[06:59] good morning :) === tvoss_ is now known as tvoss [07:02] mandel, around? [07:02] davmor2_, around? === dbarth is now known as dbarth-afk [07:42] tvoss, yes, tel me [07:42] tvoss, have we landed anythign? [07:42] anything* [07:42] mandel, I patched out build-deps that are not in main, yet. Can you give the ppa a final spin? [07:42] tvoss, sure [07:42] mandel, thanks [08:25] sil2100, i have a DSL technician coming within the next hour ... so i wont be able to attend i guess (i will try but i thik i saw the car on the street already) [08:25] ogra_: ACK [08:26] stgraber, good morning [08:27] tvoss: Stephane will most probably be up much later [08:28] sil2100, ack and thx [08:49] davmor2_: hi! Are you on holiday today? [08:57] sil2100: https://bugs.launchpad.net/ubuntu-app-launch/+bug/1335778 for qmlscene crash [08:57] Ubuntu bug 1335778 in Ubuntu Application Launcher "qmlscene crashed during unity8 tests with image 105" [Undecided,New] [08:57] psivaa: thanks! [08:58] yw [09:02] sil2100: where do bugs in snap decisions go? [09:08] popey: hmmm... depends, unity8 is responsible for rendering those I guess, but it's not always true to target it instead of the project generating the snap-decision notification [09:08] popey: what problem do you have? [09:09] the colour of text in them is unreadable [09:09] cjwatson, where to eth build logs hide nowadays (seems they are not mirrored anymore to people.u.c) [09:09] *where do the [09:09] (my god) [09:09] popey: than I would target unity8 - I just hope nothing changed with that [09:09] cjwatson, ah, ignore ... Laney helped [09:10] sil2100: no, long standing issue I think, not regression [09:22] sil2100: bug 1335787 [09:22] bug 1335787 in unity8 (Ubuntu) "Coloured text in snap decisions is hard to read." [Undecided,New] https://launchpad.net/bugs/1335787 [09:23] I know you're a heavy social media user, would be good to get a confirmation ☻ [09:23] Me ;p ?! === sil2100 changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cprov | CI Train Status: #87 promoted | CI Train Sheriff: sil2100 | Known issues: TRAINCON-0 ! The spreadsheet is extremely slow, http://people.canonical.com/~rbpark/citrain/ knows the statuses before the spreadsheet does. Long mako wait times on s-jenkins. === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [09:45] popey: in case davmor2_ is not around today... how busy are you today? Do you think you could do promotion-wise dogfooding in davmor2_'s stead? [09:50] sil2100: if he's not about, sure. [09:50] sil2100: which image? [09:51] popey: I would say #105, it *should* have all the fixes for our blockers [09:52] brendand: could you then do a quick look at the AP failures we're having and give us a sign if you think we shuoldn't promote the image because of those? [09:53] sil2100, ok [09:54] sil2100: ok [09:56] popey: I think that bug is specific to friends; they choose the color [09:56] interesting [09:56] thanks sergiusens [09:56] hey, since a recent build (between 102 and 104) system logs are full of messages from healthd. I filed bug 1335748 . The message itself might be harmless but it makes the logs grow pretty quickly [09:57] bug 1335748 in android (Ubuntu) "healthd spams system logs" [Undecided,New] https://launchpad.net/bugs/1335748 [09:57] sil2100, there seems to be a few tests missing. the normal test count is 857 [09:57] jibel: right; there's a patch for that coming [09:57] sil2100, there are only 826 run [09:57] sergiusens, ok, thanks [09:57] Morning all === davmor2_ is now known as davmor2 [09:58] sil2100: we looking at any image for promotion? [09:58] sil2100: looks like I'm not needed? ^ [09:59] sil2100: I forgot to say I wouldn't be back in time for the meeting but would for the start of my normal hours [10:01] ;) [10:01] popey: yeah [10:01] davmor2: hey! Yeah, #105 would need some love from a dogfooder [10:01] sil2100: no worries [10:02] brendand: hmmm, maybe the health tests are missing now? === Chipaca` is now known as Chipaca [10:02] davmor2, ping [10:03] tvoss: hello [10:04] davmor2, could you give silo11 another spin? I patched out all the build-deps