[00:06] kgunn: already asked for landing the pre-start thingy? === veebers_ is now known as veebers === robru-sick is now known as robru-sick-eod [07:08] hi. unfortunately powerpc among else is slowing down things. hopefully after qtdeclarative it'll become smoother, but I expect that a person in US timezone able to land the non-qt packages would be beneficial [07:09] didrocks: ^ I already spend last evening occasionally pushing things further, but it would have helped also with Qt if someone during my night could have pushed the buttons [07:09] also, it seems armhf builds in Debian are not all done, so I need to upload qttools ubuntu version for its symbols again, after qtdeclarative rebuild finishes [07:10] Mirv: if you can add at your end of day some infos, I guess we can have someone doing that [07:10] didrocks: yeah, will do [07:10] Mirv: ack on qttools [07:11] didrocks: in other news, http://pad.ubuntu.com/qt52-dependencies should be now ready - ie it includes (bolded) all the packages that need something else done than re-building trunk or current archive version [07:11] I guess for non-CITrain packages a manual upload would be done, and for CI Train packages a merge request [07:12] Mirv: can we have a trial without any rebuild on our packages, just the Qt part? [07:13] Mirv: like, running all AP tests [07:13] (do you think that will be ready soon?) [07:13] (on armhf) [07:13] I guess that will give some data on management who wants to do how Qt handled minor release, even with this ABI break [07:14] didrocks: yes, the combination of the silo and the beta2 PPA should give something to test where the Qt of the silo is used instead. I think after qttools and qtmultimedia are ready quite many tests should use the newer builds [07:14] qtwebkit won't be ready by my EOD, but hopefully I get far enough (qt3d, qtsensors, qtlocation) that I can push a build of it [07:15] Mirv: ok, keep us posted :) [07:16] (nice to have that bar on the pastebin) [07:17] didrocks: do you by the way know any way to get the builder output all changed symbols, since it tends to stop after the first symbols file? that's why I've now the second rebuild of qtdeclarative for powerpc symbols [07:17] Mirv: you can lower the warning so that it doesn't trigger any error [07:17] but that's about it [07:17] right [07:17] (and it will then need another rebuild) [07:42] qtdeclarative finished on powerpc \o/ [07:43] that's actually the first time ever, since V8 wasn't available for powerpc before 5.2 [07:44] yeah ;) === popey_ is now known as popey === didrocks changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cihelp | CITrain support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone | Landing instructions: http://goo.gl/8H1Du3. Landing in degraded mode (see http://goo.gl/J1EqPW) | Known Issues: - [08:14] sil2100: hey, how are you? [08:22] didrocks: hey! Upgrading my system ;) [08:22] How about you? [08:23] sil2100: I'm good, thanks (even if it's rainy here) [08:23] sil2100: can you prepare an order of safe landings for the meeting that you and Mirv will be able to proceed today? As you are going to test them as well, try maybe to match them in logical chunks? [08:25] Ok, on it then [08:27] thanks! [08:55] Damn, my phone is acting strange [09:20] \o/ look at line 23 [09:20] Wooooo [09:20] Didn't know we had that feature! [09:20] well, we didn't have it 5 minutes ago :p [09:20] ok, let me get that triggered every 5 minutes [09:21] You did-*rock* ;) [09:21] heh ;) [09:21] (there are a lot of checks to ensure we don't overwrite if a job is in progress or rollback or whatever) [09:21] and it's checking every silos [09:21] Let's just not trigger the spreadsheet breakage again! [09:23] heh, I guess it won't make that many changes [10:02] grrr, delivery of a 20 kgs pack between 8am and 7pm [10:03] and of course, they choose the meeting time [10:10] === Image 212 Building === [10:15] ;) [10:30] I might be able to eat some very random "lunch" now while the phone is flashing [10:36] ok, automated migration check done [10:36] and deployed [10:39] didrocks: okay so looking at 211 it is pretty much on par with 210, I'll carry on testing it but I've not had any crashes yet [10:39] davmor2: great! keep us posted :) [10:39] bzoltan1: hmm.. one of your merges doesn't seem to exist anymore [10:39] bzoltan1: in the list of 13 landings [10:40] bzoltan1: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/tabbar-apfix-unit/+merge/208368 <- the branch is there but no merge request, can you clarify this? [10:40] didrocks: well other than the camera app :) which we know and blame rsalveti for whole heartedly right ;) [10:40] sil2100: that must be renato's one [10:40] davmor2: sure sure ;) [10:40] sil2100: ohh that one... sure, a sec [10:41] sil2100: ah, ensure renato one (if still listed) is removed btw, it's already listed [10:41] landed* [10:41] bzoltan1: ah, ok ;) [10:41] didrocks: I mean: ah, ok [10:41] ;) [10:41] bzoltan1: could you also check for renato one? ;) [10:44] psivaa: ok, so gallery_app is the v8 crash [10:44] psivaa: so, what I have done to quickly check [10:45] wget [10:45] apport-unpack foo [10:45] gdb gallery-app foo/CoreDump [10:46] then I'm seeing: #0 0x50b0d8f8 in ?? () [10:46] (on the phone of course) [10:46] didrocks: ohh on the phone. [10:46] phew [10:46] didrocks: thanks for the steps. [10:46] psivaa: yeah, I tried on the desktop, but symbols don't match [10:46] yw! :) [10:46] no need to install any dbgsym, nothing :p [10:46] sil2100: that MR was replaced by an other one [10:46] just work on /tmp [10:47] sil2100: renato's legendary MR has landed on the trunk already and so it is not in my MP [10:48] bzoltan1: could you remove/replace what's needed in the list? I'll assign a silo then ;) [10:50] sil2100: done [11:02] sil2100: err, unity-scope-scopes isn't installed by default on the image? [11:02] so what exactly needs testing I'm wondering [11:03] Mirv: if it's not installed (it's for the new scope API), just get it in [11:03] Mirv: the real testing will come when we activate them [11:06] thanks, just double checking because of the toll this context switch is having on my brain :) [11:09] heh ;) [11:09] I'm pretty sure it's not activated anyway [11:09] (it's on the mwc image) [11:10] I just love my ISP [11:19] I kind of love my ISP but I'm currently pondering if I love it so much I'd get a two year fixed contract - the contract terms are usually so that even if in reality it would start to suck, there's no real chance to abort the contract [11:19] I'd just get 100/10 instead of 24/10 and for slightly less price [11:20] but I generally hate all "bind yourself to us for N years" deals [11:30] Yeah, I'm also bound to a 2 year contract here [11:41] sil2100: so based on your experience it'd be an valuable option to be able to switch :) [11:49] didrocks: packaging ACK needed! Looks ok, although the description seems to be doubled ;/ python 2 and 3 ones have the same desc... not sure if we should block on that? http://162.213.34.102/job/landing-009-2-publish/lastSuccessfulBuild/artifact/packaging_changes_window-mocker_1.4+14.04.20140220.1-0ubuntu1.diff [11:49] === Image 212 DONE === [11:50] (yay, and my desktop build notification tool works) [11:50] didrocks: did you want me to dogfood 212? [11:51] sil2100: +1, I think the package name helps knowing what it is [11:52] popey: weather-app only + everything related to sound [11:52] popey: if you can rerun the weather-app AP test btw, just in case :) [11:52] definitely dogfood it [11:52] make sure calls work again [11:52] k [11:52] thats the importand change in this image [11:52] so, after that one, only: [11:52] *important [11:52] - the qt crash [11:52] - camera app [11:53] - clock apps [11:53] should