[00:00] thomi, actually yeah, it's showing https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/3118/console here the version is lacking the necessary .1 that fixes everything. hrm [00:01] robru: :-< [00:02] thomi, no, wait [00:02] thomi, was looking at the wrong thing. this one isn't just missing the .1, it's actually grabbing 1.3... [00:02] hmmm, weird, the log is showing that it has the PPA... [00:03] fginther, can you see why this one has the PPA but isn't grabbing autopilot from it? https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/3118/console [00:03] robru: which MP is that for? [00:03] thomi, that's for the friends-app one [00:03] https://code.launchpad.net/~chris.gagnon/friends-app/autopilot-1.4/+merge/193293 [00:03] robru: looks like robotfuel forgot to update debian/control perhaps [00:04] so it's pulling in 1.3 of those packages [00:04] robotfuel: are you able to update & re-push please? ^^ [00:07] sure [00:10] cihelp, how do I create an account here? 91.189.93.70:8080 I only see the option to log in. [00:11] thomi: done, will you re-approve? [00:11] robotfuel: yup [00:11] robru: Can you cancel the old friends-app run please? [00:11] robru: or better yet, kick it off again? [00:12] which run, the CI one? [00:15] thomi, hmmmm, not sure. this CI stuff is more fginther's territory. i cancelled one job but i'm not sure how to restart it [00:16] thomi, re-approved anyway, hopefully jenkins picks it up soon [00:16] robru: the autolanding [00:16] but that's OK, I guess it'll get picked up again in time [00:29] our the maguros having problems again? http://10.97.0.26:8080/job/generic-mediumtests-runner-maguro/ [00:32] thomi: autopilot-trusty-master is stuck. [00:32] :( [00:37] fginther: Can you please make the unity8 autolanding use the daily-build PPA, since otherwise it picks up autopilot 1.3 still [00:39] http://s-jenkins:8080/job/unity8-autolanding/648/ [00:39] thomi, fginther added the hook ↑ [00:39] Saviq: gonna kick it off again then? [00:40] thomi, already running [00:46] thomi, yup, picked up 1.4 [00:46] damn, just got my mako by fedex. I now have mako, maguro, and grouper on my desk. I'm a testing machine! [00:52] thomi, seems we got green on amd64, let's see if devices are happy, too [00:52] \o/ [01:07] fginther: looks like the maguro's are stuck again [01:08] can we disable them please? They're holing up the landing process more than is healthy IMO [01:08] cihelp - maybe smoeone else can help with that? [01:11] yeah, two maguros stuck in flashing again :/ [01:11] veebers, you disabled them yesterday didn't you ↑? [01:12] shame all the jobs that are running will break again [01:12] Saviq: I disabled the job in mediumtests-touch, not the actual devices (as yesterday they were all down) [01:12] looks like it was added back to the jenkins job [01:12] veebers, yeah yeah, two of them still running now (at least until they're taken up to flashing...) [01:13] right, at which point they will have issues, perhaps we should disable it from the job again [01:13] veebers, we've already a significant queue http://10.97.0.26:8080/job/generic-mediumtests-runner-maguro/ :/ [01:14] and now the other two are flashing... wonder if they'll come up... [01:16] cihelp: can https://code.launchpad.net/~xnox/ubuntu-keyboard/libpinyin4/+merge/193859/comments/447347 be retried? the i386 build failure seems spurious. [01:22] thomi, ok, so touch couldn't pass, as we don't have >= 1.4 in debian/control for autopilot :/ [01:23] Saviq: for u*/ [01:23] err, for u*? [01:23] dammit [01:23] thomi, yeah, unity8 [01:23] u8? [01:23] sticky shift key :) [01:23] robotfuel: still awake? [01:24] Saviq: if robotfuel isn't arund I'll propse a new branch with that in it as well [01:24] thomi: which one? [01:25] robotfuel, https://code.launchpad.net/~chris.gagnon/unity8/autopilot-1.4-wip/+merge/193529 [01:25] robotfuel: unity8 [01:25] needs >= 1.4 for the autopilot deps [01:25] robotfuel, autopilot >= 1.8 please [01:25] you mean 1.4? [01:25] 1.4 that is ;) [01:25] Saviq: >= 1.4 :) [01:25] yeah, too many 8s [01:26] oh, looks like the two maguros came back from the dead [01:27] scratch that, the two are still dead, the two that were just flashing seem to have flashed fine [01:27] thomi: Saviq done [01:27] robotfuel, cheers [01:28] * Saviq kicks the build again... [01:28] thanks Saviq [01:37] jeez they're being flashed after every job? no wonder they're dying... [01:43] xnox, the build is retried [01:43] fginther: thanks. [01:43] fginther, two maguros stuck again in flashing :/ [01:44] Saviq, saw that, our attempts to switch hosts and hubs this week have not helped :-( [01:44] Saviq, I'm considering removing the maguro job so they don't get wedged while I'm sleeping [01:45] fginther, yeah, it helped yesterday [01:45] fginther, we're not getting anything over what makos are giving us, either [01:45] Saviq, they can't be fixed without someone physcially resetting the devices :-( [01:45] fginther, you've got +1 from me and thomi on just dropping them from mediumtests-trusty-touch [01:46] Saviq, thomi done [01:46] the current jobs will hopefully trickle down through the two (assuming they don't get stuck, too) [01:46] Saviq, that's my hope [01:47] fginther, we should try the system-image-cli approach I mailed about [01:47] it might be less error prone since it only reboots the phone into recovery and back [01:47] CI will have to revisit the continued use of these devices. If we can't get them to be stable, they do more harm then good [01:48] in earnest we shouldn't even do that, the system.img file could just be duplicated and reused instead of flashing [01:48] Saviq, I saw that, how does that method work with manually installed deb packages? [01:48] fginther, it just replaces the rootfs [01:48] fginther, with the pristine image [01:48] Saviq, hmm [01:48] fginther, what phablet-flash gives you over that is it wipes userdata, too - system-image-cli does not [01:49] but if our tests modify userdata, they're broken anyway [01:49] fginther, but, to not download the image over and over again - we should have two modes - "hard" and "soft" flash [01:50] fginther, "hard" could be just phablet-flash [01:50] fginther, or system-image-cli -b0 [01:50] fginther, we should then copy system.img on the device [01:50] and subsequent "soft" flashes would just replace the system.img used with the copy [01:50] thus not requiring networking to bring the device to clean state [01:51] Saviq, that's reasonable, I'll investigate [01:51] gotta go for a bit, will check back in later [01:51] fginther, o/ [01:54] fginther, fwiw, such "go to dev mode, revert to read-only mode" could be built into system-image-cli itself [01:55] fginther: thanks man [01:56] fginther: do those old jobs need to be kicked off again? [01:58] thomi, alright, I'm EOD. I'll probably be around for a few more hours, but slow to respond. ping if you need anything (but ping fginther about that friends-app merge, i'm not sure how to fix that) [01:58] robru: ok [01:59] fginther: are you going to be around for a bit longer? [02:00] thomi, there are still two maguros in play, so hopefully they'll trickle through [02:02] thomi, looking at friends-app merge... I'd say those are real test failures https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/445/? https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/3122/ [02:07] Saviq, nah, if you click through to the console output, it says "Class 'friends-app' has no attribute 'wait_select_single'." which implies autopilot 1.3 is in use, not 1.4. [02:07] robru, you can see autopilot 1.4 being installed there... [02:08] python-autopilot_1.4+14.04.20131105.2-0ubuntu1_all.de [02:08] Saviq, "00:01:39.805 INFO autopilot:141 - Autopilot Source Version: 1.3.1" [02:08] robru: I bet it's that borked autopilot versin again [02:08] robru, ah, that's on device? [02:08] Saviq, i'm looking at https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/3122/console [02:08] robru, there's no autopilot (>= 1.4) there, then? [02:09] Saviq, in that log i can't find any evidence that it's installing autopilot from the PPA, it seems to just be using whatever's on device already, which would be old [02:09] robru, yeah, missing version dep in the merge [02:09] same we just had in unity8 [02:09] robru, but https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/445/? is using 1.4 - and still one test is failing [02:09] Saviq: which merge? [02:10] thomi, https://code.launchpad.net/~chris.gagnon/friends-app/autopilot-1.4/+merge/193293 afaics [02:10] thomi, the autolanding job didn't use the latest revision [02:10] Saviq: debian/control is up-to-date [02:10] robru: can you double check that the borked AP package is really deleted please? [02:11] thomi, there's a new autolanding running http://s-jenkins:8080/job/friends-app-autolanding/26/ with the correct rev [02:11] thomi, well it would have never been 'deleted', just superceded by the newer one. there's now two newer ones that replaced it, .1 and .2 [02:11] hmm, ok [02:11] Saviq: ahh ok [02:11] so we just gotta wait :) [02:12] thomi, yeah http://s-jenkins:8080/job/friends-app-autolanding/26/console [02:12] thomi, we'll know about unity8 on mako soon enough, too [02:13] Saviq, https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/445/console does look like a 'real' failure, but i don't understand it. friends-app test is practically a no-op, i'm not sure what could possibly fail there. [02:13] robru, self.assertThat(toolbar.state, Equals('')) [02:13] Saviq, thomi: seriously, the friends-app test is so simple, all it does is launches the app, confirms it's launched, and then tests that the toolbar can be revealed. [02:13] robru, but it remains at 'spread' [02:13] robru, maybe it's just missing an Eventually()? [02:14] Saviq, nope, it wouldn't be an Eventually [02:15] Saviq, that test is first testing that the toolbar defaults to hidden, it's testing that the initial state is as we expect [02:15] hmmm [02:15] thomi, do you know anything about toolbar.state? it should be '' when hidden and 'spread' when revealed, right? that's what we test for. but now it seems to start as 'spread' in this test failure [02:16] robru: I don't know, sorry - elopio would be a better person to ask I think [02:16] elopio, ^^ ? [02:16] hey, I'm just getting back. [02:17] I'm not sure about spread. Is that a property? [02:17] robru, maybe it's just that the UITK changed the behavior to be shown initially? === bjf is now known as bjf[afk] [02:17] robru, there was talk of doing that [02:17] elopio: 'toolbar.state' property [02:18] robru, and when you launch the app, it's indeed there for a few seconds [02:18] sounds like we need an emulator for the panel (isn't there one yet?) [02:18] Saviq, hmmm, yeah, it seems so [02:18] robru: the toolbar properties is something you shouldn't be testing on the apps, that should be tested on the toolkit. [02:18] robru, http://pad.lv/1207369 [02:18] Launchpad bug 1207369 in Ubuntu UX "Automatically reveal toolbar when switching pages" [Critical,Fix committed] [02:19] elopio, yeah, i agree, but we weren't sure how to get any mock accounts to be able to test friends itself, so the toolbar is really the only thing we can fiddle with there ;-) [02:19] yeah, at most it should be using the emulator [02:19] on the apps you should just call the emulator toolbar.opened. [02:20] eh... mako dead now? http://s-jenkins:8080/job/generic-mediumtests-runner-mako/3131/consoleFull ;/ [02:21] elopio: so.. I'mleaning towards filing a bug against uuitk, and merging this manually? [02:22] elopio: or maybe we just remove that line from the test? [02:22] thomi, uitk is correct [02:22] robru, thomi: I think you can just remove this test_toolbar. [02:23] robotfuel, still around to do that on your branch? or should i just do it? [02:24] robru: which branch? [02:24] robotfuel, https://code.launchpad.net/~chris.gagnon/friends-app/autopilot-1.4/+merge/193293 that [02:24] that's you, right? [02:24] yes [02:24] robotfuel, yeah, can you just delete the test_toolbar test? it's bad [02:24] as Saviq said, it's failing because now the toolbar starts visible and then it's hidden. [02:25] grr [02:25] thomi, if