=== jhodapp|afk is now known as jhodapp === fginther` is now known as fginther === _salem is now known as salem_ [07:30] Saviq: ping [08:11] tsdgeos, pong [08:11] Saviq: was going to ask if the test failed standaolne or inside the full test, but i got it to fail in xvfbtest quite reliably, so nothing :D [08:12] tsdgeos, ok, I couldn't get it to fail in isolation... [08:12] Saviq: i'll have a branch soon, been runing for 10 min here without a fail, i'll MR it soon [08:12] tsdgeos, awesomes [08:13] mzanetti: wops segfault :D [08:13] https://launchpadlibrarian.net/176162572/buildlog_ubuntu-utopic-i386.unity-api_1%3A7.81%2B14.10.20140520-0~134%2B201405201819~ubuntu14.10.1%2Bfix-launcher-tests-with-5.3~test1_FAILEDTOBUILD.txt.gz [08:15] saviq, guess you noticed the unity-api u8 branch landed, but if not you'll need to rebuild the existing u8 silos [08:16] mhr3, yeah, old news [08:17] * tsdgeos is happy that one landed so quickly :) [08:17] tx guys [08:18] Saviq: the fix for me is simply this one ^_^ https://code.launchpad.net/~aacid/unity8/flaky_gsv_test/+merge/220767 [08:18] tsdgeos, coolz, let's see what jenkins says [08:19] I'm worried we multiplied the time to run the tests because of xvfb... [08:19] it takes over 1.5h now :| [08:22] yeah :/ [08:22] who is my ApplicationManager man? [08:22] gerry? [08:23] * greyback suddenly remembers he has something very important to do somewhere else.... [08:23] tsdgeos: what's up? [08:23] greyback: is ApplicationManager.stopApplication blocking? [08:24] greyback: i.e. will it wait until the app is gone from the application manager? [08:24] tsdgeos: no [08:24] greyback: btw where's the ApplicationManager code? [08:25] greyback: so the loop in lines 284 here is "wrong" https://code.launchpad.net/~paulliu/unity8/logout/+merge/216373 [08:25] no? [08:25] tsdgeos: app is removed from the appManager list immediately when you call stopApplication [08:25] ah, so it's removed but the app is not stopped immediately [08:25] tsdgeos: lp:unity-mir src/modules/Unity/Application/application_manager.cpp [08:25] so the loop is not "incorrect" [08:25] tsdgeos: right. As from user's point of view, they just want the app to go away [08:26] they don't care it takes a second or two for it to stop [08:27] greyback: well actually if i''m loging out [08:27] i want to make sure the apps have really stopped [08:27] tsdgeos: greyback: I can fix the crash in unity-api's appman I guess. or is someone on it already? [08:28] greyback: i don't want to lose any work because we shutdown stuff too early [08:28] tsdgeos: I don't think there's a way to wait for that in the current api [08:28] greyback: is there any way to make sure all apps are closed? [08:28] tsdgeos: the "get" part of loop should not be necessary, as once you call stopApplication, that entry has definitely gone from the list [08:28] i.e. something like iterate over all of them, stop them [08:29] and then verify they are all really stopped [08:29] greyback: well, it needs the get to get the appId, no? [08:29] tsdgeos: oh yeah, totally mis-read it. Thought it was checking if app was removed. Duhh [08:30] tsdgeos: I've didn't code AppMan to deal with such logging off situations. On phone, I don't want a "these apps have not shut down, kill them?" message though [08:31] desktop, sure, different story [08:31] greyback: well, it should delay the shut down process enough time to let them close themselves [08:31] tsdgeos: upstart is shutting them down [08:31] upstart is the thing actually managing the applications, not AppMan [08:32] greyback: right, but what happens if unity8 is gone? does the app survive that? [08:32] tsdgeos: you'd better ask an upstart person how apps are managed when a user session is ending. I can't say [08:33] ok [08:33] but AppMan's job is done by then: it told upstart to close all apps [08:33] greyback: yes, the apps still live on when unity/mir goes away, no? [08:34] mzanetti: well they often crash if they try to render, when unity8/mir has gone away. But yes, they live on [08:34] right... yes. there might be issues with crashes, but assuming they'll be fixed eventually [08:35] so we shouldn't really have a problem with this shutdown [08:36] mzanetti: well you