be the remaining issues [11:57] popey: On the click updater I saw the new image for the calculator however once it had installed I still see the old one, also the calculator doesn't seem to open once updated even after a reboot so I'm wondering if something got broken there [11:58] didrocks: ^ other than that issue and the camera app locking up all seems good here on mako :) [11:58] davmor2: excellent, and so, no crash at all? [11:58] didrocks: not so far [11:58] great [12:00] popey: the new weather opens fine I don't know if that was the fix for the qt5.2.1 issue we saw or what :) [12:04] didrocks: packaging change! http://162.213.34.102/job/landing-014-2-publish/6/artifact/packaging_changes_unity-scope-click_0.1+14.04.20140226.1-0ubuntu1.diff <- added dependency, it's used so it should be ok [12:04] sil2100: yeah, +1 [12:06] mh, that cordova thing is a bit hard, but I guess I can downgrade and see if it looks identical with that [12:06] I've a "non-styled" app after instead of just running on device I installed it. plus the instructions on the google docs don't exactly match what I have in the SDK [12:08] ah hmm, I guess I should use the SDK PPA even though archive version is only a week old [12:08] maybe that helps with the desktop side of things and wrapping up what's going to get installed [12:13] sil2100: seems ubuntu-ui-extras had a silo but nobody checked the lander for him [12:13] so I assigned bill as he requests to flush trunk [12:14] didrocks: indeed, right, thanks! [12:14] yw === MacSlow is now known as MacSlow|lunch [12:19] eh, now the docs instructions for testing HTML5 apps are even more wrong [12:20] hum, no dbarth [12:20] Mirv: maybe just put a comment and move to the next one then? [12:20] (put your comment on red background) [12:20] ;/ [12:24] sil2100: can you check if l12 is proper now? I just added my stuff to the comments [12:25] moving to unity-mir next [12:25] didrocks: 212 seems okay for clock/camera/audio/dialer (not fully dogfooded) [12:25] popey: great, tell us if you see any crash impacting you [12:26] thanks :) [12:26] so, maybe we'll be able to promote an image once the camera+clock fix are in [12:26] didrocks: will do, i have a couple of meetings but will dogfood properly in a bit [12:26] and will update the sheet and ping you when done [12:26] popey: I count on your breakage expertise! :) [12:26] (take care of the phone glass though, not that kind of breakage :p) [12:27] * didrocks runs [12:29] ;) [12:30] no answer on the troll, even not a disdain "ahahah", so disappointing! [12:30] sil2100: more precisely, do I need to switch some testing "unsuccessful" switch or is comment enough? [12:31] :þ [12:33] Mirv: comment and you can even switch the Tested field to No if needed ;) [12:34] that's what I was wondering about. yep, I'll switch it so it's more clear it's not been (successfully) tested (by us) [12:35] AH HA [12:35] Good one [12:35] I'm testing now address-book-app and wanted to check the testing guidelines to see if I'm doing it right [12:35] I click the links and HA! Empty! [12:36] * Mirv updated unity-mir and has a black screen after reboot. is this bad? :) [12:36] NOTREGRESSION [12:36] ;D [12:36] Just kidding [12:36] Really? === alan_g is now known as alan_g|lunch [12:37] yeah. unity8 and maliit-server crashed, apport running. [12:37] I'll need to do some reiterations downgrade/upgrade I guess first [12:37] Mirv: is it a fresh flash or just added the unity8 stuff? [12:37] davmor2: fresh flash + landing-011 (unity-mir) (and well also html5 update) [12:38] hmm, another reboot and no crash [12:39] have we had random unity8 crashes lately on the images? [12:39] didrocks: terminate called after throwing an instance of 'boost::exception_detail::clone_impl >' [12:39] maybe I pretend what happened didn't really happen and run unity8 tests and do a couple more reboots after that [12:39] expecting that? [12:41] popey: not really [12:41] you do have a crash file? [12:41] not a recent one [12:43] didrocks: getting a few crashes [12:43] I'm being a bit brutal with it [12:43] ok, can you try to clean /var/crash and retry? [12:43] to see if it's reproducible [12:43] ya [12:44] had one unity crash and a bunch of app crashes / oddities [12:47] popey: so you too got unity8 crash on stock image (or whatever you're testing?) [12:47] I'm retracing one unity8 crash file now on desktop (if it works) while running unity8 tests [12:48] Mirv: this is stock 212 [12:49] I'm using 211 - landing-011 [12:53] bah, can't reproduce it now [12:53] did exactly the same thing [12:56] Mirv: don't retrace, just try to look at the address first [12:56] if it's d8f8 [12:58] didrocks: how to look at the address? [12:59] -> hangout, but AP should be soon ready and I did manual testing too already. if AP ok I'll do a couple of reboots more, and it should be ~ready [12:59] Mirv: excellent! so it's what I pasted for psivaa this morning [12:59] one sec, looking [13:00] didrocks | psivaa: ok, so gallery_app is the v8 crash [13:00] didrocks | psivaa: so, what I have done to quickly check [13:00] didrocks | wget [13:00] didrocks | apport-unpack foo [13:00] didrocks | gdb gallery-app foo/CoreDump [13:00] didrocks | then I'm seeing: #0 0x50b0d8f8 in ?? () [13:00] didrocks | (on the phone of course) [13:00] Mirv: ^ [13:00] retrace is ready, though, will look tha to [13:05] Mirv: so can we for the sake of testing back out the binary packag rename until we are ready? [13:05] makes no sense to shoot ourselves in the foot [13:05] imo [13:06] Mirv: its very very important that we can test stuff without the apps getting rebuild [13:06] all the apps that have no private headers and dont use qreal [13:07] in API [13:07] asac: it'd restart the building once again, but I guess it could be doable. but not without approval from slangasek who in the first place asked Debian to do the rename and Debian accepted it to help us [13:07] Mirv: you are the man who preps stuff in the ppa [13:07] you dont need slangasek approval for our testing [13:07] really [13:07] thats crazy. b [13:07] we are not saying that we wont do the package rename for the landing. just for testing etc. [13:07] asac: well it seems going back and forth with different solutions === MacSlow|lunch is now known as MacSlow [13:08] its not going back and forth. its just operationally we cant rename it right now [13:08] asac: then for landing it would be another 1-2 weeks of rebuilds, if now the rename would temporarily removed [13:08] we can do the rename before we do the push to archive [13:08] Mirv: ok, what you can do is include a compatibilty package for the testing [13:08] so we just need to drop that [13:08] so I think it needs discussion before I remove the PPA [13:08] and no rebuild after [13:09] can we screw apt? [13:09] and tell to --force-depends [13:09] or something like that [13:09] we can just shiup empty package named like old and depend on the new package [13:09] didrocks: I think for testing purposes that would be possible and a more sensible path [13:09] while we test [13:09] whatever [13:09] find a solution [13:09] some dummy libqt5core5 package could work too [13:09] without infrsatructure work [13:10] I would prefer the dummy libqt5core5 package approach [13:10] but now back to -> hangout [13:10] please do the dummy one for the ifnal testing [13:10] thanks. please upload that asap so we can do the testing we need [13:15] didrocks: http://pastebin.ubuntu.com/7010279/ ? [13:16] Mirv: should be arch:all [13:16] section oldlibs [13:16] (if we want to keep it) [13:16] otherwise +1 [13:17] yeah this multi-tasking is quite hard [13:26] Time for some lunch === alan_g|lunch is now known as alan_g [13:40] popey: if I install manually a click package, it should appear in