[06:17] Mirv: hey, how are you? [06:34] didrocks: morning, fine if somewhat tired. [06:34] didrocks: how're you, you're working today? robru thought you'd visit only for a day. [06:34] the image #92 seems great on mako [06:34] at least compared to how it was before [06:34] Mirv: no, if was 2 days (I came back yesterday late, at 10PM, and then an hour of non stopping requests ;)) [06:35] Mirv: so, we have quite some good stuff flowing [06:35] Mirv: we'll need to restart building Mir as soon as duflu is done with his branch [06:35] I'll handle that one [06:35] Mirv: gush is reviewing https://code.launchpad.net/~dandrader/ubuntu-keyboard/improve_kbd_info_ipc/+merge/190418 [06:36] Mirv: as soon as it's done, can you handle ubuntu-keyboard? [06:36] we need both in the first image this morning I guess [06:36] would be nice to have this image start building in ~1h [06:36] dbarth would like to have qtmultimedia upload, I've prepared it this morning although I'm not sure about all the hoops [06:36] right, ok... [06:36] (req 184) [06:36] Mirv: after this image [06:36] (so 93) [06:37] Mirv: we'll handle all those requests after 93 to catch up I guess [06:38] ok, so refreshing that keyboard branch page [06:50] Mirv: maybe check with gush if we can't merge manually to speed things up? [06:53] ok [07:04] vila: hey! [07:04] vila: I don't get 36+ line = line + ' ' # Add a space to preserve lines in the diff [07:04] also: [07:04] === modified file 'tests/data/bzrlogs/classiclogdiff' [07:04] === modified file 'tests/data/bzrlogs/logswithlaunchpadcommits' [07:04] those were juts bzr logs output from bzr directly [07:05] (so not sure why they need modificatoin) [07:05] didrocks: the split later needs a separation otherwise lines are concatenated [07:05] didrocks: almost from bzr log, some lines were arbitrarily split in the middle of words [07:05] vila: they need to be concatenated for dch, right? [07:06] didrocks: they will be re-concatenated but they need to be properly split first [07:06] vila: not sure to follow you :) [07:06] achhh several splits are involved... [07:06] the root fix is to remove the leading spaces added by bzr log, [07:07] yeah, that one, I got it :) [07:07] but that triggered a bug (shown in the bzrlogs/* fixes) [07:08] vila: do you have a good example? if we remove the leading space, the words will be cut in the middle? [07:08] from there we have a clean commit message from bzr but we've lost the spaces that were preserving the line in the join/split/Fixes:/... damn can't continue just from memory, let me get the mp back ;) [07:09] vila: thanks :) just wanting to understand, I trust you, but better for me in the future ;) [07:09] * vila nods [07:11] ha right, the split I'm referring to is not shown in the diff [07:11] didrocks: _extract_commit_bugs [07:12] didrocks: the rsplits get rid of stuff we don't care, the final split needs spaces [07:12] didrocks: the trailing one I'm adding sort of replace the leading ones I'm removing [07:13] vila: ok, got you ;) [07:13] let me merge it then [07:13] didrocks: look closely at the diff in the bzrlogs... ok [07:14] didrocks: gusch has just identified a problem/crash with the keyboard MP [07:14] finishing anyway ;) : the fixes in the bzrlogs files shows 'of' (in 'of the barrier') being cut leading to a line with no leading spaces [07:15] vila: yeah, this is because of this leading whitespace removal? [07:15] ah, saw now on other channel [07:15] Mirv: yep ;) [07:15] Mirv: ok, I'm handling Mir, prepare the other ones please meanwhile, please [07:15] Mirv: or you need more to be able to prepare? (I can add some) [07:15] (fixing the arm64 issue meanwhile) [07:16] didrocks: cumulative effect of the fixes around the spaces yes, a line with no leading spaces in bzr log became a bogus input since bzr cannot produce such lines [07:17] didrocks: can you specify "other ones", to be sure? the qtmultimedia for dbarth, and/or qtdeclarative for tsdgeos? [07:18] vila: ok, sounds good :) [07:18] Mirv: can you do indicator + unity stack? [07:19] (unity7) [07:19] didrocks: the whole stacks, of both? is there a landing plan item for those? [07:19] Mirv: I guess unity will have no impact on the phone, so good to refresh it and stage that in proposed [07:19] Mirv: I'll add the landing plan in a few [07:19] aha, ok [07:19] Mirv: indicators will impact phone, so needs testing first :) [07:19] (on phone) [07:20] we also don't have desktop autopilot tests, still [07:20] Mirv: yeah, I know :/ I think there is a pending dbus fix [07:20] Mirv: we'll get that one today [07:20] didrocks: it got uploaded last night, I updated the container + host but still getting http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2511/console [07:21] Mirv: can you check with jibel once he's around? (or vila?) [07:21] vila is already looking at it [07:22] vila: any progress? ^ [07:22] didrocks: and I'll close the loop with jibel when done [07:22] ok [07:22] thanks [07:22] didrocks: so far the bug is not fixed but the lxc/dbus guys ruled out otto (which is good news) [07:23] didrocks: fixes are in the pipe for dbus and lxc, should have landed even, I'm unclear about the host/lxc setup now, either we missed a step and the bug is fixed or we don't and it's not ;) [07:24] oh, silly me jibel is there ! ^ otto is ruled out \o/ [07:24] vila: both dbus and lxc saw updates last night and are in release pocket [07:24] vila: they did land [07:24] didrocks: yes they did [07:24] didrocks: the question is: do we really use them [07:24] I did update both intel and nvidia-gtx660, both host and inside lxc-start (dist-upgrade) [07:25] if there's some service restarting/rebooting needed additionally, those I didn't do [07:25] Mirv: right, that left the iso, can't see why dist-upgrade won't cover that though... [07:25] vila: I saw them getting upgraded on both, so the packages are installed [07:25] Mirv: but I'd rather make sure we have a clean plate before bouncing to the lxc/dbus guys [07:25] *make triple sure even [07:26] vila: what do you mean? [07:26] vila: feel free to log in, look around and do reboots [07:26] I think Mirv upgraded + reboot? [07:26] Mirv: but I'm not super optimistic it will change anything [07:26] vila: you know it's blocking the whole production for quite some days and we have to deal with everything by hand? it's really a priority [07:26] (and should be fixed when it happens in the hours) [07:27] didrocks: you know I'm been focusing on it for the last two days ;) [07:27] vila: I wasn't there before yesterday, just giving you the context [07:27] basically, the consequence right now is "ok, let's dimiss the tests" [07:27] didrocks: sure, no worries :) [07:29] didrocks: the only updated thing from unity stack affecting touch is unity-scope-mediascanner, so testing that too in addition to indicators [07:31] vila, to be certain new dbus and lxc are used I'd reprovision a new