[02:04] === trainguard: IMAGE 294 building (started: 20140414 02:05) === [03:29] === trainguard: IMAGE 294 DONE (finished: 20140414 03:30) === [03:29] === changelog: http://people.canonical.com/~ogra/touch-image-stats/294.changes === [07:05] good morning [07:40] hola [07:40] hw does the image look like today? :) [07:41] o/ [07:41] asac: quite good, but a lot of images in [07:41] asac: seems, that all blockers but the Qt issues are vanished, we'll need dogfooders to confirm though for the webapps stuff [07:41] asac: however, seems we still have the dialer-app flaky test for instance [07:42] (so the branch that was supposed to fix it doesn't) [07:46] On latest image smoketesting I also see some clock app flakyness [07:46] yeah [07:46] sil2100: hey! [07:47] sil2100: I was wondering btw, do you mind looking at them before the meeting? [07:47] like, ensuring it's real flakyness and see how they goes? [07:47] (if you can reproduce, and so on) [07:47] didrocks: ok, sure, I'm upgrading my device anyway [07:47] So I will have a look [07:48] thanks! [07:56] there was even address book app failing test in #290 [07:58] I'll run a couple of test runs two for those three [07:59] I'm running the test suite for dialer already if anything, but you could check the other two [08:00] thanks Mirv! [08:01] I think there's time to run all of those quite easily about ~two times before the meeting [08:14] dialer seems really flaky ... i had random tests failing on and off over the weekend [08:14] i wonder if its the new AP [08:16] I just ran the whole suite 3 times just now and didn't get any failures, all the time just OK [08:16] * sil2100 has a lucky phone [08:16] ;( [08:19] Anyway, latest autopilot upload got rid of the python-evdev dependency (leaving only python3-evdev), but it's still required as python2.7 is used for dialer tests still [08:20] sil2100: hum, so it's reinstalled? [08:20] when installing the dialer-app AP test [08:21] ogra_: I doubt that's the new AP then [08:21] ogra_: only commit: http://bazaar.launchpad.net/~autopilot/autopilot/trunk/revision/480 [08:21] and so it's reinstalled :p [08:22] bah [08:23] I had to install python-evdev manually, since dialer-app-autopilot's deps didn't pull it in itself - I wonder how it's done on smoketesting infra [08:24] sil2100: hum, interesting [08:24] psivaa: would you know? ^ [08:24] something to bring up [08:24] what happens if you dont ? [08:24] are you sure it's not running the python3 flavor? [08:25] ogra_: all tests fail straight away [08:25] ok [08:25] didrocks: I checked the logs, it says Loading tests from: /usr/lib/python2.7/dist-packages [08:25] so either "something installs it beforehand" [08:25] or they are using another path [08:25] hum… ok [08:25] so something should install it [08:26] psivaa: can you check on the latest mako device if python-evdev was installed? [08:26] didrocks: sil2100: looking [08:26] thanks :) [08:26] It has to be installed somehow since the tests start running - on my device without evdev bindings it was just barfing on the start [08:27] didrocks: sil2100: 'Installed: 0.4.1-0ubuntu3' [08:27] ok, maybe psivaa can tell us when it was installed [08:28] (I guess it's at the global setup) [08:30] that's being installed as a dep for unity8-autopilot [08:30] Ah [08:31] thats not python3 either yet [08:31] ? [08:31] Interesting [08:31] i thought the tests were all switched when we dropped it [08:31] cihelp, hey, we're having issues in otto runners... for some time now no unity8 autopilot test completes - the tests can't find the process, but there's no crash or nothing... [08:31] ogra_: seems all that is puzzling [08:31] oh, meeting time [08:31] coming in 2 minutes [08:31] Saviq, still the timeout ? [08:31] ogra_, "the timeout"? [08:32] Saviq, the CI bot had timeouts in the logs last week [08:32] sil2100: didrocks: http://pastebin.ubuntu.com/7248624/ is the relevant snippet [08:32] I can see "GLib-GIO-CRITICAL **: g_bus_watch_name: assertion 'g_dbus_is_name (name)' failed" in unity8 logs, that would eject into oblivion it wouldn't it [08:32] ogra_, that's different, we're testing fine on devices, not on desktop, though [08:33] ah [08:33] Saviq, ogra_: weird, I thought fginther diagnosed it as requiring the otto nodes to be upgraded to trusty... [08:33] psivaa: thank a lot! [08:33] yw :) [08:33] vila, yeah, seems i was wrong [08:33] ogra_: not at all, the timeouts was the symptom [08:33] were [08:34] right, but they are not the issue here apparently [08:34] ogra_: well, not the issue per se indeed [08:43] sil2100, icanhassilo for row 32 please? [08:44] Saviq: hi! It has some features in it, yes? [08:44] sil2100, one, yes [08:44] sil2100, test tweaks and other UI fixes otherwise [08:44] sil2100, the one is real small, too [08:46] additionally, dialer seems to sometimes work (all 9 AP:s just passed) [08:46] Oh, ok - will it be easily 'revertable' in case something bad happens? SInce I see you have 2 components in the silo [08:46] Mirv: same for me... [08:46] I've the .crash files too, yes [08:46] telepathy-ofono + dialer [08:48] Saviq: for now I set QA sign-off needed to 'No', but I will consult with didrocks if it's the case here [08:48] Assigning [08:48] sil2100, thanks [08:51] Saviq: had to correct the last merge URL as it was missing the last character ;) But besides that it should be assigned now to 001 [08:51] sil2100, craps, thanks [08:52] ooh and a prime silo, too ;D [08:52] ;p [08:56] remember to use the premium silo only for premium landings! [09:02] not getting clock to fail for me [09:10] Saviq, aren't the indicator changes going to break the preview session? [09:11] mhr3, no, the original manager only emitted the same signal, too [09:11] mhr3, ah you mean that the session has to emit them, too? [09:11] Saviq, right, but the preview session upstart job doesn't emit them now [09:11] mhr3, right, forgot it's using a different job... :| [09:11] it shouldn't, btw... [09:12] ok /me drops that MP for now [09:12] mhr3, thanks [09:12] or add one to fix the preview session job ;) [09:13] mhr3, yeah, will do that in the next instalment [09:13] want to land this asap [09:13] k [09:14] * Saviq needs to look into what the desktop session does... [09:14] ogra_: do we block ping on the phone!? [09:14] popey, so i was trying to land the gcmanager vs lxc-android-config change you tested ... but i couldnt even get the device to boot when i added the changes stgraber proposed in the bug ... do you remember if you had to do something special ? [09:14] http://popey.com/~alan/phablet/device-2014-04-14-101410.png [09:14] ogra_: i did [09:14] popey, probably confinement blocks the teminal app ... i checked from adb last time it happened to me [09:15] ogra_: http://paste.ubuntu.com/7219206/ [09:15] ogra_: i cant use adb if I am out and about ☻ [09:16] popey, sudo ping ... [09:16] the user doesnt have permission to access the ICMP socket [09:16] s/socket/protocol/ [09:17] (and the sudo in the terminal will reveal a beautiful bug to you too :) ) [09:17] (autocorrection printing the pw in clear text) [09:17] would that be bug 1307386 [09:17] bug 1307386 in Ubuntu Terminal App "Terminal should not capitalise first letter" [Undecided,New] https://launchpad.net/bugs/1307386 [09:17] which I just filed [09:17] oh, related [09:18] well i'm more concerned by it printing the full password at the sudo password prompt [09:18] do you have a bug for that? [09:18] :) [09:18] i dont think so [09:18] I'll modify mine [09:18] i noticed it last week the first time but forgot about it again [09:19] there's a hint I think we can give the kb to switch those off I believe [09:20] can you confirm bug 1307386 pls [09:20] bug 1307386 in Ubuntu Terminal App "Terminal should not use assistive technologies" [Undecided,New] https://launchpad.net/bugs/1307386 [09:46] didrocks, Mirv, psivaa: sadly, even with a cleared out /var/crash I still succeed