might have a point there, might be good to block mir shutdown until apps have disconnected from mir [08:36] otherwise we'll get crash reports for things that weren't really the app's fault [08:37] greyback: fair enough. But wasn't the ideas that we fix the mirclient qpa plugin in a way that we could even restart mir and the same instance of th app will just reconnect? [08:38] mzanetti: eventually yes. That would probably do it [08:38] but a more immediate solution might be what tsdgeos was thinking [08:38] not saying were there yet... [08:38] we get random crashes on shutdown of LVPWH in test :/ [08:38] http://paste.ubuntu.com/7504702/ [08:38] 5.3 at least [08:39] tsdgeos: due to paulliu 's change? [08:39] greyback: no no [08:39] i'm just multi-tasking :D [08:39] sorry [08:40] heh np [08:40] Mirv: hey ho, can you give this one a shot in the ppa? https://code.launchpad.net/~mzanetti/unity-api/fix-appman-tests-with-5.3/+merge/220768 [08:41] mzanetti: ok, merging that on top of the other [08:41] tsdgeos: for iteration 1, I think paulliu's approach will do nicely. Then if we get a series of bugs on apps shutting down badly, we can try a more blocking approach [08:41] greyback: mzanetti: so what's your suggestion for now for that? [08:41] xnox, another problem we noticed: systemtap-sdt-dev is Arch: all, liblttng-ust-dev (Arch: any) depends on the former, so tries to pull :armhf, which doesn't exist [08:42] greyback: ok, i'll review it and see what's going on [08:42] xnox, what would be the correct fix there? make systemtap Arch: any? [08:42] greyback: note that this is more for the desktop i think [08:42] xnox, or adapting the liblttng B-D? [08:42] tsdgeos: well you'd better test it there so ;) [08:42] greyback: ok, i'll leave it for you then :D [08:42] greyback: I would check if there even is an issue with pauls branch in practice, if yes, probably try to fix the crash when mir goes away [08:43] i can't get unity8 to run on my desktop [08:43] greyback: only if that turns out to be too troublesome, add something to allow us to wait for apps to be closed [08:43] tsdgeos: oh poor baby [08:43] :D [08:46] tsdgeos: same here. I got black screen. [08:48] Mirv: yep, looks better i386 passed [08:48] tsdgeos: what I've been told is you have to shutdown your machine. And login with unity8 at beginning. [08:48] tsdgeos: if you logout and then switch, that will be failed. [08:48] paulliu: didn't work here last time i tried [08:49] mzanetti: so it did, awesome! [08:50] tsdgeos: paulliu: first check, install "mir-demos", then switch to VT1, run "sudo mir_demo_server_shell" , then switch to VT2 and run "sudo mir_demo_client_egltriangle" [08:50] *definitely use sudo or you'll be locked out of your machine* [08:52] if you switch back to VT1, you should see a rotating rectangle. That means Mir is working on your machine [08:52] Cimi: lol, just came by this: https://bugs.launchpad.net/unity8/+bug/1246688/comments/1 [08:52] Ubuntu bug 1246688 in Unity 8 "[Launcher] Icons don't fill the UbuntuShape" [Low,In progress] [08:53] ahahah [08:53] mzanetti, tempted to write a new comment [08:53] Saviq: I'll fix that. [08:53] xnox, awesome, thanks [08:53] Saviq: what was example package to test? [08:53] UPDATE: Waiting new ubuntushape rework to land in future centuries [08:53] xnox, unity-scopes-api [08:53] :D [08:53] Saviq: ack. [08:54] Cimi, mzanetti, we still didn't list the features we're missing most from the old shape... [08:54] Saviq: I sort of did to kgunn [08:55] mzanetti, yeah, we need to put it down on paper [08:55] mzanetti, and send up to Pat [08:55] Saviq: but I don't think its an issue that they don't know /what/ to do, the issue more they don't know /when/ to do it [08:55] mzanetti, they asked me for a prio list to backport to the old shape [08:56] uh [08:56] xnox: if it helps, here is my buildlog of that build failing http://paste.ubuntu.com/7504758/ [08:56] mzanetti, yeah, my reaction as well :| [08:56] I saw the fooking shape working... over a year ago in London [08:56] (new) [08:56] yeah... I mean... how hard can it be? [08:56] not to write it, to release it [08:57] when you're a perfectionist like Loïc, it might be really hard... [08:57] Saviq: pete-woods: tah. It just needs marking to be M-A:foreign. I'll upload that & forward delta to debian. [08:57] ok yeah... fair point [08:57] xnox, coolz [08:57] Saviq: ok, I can write you my list [08:58] Cimi, ↑ please do the same (I can find the log of our discussion about it if you want) [08:58] Saviq, mzanetti just sit down in malta on monday [08:58] Saviq, I will write a post it note today [08:58] Cimi, sure, thinking about it before then is still useful [08:58] Saviq, next week we discuss what has higher priority [08:58] Cimi, remember to bring the note over! ;) [08:59] or use the reminders app for it :P [09:00] mzanetti, you? recommending sending your private data to Evernote? ;) [09:00] mzanetti, you should've made a backend for ownCloud ;P [09:00] Saviq: well, given this ends up on Launchpad... [09:01] Saviq: well, they wanted to drop the notes app and make reminders the default notes app. I vetoed that for this reason, yes [09:01] mzanetti, well, it's probably owned by Evernote before that [09:01] (the data) [09:02] mzanetti, good, I did complain about this, too (I'd gladly have only one app, but being forced to do Evernote to use it... nowai) [09:02] yep, exactly my words [09:24] xnox: in addition to the system package, it looks like jsoncpp has an invalid non-multiarch .pc file (http://paste.ubuntu.com/7504825/) [09:24] (I think) [09:25] Saviq: apparmor fix is in https://code.launchpad.net/~xnox/apparmor/xnox-citrain/+merge/220770 [09:25] Saviq: systemtap fix in https://launchpad.net/ubuntu/+source/systemtap/2.3-2ubuntu1 [09:25] xnox, awesome [09:26] pete-woods: that would not be a problem. [09:26] Saviq, if I want to edit tests/mocks/CMakeLists.txt to add tests/Mocks/Infographics/ (which has Infographics.qml and qmldir) what shall I add? [09:27] Cimi, add_subdirectory(Infographics)? [09:27] xnox: okay, it's just that cmake isn't finding that library in the sbuild environment [09:27] pete-woods: although it is incorrect (locations specified is not where things are) but in practice makes no difference, because it should be in the default location.... [09:27] pete-woods: do you have a log? [09:27] xnox: sure [09:27] Saviq, cointains no cmakelists.txt though [09:28] Cimi, why doesn't it? [09:28] Saviq, because it's just a folder to install [09:28] Cimi, no, it's not "just a folder to install" [09:28] Cimi, it's a plugin [09:28] Saviq, there's nothing to compile [09:28] Cimi, that needs to be copied into the binary dir [09:28] Cimi, just copy what other mocks do, it doesn't matter that there's nothing to compile [09:28] Saviq, indeed, so I thought I would have done this inside tests/mocks/CMakeLists.txt [09:29] xnox: there's the lttng stuff (but I expected that, I just wanted to see if there'd be other errors) - http://paste.ubuntu.com/7504845/ [09:29] Cimi, just do what other mocks do [09:29] ok [09:31] Saviq: top-approve https://code.launchpad.net/~aacid/unity8/nomorecardheader/+merge/220659 ? [09:31] tsdgeos, just remembered, don't we have a PreviewHeader now? are tests for it enabled? [09:32] Saviq: they are [09:32] tsdgeos, ok [09:33] ./qmltests/CMakeLists.txt:48:add_qml_test(Dash/Previews PreviewHeader IMPORT_PATHS ${qmltest_DEFAULT_IMPORT_PATHS} ${CMAKE_BINARY_DIR}/tests/mocks) [09:34] tsdgeos, done === vrruiz_ is now known as rvr [09:57] Saviq: since organicgrid test didn't fail in https://jenkins.qa.ubuntu.com/job/unity-phablet-qmluitests-utopic/63/? either i guess we can say https://code.launchpad.net/~aacid/unity8/organiccgridcrash/+merge/220651 "is good" ? [09:58] tsdgeos, yeah [09:59] tsdgeos, so looks like the zoomable tests are the last ones to fix [09:59] ah no, greeter, too [09:59] yep [09:59] and 5.3 regresses some too [10:00] there's some "html" that gets converted to text [10:00] like
to actual newlines [10:00] and stuff [10:00] so tests fail [10:00] oh [10:00] want me to run them all and see what's the diff? [10:00] tsdgeos, I will [10:00] ok [10:02] :| gotta reboot again, usb loop once more [10:20] hi Saviq! I was windering if there were any chances that we could add an easier (shorter and more intuitive) way to close apps in the phone. [10:20] la_juyis, you'll be able to swipe the app away in the right-edge switcher soon [10:21] Saviq, yay! :D [10:21] la_juyis, the app thumbnails in dash are going away, actually [10:21] Saviq, cool. I had thought of either that option, or another option in the meny of the launcher [10:21] (as it is on the computer) [10:21] launcher, long tap, close [10:22] Saviq, awesome, that's great news :D [10:22] Saviq, tnx a lot :) [10:28] la_juyis, cheers [10:38] Saviq: what's the easiest way to upgrade to U? I saw you threw a sed command replacing "trusty" with "utopic" in sources.list, I'm not sure that's the encouraged way to do it though ;) [10:38] karni, do-release-upgrade -d works now [10:38] Saviq: thanks! === salem_ is now known as _salem [10:46] xnox, liburcu-dev has basically the same issue jsoncpp has [10:59] Saviq: no liburcu-dev is not multiarch, and locations in .pc file match actual. [10:59] Saviq: even when cross-compiling, it is valid on ubuntu to link against /usr/lib/libfoo.so since we insure we install the correct arch library. [11:00] Saviq, it is not working, I cannot see what's wrong http://paste.ubuntu.com/7505139/ [11:00] Saviq, make tryInfographics cannot find Infographics [11:00] xnox, sure, but cmake doesn't find the .pc files [11:01] but make -n shows -I /home/cimi/Development/unity8/infographics-new-lightdm/builddir/tests/mocks [11:01] Saviq: then we need to fix our cmake. [11:01] xnox, it seems to have regressed indeed [11:01] and in /home/cimi/Development/unity8/infographics-new-lightdm/builddir/tests/mocks/Infographics I have the files [11:01] xnox, http://paste.ubuntu.com/7505144/ [11:01] xnox, same problem in mir (for android-properties) [11:03] Cimi, you don' [11:03] Cimi, why did you create a C++ module for infographics? [11:03] I didn't [11:03] Cimi, when you just said it's only qml + qmldir? [11:03] Cimi, for the mock [11:03] I just removed the add_library stuff and see [11:03] Cimi, you're not building a library [11:04] I know [11:04] Cimi, .qml files can't be sources for a library [11:04] Cimi, you only need the add_unity8_mock stanza in that CMakeLists.txt file [11:04] just add_unity8_mock(Infographics 0.1 Infographics) ? [11:04] yes [11:05] AH motherfooker [11:05] it's qtcreator that causes my devices to reconnect all the time [11:06] Saviq: [11:06] -- checking for module 'liburcu-bp' [11:06] -- found liburcu-bp, version 0.7.12 [11:07] Saviq, just remove the freakinc ubuntu plugin from qtcreator [11:07] I use qtcreator vanilla and I am much happier [11:07] and doesn't screw qmljsdebugger [11:07] btw [11:08] https://code.launchpad.net/~cimi/unity8/qmljsdebugger/+merge/217474 [11:08] xnox, hmm http://paste.ubuntu.com/7505169/ :/ [11:08] Cimi, yeah, I know, wanted to give it a chance [11:09] Saviq: hm, something fishy is going on. [11:09] Cimi, look at either Ubuntu.Thumbnailer or Unity.Notifications mocks, that's all you need to do [11:09] xnox, indeed [11:10] xnox, that was just "apt-get source" + "sbuild..." in a clean, updated utopic chroot [11:12] Cimi, you can always check if a plugin will import fine with `QML2_IMPORT_PATH=builddir/tests/mocks/ qmlplugindump Infographics 0.1` [11:15] Saviq, the module {} is empty here [11:20] Cimi, well, how about when you point it at tests/mocks directly? [11:20] Cimi, looks like your qml module is just broken is all [11:20] still empty [11:21] http://paste.ubuntu.com/7505215/ [11:22] module Infographics [11:22] InfographicList 0.1 InfographicList.qml [11:22] this is my cmakelist [11:22] ops [11:22] qmldir [11:23] inside I have InfographicList.qml [11:23] more than that I don't know... === MacSlow is now known as MacSlow|lunch [11:26] Cimi, push to a branch [11:27] Cimi, btw, you're missing a "typeinfo" stanza in the qmldir [11:27] ah maybe then [11:27] Cimi, but that won't change [11:28] the empty module [11:29] Saviq, pushed [11:29] Cimi, lp:~cimi/unity8/infographics-new-lightdm ? [11:29] I am going quickly to the doctor, will be back in less 10 mins [11:29] Saviq, yup [11:30] Cimi, woah, that's a close doctor :) [11:32] Cimi, and where's the .qml file? [11:34] Saviq: yeap, found regression in pkg-config. [11:35] Saviq: will fix, should unbreak a whole bunch of things that worked a month a go. [11:35] xnox, yeah great to hear [11:35] pete-woods, ↑ [11:39] hmmmm [11:40] guys can you think of a way to know when ListView finished moving because the currentItem changed? [11:40] i tried listening to moving but seems moving is never set to true in this case [11:42] tsdgeos, hmm sounds like a bug then? [11:42] Saviq: not really [11:42] tsdgeos, only stupid idea comes to mind [11:42] "due to the user either dragging or flicking the view." [11:42] the user ain't doing anything [11:42] name is bad i agree [11:43] should be userMoving :D [11:43] tsdgeos, is to compare delegate's x/y to contentX/Y [11:43] yeah that's what i have [11:43] not cool [11:44] Cimi, btw, InfographicsList is a singleton in the real plugin, you need to make it so in the mock, too - see wiki on QML singletons: http://qt-project.org/groups/qt-contributors-summit-2013/wiki/Evolution-of-the-QML-language [11:44] tsdgeos, ah wait, I think velocity still changes [11:45] tsdgeos, so moving is false, but velocity is != 0 [11:45] interesting [11:45] I think I reported this as a bug at some point [11:46] or at least considered it one ;P [11:46] Saviq, ok [11:46] Saviq: cool, thanks for the update [11:46] Cimi, but anyway, you didn't bzr add the file [11:46] Saviq, my doctor is 2 mins walk, cool eh? :) [11:46] Cimi, are you sure you have that .qml file there at all? [11:47] Saviq, it's pushed [11:47] Cimi, === marlinc_ is now known as marlinc [11:48] Saviq: weeeeird [11:48] tsdgeos, indeed [11:48] Saviq: this is my horizontal velocity log http://paste.ubuntu.com/7505281/ [11:48] why all those 0! [11:49] tsdgeos, oh [11:49] yikes [11:49] let me see if i get that in a simple test case [11:49] that's kind of bad [11:50] Cimi, ah, Module { } can be empty indeed [11:50] Cimi, but that's fine, it means the import works [11:51] Cimi, that works http://paste.ubuntu.com/7505295/ [11:53] Saviq, so I need the singletone thing [11:54] Saviq: pkg-config caused regression 0.28-1 (released in 2014-04-25) in utopic. Fix uploaded 0.28-1ubuntu1. Once that builds, everything should be tip-top. [11:54] Saviq: if you are still in 5.2 can you quick try http://paste.ubuntu.com/7505308/ and see if you get all the 0s? [11:54] Saviq: e.g. unity-scopes-api crossbuilds. [11:54] xnox, awesomes! [11:55] tsdgeos, yes, 0s here [11:55] ok, so it's not a regression [12:14] /food [12:54] tsdgeos, fyi bug #1322615 [12:54] bug 1322615 in Unity 8 "QML test failures under Qt 5.3" [Undecided,New] https://launchpad.net/bugs/1322615 [12:57] mterry, hey, you back already? sure don't want to get more sleep? [12:57] Saviq, I had trouble sleeping last night :-/ [12:58] mterry, sorry to hear that [12:58] mterry, hope it's not split greeter causing it... [12:59] mterry, where did you find the 10% CPU issue in the end? [13:00] Saviq, that is a little odd... So there was that place in qtubuntu where we were checking for "unity8" process name and disabling sensors. Seemed by visual inspection to be 100% the problem. But changing it locally didn't do anything. And adding debug statements to see if it was being run didn't show up! So I figured it was dead code at this point. But I added a branch to look for unity8-greeter too, since wh [13:00] y not in case that code did get run in a different situation [13:00] Saviq, I put it in the PPA and after having that there, the CPU bug was fixe [13:01] Saviq, so somehow I wasn't installing the deb I built right or something when testing it? :-/ [13:01] mterry, right, now that you're saying this... I think I know what's happening [13:02] mterry, you can see app+sensors taking 10% cpu whenever you have any app focused [13:02] mterry, that's most probably the accelerometer sensor for orientation eating away [13:02] mterry, I now remember we disabled sensors for unity8 since we knew it's not going to rotate [13:02] Saviq: that many? [13:02] tsdgeos, yeah it's bad - we're using the wrong sensor and generally it's bad [13:03] Saviq, in order to get InfographicList without brackets, it means is a singleton, right? [13:03] Cimi, yes [13:03] Saviq, but once I added pragma singleton to the file [13:03] Cimi, did you also add "singleton" to the qmldir file? [13:03] nope [13:05] Saviq, yeah tht's what I figured from looking at code, but was confused why changing it and installing the deb didn't work... But I must have done it wrong [13:05] Saviq, but that was a pleasant surprise after updating from PPA :) [13:05] mterry, good [13:05] mterry, so yeah, as you can see from my comment we're real close [13:05] Saviq, so how bad is the "not syncing volume" issue for your? Like percent wise it happening? [13:05] mterry, not any more at all [13:05] Saviq, keeps saying Pragma requires a valid qualifier when I run make [13:06] Saviq, is this a qt 5.3 thing? [13:06] Cimi, no, 5.2 [13:06] weird then [13:06] mterry, once I had ofono sounds gone, but music was playing, so separate issue [13:06] and couldn't reproduce either [13:06] at that point sms wouldn't show up in greeter, too, but can't reproduce either [13:07] Saviq, isn't that the same bug as we have on normal Touch? [13:08] Saviq, "missed call / sms number not recognized in greeter messaging menu" is a known issue, it's on Ted's TDO [13:08] *TODO [13:08] mterry, k [13:08] mterry, which issue known? [13:08] on normal Touch? [13:08] Saviq, well both I thought. The missed notification for calls was happening on normal Touch. Did we fix that? [13:09] mterry, yeah, it was there under the greeter [13:09] mterry, ah now you mention it... [13:09] mterry, dual notifications... one in greeter, one in session, looked weird, too, and we probably need to address that at some point [13:10] Saviq, well... yeah. that's tricky design wise. Not sure how they want that to look (especially since notifications in session could be a different set than on the greeter theoretically) [13:10] mterry, indeed [13:11] Saviq, infographic bubbles have lost their color? Odd [13:11] mterry, I saw that before [13:11] mterry, they were doing that under run_on_device for example [13:11] Saviq, ahah found out [13:12] Saviq, that wiki is wrong, pragma Singleton not pragma singleton, uppercase [13:12] had to read qt code [13:13] Cimi, indeed, it's not documented well :| [13:14] Saviq: reported https://bugreports.qt-project.org/browse/QTBUG-39226 [13:17] tsdgeos, so you can't use it anyway can you... [13:17] Saviq: correct [13:17] Saviq, so are all those issues blockers? I'm wondering if it's possible to squeeze in landing today, but probably not. At least the logind using 3% CPU is probs a blocker. Maybe some of the visual flashing is too [13:17] Saviq: mzanetti: tests will now run slowlier if we do the qmldebug thing [13:18] mterry, no, they're not blockers [13:18] mterry, that's why I said 'file bugs for indicators' [13:18] not that i mind but lowers our "response time" on merging if we need to wait for a ci that runs in 2 hours [13:18] Saviq, yeah but the other stuff I was thinking [13:18] but since i'm not the one doing the merges that's ok to me :D [13:18] tsdgeos, "qmldebug thing"? [13:18] Saviq: https://code.launchpad.net/~cimi/unity8/qmljsdebugger/+merge/217474 [13:18] tsdgeos, isn't that optional? [13:19] tsdgeos, like only for tryFoo? [13:19] no? [13:19] it should be [13:19] I think so yes, let me make sure it is [13:19] tsdgeos, it's only there for qmlscene [13:19] tsdgeos, not for qmltestrunner [13:19] ah right [13:19] sorr [13:19] yep [13:19] nw [13:20] tsdgeos, we're running 2h anyway now... we should not do xvfb in qmluitests on jenkins... [13:20] not as long as we're in a VM there [13:20] Saviq: huh? [13:20] does that make such a difference? [13:20] mterry, I think yeah, the white flash on manta would be a thing to fix - I'd like the "SIM PIN" not fullscreen to be fixed, the blank frame on mako is very difficult to reproduce [13:20] mzanetti, well, sure, everything's going on the CPU instead of the GPU [13:21] mzanetti, so if we have 3D accel there, but not a lot of CPU power, it'll be much slower [13:21] yeah... but... 10 minutes and 2 hours doesn't really compare [13:21] mzanetti, well, bug #1322202 [13:21] bug 1322202 in Ubuntu CI Services "Jenkins logs need timestamps" [Undecided,New] https://launchpad.net/bugs/1322202 [13:21] I've no idea what takes so long [13:21] ok [13:22] mterry: found issues with the left edge animation branch you did [13:22] sorry was offline, updated to utopic, forgot to unpin the wireless driver, wireless driver failed with newer kernel, hilarity ensued [13:22]