installed, right? [13:40] i tend to.... [13:40] adb push foo.click /tmp [13:40] yeah [13:41] then pkcon install-local ? [13:41] adb shell then sudo -u phablet -i, then pkcon install-local /tmp/foo.click [13:41] then search for it [13:41] ah [13:41] I'm dumb [13:41] I was root [13:41] ya, everyone does that [13:41] ahah :) [13:41] hum [13:42] ok, black scren [13:42] screen [13:42] which app? [13:42] I think it's because I need some permissions or whatever :p [13:42] popey: a dummy test app, just generated for default qtcreator [13:42] ah [13:42] (weird we don't ship the correct manifest) [13:42] I'd look in /home/phablet/.cache/upstart [13:43] or dmesg -T | grep DEN [13:43] libust[7388/7392]: Error: Error opening shm /lttng-ust-wait-5-32011 (in get_wait_shm() at lttng-ust-comm.c:881) [13:43] no other error [13:46] ok, it works from qtcreator [13:47] popey: do you still have a device with Qt 5.2? [13:48] didrocks: no [13:48] do you have the instructions somewhere so that I can upgrade? [13:49] ya [13:49] https://launchpad.net/~canonical-qt5-edgers/+archive/qt5-beta2 [13:49] == Updating on device == [13:49] obviously make it rw and reboot first [13:49] oh! :) [13:51] popey: ok, I'm dist-upgrading, seems that's all needed then, thanks ;) [13:57] asac: I did a test build locally and now before switching hangouts uploading to PPA [13:57] mhr3: line 14 ready for clean and merge [13:58] didrocks: the EOD status is actually fully documented at http://pad.ubuntu.com/qt52-dependencies in case it'd be wanted that we go forward with recompilations, which I guess we won't be doing. but just in case, that doc has all the information landing team should need with Qt 5.2 [13:58] Mirv: thanks! ... remember to drop the transitional package for the final upload i guess - unless we want to go through the motion of talking to steve again [13:58] Mirv: thanks a lot :) [13:59] sil2100: so unity-mir and the scopes published [14:02] morning [14:03] * rsalveti reading backlog [14:04] didrocks: what controls rotation in apps? I mean, which system component says "ok, we're 90 degrees from normal now, rotate!" ? [14:04] popey: I would say qtubuntu-sensors, but TBH, I'm not 100% sure [14:05] looks plausible [14:05] popey, right now, it's qtubuntu-sensors [14:05] popey, but that's soemthing that will change over time [14:06] ok, i need to file a bug for qt5.2 [14:06] oh, no I don't i mean for image 212 [14:06] i didnt say 5.2, ignore that [14:06] you did say it [14:06] I read it! [14:06] * popey edits the logs [14:06] heh :) [14:06] and my memory as well? [14:07] * ogra didnt see 5.2 anywhere [14:07] :P [14:07] *** Problem in qtubuntu-sensors [14:07] This is not an official Ubuntu package. Please remove any third party package and try again. [14:07] dammit [14:07] but why is popey working with qt212 [14:07] oh, i have way too many devices [14:17] didrocks: bug 1286150 [14:17] bug 1286150 in qtubuntu-sensors (Ubuntu) "Rotating detection is very sensitive mako #212" [Undecided,New] https://launchpad.net/bugs/1286150 [14:17] just nudging or tapping the phone causes it to try to rotate [14:18] popey: hum, I don't see anything in qtubuntu-sensors that could have caused it [14:18] could be android/hybris? [14:18] rsalveti: any idea? ^ [14:18] it could well have arrived earlier than 212 [14:19] I can check [14:19] popey: let me flag 202 just in case [14:20] popey: can you try 206 meanwhile? [14:20] just to know if it's pre/post-android 4.4 [14:20] didrocks: does that need a 4.4 downgrade? [14:21] popey: not for 206 [14:21] ok [14:27] ugly recovery logo on 202 :p [14:33] popey: I don't see real difference between 202 and 212 [14:33] on the titling [14:33] bang the side of the phone [14:33] it was always very sensitive for me [14:33] it's as sensitive [14:33] as it was [14:33] to me [14:33] I can try 209 [14:33] to compare [14:34] (also i dont bang my