[00:03] hello, where are mouse settings stored for unity? === nudtrobert1 is now known as nudtrobert [07:16] moin [07:18] Saviq: nice bluetooth [07:20] tsdgeos, yeah, with that (once we migrate) we'll actually be green in CI [07:20] but [07:20] tsdgeos, does testSessionBackend pass for you? and for that matter, am I daft or is testLogin1Capabilities there dumb [07:20] Saviq: you have tags in lp:~saviq/unity8/wily-skip-bluetooth [07:20] tsdgeos, d'oh [07:20] plz ./strip-tags.py lp:~saviq/unity8/wily-skip-bluetooth [07:21] Saviq: code not merged upstream yet, is that because we have not migrated yet? [07:22] tsdgeos, yeah, KDE (hint hint) clogged our armhf autopkgtest queue [07:22] tsdgeos, and also boottest unreliable, need someone to kick those [07:22] hmm ok [07:23] it feels weird to have the .deb available but the code not upstream :D [07:24] well, yeah, that's because migration only happens to wily, vivid just publishes in the ppa [07:24] Morning o/ [07:25] tsdgeos, re testSessionBackend ↑, testLogin1Capabilities has a condition to even test anything (dbus iface .isValid()), which seems wrong in the first place [07:25] but on top of that [07:26] it talks to the real login1 interface afaict [07:26] (this test fails for me and failed in migration) [07:31] yeah I'm declaring these tests broken [07:33] they only pass on jenkins becuase of the isValid()s [07:34] there's a fire on my street [07:35] if i disconnect is because i've been told to evacuate [07:35] but it's like 4 buildings right, so should be ok [07:35] yikes [07:35] :D [07:35] only if they as you to evacuate? :D [07:38] well it doesnt' look like at the moment it'll jump to other buildings [07:46] internet cut for a minute === tsdgeos_ is now known as tsdgeos [07:50] Saviq: so, yes testSessionBackend passes here locally [07:52] and the isvalid() returns true to me [07:52] it doesn't for you? [08:03] Saviq: ↑ [08:36] tsdgeos, it does, but I can't hibernate [08:38] Saviq: you mean "in real life"? [08:42] tsdgeos, which shouldn't make the test fail, but does, but regardless, the test should never talk to my real login1 on the system bus [08:42] agreed [08:42] i mean [08:42] it has it's own bus [08:42] since it's run with dbus-test-runner [08:43] or at least should have, no? [08:43] but that's only session [08:43] not system bus [08:43] ah [08:43] ok [08:43] makes sense [08:43] since if not who would be answering those calls :D [08:43] yeah we need a mock [08:43] yup [09:32] ltinkl, https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1498825 [09:32] Ubuntu bug 1498825 in unity8 (Ubuntu) "SessionBackend tests unreliable, talk to real system services" [High,Triaged] [09:32] morning, btw [09:38] tsdgeos: surfaceItem.surface = null in QML, does that equate to a NULL assignment in the C++ object? [09:39] It looks like we're leaking because of bad QML logic, maybe. Which is something I've never experienced [09:39] Also Mir is buggy too [09:40] duflu, http://bazaar.launchpad.net/~mir-team/qtmir/trunk/view/head:/src/modules/Unity/Application/mirsurfaceitem.cpp#L570 [09:41] Saviq: Thanks [09:41] duflu, it's the setter that decides what to do with a null passed to it [09:42] Argh. I was fooled by ::MirSurface != qtmir::MirSurface [09:42] They're different things [09:43] duflu, FWIW null passed from QML is QObject(0x0) in C++ [09:44] duflu: surfaceItem.surface = null means qtmir::MirSurface.setSurface(nullptr) [09:45] ah right, it's cast before it's printed, hence the QObject(0x0) in the log [09:45] hi how come i don't have com.canonical.Unity.panel, because i have trayicon and it doesn't show and for the google chrome it shows [09:46] actually i think it's mirsurfaceitem [09:46] not mirsurface [09:46] Q_PROPERTY(unity::shell::application::MirSurfaceInterface* surface READ surface WRITE setSurface NOTIFY surfaceChanged) [09:46] defines the property and the functions backing it up [09:46] in /usr/include/unity/shell/application/MirSurfaceItemInterface.h [09:46] duflu: ↑ [09:47] i dont have directory shell [09:48] sry unity dir i dont have [09:50] hi how come i don't have com.canonical.Unity.panel, because i have trayicon and it doesn't show and for the google chrome it shows [09:51] Silentlord, where are you looking for