container [07:31] Mirv: oh, please ignore mediascanner [07:32] Mirv: there is a whole set of update for mediacsanner [07:32] vila, can I abort current test ? [07:32] jibel: I think so [07:32] didrocks: aha, ok [07:32] Mirv: FYI, code for arm64 ignoring -> done [07:33] jibel: be aware that the host is using the -11 kernel currently [07:33] didrocks: thanks, helpful, I skipped indicators arm64 again today [07:33] Mirv: if you see anything broken, ping me :) [07:33] sure [07:33] vila, I'm marking saucy nodes offline [07:33] vila, abort the tests and reprovision [07:34] jibel: ok, let me know when there is a run that can confirm the fix or not [07:34] vila, I'll also revert the kernel to latest version [07:34] jibel: great, thanks [07:34] vila, kernel on the host and container must always be in sync otherwise graphics driver do not match [07:35] Mirv: indicator-network is in as well? so it will affect touch (but fine for those one) [07:35] jibel: right, stgraber wanted to test a theory requiring the -11 driver [07:35] vila, jenkins configuration must be fixed. I can disable nodes as anonymous [07:36] Mirv: let me note all that in the landing plan [07:36] jibel: cant ? [07:36] vila, I *can* without being logged in [07:36] it shoudn't be allowed [07:36] didrocks: indicator-network, indicator-bluetooth, indicator-messages, indicator-sound, indicator-location and their support packages are all in touch [07:36] + ido and qmenumodel [07:36] jibel: errr, weird... yeah, but in the present case, good for you ;) [07:37] didrocks: if you don't want all of those, tell [07:37] vila, uh, i can login, I just didn't realize I wasn't [07:38] jibel: I pinged retoaded and fginther about that in ci and will followup [07:38] didrocks: and bamf as well [07:38] (on desktop) [07:38] Mirv: we do want those, I'm trying to list ;) [07:38] jibel: may be that's expected based on vpn access being required anyway [07:40] vila, I don't think doing everything as anonymous, or root or other generic user is a good policy even behind a vpn but well [07:40] I can also try the desktop testing by moving to a backup machine and letting this main machine churn AP (for some reason I has some problem the last time although usually it's been fine) [07:40] jibel: agree, just trying to understand why it's setup this way, there may be hysterical raisins I'm not aware of [07:41] Mirv: ok, should be updated now [07:41] Mirv: tell me if that makes sense (trying to list some dogfooding for touch) [07:41] jibel: *I* wasn't able to look at job configurations 2 days ago... [07:41] yesterday even [07:42] jibel: I blame lack of tests all over the place ;) [07:43] Mirv: indicator-messages is also in 172 [07:43] didrocks: should that be all indicators + unity except mediascanner (which is in unity) [07:43] Mirv: all indicators + all unity (so compiz, nux…) expect mediascanner, right [07:43] and desktop AP:s on those, and then touch specific tests [07:43] right, makes sense [07:43] except* [07:43] jibel: only put autopilot-intel is offline, shouldn't autopilot-ati and autopilot-nvidia receive the same treatment ? [07:43] Mirv: exactly! [07:43] grr [07:44] jibel: only autopilot-intel is offline, shouldn't autopilot-ati and autopilot-nvidia receive the same treatment ? [07:44] Mirv: please talk with the release team to got most of the things in proposed at least [07:44] Mirv: will be nice to have the touch ones in distro (system-settings depends on those indicators) [07:44] vila, -ati is not used and -nvidia is for raring [07:44] jibel: ok [07:45] vila, nvidia running saucy is gtx660 [07:45] jibel: qa-nvidia-gtx660 ? [07:46] right, as seen in http://10.97.0.1:8080/job/autopilot-saucy-daily_release/configure [07:49] didrocks on #90 I am seeing the issue ogra_ saw the other day, memory leak using mir. init is now using ~1GB RAM on my device [07:49] 1816 phablet 20 0 994m 789m 1008 S 0.3 42.2 8:39.77 init [07:49] there is a fix in the landing asks page i think [07:49] ok, coolio [07:49] from slangasek [07:50] didrocks, why did you strike 181 ? [07:50] didrocks, did you just upload it ? [07:50] ogra_: wasn't that yesterday's lool upload? [07:50] no [07:50] ah, he updated with duplicate stop [07:50] thats only half of it [07:50] it still crashes when you stop it [07:51] ogra_: because of the duplicate stop? [07:51] ogra_: mind adding that to the landing plan for image 94? [07:51] lools fix makes it only start and run without crashes [07:51] yeah [07:51] thanks! [07:51] popey: ok, so confirming what we thought [07:51] let's hope slangasek's fix will do it [07:51] popey, line 182 in landing asks should help [07:52] (hopefully) [07:55] vila, no improvements, did you see denied on receive too or only send of dbus messages? [07:56] didrocks, there is another issue with autopilot and maliit, i dont see an entry on the page for it, veebers and thomi were working on it [07:56] jibel: I think they stop the diagnosis on the first send so we may be seeing something new [07:57] jibel: what are *you* seeing ? [07:57] didrocks, the unity8 autopilot test does not use upstart to start/stop unity during its runs ... that means maliit doesnt get started along [07:57] vila, can you remind me when this error started exactly? [07:58] jibel: opinions diverged on that topic, mine is based on the job history... [07:58] ogra_: I'm just passing though, but there is this: https://code.launchpad.net/~veebers/unity8/ap_launch_unity_with_upstart/ but it's being held up by this: https://bugs.launchpad.net/mir/+bug/1238417 [07:58] Ubuntu bug 1238417 in Mir "Unity does not process events from evdev device created before unity is restarted (autopilot tests)" [Critical,Confirmed] [07:59] ogra_: I'm hopeing the first link (the branch) can be worked on/merged during my night/weekend (which is now :-) ) [07:59] veebers, oh, thanks ! [07:59] jibel: so circa 8 October, http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/ run 2442 ? [08:01] ogra_: interesting, I don't see anything as well [08:01] * didrocks looks at autopilot trunk [08:02] didrocks, http://bazaar.launchpad.net/~veebers/unity8/ap_launch_unity_with_upstart/revision/435 but blocked by https://bugs.launchpad.net/mir/+bug/1238417 [08:02] Ubuntu bug 1238417 in Mir "Unity does not process events from evdev device created before unity is restarted (autopilot tests)" [Critical,Confirmed] [08:02] jibel: nah, 2443 was still good [08:02] didrocks, see abiove, what veebers said [08:03] jibel: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2449/console seems to be the first with denials [08:03] jibel: so Oct 8 13:07:07 [08:03] veebers: ok, keep us posted please [08:05] vila, http://10.97.0.1:8080/view/cu2d/view/Autopilot%20Release%20Testing/job/autopilot-saucy-setup_otto/label=autopilot-intel/144/console [08:06] is the first time the error appears [08:06] during provisioning [08:06] vila, we want to diff the manifests between ubuntu