that are not in main (yet). Just to make sure that everything still works :) [10:05] tvoss: I can once the testing for the dogfooding is done so it will be after Lunch [10:05] sil2100, hey, can silo 2 be published, please? [10:05] davmor2, ack [10:06] sil2100, oh yes, indeed === lool- is now known as lool [10:10] sil2100, i don't see anything specifically to block promotion. just ask davmor2 to do some particularly good dogfooding on terminal-app and address-book-app [10:10] oSoMoN: one moment [10:11] brendand: you're just not trying to break things then are you ;) [10:11] davmor2, no [10:12] brendand: hahaha :) [10:12] davmor2, you're the one with the hammer [10:14] brendand: I'll have you know I'm a professional I have a selection of hammers, toffee→Mjölnir :) [10:14] oSoMoN: ok, I'm back from the phone [10:14] oSoMoN: is this like a feature, or a bug-fix ;)? If it's a feature, is it a big one or a small one? [10:15] (asking because we're in TRAINCON-0 [10:15] ) [10:15] sil2100, it’s a feature, that was there but had been disabled since the switch to oxide, and is now being re-enabled [10:15] damn this stupid traincon0 state [10:15] oSoMoN: how would you assess the regression risk of this? [10:16] Yeah... we should get rid of it soon, if we'll be able to promote #105 [10:17] sil2100, given that the feature is fully autopilot-tested, and that the overall test coverage of the app is high, I would say the risk is low, but obviously I’m biased… [10:18] davmor2 rhymes with thor [10:18] kind of [10:18] if you drop the 2 [10:19] brendand: that's a stretch even for you :) [10:19] davmor2, what's that supposed to mean! [10:21] brendand: You're QA so you are used to thinking outside the box but that is so far outta the box that the box is pinhole in the sky ;) [10:21] davmor2, what box? [10:21] ;) [10:23] brendand: nice [10:23] davmor2: hello [10:23] asac: hello [10:23] davmor2: is it true that the webbrowser AP failures are not reproducible on local device? [10:24] brendand: ^ [10:24] asac, i haven't been able to reproduce them yet [10:25] asac: I've not seen any issues using the device day to day [10:25] davmor2: well, i wondered if you managed to run the APs [10:25] and have them fail [10:25] you are not doing these at all? [10:26] is selene still doing AP tests? [10:26] asac: I don't run the AP's at all brendand is though hence pointing him at the question [10:27] asac: brendand and elopio are the 2 main guys spearheading the AP test failures [10:27] brendand: its odd that you dont see them [10:27] asac, are you expecting someone to be running the AP tests outside of CI? [10:27] brendand: they are happening every day [10:27] yes [10:27] if folks claim that the APs are not reproducible [10:27] then we have to double check that claim [10:28] asac, well the problem is we can't reproduce the exact environment they are run in [10:28] i know, but its very close [10:28] so we always should continue figuring the difference [10:28] and close th gap further [10:28] asac, from what i understand there is a mere reboot between test suites [10:28] right [10:28] so that might be an angle [10:29] what is key is that we get a way to reproduce them to the devs [10:29] so they can fix [10:29] asac, so if the test suites aren't cleaning up properly the state then it could cause issues [10:29] i could [10:29] yes [10:29] but is that the reason ? :) [10:29] this thing is long enough broken reliable that we should really check out whats going on [10:30] ogra_: there? [10:30] brendand: so if the theory is that they fail because of dirt we can ask plars to make a special run that just makes a clean phone install and just runs this [10:31] if we know its dirt then thats a step forward [10:31] asac, that would be a great idea [10:31] asac, best would be for each suite to have it's own device :) [10:31] always continue to find out [10:31] not cheap i know [10:31] well, you can dream, yes, but reality is what we have to deal with first [10:32] asac, but more seriously we can isolate the misbehaving ones [10:32] if you see it not reproducible, the way to fix it is to debug and plars_ is at your command for that [10:33] plars_: psivaa: can you guys can makea run of the webbrowser-app AP happening [10:33] on a clean install? [10:33] e.g. just boot-and-install -> reboot -> webbrowser-app [10:33] asac: ack, will do [10:34] psivaa: cool. please let brendand and me know [10:34] asac: ack [10:34] psivaa, we mean to have that permanently, not just once [10:34] brendand: ohh? [10:34] psivaa, so on every run, webbrowser will have its own clean install and run [10:34] psivaa, yes [10:35] asac, correct ? ^ [10:35] Permanently? [10:36] I wouldn't want that to be permanent, as this would mean we have a separate way of running webbrowser-app than others [10:36] Which doesn't give good final results in comparison to other suites [10:37] sil2100: I just noticed a minor regression in dialer app with dual calls but I'm assuming that isn't so important one call works fine :) [10:38] brendand: we dont have that permanently right now. [10:38] but doesnt matter for this case [10:38] if in doubt you can always work with CI vanguards to get a clean rerun [10:39] to check if thats the different [10:39] its not perfect, but rest assure it's in the backlog somewhere and will get done eventually :) [10:40] anywya, we want test suites to run through with and without clean or reboot in between [10:41] * sil2100 jumps out to the post office real quick [10:41] sil2100: you like snail mail? :) [10:41] asac, semi "there", yes [10:42] ogra_: so you wear only 25% of your cloth? :) [10:42] or lost your tobacco :P? [10:42] lol [10:42] heh, no, but i'm distracted by t-online [10:43] (they messed up my SDSL two weeks ago ... had a tech. here but he didnt manage to get it back up yet) [10:43] lock them in your cellar until its fixed [10:43] :) [10:43] heh, well, he is gone ... got to get a new one now :( [10:43] see ... you should have locked him in right away [10:44] yeah, to late [10:44] ogra_: have you ever run APs lately? [10:45] the last time was a while ago (and my test device currently has the new developer mode code on it, i wouldnt like to re-flash until i could land that after TRAINCON-0 is over) [10:46] asac, bug 1334676 will get you weird results though [10:46] bug 1334676 in qtubuntu-sensors "Applications have a dead open dbus connection (since image #96)" [Critical,New] https://launchpad.net/bugs/1334676 [10:46] ogra_: i have 95? [10:46] (devel) [10:47] oh we are in TRAINCON-0 already [10:47] sil2100: is QA helping testing silos etc.? [10:47] asac, you should have 87 i think [10:49] ogra_: so since that you can reproduce AP test failures locally? [10:49] since dbuss bug [10:50] you need to use a special option that UTAH uses too [10:51] asac, phablet-test-run -A '--timeout-profile=long' ... === gatox_ultra is now known as gatox [11:06] * ogra_ gets off the phone and feels like killing someone [11:07] brendand: asac: so the webbrowser test alone was run on a freshly flashed mako and the tests are still failing: [11:07] http://q-jenkins.ubuntu-ci:8080/job/utopic-touch-mako-smoke-daily/373/consoleText [11:07] right [11:07] :) [11:08] http://ci.ubuntu.com/smokeng/utopic/touch/mako/105:20140630:20140625/8793/webbrowser_app/ [11:08] i am sure if you download and run it locally it also fails [11:08] likely [11:08] :) [11:08] well, folks tell me the are all not reproducible [11:08] but dont forget -A '--timeout-profile=long' [11:08] else it will fail for sure [11:08] ogra_: whats that? [11:08] see above [11:09] why isnt that the default for phablet-flash? [11:09] err phablet-testrun [11:09] the workaround for bug 1334676 [11:09] bug 1334676 in qtubuntu-sensors "Applications have a dead open dbus connection (since image #96)" [Critical,New] https://launchpad.net/bugs/1334676 [11:09] because then your tests take 10x as long [11:09] we dont want that as default [11:09] but it works around the dbus timeouts [11:10] (it is the default in UTAH for whatever reason) === MacSlow is now known as MacSlow|lunch [11:37] asac: yeah, I have asked om26er to fill in as our QA-sign-off person, but currently we seem to be ok [11:38] As not many landings are ready to be released [11:38] davmor2: how's the testing going so far? [11:40] sil2100: dual calls has a regression you now can't seem to hang up on them other than that pretty good so far. [11:40] davmor2: yeah, that one I saw you mention earlier, but well ;) I guess it's not super big - we would need a bug for that though [11:41] sil2100: yeah as soon as I finish I'll write it up [11:41] Thank you :) [11:42] sil2100: might of just noticed another slightly more serious one though. [11:43] popey: can you play a piece of music from the dash and tell me which speaker it plays out of please [11:43] Crap... [11:43] popey: same in the music app [11:43] davmor2: the right one [11:43] loudspeaker [11:43] for me it seems to play out of the earphone [11:44] same in music app, works as expected [11:46] okay that was wierd so ear speaker, plug in headset remove headset no loud speaker so it was set to play through loud it just wasn't [11:47] I wonder if switching loud speaker on and off in call were the cause of that [11:48] * davmor2 reboots and tries again [11:48] nope reboot and it's playing in the ear speaker again [11:51] sil2100: allright. dont relax the -0 approach [11:52] if QA is a bottleneck its partly intended, but if its too much slow down let me know [11:52] asac: ACK :) [11:52] sil2100: at best coordinate with jfunk every time we start traincon-0 [11:52] or at least forward announce just about that to qa list [11:52] thx [11:55] asac: the traincon-0 was announced in the landing team on Friday, so it's not a shocker [11:55] sil2100: right, its part of the normal mail [11:55] we should send special mail just about that [11:55] so i can forward and get proper attention etc. [11:55] sil2100, hey, regarding silo 2, the packaging changes were acked by didrocks (see the MR) [11:56] also since jfunk has to contribute significant engineers we should alywas coordinate with him that we do it etc. [11:56] asac: ok, will send it out now, since basically today is the first day of TRAINCON-0 [12:01] cool [12:01] sil2100: just a nice mail: TRAINCON-0 in effect [12:01] :) === alan_g is now known as alan_g|lunch [12:10] psivaa, it would be good to try and reproduce the conditions in which those tests were run locally [12:10] psivaa, is there a script or something that will do all the steps, including provisioning, setup etc [12:10] brendand: yes, just a sec [12:16] brendand: http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/scripts/provision.sh should do the provisioning [12:18] and then http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/scripts/run-autopilot-tests.sh with -a webbrowser_app will run the webbrowser test [12:18] brendand: http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/README-cli.rst has some information === MacSlow|lunch is now known as MacSlow === renato_ is now known as Guest10805 [12:48] psivaa, provision.sh doesn't run without a file called magners-wifi? [12:51] brendand: yes, that was for the lab, in your local setup you could either modify provision.sh or run 'phablet-network -n $NETWORK_FILE' [12:51] psivaa, ok [12:52] brendand: sorry, dint mention that [12:52] sil2100, are you not assigning silos during TRAINCON-0 or is that indeed ok? I have one pending... row 25 :-).. [12:53] camako: hi! I'm starting lunch now, that's why I didn't assign yet ;) [12:54] camako, silos get assigned but need special QA team signoff and testing [12:54] sil2100... was just curious... no hurry... [12:54] ogra_, sure... was just gonna get a head start on testing etc... hoping TRAINCON-0 will not last very lng [12:55] * ogra_ hopes so too [12:55] :) [12:57] psivaa, why would i get 'unknow flag developer mode'? [12:58] brendand: just a sec [12:58] psivaa, actually i probably need to upgrade ubuntu-device-flash [12:59] brendand: that'd run 'ubuntu-device-flash --bootstrap --developer-mode --channel ubuntu-touch/utopic-proposed' [12:59] which is what we are running in the lab. so yea you might need to upgrade that [12:59] psivaa, yeah got it now === alan_g|lunch is now known as alan_g [13:01] davmor2, ping [13:01] tvoss: hello [13:02] sil2100: okay so everything seems to work. I don't understand the issue with the loud speaker/earpiece switch but I can't reproduce it [13:03] sil2100: so just the hangup issue on dual calls that I will bug now === plars_ is now known as plars [13:09] brendand: yeah, you'll need a somewhat recent version of phablet-tools and ubuntu-device-flash [13:10] brendand: psivaa: so what's the deal you and asac were talking about earlier with wanting a permanent clean run of webbrowser-app? It gets run on every single new image. If some other test is interfering with it, we should figure out how and fix that [13:10] but I would doubt that a bit [13:11] plars: yea, webbroser failed when ran on its own in tha lab [13:11] plars: i dont want permanent clean run :) [13:11] plars: i just wanted to get one clean run to eliminate FUD [13:11] :( [13:11] :) [13:11] * ogra_ wants permanent clean runs of all tests all the time [13:11] ok, that makes more sense :) [13:11] plars: do you know when that reproducible failure started? [13:11] pretty recently [13:12] thats exactly not the answer i want :P [13:12] asac: that was on image 92 [13:12] ok [13:12] and we see a number of web related updates on that [13:12] psivaa: and before webbrowser was "flaky"? [13:12] asac: not really [13:12] sil2100: ^ [13:12] psivaa: thanks [13:12] we had failures before but not to this extent [13:13] http://people.canonical.com/~ogra/touch-image-stats/92.changes [13:13] asac: I mentioned last week that they had been failing for a bit, but due to the recent autopilot issues we decided on the landing call that it made sense to look at it after that fix goes in [13:13] psivaa: did thowse failres before happen reliably? [13:13] psivaa, i think it started in 91 already [13:13] ah, no [13:13] i'm wrong [13:14] psivaa: have one example of the failures that happened for a while before? [13:14] asac: let me check [13:14] i dodnt think we had any for a while before 92 [13:15] 92/93 had a few failures, then 94 went back to 100%, then ever since then random numbers of webbrowser tests failing [13:16] plars, right, but beofre 92 we didnt have any webbrowser failures for at least ten images (i didnt go back further) [13:16] plars: so 94 was either a lottery win or we have the current problem starting 95? [13:17] asac: I'd say the flakiness started at 92 [13:17] asac: 83,82,81 had failures before this, but they appear to be flaky [13:17] http://ci.ubuntu.com/smokeng/utopic/touch/mako/83:20140616:20140530/8570/webbrowser_app/ [13:17] plars, asac, i wouldnt trust 94 results at all ... see the UITK test in there, it totally misbehaved [13:17] asac: plars: since 92 the failures might be with different tests but atleast one fail everytime since then [13:18] we got a new browser in 92 ... and with that it started to fail pretty reliable [13:18] * asac wonders where thomis cool dashboard feature ended up be [13:18] asac, ++ [13:18] we really need that asap ... would be soo helpful right now [13:18] oh i think it didnt allow to drill down to indiviudal tests? [13:19] "where is thomi"? :) [13:19] sil2100, re- silo 2, have you seen my message about packaging changes? [13:19] it showed exactly the overview we want right now [13:19] this guy has to move i am sure [13:19] ChickenCutlass: do you have a big house where thomi could live until october? :) [13:19] * ogra_ doesnt mind if he has to do the drilling down on the old dashboard [13:19] oh i know the answer: ogra's house :P [13:19] lol [13:19] ha [13:20] thats big and you can even do evening exercise [13:20] he can have a tent in my garden (if he mows the lawn at least twice a week) [13:20] moving things from A to B to T (trash) :) [13:20] i think OZ folks know best how the sheep keep the gras low :) [13:20] oh [13:20] he might gain some weight finally :P [13:21] he could bring a sheep or two ... susie would forever love him [13:21] or be on sheep-diet :P [13:21] you could offer that as a survival tour to rich people :P [13:21] heh [13:25] how is the sqlite regression going? [13:25] ogra_: ? [13:25] no idea [13:25] not on my plate (i was busy the whole weekend with developer