=== nudtrobert1 is now known as nudtrobert [06:03] hello [06:12] good morning [06:21] morning [06:45] Good morning [06:56] good morning desktopers [06:56] hey pitti, wie gehts? [06:56] hey seb128, Pici [06:56] pitti* [06:56] tabfail [06:56] re didrocks ;-) [06:57] ça va bien, merci ! j'essaie me lève tard et aller dormir tard, pour la semaine prochaine à Austin [06:57] aller dormir tard va bien, se lèver tard ne va pas :) [06:57] héhé ;) [06:58] snappy sprint in Austin ? [06:58] comment allez-vous ? didrocks, ça va mieux ? [06:58] didrocks: mon dieu, pas encore ! foundations+kernel+security sprint [06:59] pitti: un peu mieux, mais toujours en train de tousser et de me moucher [06:59] pitti: plus de mal à la tête au moins [06:59] pitti: ah, cool :) === nudtrobert1 is now known as nudtrobert [07:59] hey hey [08:00] tjaalton, so far so good, no crash with wily's intel driver [08:00] hey dholbach [08:02] dholbach: good, so next I'll backport the dri2 fixes [08:03] tjaalton, if anything changes I see random explosions, I'll let you know [08:07] sure, thx [08:24] running the medium tests (and docker) directly on the same machine that is used by jenkins works… [08:24] not sure what is changing creating those failures on the jenkins infra [08:24] * didrocks adds some debug [08:42] * pitti fait une réservation à "Steak Manufaktur" [08:42] didrocks: avez-vous mangé le steak de Kobe à Japon ? [08:42] didrocks: on va en essayer ce soir :) [08:42] pitti: oui, un peu (pas un grand steak, juste un petit morceau) :) [08:43] didrocks: ils sont toujours petit [08:43] pitti: on est allé à Kobe, donc c'était l'occasion :) [08:43] didrocks: avez-vous l'aimé ? [08:43] "petits" [08:43] pitti: c'était assez tendre, oui ;) [08:44] on n'a pas vu les vaches se faire masser par contre [08:44] .. parce qu'ils sont chers, mais aussi ... they are said to fill the stomach very well [08:45] didrocks: et tu n'as pas obtenier un massage soit ? :-) [08:46] pitti: non plus, mais je ne suis pas allé à l'abatoire non plus, donc ça va :p [08:47] didrocks: ah oui, je ne voudrais pas vu ça non plus :) [08:48] "voir" [08:49] ;) [08:54] * didrocks loves pycharm [08:54] instead of running a command on the jenkins machine, I did run it locally [08:54] knowing that first command is cd ubuntu-make, rm -r * [08:54] deleted all the uncheckouted work from yesterday… [08:55] but pycharm was opened [08:55] and it has a local history :) [08:55] reverted back -> done \o/ [09:03] good morning! [09:03] hey larsu [09:04] hi didrocks - ça va? [09:04] hey larsu, happy friday! wie gehts? [09:04] how is Italy? [09:04] seb128: good! sunny day in Milan [09:04] how are you? [09:05] larsu: feeling a little bit better (no more headache), but still sick [09:06] didrocks: sorry to hear :( [09:07] larsu: could have been worse, almost lost all my work from yesterday [09:08] larsu, doing good, though I wonder if I'm starting getting a cold as well :-/ [09:08] didrocks: Ctrl+D in bz-do? :P [09:08] seb128: dude watch out! Don't go to close to Lyon ;) [09:08] used git? [09:08] haha [09:08] larsu: nope, just run the jenkins script localy… [09:08] larsu, saw xkcb today? :p [09:08] no [09:08] * larsu checks [09:08] larsu, did dedicated it to me :p [09:08] didrocks even [09:09] seb128: ya, that is *so* true [09:09] :-) [09:09] dholbach likes it as well! [09:11] seb128: just fyi: I'm on the hidpi background problem (it's a bit more effort than I thought - the patch knows nothing about hidpi) [09:12] ok, thanks [09:12] so you can reproduce on your box? that's a good first step ;-) [09:13] ya, no problem [09:13] don't even need to logout/login [09:13] k [09:13] set scale=2; restart nautilus; set scale=1 [09:13] we had workitems to migrate the background rendering to compiz some cycles ago [09:14] but unsure if we should try to go to that this cycle [09:14] feels like we have something that works and it might not be the top priority thing to invest work into [09:14] with risks of creating new problems [09:14] also we probably need to fix the nautilus patch anyway since the oem request is for trusty [09:15] we still need the desktop icon part, don't we? [09:15] right, but that's upstream [09:15] compiz already draws the bg for me [09:15] when I quit nautilus [09:15] does it? [09:15] yes [09:15] do you have the plugin enabled? [09:16] fuck if I know [09:16] I don't even have ccsd installed [09:16] ccsm* [09:16] so probably I have something default [09:16] didrocks: indeed :) [09:16] * larsu dealt with client side decorations for too long [09:17] larsu, if you quit nautilus and change background in ucc, does it reflect? [09:17] seb128: ah, you think about stuck rendering [09:18] seb128: haha no it doesn't :) [09:18] ok, so you don't :) [09:18] but restarting nautilus and then quitting it again works [09:18] right [09:18] no rendering issues at all [09:18] compiz doesn't draw anything [09:18] it's like the image is in the xorg buffer or something [09:18] yep [09:18] right, same here [09:18] dunno why [09:18] I guess because nautilus draws it into the root window [09:18] ok [09:18] yeah, compiz keeps the root window untouched (it's for perf reason) [09:19] wasn't the case in the past [09:19] don't know if we should move it to compiz - that's a question for hikiko-lpt / Trevinho / andyrock [09:19] I'll update the patch for now [09:19] * larsu has the feeling this is a bit high prio [09:19] well, you have a lot of transformation to take into account and such, doesn't feel good for a LTS IMHO [09:19] and we need the patch for trusty anyway [09:19] so… :) [09:19] right [09:20] larsu, yeah, apparently it's blocking dell models to ship with ubuntu [09:20] oh wow [09:21] I don't know why they would change the scaling on the fly [09:22] the issue is that you buy an hidpi machine, log in and go play with the scaling factor and you hit the issue, it's quite visible [09:22] "quite" [09:22] larsu, well, it's not blocking new models I think, they just want the image fixed before doing a new batch of production I think [09:22] -I think [09:23] still defining as rather high priority though ;-) [09:24] hey :) what's the problem with nautilus? I didn't understand: you quit nautilus and you change background and it doesn't update? === hikiko-lpt is now known as hikiko [09:25] hikiko: inly if you change scaling factor (but don't worry, I'm fixing it right now) [09:25] hikiko: pinged you becasue we were discussing switching to the compiz background plugin [09:25] hikiko: also, good morning :) [09:25] oh, cool :D thanks larsu [09:25] +good morning :) [09:37] hikiko, Sam back then did some work on that, e.g https://code.launchpad.net/~compiz-team/compiz/compiz.fix_1159430/+merge/170822 [09:37] see also the bug report listed on that changeset [09:37] Trevinho, andyrock, ^ [09:38] no willcooke today? [09:38] unsure if that's something we should look at doing this cycle or just keep the nautilus patch instead [09:38] larsu, no, he's on vac wed and fri [09:38] c.f monday's email [09:38] ah, right, thanks [09:38] yw [09:41] larsu, looking at the new gedit, did we discuss back then not making the headerbar the titlebar? [09:41] or was that looking too weird? [09:42] you mean toolbar? [09:42] that looked vwry weird because it only has those few buttons [09:43] I mean not using csd basically [09:43] as we do for e.g evince or eog [09:43] right. that looked weird [09:43] more than eog? [09:43] yes, way more [09:43] like evince :) [09:44] evince looks fine [09:44] once you have a document open [09:44] turns out gtk sends configure events when changing the scale factor [09:44] awesome. patch very small [09:44] eog has 2 icons which looks a bit more weird [09:44] hm, indeed [09:44] larsu, great! [09:44] ok, going to upload gedit like that [09:45] it works good enough, ship it [09:45] then we can collect feedback and fix issues [09:45] \o/ [09:46] Morning [09:47] Trevinho: hi from Milan! Very nice city! [09:47] (also, good coffee of which I've already had to much) [09:47] hey Trevinho, how are you? [09:48] larsu: eh eh well, it's quite busy city, I don't love it too much, but indeed there are cool stuff there [09:48] larsu: don't exaggerate with espresso, it can make you nervous :-) [09:49] seb128: oook... [09:49] hehe will try ;) [09:50] Trevinho: I like the business. It's also pretty to look at in the sun right now :) [09:52] seb128: if you want to try: https://code.launchpad.net/~larsu/nautilus/lp1480217/+merge/276228 [09:53] ah you don't have hidpi [09:53] larsu, looking [09:53] no, but I should be able to reproduce by changing the scale factor anyway no? [09:53] though I tried the other day and didn't get it to bug [09:53] like I moved the slider in the display panel [09:53] then restarted nautilus [09:53] then moved back to 1 [09:54] yes, that should work === davidcalle_ is now known as davidcalle [09:57] As for compiz and root window, I'd avoid to change things... [09:57] It's better if nautilus does that [09:57] why? [09:57] (not that I care that much - just curious) [10:01] oh, ssh -t is different from ssh -t -t [10:01] * didrocks tries [10:01] secret hi [10:01] larsu: can try that patch on monday :-) [10:01] Laney: GET OFF IRC [10:01] hahaHAHAHAHA [10:01] i'm checking osm for walking routes [10:01] Laney: also, good morning :) How's Freiburg? [10:01] hey Laney [10:01] bit misty today [10:01] but it is lifting I think [10:02] good luck that it is [10:02] this place is attractive [10:03] how's milano? [10:03] get to explore? [10:03] hey Laney, how are you? [10:03] hey seb128 & didrocks [10:03] larsu, do you still have the patch that was setting the csd off for gedit? [10:03] feeling good! [10:04] * Laney will pass through france tomorrow [10:04] Laney: beautiful. Sunny right now. Sitting in a cafe and work today - exploring is for the weekend ;) [10:04] experience the RER [10:04] seb128: I don't think so sorry [10:04] and I think it was just a quick hack (not checking running desktop etc) [10:04] Laney: RER? [10:05] Laney: argh, "enjoy" [10:05] larsu: Paris' region train [10:05] ah [10:05] larsu, k, no worry [10:06] larsu: in the morning, at some stations, it can be: http://transports.blog.lemonde.fr/files/2014/01/RER-cop-@_ImAnndy_.jpg [10:06] woah! [10:11] oh RER [10:11] how wonderful [10:11] just be happy you're not there in the summer [10:12] the AC never works and it seems that a lot of people in france have an aversion to use of deoderant... [10:12] ...and it's approximately as full as in that picture :) [10:13] desrt: you really do hate France, don't you? [10:13] morning :) [10:14] larsu: i think in this case it's actually true. i think didrocks even said some mocking things on this topic in the other direction... [10:14] heh [10:16] it's not "always as full as in that picture", but yeah, can happen… [10:16] and not really clean [10:16] didrocks: did you mock me once for wearing deoderant? [10:17] ? not that I remember of [10:17] perhaps i misremembered [10:17] I guess you did :) [10:17] yeahhhhh, medium test pass even in jenkins now \o/ [10:17] it's really really ugly though, to force allocation of a pseudo tty there… [10:19] seb128: no packaging branch for glib either? [10:19] * seb128 shakes fist at unity handling of scale factor [10:20] larsu, no, glib we get from Debian in sync most of the time [10:20] thanks [10:20] larsu, sorry, setting scaling factor to 2 was ridiculusly complex due to unity insisting I must not to that and undo my gsettings changes [10:21] oh why that? [10:21] does it not allow it for some resolutions? [10:22] people in this city are exceptionally well dressed [10:22] right [10:22] * larsu likes his window seat [10:22] they limit it so you can't set 4 and never be able to find the controls again [10:22] that is reasonable [10:22] though it has moved the controls off screen for me countless times [10:23] having this instant-apply is also kind of bad in my opinion [10:34] larsu: that's italy... And Milan is the capital of fashion. People, (unfortunately, in some cases), gives lots of importance to the way you dress and express yourself [10:34] seb128: setting the scaling has to be done from unity scaling settings, not from gnome ones :) [10:35] Trevinho, doesn't work [10:35] weeeeird [10:37] Trevinho, ok, in fact it does work, it's just that multimonitor makes it more complex [10:37] larsu, that branch from you doesn't fix it there :-/ [10:37] bah [10:38] does it work for you? [10:38] ya [10:38] :-( [10:38] seb128: yeah, it can be defined globally only if chosen [10:38] seb128: I wonder what's different on your setup... [10:39] larsu, the way I tested is "quit nautilus; set scaling to 2; start nautilus; changing scaling to 