desktop 20131008 and 20131007 [08:07] vila, then downgrade the host to the packages of the 7 [08:07] jibel: have you seen https://bugs.launchpad.net/ubuntu/saucy/+source/dbus/+bug/1238267 ? It describes the understanding of the root cause for the last fix (which apparently is not enough but should still be relevant) [08:07] Ubuntu bug 1238267 in dbus (Ubuntu Saucy) "AppArmor initialization code should open a file in apparmorfs instead of stat'ing it" [High,Fix released] [08:07] hey popey, already playing with 93? [08:07] 92* [08:08] didrocks: yes [08:08] jibel: now if you see a new message I'd like to update the bug with it [08:08] vila, yes, but my goal is to make the system work, not finding a fix [08:08] vila, once it is back to work, you have all time to find a fix [08:09] didrocks: nothing alarming here [08:09] jibel: I too would like to have a working system and so far the fix has been outside of any source I control (or even know about );) [08:09] didrocks, 92 LGTM, I still have problem to dismiss OSK but it was already on previous build [08:10] vila, are manifests stored somewhere in the lab? [08:10] ogra_: want to play with it on maguro? [08:10] jibel: you have a mako or maguro? [08:11] jibel: I have no idea [08:11] didrocks, mako [08:11] vila, nm, I'll figure from the machine's logs [08:12] jibel: but I'm willing to learn if you give me enough info :) [08:13] vila, or not, there is a gap in the logs between sept 25 and oct 10 :/ [08:13] jibel: so far I've been trying to relay info being people knowing more than me about almost every piece involved. [08:13] didrocks, i can, after the meeting though [08:14] jibel: I'm slowly getting my way about what is involved and which people knows about the various pieces but I'm still far from being able to fix it myself [08:16] vila, I purged the gazillion of kernels installed on the machines, that'll make some room on the disk and speed up upgrades [08:16] morning [08:16] jibel: how ? apt-get autoremove ? [08:17] vila, sudo apt-get autoremove --purge list_of_kernels [08:21] ogra_: hey sil2100! [08:21] argh -ogra [08:21] :p [08:27] ;) [08:27] Hello! [08:29] heya [08:31] asac: joining? [08:31] sil2100: joining? [08:31] coming [08:48] vila, dx-autopilot-intel resumed normal operation [08:48] jibel: seeing http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2514/console [08:48] vila, nvidia is still offline, graphics driver fails to load [08:49] jibel: this will work until http://10.97.0.1:8080/job/autopilot-saucy-setup_otto/ triggers again right ? Should we disable it until we have a fix ? [08:49] jibel: if I'm guessing correctly you forced an old container to be used right ? [08:50] jibel: what needs to be monitored to ensure it stays this way ? [08:50] vila, it will work even if setup run [08:50] vila, I disabled the purge though [08:50] vila, disk space :) [08:50] which purge ? the otto triggered one ? [08:50] ^ this is what you need to monitor [08:51] vila, the update script, updates the host, provision a new container based on lastest desktop image and purges old containers [08:51] you're giving me fish... I won't be able to diagnose anything without a better understanding :-( [08:51] vila, it runs when it detects a new image [08:52] by update script you mean ~jenkins/bin/update_host ? [08:52] vila, yes [08:52] jibel: yeah I've seen the trigger in http://10.97.0.1:8080/job/autopilot-saucy-setup_otto/configure hence my question [08:52] jibel: what I don't get is how the container is selected [08:52] vila, the script that is called in the jenkins jobs you mentionend before [08:54] jibel: ha ! Saw your patch in update_host, ack [08:54] jibel: still miss the container selection though [08:55] vila, also add yourself to the notifications of this job [08:55] jibel: thanks, will update that to use the ci-notifications email when I get to it [08:55] vila, it failed for a while with "E: dpkg was interrupted, you must manually run 'sudo dpkg --configure -a' to correct the problem. " [08:56] vila, in this case it needs a hand [08:56] jibel: ECONTEXT [08:56] update_host ? [08:56] vila, yes [08:57] vila, http://10.97.0.1:8080/job/autopilot-saucy-setup_otto/143/label=qa-nvidia-gtx660/console near the end [08:57] jibel: ha, you switched to debug the nvidia one [08:57] ? [08:58] jibel: sry in the landing ho, hard to keep up with both [08:59] vila, I didn't, I'm trying to figure why the driver doesn't load [09:01] didrocks: didn't see the libhybris upload either [09:01] win 17 [09:01] lool: yeah, it's not in [09:01] vila, I might need to downgrade nvidia to -11 [09:05] jibel: ack. So back to container selection, where does that happen ? I want to make sure I can assist the dbus/lxc guys debugging on autopilot-intel (probably taking it offline when needed) and don't interfer with normal business more than strictly required [09:05] ogra_: mir is in, please kick #93 [09:11] vila, http://10.97.0.1:8080/job/autopilot-saucy-daily_release/ calls $HOME/bin/run_otto_job which selects the most recent container, sets the environment from what jenkins defined and pass everything to otto-run [09:11] jibel: yeah, just found it, so the trick is in the container name [09:11] cyphermox: bluetooth-touch landing >> upload anytime! :-) [09:12] cjwatson: click landing >> upload anytime! [09:12] jibel: so as long the lxc/dbus guys don't need too much stuff on the host they can even work in whatever container they need without even interrupting ci correct ? [09:13] vila, if no other job is running yes, otherwise it is not possible because access to the HW is exclusive [09:14] jibel: ha right, so that means starting another container can't be done [09:14] vila, but they can use a specific container to debug, and you use another one for CI [09:14] vila, as long as they don't run at the same time [09:14] jibel: ack, so either I monitor failures for retries (ugh) or I take the node offline when needed [09:15] vila, offline is safest, jenkins will just queue jobs until it is back up [09:15] * vila nods [09:16] jibel: thanks for the help unblocking that, I think I know how to assist the lxc/dbus guys know, will keep you informed about progress [09:16] didrocks: another question, do we promote 92? [09:16] s/guys know/guys now/ obvisouly ;) [09:17] asac: waiting on ogra's feedback on maguro, if good, yes [09:17] didrocks: let me see if QA also is happy [09:17] jibel: ^^ [09:17] vila, it is not completely unblocked, there is still this problem with proprietary graphics driver [09:17] asac: already got those confirmations [09:17] jibel: does QA have positive things to say yet about 92? e.g. stricly better and awesome? [09:17] ah nice [09:17] let's try to not go in too much useless ping crazyness please ;) [09:17] * asac wont bother about that part of the process then [09:17] we have enough to do :p [09:18] didrocks: well, i think its important to get QA involved. if its now always in our process then its good [09:18] it is [09:18] and i wont ping in the future anymore :) [09:18] jibel: on nvidia not on intel right ? [09:18] (uncoordinated) [09:18] vila, right, intel drivers are open [09:19] nvidia are not [09:19] well, we could use nouveau instead that'd fix the problem :) [09:21] jibel: ack, different issue then. But then how should I read http://10.97.0.1:8080/job/autopilot-saucy-setup_otto/label=qa-nvidia-gtx660/148/console ? Doesn't it mean it works now ? Or is it because you tweaked something there ? [09:22] nice, top segfault on mako :/ [09:23] in the final release will we still have the option of deleting that file and going back to SF? [09:24] vila, update of the host worked because Mirv fixed the upgrade manually yesterday. [09:24] jibel: ha, same trick with the container name, will be nice if you could just tell me so I don't have to guess :) [09:25] vila, but the container doesn't work, you see the dbus denials at the beginning of the log [09:26] Mirv: while you are testing: what's need to be done for qtorganizer5-eds and qtdeclarative-opensource-src? [09:26] Mirv: those were done (testing done) and just need sponsoring? [09:26] (did you retry qtdeclarative-opensource-src under Mir?) [09:27] jibel: right, but that's the message without the dbus.*-0ubuntu10 fix while building a new container, pfew, getting there [09:28] didrocks: I think sil2100 was working on qtorganizer5-eds, I don't know about that update. qtdeclarative was fully tested on non-mir to not cause regressions, plus fixes what tsdgeos wants, so would only need upload unless wanted to be tried to somehow AP tested on Mir (I've been running it on my device, but not ran much tests) [09:29] the unity8 AP still fails for me if I try to run in one run [09:29] Mirv: can you try the calendar-app + notes-app integration tests? [09:29] Mirv: if good, let's push that one [09:29] sil2100: qtorganizer5-eds, AP tested, all good? [09:31] didrocks: ok [09:32] jibel: so, in summary, workaround in place to use KEEPME_ILOVEAPPARMOR on intel and nvidia. Once the bug is fixed for intel, nvidia will still need to be fixed for the driver issue. Correct ? [09:33] jibel: and there is an autopilot-saucy-daily_release in flight to check the workaround works ? [09:34] jibel: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/2514/console ? [09:34] sil2100: i have a 92 image on maguro.. please let me know if everything needed has been built so that I could update [09:36] ogra_: so, image 92 on maguro is fine, promoted? [09:40] still flashing, sorry [09:40] it is rebooting now, gimme 10 [09:40] ogra_: waow, you should start downloading the image even before it's built I guess! :) [09:40] jibel: apparently that summary wasn't correct, saucy-i386-KEEPME_ILOVEAPPARMOR_20131007-0916 is gone on qa-nvidia-gtx660 ? [09:41] didrocks, there was sdome other stuff i had to finish first that blocked it [09:41] sil2100: ok, done with my stuff, waiting for you on qtorganizer5-eds [09:41] ogra_: ah ok ;) [09:44] didrocks: good news, indicators + unity desktop AP better than ever (so they've sneaked some AP fixes in) [09:45] didrocks: I've also gone through the touch manual testing, except that I've only used wifi dialog, not sim password unlock. if needed, I do have a tablet with a microsim card which I guess could be used. [09:45] Mirv: \o/ awesome awesomeness of awesome :) [09:45] popey: hey, around? [09:45] ya [09:45] popey: mind testing latest indicator-network in the daily-build ppa for this password unlock? ^ [09:46] sure [09:46] Mirv: so, you got everything with latest nux, compiz, and all the jazz? ;) [09:46] didrocks, Mirv: I was only dogfooding that in the past, as when I was assigned to it AP tests were bad - but calendar-app autopilots went fine all of them [09:46] sil2100: just tested all of them? [09:46] didrocks, Mirv: by 'only' I mean I just ran the few calendar-app APs [09:46] so I can push? [09:46] didrocks: where's the daily-build ppa? [09:46] didrocks: yep latest nux, compiz, unity, bamf, indicators*, checked with synaptic (I like the sort by version number where I can make sure I'm not missing some support packages) [09:46] Tested yesterday, probably not broken from yesterady [09:46] Mirv: marvellous! ;) [09:47] sil2100: ok, I'm going to push it [09:47] Mirv: so, let's wait for popey's feedback, and if all good, let's push ;) [09:47] didrocks: cool, ok [09:47] Mirv: have you tested on phone the indicator part? [09:47] didrocks: yes, so updated to the chart [09:47] Mirv: excellent! so be ready to push publish :) [09:48] didrocks: yep [09:48] Mirv: now, qtdeclaractive-*? [09:48] yes [09:48] sweet! [09:48] * didrocks pushes while reflashing the qtorganizer one [09:49] popey: apt-add-repository ppa:ubuntu-unity/daily-build , apt-get update, apt-get install libido3-0.1-0 indicator-bluetooth libqmenumodel0 qmenumodel-qml indicator-sound indicator-network indicator-location indicator-datetime indicator-messages libmessaging-menu0 [09:49] reboot [09:49] ta [09:49] sil2100: oh, it's under dailies, feel free to push it yourself then [09:49] didrocks: aye! [09:49] It's in misc [09:50] didrocks, dialker app still doesnt come to foreground when answering a call and i cant save contacts [09:50] didrocks, beyond that i see no regressions (and both of the above are older ones iirc) [09:50] ogra_: not a regression though? [09:51] (well, regression from Mir, but nothing else) [09:51] geez [09:51] * ogra_ just adb'ed in for the first time and ran top [09:51] load is >10 [09:52] ogra_: getting down or up? [09:52] ogra_: I guess we can publish 92 anyway :) [09:52] down now [09:52] didrocks / Mirv what am I testing btw? Just looking for regressions or is there new crack in this? [09:52] around 4 [09:52] * didrocks hopes 93 is soon finished to build [09:52] but stays there [09:53] didrocks, it just did this second [09:53] popey: SIM unlock at least. it has indicator updates as mentioned, so just no regressions [09:53] ogra_: \o/ [09:53] give system-image a few mins to pick it up [09:53] I am missing an indicator [09:53] * didrocks just finished flashing 92, already on a deprecated image :p [09:53] i see no network indicator [09:53] i have it here === alan_g is now known as alan_g|vt [09:53] popey: hmm, I do have it just fine [09:53] want mine ? [09:53] :P [09:53] it is there, i can pull down and see it [09:54] I've the icon as well [09:54] but it says "No signal" [09:54] and I think search is overlaying it [09:54] so it only appears when i pull down [09:54] there.. once I get signal it "appears" again [09:54] I've wifis listed and I can connect [09:54] aha, I don't have experience on the GSM side of things [09:54] ditto [09:55] ooh, i see an access point I never saw before [09:55] blame it then [09:55] muhahaha one of my neighbours has an open access point [09:55] Mirv: i have no lock on my sim.. [09:56] didrocks: i see no regressions from updating those packages [09:56] popey: great! [09:56] Mirv: please, publish both then and ping/reassure the release team on the desktop part ;) [09:57] maybe we can get Laney involved if needed :) [09:57] vila, I reprovisioned again with ubuntu desktop 20131011 which has dbus 1.6.12-0ubuntu10 [09:57] he loves Friday uploads :p [09:57] popey: thanks a lot! ;) [09:57] vila, and it should just work [09:57] ;) [09:57] is this some new definition of "love" that only the french know? ☻ [09:57] didrocks: ok then [09:57] vila, unless bzr breaks :/ [09:57] popey: probably, you know my level of english… I keep applying it though :p [09:57] jibel: for nvidia ? [09:58] psivaa: ok, we seem to be having everything rebuilt now - I guess it's time for testing! [09:58] vila, can you retrigger a check without a build of the apps stack? [09:58] jibel: EPARSE [09:58] popey: Alan! Are you free for testing right now? [09:58] jibel: still trying to figure out where my mental model diverges from reality 8-/ [09:59] vila, and also fix the definition of the stack, there are missing packages [09:59] /var/log/upstart/otto-setup.log: +content-hub [09:59] sil2100: for you, always [09:59] /var/log/upstart/otto-setup.log: +libgsettings-qt1 [09:59] I guess popey will be free once he finishes hacking his neighbour :) [09:59] How nice ;) [09:59] ☻ [09:59] Well, I'll pop round there later and "educate" them on security matters. [09:59] heh [09:59] jibel: I can't do that, I have no idea what you're expecting [10:00] It's almost the same as when I see someone around with WiFi with WPS enabled, so sad... [10:01] jibel: I'm not even sure I understand what you did. Can we serialize a bit so I can catch up ? [10:01] didrocks: pkg acks http://10.97.0.1:8080/view/cu2d/view/Saucy/view/Indicators/job/cu2d-indicators-saucy-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_indicator-network_0.5.1+13.10.20131011-0ubuntu1.diff (libnotify dep add), http://10.97.0.1:8080/view/cu2d/view/Saucy/view/Unity/job/cu2d-unity-saucy-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_compiz_1%3A0.9.10+13.10.20131011-0ubuntu1.diff (ubuntu settings patch grid 1 new s [10:01] jibel: " I reprovisioned again with ubuntu desktop 20131011 which has dbus 1.6.12-0ubuntu10" on nvidia ? You expect the dbus issue we see on intel to not occur here ? [10:02] popey: I'm sure you can come up with education methods that don't need you to even visit them in person [10:02] ooh! Good call [10:02] Mirv: your last sentence was cut after "(ubuntu settings patch grid 1 new s" [10:02] "etting)" [10:02] ;) [10:02] ok +1 on both [10:03] sil2100: what do you need testing? [10:04] popey: let me prepare a list maybe [10:04] ok [10:08] didrocks: 92 getting released soon? [10:09] popey: already done and promoted :) [10:09] or under promotion rather [10:09] oh [10:09] ogra_ is finishing up on maguro [10:09] so be ready for that email [10:09] its ready, just wanted to hit send [10:09] yeah, maguro is fine i'd say [10:10] popey, do it ... i'll publish alongside [10:10] nah, i always wait to send the mail until after i have OTA updated my daily driver [10:10] so i test both flashing clean and OTA updating [10:10] popey: this should be it: http://paste.ubuntu.com/6221772/ [10:10] sil2100: ok [10:11] sil2100: laptop or phone? [10:11] popey: phone, we need to run all the AP tests related and serious dogfooding [10:11] and I take it that's from the daily build ppa I already have enabled [10:11] psivaa: ^ http://paste.ubuntu.com/6221772/ [10:11] popey: yes, just upgrade now to have the latest of the latest [10:11] sil2100: ack, thanks. installing 93 now [10:11] ==== Image 92 / 20131010.2 published ==== [10:11] *all* ap tests... [10:11] erk [10:12] popey: that's what I have written on my Landing sheet :< Run *all* tests [10:12] popey: sil2100: we could split if you want https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0ApRxJIi-SaMddEdtdWtMaXNkMzYtN2puZ2YxdU9MekE#gid=0 [10:13] psivaa: sure, split it up, just let me know which ones I'm doing ☻ [10:13] psivaa: I guess, but for such a serious landing I would also like that some of us run the same tests as well, making sure we don't miss anything or do anything wrong ;) [10:14] So let's do some 'overlaps' for the most serious tests, like unity8 [10:14] didrocks, sil2100: Hold on mediascanner / thumbnailer landing [10:14] ! [10:14] didrocks, sil2100: thostr_ just updated to #92 + packages from PPA [10:14] sil2100: http://paste.ubuntu.com/6221785/ so do I seriously want all of that? [10:14] and found an issue with it [10:14] sil2100: we just found one issue [10:14] he said to hold a bit [10:14] popey, psivaa: SHTAPH! [10:14] ogra_: woot [10:14] heh [10:14] lool: can thostr_ talk to us? or here, that would be helpful [10:14] didrocks: he's here [10:14] :) [10:14] :) [10:14] thostr_: please, sync here directly… [10:14] didrocks: so, we just discovered one issue [10:15] didrocks: yup, just told thostr_ to best shoot here [10:15] didrocks: it all works fine as such but when adding a video that codec cannot be understood something fails internally [10:15] didrocks: we're on it... [10:15] thostr_: what's the ETA for that fix? [10:15] thostr_: which components will need to be update in your opinion? [10:16] didrocks: grilo-plugins, libmediascanner, mediascanner, unity8 and unity-scope-mediascanner [10:16] didrocks: those package bring thumbnailing and caching [10:16] didrocks: however, to be able to test, you need to remove /home/phablet/.cache/mediascanner [10:17] thostr_: all of them will be impacted by the fix? [10:17] didrocks: mail sent [10:17] popey: thanks! [10:17] np [10:18] indicators + unity stacks (minus mediascanner) published [10:18] Mirv: \o/ keep me posted how it's working with the release team [10:19] yep [10:20] didrocks: well, no. unity8 can still be updated [10:20] didrocks: but the others we should land as a block [10:20] didrocks: and, it's not a regression it seems, but something we need to fix nevertheless [10:21] sil2100: can you advance on the unity8 front? [10:21] sil2100: + other services but mediascanner [10:21] tyhicks, thanks for the dbus fix [10:22] vila, Mirv, sil2100, autopilot for daily release is back on intel and nvidia [10:22] media is currently running [10:23] jibel, vila: wow, that was quite some push, nice to have those back [10:23] jibel: thanks! [10:24] didrocks: ok... but won't we need to re-run unity8 tests with the new mediascanner? [10:24] jibel: great ! [10:24] sil2100: no, it's more for the code we already have in [10:24] jibel: can you elaborate a bit as I don't quite get what wasn't working, what you tried, what you fixed and why it's working now ? [10:25] thostr_: It might be a good idea to update to what we have if the fix only comes in late today or monday [10:25] thostr_: IIRC, it's a bug we already have, so unless it causes e.g. unity8 to crash when it used not to, we should update the stack to get all the benefits when it happens to work? [10:25] ok [10:25] thostr_: sil2100 was well underway in his testing [10:25] lool: ok, then let's do that [10:26] sil2100: Let's keep an eye on the ETA of the fix [10:26] lool: and we're trying to get a fix asap [10:26] thostr_: if you can get an ETA for the fix, we can decide whether to wait for it or not [10:26] popey, psivaa: ok, let's test unity8 related stuff for now - please upgrade with the line as mentioned in the pastebinit and let's run unity8 autopilot, dogfooding and normal application APs [10:26] thostr_: sil2100 is EU TZ and today is Friday, so he probably doesn't want to land this past 2pm UTC I'd guess -- sil2100? [10:27] sil2100: ack [10:27] sil2100: when your ~EOD? [10:27] lool: yes! Sadly, today I have an appoitment around 18 UTC, so I might have to EOD around that [10:27] vila: jibel: swee-et, http://10.97.0.1:8080/job/autopilot-saucy-daily_release/2519/ we has results [10:27] asac: http://paste.ubuntu.com/6221814/ [10:28] vila, dbus was not working but upgrading in the container was not enough, so I decided to just rollback the container to the latest known good version to unblock the situation [10:28] vila, I didn't expect that nvidia wouldn't load in this configuration [10:28] * lool lunch & [10:28] didrocks: notes_app + calendar_app 100% pass with qtdeclarative update [10:28] vila, during that time on new image has been publish with the dbus fix [10:28] yeah I haven't eaten anything either and it's soon 2pm, hmm [10:28] s/on new/a new [10:29] this would probably be a good(ish) time [10:29] jibel: ack so far [10:30] vila, I restarted apps but it failed on missing deps in the stack definition [10:30] vila, and re-ran media which seems to be ok, so I removed the warkaroung [10:30] *workaround [10:30] jibel: ack (that bit I couldn't figure and start diverging from reality ;) [10:31] jibel: from there new containers were created and we get succesful runs [10:31] sil2100: ok [10:31] jibel: so basically despite dist-upgrading the containers something was still missing that the new image gave us. Correct ? [10:32] jibel: so basically despite dist-upgrading the containers (which Mirv did) something was still missing that the new image gave us. Correct ? [10:32] sil2100: popey: i'm thinking of taking the click app tests + unity8 +webbrowser so feel free to pick up the rest :) [10:33] unity8 and webbrowser could do better with some overlapping [10:33] psivaa: will run unity8 [10:33] psivaa: sil2100 can i still use phablet-test-run for this or have things changed? [10:34] popey: that still works :) with -n i suppose [10:34] k [10:35] Mirv: \o/ so sponsoring? [10:35] didrocks: yes please [10:36] doing in few minutes [10:37] popey: I'm using phablet-test-run for for instance unity8 testing [10:38] psivaa: ok ;) [10:38] Since I want everyone of us to run unity8 [10:38] I guess there might be failures anyway, but let's compare to the latest smoketesting [10:49] Mirv: are you still around for some more stacks then? [10:52] sil2100: http://paste.ubuntu.com/6221897/ any use? [10:53] popey: i *think we need to run 'stop powerd' before running the tests but not really sure [10:53] psivaa: sil2100 can you give me a set of steps to run to reliably run the tests pls? [10:54] sil2100: you'd be the best for this. i am also doing trial and error.. got 21 failures last [10:58] Can you guys see the mediaplayer-app in the list of apps? [10:58] I cant [10:58] it's in the image [10:58] in fact the installed list is quite small [10:59] popey: shit ;/ [10:59] popey, psivaa: ran the tests as well twice, 22 failures [11:00] sil2100: popey: i ran 'stop powerd' before running for the second time and saw 'OK' in one of the tests.. did not complete yet [11:00] will let you know the outcome in a bit [11:00] psivaa: you think that's the way to go? [11:00] sil2100: i think in the smoke tests that's being done [11:01] psivaa: if that's the case, I'll try that as well [11:01] popey: ^ [11:01] ok [11:01] sil2100: popey: or if you could wait for 10 more minutes i'll confirm if it really worked :) [11:02] did we get 92 out? [11:02] yes asac [11:02] yes [11:02] * popey runs the test with powerd stopped [11:02] amazing! [11:02] stopping powerd? [11:02] asac: ignore me [11:02] which tests are that? [11:02] unity8 [11:02] 93 looks nice so far [11:03] ok ... maybe we have to put the screen in max brightness :) [11:03] ooh, 93 is out? [11:03] (not that the tests are very far yet) [11:03] popey, on the dashboard [11:03] will update once these tests finish [11:03] ogra_: 92 looked similar from tha pass/fail at this stage last time. do we see less crashes as well? [11:04] ok as expected unity8 is still dying [11:04] but the maliit thing isnt everywhere ? [11:05] didrocks: right now, food. after that yes some more. [11:06] popey, psivaa: with powerd disabled it doesn't look any better here at least in some way [11:06] asac, but no maliit crashes yet ;) [11:06] sil2100: yea i get the same 21 failures.. damn [11:07] The screen doesn't unlock itself ;/ [11:07] ogra_: thumbs up! [11:07] yeah [11:07] sil2100: if you disable powerd yuou first have to unlock screen [11:07] otherwise its dead [11:07] sil2100: the phone basically will stay in whatever state you have it when turning of powerd [11:07] * cjwatson finally gets time to test click. landing now [11:07] asac: but we restart unity8 every time [11:08] sil2100: sure, just sayuing you need to turn the screen on before stopping powerd. otherwise it will never come back [11:08] Mirv: excellent! qtdeclarative sponsored btw [11:08] psivaa: can you somehow learn how the smoketests run unity8 tests? [11:08] asac: did that I guess [11:08] kk [11:09] asac, sil2100, well there is still bug 1238417 and https://code.launchpad.net/~veebers/unity8/ap_launch_unity_with_upstart/ [11:09] bug 1238417 in Mir "Unity does not process events from evdev device created before unity is restarted (autopilot tests)" [Critical,Confirmed] https://launchpad.net/bugs/1238417 [11:09] i wouldnt expect much to work without that [11:09] yeah, we need that reviewed [11:09] veebers is working on it this week-end (now then :p) [11:09] ogra_: but how was that working like yesterday? Since I see some smoketesting tests from 10.10 that work [11:10] :< [11:10] sil2100: it should at least starts and pass some tests [11:10] didrocks: how are we supposed to run unity8 tests? We'll have to just do dogfooding + apps AP tests [11:10] We cannot run any test successfully it seems sadly [11:10] sil2100, it was working but will not start maliit if it isnt started via upstart [11:10] sil2100, so all tests that involve OSK will be broken [11:11] So we're blocked by something else then [11:11] sil2100, no idea how the evdev part of this manifests though [11:11] psivaa: can you check how they run the tests? [11:11] sil2100: doing that [11:11] psivaa: thanks! [11:12] sil2100: so unity8 APs are run with phablet-test-run -n unity8 [11:12] with using powerd-cli screen on [11:12] or something before [11:12] asac: we do that, 21 failures [11:12] right [11:12] So only 1 test passing [11:12] ic [11:12] vila: around? [11:13] didrocks: otp [11:13] sil2100: so i think in worst case we can use utah scripts rather than phablet-test-run [11:13] those should have ALL we do in automation [11:13] vila: anyway to break-lock lp:~ps-jenkins/webbrowser-app/latestsnapshot-0.22%2B13.10.20131011-0ubuntu1 without being ps-jenkins? [11:13] sil2100: sergiusens knows how to set things up [11:13] sil2100: utah basically does everything for you: flash phone etc. [11:13] sil2100: do you test on mako? [11:13] sil2100: on maguro the crashes will impact APs more because its slower [11:14] didrocks: not that I know of without diving more, #launchpad ? [11:14] asac, hey, morning; for what? [11:14] vila: I can workaround I guess [11:14] oh [11:14] sergiusens: how to run the stuff exact as utah does it. [11:14] utah [11:14] :) [11:14] i feel if you use utah rather than phablet-test-run we get very close [11:14] sergiusens: thats on image93? [11:15] etrr [11:15] sil2100: ^^ [11:15] sil2100: did you try without our new staged changes? [11:15] * asac installs unity8 AP from archive on maguro and runs [11:15] Get:2 http://ports.ubuntu.com/ubuntu-ports/ saucy/universe unity8-autopilot all 7.82+13.10.20131010-0ubuntu1 [17.6 kB] [11:16] doanac: plars: sergiusens: so whagever we do we have to consolidate utah and phablet-test-run ASAP... all those hacks in utah have to move over or something [11:16] it cant be that we dont know how stuff is run in utah [11:16] noone knows it seems [11:16] ev: ^^ :) [11:17] i am mostly after the setup logic that seems to be doing things in utah that noone knows ohw its done [11:17] a) unlock [11:17] b) screen on [11:17] c) other special stuff for utah etc. [11:17] sil2100: psivaa FAILED (failures=21) [11:17] with powerd stopped [11:18] Same here [11:18] asac: trying with the display on thing [11:18] sergiusens: so if i run phablet-test-run -n unity8 [11:18] it doesnt seem to even kill unity [11:18] at all [11:18] something is broken for sure with phablet-test-run [11:19] i guess since mir it doesnt work with -n anymore [11:19] hmmm! [11:19] well, -n should just call initctl stop unity8 [11:19] i think it does that [11:19] display on might help! [11:19] thats not Mir specific [11:19] but i saw the tests started according to adb/cli [11:19] right [11:19] but i still saw the unity with all those apps in there [11:19] etc. [11:19] Disabling shell [11:19] unity8 stop/waiting [11:19] it says that [11:19] you definitely need to switch the display on before killing powerd [11:19] ...or not... [11:19] but unity8 just continues to be on screen [11:20] asac, Mir just doesnt wipe it [11:20] No idea what to do to get the tests working [11:20] i am sure powerd is a red herring again :) [11:20] try touching it ;) [11:20] now the screen is blank [11:20] didrocks: lool: coming back to the mediascanner issue: seems like some low level stuff still somehow using SF or trying to which causes time outs in gstreamer. so, this only happens under mir... we're investigating more [11:20] asac, is it being bloecked by maliit-server? [11:20] sergiusens: how can i find out ? [11:20] thostr_: ok, we publish the rest but mediascanner as soon as sil2100 finishes testing [11:21] initctl list [11:21] or status [11:21] one sec [11:21] sergiusens, thats fixed [11:21] (or supposed to be) [11:21] didrocks: yes, as it seem proper fix needs to happen somewhere else [11:21] (in #93) [11:22] i dont know ... i think i have 93 [11:22] * asac updates [11:23] heh, if you can update you dont have 93 [11:23] i know i dont have 93 [11:23] still i see good unity8 results on 92 [11:23] in dashboard, so ... [11:23] but also a maliit crash when stopping it i bet [11:26] sil2100: popey just to let you know, when i ran one test alone it passed: http://pastebin.ubuntu.com/6221958/ [11:26] still working on to make all run together === gatox is now known as gatox_brb [11:30] psivaa: ;/ [11:54] asac: can we shift this Mir meeting half an hour later? [11:54] didrocks: of course === gatox_brb is now known as gatox [11:55] didrocks: but the earlier the more time thostr_ and rsalveti have to clear the road for monday :) [11:55] asac: I think thostr_ is busy on mediacsanner :) [11:55] asac: just need to run outside now that everything is under control for my own sanity ;) [11:55] didrocks: rigfht. but its part of the media [11:55] MM stry [11:55] didrocks: ok talk later [11:55] will move by 30 [11:55] thanks! [11:55] (meanwhile apps published) [12:00] didrocks: did you sort out the lock issue ? Probably too late for me to try digging so just on the post mortem pov: what was the problem ? A pending lock ? Or a bzr operation hanging ? (I think both should recover automatically) Or something else ? [12:01] vila: a bzr operation hanging and cancelled [12:01] asac: Sorry, was on a call. I'll kick off a discussion about getting that resolved. [12:01] then, not possible to push again [12:01] ev: yeahg. very important topic :) [12:01] *nods* [12:01] like all this consolidation and getting some automation at the daily-release stage :) [12:02] anyway [12:02] * asac goes for lunch [12:02] didrocks: any job I can look at for reference ? I kind of understand this is not supported if the client is killed with prejudice but could use some more concrete data ;) [12:04] didrocks: in any case, if you have write access I think you can just 'bzr break-lock' if you know what you're doing [12:07] sil2100: popey: so this is working for me: http://pastebin.ubuntu.com/6222107/ [12:07] just in case if you are still waiting :) [12:08] if you get a device not found error, please retry :) [12:09] * vila breaks for lunch [12:12] didrocks: what's the status on updating click package scope? [12:16] sil2100: popey: i got 16 passes (including 2 system settle) and 8 failures in this way. [12:18] is anyone aware of any other UI Toolkit branch that is wanted to be manually merged than lp:~jpakkane/ubuntu-ui-toolkit/video-image ? see https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/trunk/+activereviews [12:18] if not, I'll merge only that and build [12:21] @ci hey, this has been stuck for a while http://91.189.93.70:8080/job/generic-update_mp/799/console [12:24] sergiusens: aborted that. the rest have run [12:24] cjwatson, do you think that would work ? http://paste.ubuntu.com/6222163/ [12:26] ogra_: That's probably OK if you check that the formats match up, but please drop the separator [12:27] cjwatson, the heading you mean ? ok [12:27] ogra_: I don't think it's helpful to have a separator in there breaking the format - better for consumers to be aware that the package names might be either .deb or .click [12:27] yeah, sorry, i