http://s-jenkins:8080/job/generic-mediumtests-runner-mako/3131/console does not recover, just cancel, re-approve, restart http://s-jenkins:8080/job/unity8-autolanding/649/ [02:26] past 3am here, gonna go now [02:26] if I find the above to still be hanging in the morning, I'll tackle it first thing [02:26] Saviq: thanks man [02:26] this is starting to get seriously frustrating :-( [02:26] it does [02:27] * Saviq wants the emulator [02:27] have fun peeps o/ [02:28] I've just approved the file manager app. It has passed on autopilot-trusty-master [02:28] robru: is there the friends app passes locally. [02:28] elopio: thanks man [02:28] robotfuel, yeah, you probably don't have the updated ubuntu-ui-toolkit yet [02:29] maguros are coming back :) [02:29] keyboard is running atm. [02:31] robru: can I get a link to a failing test? maybe there is a simple fix? [02:32] robotfuel, yeah, the simple fix is to delete test_toolbar. it's a stupid test, it's not testing the app but tests ui-toolkit, and ui-toolkit changes, to the test is broken. it doesn't need to be fixed, just deleted [02:33] robotfuel, the test failure is somewhere in the scrollback, sorry I don't remember which one. [02:35] robru: done [02:37] robru: I don't even think we should have ported that one to 1.4, there are no real tests. [02:43] down to one maguro [02:45] and down to two makos [02:49] robotfuel, well, the idea is that we have that stub there so that it will be easier to write more tests in the future. it definitely needs to be ported to 1.4 [02:49] robotfuel, and thanks [03:07] fginther: any idea WTF is going on here? https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/2982/console [03:11] robru: fginther: Are either of you able to kick off the UI toolkit autolanding? It seems it's not enabled by default [03:13] well, I may have done it correctly, maybe not: http://s-jenkins:8080/job/ubuntu-ui-toolkit-autolanding/427/ [03:38] looks like some new failures with the new image [03:38] webbrowser_app.tests.test_title.TestWindowTitle.test_window_title for one [03:41] thomi, hey [03:41] hey fginther [03:41] thomi, looking at phone-app now [03:42] so... I'm not a happy man- the CI hardware seems totally fucked from my POV :( [03:42] soooo many failures, and one job that seems permenantly stuck here: http://10.97.0.26:8080/job/generic-mediumtests-runner-maguro/ [03:43] and I'm seeing some strange otto-related issues here as well: https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/453/console [03:48] fginther: TBH, It must be getting late for you, I think the problems we're seeing are more systemic, and it's probably best if you get some sleep and attack this tomorrow [03:49] I'll send a status update tomorrow [03:49] err, today [03:49] thomi, I see you found the package name issue with friends-app [03:50] thomi, is phone-app depricated? [03:50] I have no idea... I would hope that the project would be deleted, if that were the case [03:51] It's not too late yet, I'll try to help if I can.. [03:52] OK, I'm just going to approve all the branches, and hope that a few make it through [03:52] thomi, phone-app is not on my device and hasn't been updated in months [03:52] OK, I'll delete that then [03:52] ugh, why wasn't it deleted :-( [03:53] yeah [03:53] Words cannot convey my emotional state right now... [03:54] fginther: quick query: says in this log that there was a bunch of files archived (i.e. the ogv) but the jobs Build artifacts only contains the xml: http://10.97.0.26:8080/job/generic-mediumtests-trusty/488/console [03:54] fginther: oh, is that the log of the child job that was fired off, because I've found the files on that job [03:54] veebers, right, walk down to the child job, they should be there [03:55] fginther: awesome thanks [03:58] thomi, agreed, we have many systemic issues [03:58] robru: are you able to top-approve this please? https://code.launchpad.net/~chris.gagnon/webbrowser-app/select-many-noorder-ap1.4/+merge/194031 [03:59] thomi, hmmm, oddly not. [03:59] thomi, osomon should be up soon i think [04:01] lool, around? i think you can approve this for us: https://code.launchpad.net/~chris.gagnon/webbrowser-app/select-many-noorder-ap1.4/+merge/194031 [04:04] elopio: still around? [04:10] thomi: will that fix the error I mentioned earlier? [04:11] plars: I'm not sure which error that is? [04:11] plars: oh [04:11] webbrowser_app.tests.test_title.TestWindowTitle.test_window_title failing in the smoke tests on the latest image [04:11] plars: I'm not sure. was the image built with autopilot 1.4? [04:11] was supposed to be... checking [04:12] thomi: no, looks like 1.3 still [04:12] so, nm [04:12] thomi, ok, I'm out for reals. I think friends-app will land shortly, anything else i'll help out with in the morning. [04:13] robru: ok, thanks for your help [04:13] thomi, you're welcome === Ursinha_ is now known as Ursinha [04:38] thomi, ubuntu-rssreader-app and ubuntu-weather-app have merged [04:38] wooo! [04:40] fginther: ui toolkit merged as well - sweet! [04:48] thomi, sudoku-app doesn't look good. Not sure how important it is as the results are not included in the dashboard [04:48] thomi, also ubuntu-filemanager-app fails, but it also fails on the dashboard, appear to be some existing issues [04:48] fginther: OK, can we manually merge the filemanager app branch in that case? [04:49] thomi, yes [04:49] do you do that, or...? [04:49] thomi, I'll take care of it [04:49] thanks [04:53] thomi, done [05:12] let's see... [05:15] * Mirv looks at the redness at cu2d [05:18] thomi: I'm here. [05:19] elopio: it's ok, I sorted it [05:20] thomi: ok, it seems we are almost done. [05:20] elopio: well, I doubt that very much :) [05:22] thomi: well, 10 branches landed, that's almost all of them. Where do you need help now? [05:23] elopio: I imagine there will be further issues we need to resolve - tests failing ec. [05:23] *etc [05:23] but I'm done for today [05:25] I see tomorrow as the day we can start cleaning tests, so it's not a problem if some fail, those will be the place to start. [05:26] yeah [05:34] thomi, good luck on the remaining work for today, I'm off to bed. [05:34] cheers! [05:34] I'm off to cook dinner, will BBL [05:51] lucky shot, the mediaplayer-app works fine on autopilot 1.4. [05:51] now there are no apps TODO. [05:58] QA stack now green again [05:59] (needed update to xpathselect1.4 in package list) [07:08] elopio: awesome! [07:08] Mirv: so, there's nothing more I need to do urgently today? [07:14] thomi: if there is, we'll find out about it only later today. I'm now trying to ensure trevinho's unity branch is ok and will then merge it manually if it is [07:14] if it works properly it will resolve many of the remaining problems [07:15] thomi: but already having autopilot itself, unity8 and ui-toolkit in releasable state is promising [07:16] Mirv: OK, I'll EOD then. Can you email me with whatever you find please? [07:17] thomi: yep, I'll keep up with the thread [07:17] Mirv: oh, and if you can find someone with the permissions to merge the webbrowser-app branch, that'd be awesome [07:17] thanks [07:17] thomi: ok, I'll see about that [08:50] Mirv: ping [08:51] Mirv: I've lost track of what needs to happen regarding https://wiki.canonical.com/UbuntuEngineering/CI/IncidentLog/2013-11-04-otto-outage [08:51] Mirv: we've reverted the kernels but is someone looking at the corresponding bug ? [08:51] Mirv: is there even a bug filed for that ? [08:57] Mirv: so the cmake bit is still not merged in? We might need to help finding someone who can top-approve that branch then [08:57] vila: I don't know what was worked on that, but the autopilot machines currently seem to be happy. the last I remember was that it was asked to revert the kernel. [08:57] Since it's slowing things down [08:58] sil2100: yes, I suspect didrocks can help us here, that's why I sent the e-mail. [08:58] hey Mirv [08:58] hey didrocks [08:58] * didrocks looks at the discussion [08:58] Morning! [08:59] I've also pinged osomon to get the webbrowser AP 1.4 branch in today [08:59] Mirv: argh, yeah, they are happy but they are not testing the