1" [10:40] * larsu nods [10:40] I wonder if I messed up moving the patch [10:40] but it looks cirrect in the diff [10:41] *correct [10:42] is there any function that I should check is called when the scaling factor changes? [10:42] configure event [10:43] widget_configure_cb in the patch [10:44] bah! Waiting for glib to build and the tests don't pass [10:44] hm? [10:44] how do we upload packages like this? [10:44] * desrt eyebrow raise [10:44] ERROR: actions - too few tests run (expected 11, got 10) [10:44] ERROR: actions - exited with status 139 (terminated by signal 11?) [10:45] is your dbus environment whack? [10:45] it's usually the export/threaded tests that are failing there [10:45] (on account of dbus) [10:45] whack? no [10:45] stock wily [10:46] and you just run it via 'make check' from cmdline? [10:46] no, building the package [10:46] which is annoying - I want to test my patch [10:46] I'd rather not run the tests [10:46] * larsu wonders if he can ask debuild to do so [10:46] if you have a copy of vi that you can point at the rules, i imagine so :) [10:47] --disable-tests? [10:48] larsu, you can probably edit debian/rules and comment DEB_MAKE_CHECK_TARGET = -k check || true [10:48] going for lunch but I'm going to try the nautilus scale thing a bit more after that [10:49] seems to not call the configure_cb here :-/ [10:49] tanks [11:04] seb128: built and installed the package with my patch (worked from source before) - works as well [11:19] sigh, time is wrong again [12:11] hey desktopers [12:12] seb128, hi, did you still have the MIR of vala on your list -- https://launchpad.net/ubuntu/+source/vala -- it likely creates some build failure already === rickspencer3_ is now known as rickspencer3 [12:32] hrhr, nice. LibreOffice repositories are now replicated live on launchpad: https://code.launchpad.net/~documentfoundation/+git [12:32] (along with freedesktop, github as alternative mirrors to the prestine repo at libreoffice.org) [12:43] Sweet5hark, hey, was the release of 5.0.3 postponed? [12:47] ricotz: http://nabble.documentfoundation.org/minutes-of-ESC-call-td4164709.html [12:48] Sweet5hark, ah alright, did you read my pm? [12:48] ricotz: the one about boost? [12:48] Sweet5hark, while the mentioned commits breaks building with older boost versions [12:49] so the interal orcus is broken [12:49] ricotz: just revert the commit on backports? [12:49] obviously I did that, but doesnt this create issues everywhere? [12:51] seb128: btw I see 0ubuntu4 in xenial/proposed, great! Is 0ubuntu3/wily on track or blocking on anything from me? [12:55] morning === alan_g is now known as alan_g|lunch [13:04] hey andyrock [13:08] ricotz, the needed binaries were promoted this morning [13:08] hey andyrock [13:08] Sweet5hark, https://launchpad.net/ubuntu/wily/+queue?queue_state=1 is an useful url ;-) [13:10] seb128: thx [13:15] seb128, ok, launchpad website seems to lack behind then [13:15] seb128, can you retry https://launchpad.net/ubuntu/+source/gnome-control-center/1:3.16.3-0ubuntu2 [13:16] yes [13:16] launchpad doesn't lack behind, but I promoted the binaries, forgot the source [13:16] thanks [13:56] desrt: you maintain gdbus now, right? Want to have a look at this? https://bugzilla.gnome.org/show_bug.cgi?id=755421 [13:56] Gnome bug 755421 in gdbus "GDBus ignores NO_REPLY_EXPECTED flag in messages, leading to warnings on system bus" [Normal,New] [14:00] seb128: are you still seeing that background scaling error? I cannot reproduce it anymore :/ [14:00] (with my patch, I mean) [14:00] larsu, yes :-/ [14:01] that function is never called for you? [14:01] latest wily/xenial? [14:02] seems it's not, or the output doesn't end up where I expect it [14:02] like me try again with a g_warning === alan_g|lunch is now known as alan_g [14:03] should output to the terminal if no other instance of nautilus was running before (otherwise it just activates that [14:03] ) [14:03] how did you quit it? nautilus -q? [14:04] yes [14:04] nautilus -q [14:05] it's just that I started it again by using the launchpad [14:05] because the scaling x 2 on my lowres laptop makes things not fit well on screen [14:05] so I close most windows [14:06] you can scale 1.1× [14:06] same bug [14:06] hum? [14:06] I though gtk scaling was ints [14:09] it is [14:10] but setting 1.1 changes the screen geometry [14:10] oh [14:10] wallpaper will not line up, but you'll see the same artifacts [14:10] so just writing the key should trigger the callback [14:10] even if gtk widgets are not scaled [14:10] another reason for not having factors other than 2 ... sigh [14:13] larsu, no, widget_configure_cb is never called [14:14] nor on start neither when changing the scaling-factor gtk key [14:15] * Trevinho spamming anyone by renaming rls-w-incoming in rls-x-incoming :P [14:15] I'm launching this only for unity, do you guys want me to do this also for other packages? [14:16] Trevinho, desktop ones yes please, but maybe avoid for other teams? or email ubuntu-devel@ first to say you want to do that and ask if there is any objection [14:23] pitti, could I ask you to take a moment to look at #1511047 it's blocking our landing stuff in xenial [14:23] larsu, that callback never gets called, whatever I do, including connecting other screens or changing resolution [14:24] seb128: yeah same here, just tried rebuilding everything again. Don't know what I was drinking [14:24] * larsu revists [14:25] sorry [14:25] so why is it working with you now?!* [14:28] it doesn't after cleanly rebuilding my branch [14:29] :-( [14:32] hm gtk itself checks for that in realize, configure, and change screen [14:39] larsu, https://git.gnome.org/browse/gnome-shell/commit/?id=5616bbd45b13ec9b76619c2a594cdc9a930edca1 [14:39] not sure if that's useful [14:39] I also tried it but doesn't seem to get trigger when the key change... [14:40] hm interesting [14:40] trying with notify::scale-factor right now [14:41] larsu, no need for both of us to try similar things, I'm going back to the other items on my todolist but let me know if you need me to test something [14:41] seb128: ya of course. Thanks! [14:41] also what other change did you have in your unclean tree? [14:41] maybe that's some upstream update or other hack you had that fixed it? [15:13] bregma: ah, between firefighting and getting ready for travelling to the sprint; in the meantime, can you please forward this to https://bugzilla.gnome.org/ already for upstream review? [15:13] bregma: nevermind, it already is [15:13] :) [15:44] bregma: uploaded [15:44] pitti, many thanks [15:50] ChrisTownsend: hey, could you please apply http://pastebin.ubuntu.com/13010409/ to https://code.launchpad.net/~townsend/unity-lens-music/fix-icon-path.trusty.SRU/+merge/265656 [15:50] otherwise it doesn't build... [15:50] (and ci-train generates a bad changelog) [15:50] Trevinho: Ok, sure [15:50] ChrisTownsend: thanks [15:51] seb128: ok, got it now I think. totally different from what I thought. Pleae retry when you have time (force pull) [15:52] * larsu found out that nautilus' desktop handling code is insanely complex [15:53] Trevinho: Ok, done. [15:54] ChrisTownsend: thanks [15:54] Trevinho: any time [16:16] larsu, seems to not work but I did a local hack and binary cp so doing a clean build now to be sure [16:25] I suppose archive for x-series is open? [16:26] yep [16:26] great [16:26] happyaron, not only suppposed https://lists.ubuntu.com/archives/xenial-changes/2015-October/date.html [16:26] ok quite a lot already, :) [16:26] though it seems there is a slow update to do merges this cycle [16:27] uptake [16:27] I hope Laney gives an hand once he's back next week ;-) [16:28] :) [16:28] larsu, ok, so full build seems to work mostly, at least the bg image is correct now [16:28] but I've an issue with desktop icons [16:28] they stay scaled up until they are moved [16:28] which is a bit weird ;-) [16:29] do you see that as well? [16:31] seb128: I do indeed :) [16:31] this is cool [16:31] :-) [16:31] problem is that upstream has quite some refactorings [16:32] so we can't just look what they do and backport [16:32] yeah, we are looking at updating for xenial [16:32] but we need to fix trusty still... [16:32] ya [16:36] and medium tests runs in prod now \o/ [16:54] ok, time for week-end and rest :-) [16:54] have a good one everyone! [16:54] didrocks, enjoy! [16:54] thanks, you too seb128 === alan_g is now known as alan_g|EOW [18:43] Night! [18:43] Night! [19:55]