in dialer-app tests, tried 3 times already! [09:47] :( [09:47] sil2100: ok, and always the second one? [09:47] ah [09:48] but yeah [09:48] that's prooving the point :) [09:48] or not [09:48] can be a different network as well [09:48] while it's uploading [09:48] let's see once psivaa has his results [09:48] sil2100: curious if the the crash file generated during the tests after you clearing /var/crash ? [09:48] in the same network conditoin === vrruiz_ is now known as rvr [09:49] device is flashing here btw [09:49] thanks ;) [09:52] psivaa: yes, on every test suite run I then have the dialer-app crash present again [10:07] o/ sil2100: i have silo 006 ready for publication [10:07] sil2100: this is a set of desktop fixes, so let me know which route it takes (archive or -updates) [10:08] dbarth: oh, it's desktop only, yes? [10:09] Awesome [10:12] didrocks: oh, the 'packaging diff' generated by citrain/cu2d doesn't include debian/copyright changes? [10:13] didrocks: anyway, two packaging ACKs needed for a desktop landing: https://ci-train.ubuntu.com/job/landing-006-2-publish/5/artifact/packaging_changes_webapps-applications_2.4.17+14.04.20140411-0ubuntu1.diff and https://ci-train.ubuntu.com/job/landing-006-2-publish/5/artifact/packaging_changes_unity-webapps-googledocs_2.4.16+14.04.20140411-0ubuntu1.diff [10:14] didrocks: one of them also has some debian/copyright changes which are not visible here, but it seems fairly alright anyway [10:20] sil2100: it should contains, are you sure it's missing? [10:20] sil2100, didrocks: can I get a silo for l34? [10:20] it's a security issue for unity lock screen [10:20] I hope it's working fine, it didn't get a code review yet, but trying to get some head start since nobody from the unity team seems to be around yet [10:21] didrocks: yes, since see the webapps-googledocs diff and the merge https://code.launchpad.net/~abreu-alexandre/unity-webapps-googledocs/fix-pattern-and-store-cookies/+merge/214864 [10:21] seb128: assigning [10:21] sil2100, thanks [10:21] sil2100: interesting, I may have some filters for that [10:22] sil2100: +1 on both [10:27] we did take the same terms as the previous google webapps we had packaged [10:27] and so still going to the archive or updates? [10:34] sil2100, still working on assigning that silo? [10:34] sil2100, that might be a candidate for respinning trusty iso, so would be nice to get thing moving [10:36] davmor2: "Play in Grooveshark" button in the grooveshark scope does nothing here... [10:36] sil2100, Mirv, didrocks, could I have a silo for row #35 please? [10:36] seb128: seems he didn't see the conflict: https://ci-train.ubuntu.com/job/prepare-silo/197/console [10:36] seb128: unity7 is in landing-008 [10:36] should we flush/join to that one? [10:36] * didrocks looks [10:37] interesting… doesn't seem landing-008 is assigned in the spreadsheet [10:37] I guess someone removed a line… [10:37] Mirv: can you handle Chipaca's request meanwhile? [10:37] popey: I justed typed in Happy clicked on it, click on play in grooveshark an now I am playing the track [10:38] davmor2: works now I have rebooted [10:38] seb128: someone added that one:https://code.launchpad.net/~beidl/unity/unity-lockscreen-gestures/+merge/215569 [10:38] didrocks, no, landing 008 is the mp I was asking to get, seems like it's the mainscreen that just didn't get the info reflected [10:38] ignore me [10:38] didrocks, thanks [10:38] seb128: hum [10:39] oh right [10:39] misread [10:39] didrocks, c34 seems to be empty [10:39] seb128: it's assigned since longer [10:39] e.g not formula [10:39] but I guess it was assigned too long [10:39] seb128: ah, the field is empty... sometimes that happens [10:39] argh [10:39] didrocks, sil2100: thanks [10:39] indeed [10:39] didrocks: I noticed that happening from time to time ;/ [10:39] sil2100: why people remove the formula? [10:40] didrocks: sure [10:40] I have no idea ;/ [10:40] sil2100: that's annoying… [10:40] But