phone)) [14:34] to simulate [14:34] i mean, i notice it flip when i put it down on the desk, even on an incline [14:34] yeah, they were also sensitive here as well [14:34] i had that since ages [14:34] hmm [14:34] okay [14:34] ignore me :D [14:35] pretty annoying if you put it down while reading a text [14:35] just to have your hands free [14:35] text [14:35] right [14:35] ☻ [14:35] ogra: I'm trying 196 [14:35] (and you delay my break! ;)) [14:35] i didnt claim it was broken ! [14:35] blame popey [14:35] argh [14:35] popey: ^ [14:36] http://blamepopey.com/ [14:36] sorry "all the same" :) [14:36] ++ [14:36] popey: you lack blinking tags :) [14:36] I apologise most sincerely from the bottom of my heart. Can you ever forgive me? [14:37] popey: thanks, I just wanted you to feel guilty :) [14:37] good luck with that [14:37] roh [14:39] didrocks: so ubuntu2 of qtbase compiling in the PPA. transitional package to be dropped for final release, and at the moment (after the meeting discussions) it does not seem anyone is actually going to investigate mixing of packages / ABI breakage right now. but the transitional package is now there for time being. I can remove it with ubuntu3 upload on Monday or something. [14:40] Mirv: ok, so let's see… keep it until next week, yeah :) [14:40] do you guys know when l10 is planned to land? [14:40] the fixes for the concurrent checks of upgrades [14:40] didrocks: sil2100: secondly, it would be "go ahead" about continuing now as listed at http://pad.ubuntu.com/qt52-dependencies for example by starting by giving CI Train the multimedia-touch link, then dee-qt/libqtdbus* etc in order. it's up to the landing team I guess to organize whether I'll do it alone or not. [14:40] we start having a stack of ubuntu-system-settings changes in the queue, one blocking toolkit changes [14:41] the qtwebkit finished building, Qt is essentially done now in landing-006 [14:41] seb128: do you know what is "We want manual testing first" [14:41] about? [14:41] didrocks, no, I want to know if anyone is actively working on making that happen [14:41] Mirv: we are tha landing team :) [14:41] e.g ralsina_ [14:41] ;-) [14:42] didrocks: yes, so up to us/you to decide eg. in the evening meeting :) [14:42] I'm EOD now, but the plan should be ~clear [14:42] in the pad [14:42] Mirv: right, depends on how much we need to land for the rest, I guess, but yeah, it's clear ;) [14:42] yep [14:42] seb128: yeah, we need him I guess to know on that one [14:42] seb128: yes? (/me reads backlog) [14:42] ralsina_, hey [14:43] ralsina_, I'm wondering when the fixes for the concurrent upgrade checks is going to land (l10), we have a mp for settings there and it's blocking any other setting work to get a slot [14:43] I am not really the right person to land system settings, my team only did a plugin for it. The "we want manual testing" was laney [14:43] * sil2100 prepares for testing unity8 [14:44] Laney: ? ^ [14:44] The u-d-m and ubuntu-system-settings one is a bit fishey [14:44] ralsina_, u-s-s is only a small part of that landing ask [14:44] right [14:44] So I wait with that one still [14:44] sil2100: what is fishy? [14:44] both barry and mandel have tested the other parts a lot [14:44] ok [14:44] I'm going to test that ppa on my n4 which is on 209 [14:45] let's see how that goes [14:45] seb128: if you +1 on that one, I'm happy that you publish it [14:45] (with your testing) [14:45] didrocks, thanks [14:45] popey: ok, same sensitivity here [14:45] with 198 [14:46] sil2100: going for a run, once unity8 is tested, can you get (if you have packaging changes) other core-devs +1? [14:46] asac: line 26 on sheet are build only changes for qt5.2. no functionality added/changed [14:46] sil2100: so that we can start another image soon [14:47] didrocks: any chance we could get a silo for line 20 today? [14:47] bfiller: sounds legit, assigning now [14:48]