this? [09:51] gsettings get [09:52] how about 'gsettings list-recursively com.canonical.Unity' ? [09:52] com.canonical.Unity integrated-menus false [09:52] com.canonical.Unity double-click-activate true [09:52] com.canonical.Unity minimize-slow-duration 800 [09:52] com.canonical.Unity minimize-fast-duration 300 [09:52] com.canonical.Unity form-factor 'Automatic' [09:52] com.canonical.Unity always-show-menus false [09:52] com.canonical.Unity home-expanded 'Expanded' [09:52] com.canonical.Unity minimize-count 100 [09:52] com.canonical.Unity minimize-speed-threshold 100 [09:52] Silentlord, use pastebin next time, also, I believe the panel schema is gone these days [09:53] it was only systray-whitelist there, and that setting's deprecated [09:53] so why only my trayicon is not showing [09:53] i am using qt [09:54] I think the new tray protocol wasn't ported to Qt yet [09:55] sil2100, you were looking into this some time in the past ↑? === nudtrobert1 is now known as nudtrobert [09:55] no i have just search to try and change the whitelist [09:55] Silentlord, the whitelist is deprecated [09:56] it was only there temporarily [09:56] so qt applications cannot work with trayicon [09:56] ? [09:57] hmmm, my memory is a bit weary [09:57] They should, I remember even some recent commits from mitya57 on those [09:58] Silentlord: works for me here [09:59] works for on what app? [10:00] I have an example Qt app from one of the developers that adds an app indicator to the systray, although I see it's using actually GTK to create the tray menu [10:02] no i am using qsystemtrayicon [10:03] Silentlord: on what series are you right now? [10:04] serie? [10:04] series? [10:04] Silentlord: vivid? wily? saucy? etc. ;) [10:04] 14.04 [10:05] Ok, that explains it [10:05] why? [10:05] The support for Qt5 systrayicons was added by mitya in wily... [10:06] Might be a good thing to backport [10:06] Let me look into SRUing it to 14.04 and 15.04 [10:06] but i am not using qt5 [10:06] i am using 4.6 [10:06] 4.6.1 to be exactly [10:07] Ah, ok, then it's not my code in that case - but from what I remember in the old appmenu-qt (used for Qt4) systrays weren't really supported [10:08] so in qt5 works fine? [10:09] Silentlord: in Qt5 it only works currently for 15.10 IIRC, but I'll try to SRU the change to the previous series [10:09] ok thanks === alan_g is now known as alan_g|lunch === pat_ is now known as Guest88770 [12:51] tsdgeos, had a miserable day yesterday trying to have a working setup with a qt built by myself [12:51] tsdgeos, have you recently built your own qt? [12:52] tsdgeos, it used to be a breeze when I did it last year. but now it's pretty broken [12:53] tsdgeos, by that I mean running apps in a qtbase built with -debug -developer-build [12:54] ltinkl, replied to you qtmir/mousePointer comments [12:55] dandrader: hmmm, qtbase or declarative? [12:55] dandrader: i built declarative this month, base maybe 2 months ago [12:55] tsdgeos, both [12:55] didn't have any specially hard [12:56] dandrader: but for base make sure you pass all the long list of options we have in debian/rules [12:56] since otherwise the plugins and stuff won't be found [12:56] dandrader: what problem do you have? [12:56] tsdgeos, qtbase builds fine with -debug -developer-build if you keep it in the dir where you built it, which is fine until you try to build a qtdeclarative against it [12:56] sure [12:56] tsdgeos, qtdeclarative include paths will be fucked up [12:56] but then you don't get the include paths [12:56] correct [12:57] you can try doing some symlinks [12:57] it did work for some of the stuff i tried a while back [12:57] tsdgeos, but then if I built qtbase with prefix=/usr etc and install it on trunk, its cmake files will miss some things [12:58] tsdgeos, so you won't be able to build untiy8 with it, for instance [12:58] yeah i'd go with the symlinks first probably [12:58] build with prefix /myhomedir/something [12:58] tsdgeos, then I resorted to build the package. then it's stream of endless pain [12:58] yeah don't do that, that's very painful === alan_g|lunch is now known as alan_g [12:59] we can try to figure out what's wrong after the call if you want [12:59] tsdgeos, symlinks to workaround the wrong include paths is a good