes, and I don't want them to :) [09:36] Laney, so just this once, can you do it? [09:37] I tried disabling debug symbols in the past, but we have dbgsym's enabled on the ppa's [09:42] it makes me edgy, the PPA you're talking about is for doing armhf test builds [09:42] let me think about it [09:49] ubiquity looks very weird with new gtk [09:50] it's like 1.5 wider than it used to be [09:51] Laney, ok, let me know [09:55] xnox: I guess file a bug, attach a screenshot and let larsu know its # [09:56] Laney: i don't exclude that ubiquity is setting style / properties / etc wrong. Ubiquity code did grow since forever, and we do try hard to center re-position the main window. [09:57] * xnox boots it in !chinese locale first [09:57] btw Ubuntu Kylin default wallpaper is pure beauty. [10:06] xnox, I also have had reports that the partition resize slider is broken [10:07] darkxst: i want a screenshot of that. because we did change in saucy the day before release, so it actually might be what it suppose to be. [10:07] * xnox spins up a large vm to test that. [10:08] xnox, these are recent, since 3.10 landed, no slider shows up at all apparently [10:10] xnox, Bug 1260473 [10:10] Launchpad bug 1260473 in ubiquity (Ubuntu) "The slide bar that allow to rezise the partition does not appear" [Undecided,Incomplete] https://launchpad.net/bugs/1260473 [10:10] darkxst: .... and no screenshot [10:11] * xnox is not Neo to read visual data based on green scrolling text ;-) [10:11] no I don;t have a screenshot [10:12] numerous people are still complaining though.... [10:12] although I guess thats about normal for the "maybe" install along-side option [10:14] nah, it really is broken worse than it normally used to be. [10:22] xnox, ok, its hard to know sometimes, we get a lot of noise on g+ and mailing lists, from people that refuse to actually file bugs [10:24] darkxst: who is "we"? [10:25] Ubuntu GNOME [10:25] darkxst: well, i believe it's a gtk+3.10 regression, and as GNOME flavour you are best suited to take it up upstream =) [10:28] ;-) [10:28] you buck passer [10:29] giggley giggley gi =) [10:29] i've been fixing theming bugs in gtk/ubiquity since I started hacking on ubiquity [10:30] i honestly fell like switching to qml/qt, but maybe that simply has it's own set of UI/themeing bugs. [10:30] we should make sure to test ubiquity with new gtks really [10:31] i don't think gtk based flavours will be happy, but then again, they'll just pick up the frontend maintainance, like kubuntu did for the qt frontend since forever. [10:31] Laney: ubiquity - the gtk stress test, caught like 3 grave bugs in glade/gtkbuilder alone.... [10:32] xnox, seems like a larsu thing really, I don't really have much to do with the gtk update (other than wanting it)... themeing is well beyond me [10:34] darkxst: well, if one wants gtk update one gets bugs & regressions that come with it. So please help fix them, instead of complaining about the complaints you get =) [10:34] anyway [10:34] with the width bug, it feels like some rough widgets blows out of proportion. [10:34] which is probably the full-ubiquity-window-width progressbar. [10:35] xnox, not complaining, just forwarding! [10:36] darkxst: that doesn't help either. triaging bugs, usually involves filtering. e.g. marking the "missing diver bug" as confirmed and attaching screenshot. [10:38] xnox: is it upstart-ish to have 'start ' in a pre-start script? [10:39] Laney: yes. [10:39] interesting [10:39] Laney: it's a non-blocking way to do inverse of "start on starting me" in the "somejob" [10:39] xnox, I don't track ubiquity bugs, unless I happen to see them reported via iso tracker [10:39] Laney: such that "somejob" doesn't need to know about all the reverse-alternatives it should be starting on. [10:40] Hmm, I think you'd want it to be blocking usually [10:41] like, you put it in pre-start because you need the job to run your main program [10:42] but I do see noise form the 'report-a-bug-phobes' [10:43] who when you tell them to run 'ubuntu-bug' say Its too hard [10:44] https://code.launchpad.net/~unity-team/ubuntu-system-settings/welcome-wizard/+merge/186862 [10:44] the upstart stuff in there seems smelly to me [10:44] wizard/main.cpp and the user session job [10:44] xnox: please look [10:50] Laney: you really don't want it blocking, otherwise one can easily end up in a dead-lock. [10:50] Laney: if blocking is needed, one should use "start on starting" for one thing to block starting the other. [10:51] how should one job make sure another one is running without the cooperation of that one? [10:52] wait-for-state [10:52] NEVER HEARD OF IT! [10:52] * Laney rtfms [10:52] typically one does "start foo" and then checks that "foo" is start/running, after a timeout [10:53] one can write it manually, or use the magic job wait-for-state [10:53] let me show you a good example. [10:53] So in https://bugs.launchpad.net/ubuntu/+source/plymouth/+bug/967229 [10:53] Launchpad bug 967229 in plymouth (Ubuntu) "Text mode shown briefly with various "cryptic" texts when logging out or shutting down" [High,Triaged] [10:53] i proposed this: f [10:53] https://launchpadlibrarian.net/158609890/message-less-shutdown.patch [10:53] that should be in the cookbook [10:53] it's a bit akward, because I first check that the job exists, since it might not. [10:54] otherwise ti's straight forward. [10:54] e.g. plymouth-shutdown, waits for lightdm to stop [10:54] and rc [016] waits for plymouth-shutdown to start [10:54] such that one gets "cryptic message less" shutdown. [10:55] "flickerless" is a better term, I guess. [10:55] ta, that's interesting [10:55] also note https://bugs.launchpad.net/ubuntu/+source/upstart/+bug/962047 [10:55] Launchpad bug 962047 in upstart (Ubuntu) "document wait-for-state" [Medium,Confirmed] [10:56] might want to change that assignment :-) [10:56] Laney: i think the right solution is for upstart to implement states [10:56] e.g. "when runlevel=2 and networking IFACE!=lo" thus start me when those conditions are satisfied and stop me when they are not [10:57] such that e.g. restart networking works and doesn't bring down the machine to ~= runlevel 1.5 [10:59] well, in this case you want to be able to cause some state transitions to happen too [11:00] when does osk keyboard start? [11:00] apparently mterry knows this is a pile of hacks apparently [11:00] from the comments [11:00] * Laney is spooked by this [11:00] "let the technical debt in and fix it later" is a model which scares me === vrruiz_ is now known as rvr [11:01] 'started unity8' [11:04] Laney: I don't see how wizard will work without unity8, so ubuntu-system-settings-wizard should be: start on started unity and started maliit-server [11:04] oh, but it's starts it's own session?! [11:05] i thought ubuntu-touch-session is gone already. [11:05] no idea [11:05] how is this meant to work? or e.g. how does the existing tour-guide work? [11:05] nah that still exists [11:05] the normal greeter starts it [11:05] I think this is about initial configuration [11:06] cause the inital system setup is suppose to be part of the tour-guide (either before or after it, can't remember from design docs) [11:06] seems like it should be a system job [11:06] it's per user, as per design. [11:06] not system-wide. [11:08] or maybe even just started by unity8 if necessary [11:09] ho hum === dednick is now known as dednick|lunch [14:06] Laney: sorry about https://code.launchpad.net/~laney/python-dbusmock/timedated-template/+merge/198976 [14:06] Laney: that's a bit strange, does timedatectl get the date itself, not from timedated? [14:23] s/date/time zone/ === dednick|lunch is now known as dednick [14:42] pitti: aha, of course, I even noticed that in the second testcase [14:43] It gets the offset from localtime() or similar [14:48] pitti: try now [14:50] Laney: cheers [14:51] Laney: merged [14:51] \o/ [14:51] thanks, I'll leave it alone for now :P [14:53] Laney: poke me when it gets urgent, otherwise I'll finish my ofono template first, and upload towards the end of week [14:54] pitti: I'll need to finish the testcases themselves first, might not be done this week (= year) anyway; take your time [14:55] vim .procmailrc [14:55] * pitti nudges Laney's mouse pointer in position [14:56] uh no, you aren't a shell prompt [14:56] damn you FFM :P [14:56] FFM ♥ [14:57] huh, a missing colon. I wonder how many years that has been there... [14:58] GNOME announce mails have always hit my inbox and I never bothered to look until now [15:16] FFM? [15:16] i thought we'd have FFE by know based on the builtin webcams. (Focus Follow Eyesight) [15:24] oh === penghuan_ is now known as penghuan === BigW is now known as BigWhale === Saviq_ is now known as Saviq === balloons_ is now known as balloons === gatox is now known as gatox_lunch [16:22] Laney: subtle [16:22] I tried unsubtle and it didn't work :P === Saviq_ is now known as Saviq === happyaro1 is now known as happyaron === balloons_ is now known as balloons === sarnold_ is now known as sarnold === ken_ is now known as kenvandine === gatox_lunch is now known as gatox [17:23] Laney, thanks for the review, i just fixed those issues [17:24] great, I'll look again tomorrow [17:35] attente: hi, would you mind having a look at https://bugs.launchpad.net/indicator-keyboard/+bug/1260226 ? [17:35] Launchpad bug 1260226 in Indicator keyboard "Xorg layout switching shortcut doesn't update indicator" [Undecided,New] === hggdh_ is now known as hggdh === fginther` is now known as fginther === csslayer_ is now known as csslayer [20:42] desrt, are we using dconf 0.19 in trusty? [20:47] we ought to be [20:47] i don't know if Laney uploaded it yet, though [20:48] it appears not [20:51] nein [20:52] I will do [20:52] is there a release this time? === Ursinha-afk is now known as Ursinha === mjohnson15_2 is now known as mjohnson15