latest kernel so if new bugs appear there... worth bringing to attention during the hangout ? [09:00] Mirv: what about gallery-app? Since I see many many tests failing, and the merge is still not in [09:00] Mirv: CI seems to be failing [09:00] vila: ah, I see (reading the page further), indeed I don't know how the 3.12 should now be tested/fixed without causing problems on the machines themselves [09:00] Mirv: ok, you need a cmake change for unity7? [09:00] Mirv: outside the ci lab ? :-D [09:01] Mirv: should we merge in manually? I'm looking at the failures now [09:01] sil2100: I've not looked at that, my time has been consumed by cu2d + unity7 [09:01] psivaa: hey, are you looking at why image 13 as a bad pass rate? [09:01] didrocks: yes, with the cmake change proposed unity7 starts working [09:01] Mirv: ok, I'll pick that up [09:01] didrocks: yes, that's the fix bregma did to take care of the annoying FTBFS for unity7 [09:01] But getting it merged in takes certainly too long already [09:02] sil2100: the cmake MP however didn't have any approve before [09:02] sil2100: so I tested compiling with it and approved it (needs top-approve / sponsoring now) [09:03] Mirv: right, that's what I meant basically, no one besides you had even made a proper review, not to mention anyone from the cmake maintainance guys [09:03] argh, bregma didn't base on latest upstream releas :/ [09:04] and I see inline changes [09:04] Mirv: sil2100: I think we need to wait for bregma for unity7 [09:04] just handle the rest [09:05] didrocks: ACK [09:05] sil2100: Mirv do you manage to track what's release/what's still needed releasing? I don't see a landing ask for tracking that, are you ok? [09:05] I don't see latest landing ask updated as well, doing [09:05] didrocks, Mirv: since veebers doesn't seem to be around anymore, I'll merge in the gallery-app 1.4 transition branch manually to avoid CI failures and we can then get someone else fixing the 2 remaining failing tests [09:05] seems the lxc-android-config from ogra_ landed as well. Updating the spreadsheet [09:06] sil2100: yep [09:06] didrocks: isn't it just the patch applied, correctly on top of current ubuntu release? [09:06] Mirv: no, it's not (for cmake) [09:07] Mirv: sil2100: do you know what's still needed to be released for autopilot 1.4? [09:08] I guess Mirv is more up-to-date ^ [09:08] didrocks: I don't think anything has been released yet for AP1.4 to archives. I fixed QA stack in cu2d only this morning, and unity8, and most of the other stacks depend on unity7 getting fixed [09:09] didrocks: i'll take a look now [09:09] Mirv: did veebers and thomi prepare 1.4 changes for unity7? [09:10] didrocks: for sure dialer-app needs fixing... many test failures with 1.4 and no 1.4 AP branch ;/ [09:10] Mirv: sil2100: can you list all the components somewhere, like in the landing spreadsheet so that we can track? [09:10] sil2100: they didn't have unity7 on their radar it seems, so I'm working as described on the e-mail [09:10] sil2100: working with upstream? [09:11] didrocks: will do, but upstream starts later for these components usually, but keeping that on the radar [09:11] I'll update the Landing Plan [09:12] thanks [09:12] didrocks: ok [09:13] well sil2100 is writing now something, so not interrupting that [09:13] ;) [09:13] I'll try listing the components that need 'tweaking and releasing' [09:16] didrocks: and true, the cmake thing is not completely clean, although the only thing I see is not from applying the added patch is the 12 lines in Tests/CMakeLists.txt [09:16] Mirv: I updated it a bit, you can add the component status that you handled more or less - I only included unity8 and autopilot from yours, but make sure it's the right status [09:17] Mirv: yeah, but there are other changes, better to not urge on unity7 anyway [09:17] sil2100: I don't think autopilot is done if it's not released, or 'done' is maybe not the correct word? [09:17] yeah, please do not forget anything :) [09:18] Mirv: right, maybe let's use READY for those [09:19] sil2100: or maybe ready for