idea [13:00] tsdgeos, at first I was manually changing the header files themselves. but gave up because there are way too many of them to fix :) [13:48] uh, webbrowser-app dies with nouveau_pushbuf_data: Asertion `kref' failed. (wily/unity8) [13:52] nouveau: kernel rejected pushbuf: no such file or directory [14:01] Guest2454, I think we all here run it in integrated intel graphics [14:01] http://paste.ubuntu.com/12530985/ [14:01] dandrader: ah, ok :( this started to happen after the last mesa update 0.11 something [14:01] today [14:02] Guest2454, don't think nvidia driver support for mir is in a good shape. might ask in #ubuntu-mir what's the situation [14:03] dandrader: ok, thanks for help :P === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [15:03] yay it migrated [15:07] hmmm [15:07] bad tags? [15:07] yep [15:07] cleaning them [15:07] http://paste.ubuntu.com/12531652/ [15:08] I think that's my fault [15:08] need to nuke my repo, 'cause it thinks those are fine locally [15:08] ah you have one of those collocated repos? [15:08] may confuse it? [15:12] yeah [15:18] seb128, is desktop-next preview image/seed a thing still? I seem to recall someone speaking a eulogy to it? [15:20] Saviq, yes, though it's not actively being worked on/changed [15:23] seb128, is that by design? we could have a good addition to it (proper greeter) [15:23] Saviq, not by design, snappy desktop isn't just a focus atm [15:23] there is too much to do so need to work in priority order [15:23] if you want to pick it up feel free [15:25] oh there's snappy involved already [15:26] Saviq, oh, right, yes that image is a snappy based one [15:26] we replaced the old desktop-next deb based iso by it [15:26] josharenson, ↑ so for now that task is moot [15:26] if you just want to target the unity8 desktop session I guess just add a depends/recommends to unity8-desktop-session-mir [15:26] Saviq: ok, thought so [15:26] https://launchpad.net/ubuntu/+source/unity8-desktop-session [15:48] tsdgeos_, mzanetti hey, i've rebuilt silo 4 with inline music playback and everything is now dual-landable [15:48] pstolowski, nice! [15:48] thanks [15:48] yw [15:49] awesomeness [15:51] there's so many media-hub & friend silos [15:51] that i don't even know anymore with what to test silo 4 [15:52] silo 47 supposedly [15:53] let me see if i can get both of them and then reproduce https://bugs.launchpad.net/ubuntu/+source/qtubuntu-media/+bug/1496736 [15:53] Ubuntu bug 1496736 in qtubuntu-media (Ubuntu) "currentSource of Playlist gets set to empty" [Undecided,Confirmed] === dandrader is now known as dandrader|lunch [16:28] tsdgeos_, isnt it 55 you want ? [16:29] pmcgowan: 47 is what jim said fixes https://bugs.launchpad.net/ubuntu/+source/qtubuntu-media/+bug/1491732 [16:29] Ubuntu bug 1491732 in media-hub (Ubuntu) "AalMediaPlaylistProvider::mediaCount locked UI over a dbus call" [Undecided,Confirmed] [16:29] i've no idea what 55 does [16:30] 55 seems to be more about the mpris indicator [16:30] that is not what i really want to test here [16:30] right ok nm === alan_g is now known as alan_g|EOD [16:49] silo 4 doesn't really work [16:49] will talk with pawel tomorrow [16:49] * tsdgeos_ waves === dandrader|lunch is now known as dandrader [18:06] dandrader, hey, can you confirm for me that unity8's dependency on qtmir should be > 0.4.6 now? [18:06] >= I mean [18:06] * dandrader checks [18:07] or should current unity8 work with qtmir pre-0.4.6 (so pre-multi-surface)? [18:08] Saviq, it's >= 0.4.6 indeed [18:08] ok so the real problem is that unity-api application is 8 [18:08] but both qtmir and unity8 still are at 6 [18:09] * Saviq notes to make those automagic [18:10] Saviq, in debian/control. yeah. The cmake files are updated though [18:13] where are they even stored in cmake files? [18:13] Saviq, pkg_check_modules(APPLICATION_API REQUIRED unity-shell-application=8) [18:13] Saviq, in the root CMake file [18:14] right [18:21] dandrader, https://code.launchpad.net/~saviq/qtmir/bump-application-api/+merge/272155 and https://code.launchpad.net/~saviq/unity8/bump-application-api/+merge/272156 please [18:23] Saviq, done [18:24] tx, we really need to make those automagic [18:32] yeah === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader