[09:36] mzanetti, poing [09:36] MacSlow: piong [09:37] has anyone run the command saviq mentioned for cleaning remote branches? [09:37] ok, i was not doing it right [09:37] D: [10:18] so we're still not autolanding? [10:30] tsdgeos, we should land manually imho [10:30] tsdgeos, the list of branches is long... [10:35] yeah that's not good [10:35] going to end up with conflicts almost for sure [10:37] Cimi, tsdgeos: I'm not sure about who is driving the landings this week due to the oakland sprint. But should definitively check with Mirv and sil2100 before turning to manual landing or things will probably be even worse [10:37] *But you should [10:38] vila: afaik we already had autolanding enabled, but something in our pkg setup was breaking the autolanders code, but sure, won't land anything manually without cheking with Saviq/kgunn [10:39] tsdgeos, read of the new scene graph with qt 5.2, is this going to improve our performance sensibly? [10:39] tsdgeos: ha, ok, sorry I'm not aware of the details, that was just a general remark to avoid generating more conflicts in a different place ;) [10:40] Cimi: don't know tbh [10:40] someone needs to sit down and see where our graphical bottlenecks are [10:40] tsdgeos: hpmf, I thought I was in #ubunu-ci-eng and misinterpreted the question... [10:40] vila: no worries [10:41] tsdgeos, let's do this soon [10:51] tsdgeos, when will we move to 5.2? [10:51] dec? after release? === _salem is now known as salem_ [11:19] Cimi, tsdgeos: hi guys, what do you mean by 'autolanding'? [11:19] sil2100, jenkins merging approved branches [11:21] Cimi: which projects don't have that enabled? My understanding was that it should have been enabled already everywhere [11:21] sil2100, think that it's not working for unity8 [11:21] Since we're not doing any 'releases' for now, but merging to trunks should be ok [11:21] Ah, hm === MacSlow is now known as MacSlow|lunch === boiko_ is now known as boiko [12:21] mzanetti, cleanup() is called at the end of every qml test and init at the beginning of all? [12:21] Cimi: both are called before/after each test function [12:22] mzanetti, what's better to do? [12:22] mzanetti, setting a value at beginning or resetting at end? [12:22] Cimi: if you need one before all the functions and after all functions use initTestCase()/cleanupTestCase() [12:22] Cimi: i.e. with 2 test functions you have this [12:22] mzanetti, just wondering what's better, think setting at init [12:23] initTestcase(); init(); test_function_1(); cleanup(); init(); test_function_2(); cleanup(); cleanupTestCase() [12:23] it certainly depends on what you want/need to do which one you choose [12:25] mzanetti, I have a test that changes a property [12:25] mzanetti, so looks like the same to me [12:26] not really, no [12:29] Cimi: that 5.2 release would be more a question for Mirv i guess [12:29] sil2100: yeah the autolander of approved branches is somehow broken for unity8 because the package does nasty stuff [12:29] mzanetti, I'll unset the values at the end of the test then [12:30] Cimi: yeah, either that or in cleanup(). but not cleanupTestcase() [12:30] mzanetti, but cleanup will call every time [12:30] Cimi: yeah... [12:30] mzanetti, still cannot see what's the difference in this case [12:31] Cimi: the difference between cleanup() and cleanupTestcase() ? [12:31] mzanetti, I want all tests to run with property a = true [12:31] mzanetti, a test sets a = false === greyback is now known as greyback|break [12:31] what's the difference in having init() ( a = true) or cleanup() (a = true) [12:31] Cimi: yeah. so if you reset it in cleanup() the next test function will have it reset [12:32] Cimi: if you do that in cleanupTestcase() the next function will still have it set to true [12:32] mzanetti, which is the same of having init [12:32] mzanetti, only difference is that init is called 1 time before [12:32] ah... between init() and cleanup(), yeah, that's probably the same in this case [12:32] mzanetti, and cleanup is useless after the last test [12:39] pete-woods: hi! [12:40] sil2100: hi [12:40] pete-woods: maybe you could take a look and see if you know what could be the problem with these failing tests? They're related to HUD - Ted was more assuming these are unity/AP failures, while the unity guys said it looks more like issues with HUD [12:41] pete-woods: https://bugs.launchpad.net/hud/+bug/1244704 [12:41] Ubuntu bug 1244704 in Unity HUD "Some test_hud tests fail on trusty desktop" [Critical,Incomplete] [12:41] So, in the end, no one can take responsibility for this bug and HUD cannot be released ;p [12:48] my interwebs are sloooow [12:49] greyback|break, ping [12:49] apt-get is giving me [12:50] 6326 B/s [12:50] boom! [12:51] * tsdgeos goes for lunch and reboots the router meanwhile === alan_g is now known as alan_g|lunch [13:07] mzanetti, greyback|break, https://code.launchpad.net/~cimi/unity8/fix-1214423/+merge/192868 [13:07] * Cimi -> groceries === MacSlow|lunch is now known as MacSlow === greyback|break is now known as greyback [13:29] dandrader: pong [13:33] greyback, would have time for a chat about the qt scene graph plans regarding input handling? [13:33] dandrader: sure [13:33] dandrader: wanna mumble? [13:34] greyback, yes [13:38] fginther, done [13:38] Saviq, thanks [13:39] sil2100, Mirv, re: bug #1244549 - I'm not sure how to proceed, really - we could kill notify-osd before every test... not sure how else we can make sure it's not holding the DBus name :/ [13:39] bug 1244549 in Unity 8 "unity8 test failures in trusty" [Critical,New] https://launchpad.net/bugs/1244549 [13:40] Saviq: yeah, saw your comment... if this is indeed the issue, then we really need to think of a way of proceeding here - why didn't we have this problem before btw.? As I'm not up-to-date in these parts [13:41] sil2100, luck, probably [13:41] Saviq: since recently in saucy we were always running unity8 tests with unity7 in the background [13:41] Saviq: and we never had such reproducible failures as now, re-runs didn't help [13:41] sil2100, not sure what changed (or if anything) - but something triggers notify-osd to start (it's dbus-activated) [13:42] sil2100, in mediumtests at some point we had it happen when camera app tests were run before unity8 ones [13:44] Interesting, ok, let me note that down [13:45] dandrader: https://qt.gitorious.org/qt-labs/qt-compositor/source/160e770efdd68125ec862b7f5efc94cc97e3935a: [13:46] uploading click package, last step, [13:46] Total uploaded: 100% [13:46] Upload failed due to an Server error. Details: BAD REQUEST code 400. [13:46] any ideas what could be the problem? [13:47] eek.. wrong tab... [13:49] blaroche: #ubuntu-app-devel would be a better place to ask for help [13:49] greyback: thank you === dandrader is now known as dandrader|afk === racarr|sick is now known as racarr|well === dandrader|afk is now known as dandrader === alan_g|lunch is now known as alan_g [14:11] greyback, but if I use one mouse area, then I have to calculate the positions of each click [14:11] Cimi: so? [14:11] greyback, doesn't sound safe as using one for each [14:12] a little piece of math is a lot more efficient that 4 MouseAreas [14:12] greyback, every button or component has it's own mouse area [14:12] if I put a row of buttons I have multiple mouse areas [14:12] because they're independent components [14:12] but this is a simple component: RatingStarts [14:12] ok makes sense [14:12] -t [14:12] ok [14:23] that's weird [14:24] I have Repeater { id: repeater } [14:24] MouseArea { anchors.fill: repeater } [14:24] the mouse area does not get any events [14:24] just like the anchors does not work [14:24] because if I fill the parent it works [14:25] repeater does not have anchors though [14:31] greyback: standup? [14:33] tsdgeos: can you hear me? [14:33] greyback: i can [14:33] tsdgeos: huh, I can't hear anything [14:40] mzanetti, re that test: http://pastebin.ubuntu.com/6318415/ [14:40] Cimi: can you give that one a quick look please? https://code.launchpad.net/~mzanetti/unity8/fix-1245482/+merge/192879 [14:40] I'm pretty sure it's a result of me changing one of the test names. [14:40] It's a bit of a cryptic message though. [14:41] nic-doffay: test_optionselector.py", line 36 [14:41] nic-doffay: that one does:self.assertThat(collapsed.selectedIndex, Equals(0)) [14:41] nic-doffay: but apparently the selectedIndex is 1 [14:41] mzanetti, got it then cheers! [15:20] greyback, I updated and pushed [15:20] Cimi: tsdgeos: I haven't had dedicated time to Qt lately because all of the interation etc. so I've