usually not only there it's removed [10:40] davmor2: now lock the phone [10:40] same with the formula [10:40] didrocks: look, not only this one, also the M column went missing [10:40] on the status [10:40] davmor2: audio in webapps stops playing when phone in welcome screen [10:40] I guess someone selected the line and set "clean" [10:40] I doubt it it's removed by someone on purpose, hm [10:40] Or that [10:41] sil2100: there is nothing touching the status column [10:41] really nothing [10:41] so it has to be manual… [10:42] hum, I already had to do re-create this twice last week, but I thought it got cleaned by accident [10:43] davmor2: but it works fine when the phone is locked/suspended [10:47] popey: that I can confirm [10:47] filing a bug [10:47] popey: no idea what is going on there [10:47] not sure where app lifecycle stuff goes? [10:47] * popey files on webbrowser-app first [10:48] unity-mir [10:48] ok, ta [10:48] popey, but this is expected behavior, webapps dont use the media-hub [10:48] (yet?) ? [10:49] dunno, if that even possible with the sandboxing model [10:49] what package is unity-mir is in? [10:49] -typo [10:49] err, isnt that unity-mir ? [10:49] Package unity-mir does not exist [10:50] yet i see uploads for it :) [10:50] not installed on my phone [10:50] libunity-mir1 I imagine? [10:50] ogra_, popey: but it works when the phone locks it's only when you hit the power button and go back to the welcome screen that it stops [10:50] https://launchpad.net/ubuntu/+source/unity-mir [10:50] davmor2: indeed [10:50] should phablet-click-test-setup create a gallery_app directory in /home/phablet/autopilot on device? [10:51] davmor2, it should stop when it locks as well [10:51] wat [10:51] ogra_: no that's why it so confusing [10:51] Chipaca: hi! Could you add a description to the landing? [10:51] that makes it useless [10:51] unless someone hacked up powerd to not stop it [10:51] ah nevermind gallery-app is in ~/autopilot/legacy-py2 [10:51] Chipaca: it is required that we know what for the landing is needed [10:51] ogra_: try it [10:51] sil2100: hi! I'm ... trying to :) [10:51] * Mirv has now executed clock app AP tests without fails 14 times... filing a bug anyhow [10:52] sil2100: "because did.rocks told me to" sounds a little bit lame :) [10:52] davmor2, i belive you :P [10:52] fginther: do you know what may cause RuntimeError: Unable to instantiate any backends [10:52] UInput: ImportError('No module named evdev',) [10:52] psivaa: was it that after a rerun is made the old logs are gone forever? I did take a note of the failing clock app test before refreshing the page, but the logs aren't there I guess? [10:52] fginther: when running gallery_app tests? see http://paste.ubuntu.com/7249001/ [10:52] ;p [10:53] Mirv: you can get them on the jenkins instance [10:53] (at least you should) [10:53] Mirv: http://q-jenkins.ubuntu-ci:8080/job/trusty-touch-mako-smoke-daily/249/testReport/junit/ubuntu_clock_app.tests.test_alarm/TestRecurringAlarm/test_add_recurring_type_alarm_must_add_to_alarm_list_weekday_/ is the one you want? [10:53] Chipaca: I guess it'd be something like "re-enable ubuntu-push for touch images"? [10:53] psivaa: ooh! [10:53] Mirv: don't assign until descriptions are there! [10:53] sil2100: psivaa: thanks! [10:53] ;) [10:53] yw ) [10:53] Mirv: sil2100: there [10:54] Mirv: how do you know if you're not assigning a silo for a time bomb?! ;p [10:54] sil2100: I'll not publish such a thing, but a fine description doesn't prevent a time bomb being planted to the silo :D [10:54] ogra_, popey: same thing happens when you play music from a scope but I'm assuming that is change in the scope as when you remove the welcome screen the scope is now on the main music scope again [10:54] * Chipaca makes a note never to put “explode ALL the things” as the landing description [10:55] ;) [10:55] davmor2, i think ricmm knows that code quite well ... probably talk to him [10:55] ogra_: infact the only way to have music play on the welcome screen is via music player [10:56] right [10:56] it should be the same for suspend though [10:56] music-player is the only whitelisted application [10:56] well, and the video player [10:56] ogra_: but they all play on a screen blank (either power button press of wait for the phone to sleep) [10:57] right, they shouldnt [10:57] only the music and video apps should [10:57] ogra_: yes they should honest :D [10:57] but not the browser [10:57] not by policy, no [10:57] davmor2: ogra_ bug 1307430 [10:57] bug 1307430 in webbrowser-app (Ubuntu) "Audio stops when on welcome screen or app scope" [Undecided,New] https://launchpad.net/bugs/1307430 [10:58] the policy is wrong then ☻ [10:58] with the media-hub landing landing this will change though [10:58] \o/ [10:58] brb, going to prepare lunch [10:58] only the media-hub will be able to keep on playing stuff [10:58] other whitelists will be dropped [11:12] libmtp seems a bit broken at the moment [11:12] causing nautilus to lock up when transferring music [11:16] Chipaca: uploaded ubuntu-touch-meta with the seed change and clicked on build package. The status is now "Packages built" once built. You can then test/set testing pass: yes. QA will be able to sign off afterwards. [11:16] didrocks: ok, will test as soon as my devices become available [11:16] (in the middle of a deep-cycle power test right now) [11:16] thanks :) [11:18] popey: blame ogra_ for that, he doesn't fix it but it helps ;) then point the finger at cyphermox [11:31] yay, ok, with the proper instructions the cgmanager stuff works [11:32] i just tested it wrong [11:43] are we green yet :)? === MacSlow is now known as MacSlow|lunch [11:47] asac: not entirely yet! [11:47] Still some flakyness being battled [11:49] sil2100: dialer still>? [11:49] sil2100: anything that blocks promotion in there? [11:49] can i have a silo for line 36 [11:49] Yeah, dialer... but at least we have some ideas [11:49] ogra_: looking [11:49] * sil2100 still lunching [11:50] take your time ... [11:50] are those crashes in dialer crashes we dont care abotu (e.g. only on shut down?) [11:51] no [11:51] they happen randomly (not all image had both of them on the weekend for example) and there is an app crash too [11:51] ogra_: so they are real crashes seen during end user use? [11:51] nope [11:51] * asac wonders if we ever have hooked up error tracker [11:52] they are real app crashes happening when testing though [11:59] sil2100: didrocks: dialer app tests when run on their own without apport had more failures: http://q-jenkins.ubuntu-ci:8080/job/psivaa-trusty-touch-mako-smoke-daily/4/ [11:59] didrocks: sil2100: i'm running the whole apps tests to see if there is any improvement. [12:00] davmor2: i see no videos in my video lens [12:00] ogra_: what crashes are those? [12:00] didrocks: sil2100: i need to go out for a lunch time appt. will update you when i come back [12:00] ogra_: do we have bugs? === psivaa is now known as psivaa-lunch [12:00] asac, there is a disaler-app crash when the test runs [12:00] popey: I do [12:00] popey: you on 3g or wifi [12:01] ogra_: no i mean real end user exposed crashes [12:01] that happen when you use the fphone [12:01] wifi [12:01] i mean, not even local ones [12:01] popey: oh that's bad. I see both local and online content [12:02] asac, i doubt it has real impact [12:02] popey: I wonder if it is a race condition in mir/scopes or something [12:02] ogra_: right. misread what you said then. onloy care for stuff that end users will see [12:02] like useing the phone for 20 minutes -> .crash file [12:03] yeah, that shouldnt happen [12:03] psivaa-lunch: thanks, keep us posted :) [12:04] popey: I've seen that on other scope once [12:04] popey: had to reboot [12:04] couldn't reproduce it [12:04] hmmm [12:04] didrocks: yeah, reboot "fixed" it [12:06] popey: there is certainly a race… Doesn't seem to be common (but I guess due to new