mode) [13:26] oSoMoN: yes, I saw those :) QA needs to sign-off the silo still though [13:26] oSoMoN: the packaging changes are ok [13:27] oSoMoN: no worries, om26er_ is on it already [13:27] sil2100, excellent, thanks! [13:27] o/ [13:28] psivaa, i think there are some steps missing. after provision, the wizard is up, and the image is not writable [13:28] did you call phablet-config --writable-image ? [13:28] sil2100: sqlite regression ... is that good now? [13:28] (or however exactly this is called) [13:29] brendand: it could be that provision.sh exited before it finish all it's steps because of the we dint setup the network config file [13:29] asac: yes, it got fixed but with a unit test disabled [13:29] asac: but thostr_ and others are working on the real fix in sqlite3 [13:29] psivaa, no it exited with success [13:30] sil2100: how did we get the sqlite3 changes that caused this? [13:30] from debian as merge? [13:30] asac: new upstream release has been imported into Ubuntu [13:30] asac, right, merge/sync [13:30] asac: doko did it, I guess it was a merge from Debian [13:30] sil2100: asac: we're working on it, but fix won't be there unitl tomorrow (jamesh working on it) [13:31] ic [13:31] sil2100: asac: as sil2100 told me earlier this shouldn't block promotion for now [13:31] thostr_: i dont care so much about when you do the fix, i care about you guys suffering :/ [13:31] :( [13:32] thostr_: thanks! [13:32] brendand: apologies, for that to be writable we should run provision with -w. now could you run 'phablet-config writable-image' pls [13:32] camako: ok, silo assigned, but please wait with building it maybe for now [13:33] camako: as platform-api should soon land from another silo [13:33] camako: and you would have to rebuild anyway [13:33] sergiusens: hi! Are you around? How's the silo 007 testing going? [13:33] sil2100... thanks for the silo ... sure [13:35] davmor2: how do you put music/video on your device? [13:35] do you use mtp or adb push? [13:35] popey: adb then change the permissions [13:35] bah [13:36] popey: sergiusens said to pick on somebody about the fact you couldn't do it via mtp [13:36] davmor2, popey: so uploading music/video through mtp doesn't work now? [13:37] sil2100: not for scripting you can manually [13:37] sil2100, i think uploading 10TB of music/video doesnt [13:37] Ah, ACK [13:37] 10TB? Damn... [13:37] these guys are crazy :) [13:37] ogra_: I only have 4.2 GB of music [13:37] * sil2100 looks at his mako phone with disappointment [13:38] I got the 'cheap' version of 16GB ;/ [13:38] you will love the BQ and Meizu ;) [13:38] they have SD slots [13:38] ! [13:38] * sil2100 wants [13:38] woah [13:38] i copy very small amounts of music [13:38] ooooooo la-de-dah [13:38] it craps out every time [13:38] serious issue, bug already filed [13:38] popey, pnly 5TB ? [13:38] *only [13:38] pffft [13:38] :) [13:39] bug 1317263 [13:39] bug 1317263 in mtp (Ubuntu) "mtp-server crash when transferring music on #15" [Undecided,Confirmed] https://launchpad.net/bugs/1317263 [13:39] Woha, #15? [13:40] NOTAREGRESSION [13:40] Ship it [13:40] ;) [13:40] ☹ [13:40] it probably just doesnt like amy winehouse [13:41] But anyway, jokes aside... damn, that's an old bug - why didn't anyone work on it? [13:41] * ogra_ guesses cyphermox_ has built in a filter so that you can only copy what he approves [13:41] ogra_: No it's popey it'll be Led Zeppelin [13:41] It was a cover.jpg in my case today [13:41] what was on it ? [13:42] A picture of my poking you with a stick. [13:42] hah [13:42] ogra_, popey: it's the porn filter killing the flesh tone on Electirc ladyland cover [13:43] popey, see, that will *never* pass the content filter :) [13:43] meanwhile.. back at the point. [13:43] electric ladyland? what kind of music do you listen to davmor2?? [13:43] :) [13:44] sil2100: it's very annoying, and while it's been around for a while, it would be very nice to get fixed [13:44] plars: Nothing wrong with Jimi Hendrix [13:44] is there a problem with gallery in devel? it wont load? [13:44] oSoMoN, which package provides webapp_container testsuite ? [13:44]