testing, I guess we can't release without testing all of that on device? [09:19] so ready for testing would mean that it's ready in cu2d with 'green light' enough to be tested and after that published from PPA [09:19] Mirv: indeed! You didn't test it on the phone yet? [09:19] ACK, sure [09:19] ogra_: hey, do you know why there is no ubuntu-ui-toolkit in http://people.canonical.com/~ogra/touch-image-stats/20131105.1.changes as well? (this one is image 13, right?) [09:20] didrocks, reload [09:20] ogra_: you cheated! :) [09:20] sil2100: I haven't had time to even attach the phone yet to my computer [09:20] yep :) [09:20] sil2100: let alone upgrade it to AP1.4 etc [09:21] didrocks, for some reason it compared 20131105 again, not 20131105.1 .... i assume something was slow still due to the wider probs last night [09:21] I need to upgrade it to the new image I guess as well [09:21] ogra_: ok, thanks for fixing! [09:21] didrocks, sadly that doesnt make 13 any better :( [09:21] ogra_: no, I'm unsure what happen, can't wait for psivaa's report :) [09:22] (the breakage started in 12 though) [09:31] sil2100: popey: coming? [09:32] sil2100: we are waiting for you :) [09:35] AAa! [09:45] hmm what happened to my hangouts [09:50] I know a cat was involved but that doesn't explain getting redirected to a hangout error url on retry [09:51] sil2100: but I guess the meeting was near end, or was there still something / being discussed? [09:51] eh, now it works [09:52] Mirv: I guess it's more or less over ;) [09:53] wow [09:53] that was a weird error I got into, but animals have the ways of finding interesting key combos [09:53] so doing a call test for image 13 i just managed to make my Nexus 5 hang hard [09:54] heh, the kitkat phone app is definitely worse than ours :) [09:54] * ogra_ manages to hang it reproducably [09:58] wow, and it is unintuitive as hell ... trying to store a contact requires you to jump through plenty of hoops [10:08] psivaa: is there a way to increase the number of loops in systemsettle idling and rerun against image 13? [10:09] to check that indicators is going quiet after a while [10:11] didrocks: let me try that [10:11] psivaa: maybe try -d 6 [10:11] as for the main run [10:11] instead of -d 2 [10:13] sil2100: didrocks: releasing ubuntu-ui-toolkit, only change the AP1.4 one and all tests succeed. I'd wish for another to test unity8 since I don't get a full pass with that (3 tests fail) [10:14] didrocks: pkging ack http://10.97.0.1:8080/view/cu2d/view/Head/view/SDK/job/cu2d-sdk-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_ubuntu-ui-toolkit_0.1.46+14.04.20131106-0ubuntu1.diff [10:15] psivaa: can you review and merge that one? https://code.launchpad.net/~didrocks/ubuntu-test-cases/fix-delay/+merge/194095 [10:16] then, if you can retry a full tests on image 13 [10:16] sil2100: are you on unity8? [10:16] Mirv: +1 [10:17] didrocks: i'll take a look, -d is the delay in between each iteration and default is 6 though [10:18] psivaa: yeah, but it's set to 2 for normal app runs [10:19] and in the global run, we see that they need 3 loops to settle [10:19] didrocks: if you meant local testing, then yes [10:20] sil2100: thanks! [10:26] filed bug #1248477 about the unity8 tests [10:26] bug 1248477 in Unity 8 "Three failing tests with unity8 + AP1.4" [Critical,New] https://launchpad.net/bugs/1248477 [10:26] Saviq: ^ [10:26] OSError: [Errno 3] No such process [10:26] didrocks, yeah, we're talking, didn't happen in autolanding [10:26] seems a crash? [10:26] ah ok ;) [10:26] didrocks, yeah [10:28] didrocks: for autopilot itself I guess the automatic testing by cu2d is enough? [10:28] didrocks: making the default run with 6 s delay make the default tests to pass. but default tests are not regressing in between image 11 and 13 [10:28] Mirv: hum, you did run the AP AP tests on the devices, right? [10:29] psivaa: yeah, we are slower to boot it seems, but I still need a run without the system settle flawing the results [10:29] psivaa: even if we commit to 6 for a couple of days and then reverting [10:30]