scopes) === alan_g is now known as alan_g|lunch [12:10] popey: put your phone to sleep, send an SMS to it. When the indicator snap decision appears click on it, does it unlock the phone and open messages? Once messages is open repeat the process does it do the same again or just display the keyboard? [12:15] davmor2: the indicator snap decision appeared and disappeared before I could tap it. it's way too fast [12:15] or i am old [12:15] you have to snap it [12:15] thus the name [12:16] popey: bit of both ;) [12:16] davmor2: confirmed [12:16] http://popey.com/~alan/phablet/device-2014-04-14-131627.png [12:16] didrocks: ^ bug [12:16] davmor2: was it unlocking before? [12:17] didrocks: yeap but I'll have to dig into when it broke [12:17] popey: mind checking on latest promoted image? [12:17] ya [12:17] Saviq: any idea if this is on your side? ^ [12:18] * Saviq looks [12:18] no unlock when clicking on a snap decision [12:18] davmor2, popey, it's bug #1233564 [12:18] bug 1233564 in mir (Ubuntu) "Wrong frame is seen on wake up/resume/unlock." [High,Triaged] https://launchpad.net/bugs/1233564 [12:19] i wonder if it is just another fallout of the blocked Qt event s [12:19] or well, yeah ↑, too [12:19] Saviq: seems it was working on latest promoted image though [12:19] from what popey is telling [12:19] didrocks, yeah, before Qt 5.0 [12:19] so a regression or not? [12:19] or maybe the "ya" was "I'm going to test" [12:19] didrocks, that had Qt 5.0 [12:19] didrocks, asac, yeah, Qt event loop [12:19] ogra_: not latest promoted image [12:19] didrocks: works as expected on 250 [12:20] and 250 is 5.2 [12:20] Saviq: didrocks: davmor2: so last prokmoted image had this? [12:20] didrocks, oh, indeed, ignore my babbling [12:20] ok, so a regression, and not linked to 5.2 [12:20] we promoted a Qt 5.2 image? [12:20] yeah [12:20] 3 weeks ago [12:20] Saviq, yeah, with the special exception for the events bug [12:21] Saviq: you really need to read the phone mailing list more ;) [12:21] it's weird they are linking that to the qt event fix [12:21] davmor2, was sprinting for 2 weeks :P [12:21] so either it's not the real one [12:21] either popey is lying :p [12:21] davmor2, popey, please file a bug with unity8 [12:21] but I doubt that the Mir fix is going to fix it then… [12:21] didrocks, whats weird about that ... phone behaves strange on wakeup ... can always be related to it replaying the events [12:22] since it queues them [12:22] ogra_: how would it be working on 250? [12:22] no, not on 250 indeed [12:22] generally though [12:22] yeah, but here, it does, trying to handle that one first :) [12:22] Saviq: so, we were nearly going to promote an image, that one is the blocker (if the Qt one is whitelisted) [12:23] if anything is weird on wakeup i usually suspect this bug ... i can understand if others do it too [12:23] didrocks, the other place where this could get introduced was right edge [12:23] Saviq: as it was working with 5.2 already, it seems that we need to dig further [12:23] ah [12:23] davmor2: you are making that test on every dogfooding? [12:24] unity8 7.85+14.04.20140401.3-0ubuntu1 was the first to have right edge, can you guys check? [12:24] davmor2: popey ^ [12:25] didrocks: normally yes but I've not done a full dogfooding for a few days with one thing and another [12:25] davmor2: urgh, do you know at least what was the last image where you tested that? [12:26] right edge is in for a week and half for instance [12:26] davmor2: how can this happen without escalation? [12:26] davmor2: i mean you were surely aware that we are trying with highest pressure to get an image baked [12:27] asac: it was testing the fixes that were landing before the silo testers were in place that was causing me not to complete a full dogfooding === josepht changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: josepht | CI Train support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone | Known issues: - [12:29]