started on 5.2 (qtbase), but I'm kind of doing it on my spare time for now :S [15:20] (now in Oakland) [15:21] or the beta1 [15:21] s/interation/integration/ [15:21] greyback, that's the diff http://bazaar.launchpad.net/~cimi/unity8/fix-1214423/revision/481 [15:22] * Cimi thinks should be starsCount not starCount [15:23] Cimi: did you discuss API with the SDK guys? [15:24] greyback, no [15:24] greyback, I guess can happen in the merge review now [15:24] s/can/should [15:25] Cimi: yeah. Please ask them, they've the experience with defining clean APIs === alan_g is now known as alan_g|tea [15:46] mhr3: hey ho [15:46] mhr3: I renamed an mp3 file in my Music dir and now I have a bunch of broken entries in the music scope [15:46] mhr3: do you know how I can reset that? [15:48] mzanetti, sadly mediascanner should handle that iirc [15:48] mhr3: I assume I can purge mediascanners somehow [15:48] mzanetti, anyway, `stop mediascanner && rm -rf ~/.cache/mediascanner [15:49] thanks :) === alan_g|tea is now known as alan_g === dandrader is now known as dandrader|lunch [15:59] mzanetti, can you see where this is failing now? https://jenkins.qa.ubuntu.com/job/ubuntu-ui-toolkit-ci/1104/ [15:59] I'm missing it. [16:05] nic-doffay: hmm... no idea right now. I guess someone from #sdk might be able to help === racarr|well is now known as racarr [16:18] Saviq: around? [16:32] Saviq: updated https://code.launchpad.net/~aacid/unity8/tabbar_dash/+merge/192505 and answered your comments, tell me if you prefer me to find someone else to continue with the review [16:32] tsdgeos, yeah, I probably won't be that useful this week [16:33] Saviq: ok [16:34] tsdgeos, I'll just try and reply [16:35] Saviq: i can ask mzanetti to do it if you prefer [16:36] ack [16:36] tsdgeos, I'll just answer where there's doubts === MacSlow is now known as MacSlow|eveningb === dandrader|lunch is now known as dandrader === MacSlow|eveningb is now known as MacSlow === alan_g is now known as alan_g|EOD [18:25] * greyback eod [18:25] o/ [18:26] fginther, http://10.97.0.26:8080/job/autopilot-testrunner-otto-trusty/128/consoleFull stuck in dep resolution? === dandrader is now known as dandrader|afk === racarr is now known as racarr|lunch [18:53] Saviq: we've a problem running unity8 AP:s even with the new Mir (in 208) [18:54] Saviq: note, with newest image + mir/unity-mir/platform-api (unity8 not updated to trunk) === dandrader|afk is now known as dandrader [19:00] Mirv, where are you? [19:03] Saviq: in 208, although stop the press :) we just found out the latest release didn't have the last two commits committed two weeks ago to unity8.. it seems better now after manually applying the diff! [19:04] (something started at least) === racarr|lunch is now known as racarr [19:32] Saviq, looking === sergiusens_ is now known as sergiusens === vila is now known as lt-columbo === lt-columbo is now known as vila === salem_ is now known as _salem === ChrisTownsend1 is now known as ChrisTownsend [22:02] * Mirv just published mir (mirserver9/mirclient4), platform-api, unity-mir, unity8, unity-system-compositor, hopefully migrating from proposed as well soonish [22:23] Mirv, awesoooome ;) [22:38] * Mirv notices that there's a slight delay with unity8, but that's ok [23:14] Saviq: note that obviously we're bypassing the process with manual testing since unity8 desktop tests don't run as indicated by that bug report you commented about (notify-osd) [23:16] and well that bypassing of the process resulted in this small messup that we now needed to rebuild unity8 and retest it [23:17] Saviq: so even though like hacks like killing notify-osd and what not might seem overly complicated, if they get us to pass desktop tests it will improve touch releasing as well as we have less points of error from manual work [23:23] Mirv, yeah, I'll prepare a branch I think, to temporarily unblock stuf [23:23] f [23:24] Mirv, no need for a new unity8 release? just a rebuild? [23:25] Saviq: there wouldn't be a need (even a rebuild need) otherwise but the test fixes weren't in yet [23:25] Mirv, k [23:37] Mirv, https://code.launchpad.net/~saviq/unity8/kill-notify-osd/+merge/192972 [23:39] Mirv, I could've used pkill / killall, but I find pidof more reliable somehow... [23:50] Saviq: ok, building soonish