looked at my diff instead of the actual manifest file :) [12:27] IMO anyway [12:28] yeah, you are right [12:28] hmmm [12:29] psivaa: ok, makes sense, let me try that - I was dogfooding in the meantime, running other tests as well [12:29] thostr_: it sounds like the fix might not be in your stack but in gst stack, and that it will take some time to land it; would you agree we should land current stack and keep tracking this issue for a later landing? [12:29] sil2100: ^ [12:29] didrocks: ^ [12:29] so I'm dogfooding the phone much more now that I've put my SIM card in it [12:29] I'm spotting a bunch of issues [12:29] sil2100: in fact the unity8 has 16 passed.. identical to the smoke tests [12:30] lool: I'm fine with that [12:30] psivaa: it's running here as well, let me see if I get the same [12:30] sil2100: let's wait for didrocks and thostr_ green lights, but then I personally think you should proceed with what we have [12:31] ogra_, popey: Do you guys see mediaplayer-app or dropping-letters in #92? [12:31] I dont [12:31] ok SDK team doesn't know of any other branches either (I'm in a hangout with them), so I guess that's it [12:31] lool, i'm on 93 and see them [12:31] lool: yes, agreeing. afterall, this is not about regressions, but things we uncovered never worked [12:32] ogra_: updating [12:33] ogra_: I see dropping letters fro mtime to time but not mediaplayer [12:33] so many images [12:33] ah it didn't pick up wifi when I walked back home [12:34] i think cyphermox has a NM upload pending [12:34] lool, what do you mean you don't see dropping letters? [12:34] sergiusens: my Installed list was very truncated, but then it got longer again [12:34] lool, launching or in the image? The click test should fail if it's not in for come weird reason [12:34] sergiusens: in the list of installed apps [12:35] does click list have it ? [12:35] sergiusens: I think it's just unity8 / scopes bugs, I've confirmed the packges are there [12:35] lool, ack; it's still listed with click list for phablet, right? [12:35] great [12:35] yeah [12:35] I see it from time to time [12:35] I never see mediaplayer-app though [12:35] they are just shy [12:35] :) [12:35] but I checked the .desktop file and it's not hidden in anyway [12:35] ah here's another bug [12:36] hmm i dont see mediaplayer either [12:36] since when should that be visible ? [12:37] (i konw it wanst in the past since it has no file selector) [12:37] * Mirv launched SDK build [12:41] ogra_: it's been made visible some weeks back becuase upstart-app-launch wouldn't want to open URLs with it if it wasn't user visible [12:42] ah [12:42] didrocks, sil2100: So either someone sneaked in something clever somewhere to rehide mediaplayer-app, or we have some scopes/unity8 bug where some apps (mediaplayer is one, there might be others) are hidden [12:42] lool, i honestly have never seen it [12:43] * ogra_ wonders how the icon is supposed to look likr [12:43] *like [12:44] lool: I must say that I never remember seeing mediaplayer recently in the scope [12:44] lool: since a few weeks I have been trying to find it in the list of apps but all I saw was music-app, and I thought that's normal actually (known 'feature'/bug) [12:44] yeah [12:44] sil2100: I remember when mediaplayer-app was added [12:45] and the debate on whether it should be visible or not [12:45] this was 10 days ago or so [12:45] [ Bill Filler ] [12:45] * remove NoDisplay=true so that it can be launched via upstart-app- [12:45] launch [12:45] -- Ubuntu daily release Mon, 30 Sep 2013 12:39:52 +0000 [12:45] but then not when it disappeared [12:46] Oh oh, missed that one then [12:46] sil2100: but perhaps it's intentional [12:46] it might be a regression or not [12:46] I see we've dropped the share button from mediaplayer-app [12:46] I guess this is to prepare removal of share-app [12:47] ogra_: who sent the request to unseed share-app? [12:47] so I found my phone off in my pocket [12:47] lool, several people [12:47] it didn't seem to be out of battery [12:47] lool, i had 3 identical MPs [12:47] this morning it was a bit hot after 20mn in my pocket with screne off [12:47] lool, bfiller, sergiusens and teg [12:47] *ted [12:47] ogra_: do you know where the decision to axe it comes from? [12:48] just curious cause I've missed this conversation [12:48] lool, it isnt used anymore ... it was dropped once and somehow came back by accident [12:48] ogra_: ooohhh [12:48] ogra_: great [12:48] ogra_: then I'm not worried [12:48] lool, it was supposed to be gone when the dialer/contacts/messages app split happened iirc [12:48] I see [12:49] ogra_: great [12:49] looong ago :) [12:51] ogra_, lool share-app was removed on 9/11 and accidentally added back [12:51] ogra_, lool it's an sdk component now [12:51] now its gone again := [12:51] has been for a while [12:55] morning [12:57] uh the screen turned black while I was using the phone and I had to hard stop it [12:57] sergiusens: 9/11 uh [12:57] that's why [12:58] sergiusens: lool: i get this when i run phablet-click-test-setup on a ppa enabled device: http://pastebin.ubuntu.com/6222284/ [12:58] any remedy? [12:59] asac: what's the issue that your seeing with utah? I'd be happy to walk through what it does but there's really not so much hackery as you think [12:59] psivaa: apt-get update on the device? [12:59] lool: i did that. and still no improvement [12:59] asac: there are 2 scripts that get run in the jenkins jobs: 1. provision.sh 2. jenkins.sh [13:01] psivaa: ah [13:01] psivaa: this is an unity8 version from PPA [13:01] psivaa: so pull-lp-source can't find it in primary-archive [13:01] asac: provision.sh is mostly a frontend to phablet-flash and phablet-network. It makes sure they both get called, it also makes sure we know what version is installed and puts that in an easy-to-find place under /home/phablet so that all the jobs can get the correct tag so the version numbers look right in the dashboard [13:01] psivaa: you'll either have to hack things badly, or to downgrade unity8 from PPA to archive version [13:02] lool: ok, downgrading unity8 will defeat the purpose of testing i am doing, which is image 93 + unity ppa packages [13:02] popey: ping [13:02] psivaa, ppa is not supported [13:02] asac: oh, provision.sh also now calls phablet-click-test-setup - which is needed for testing the click tests... so it really just handles all the provisioning steps needed for the device [13:03] psivaa, you can edit the script and remove the unity8 line from the packages to retrieve [13:03] sergiusens: ok, thanks will do that [13:03] asac: the last thing that provision.sh does, is something you may see as a hack or a bandaid - it does a dbus call to turn off the intro screen that comes up now [13:05] asac: jenkins.sh first reboots the device in preparation for testing, uses powerd-cli to ensure that the display comes on (as we discussed yesterday), and makes sure we have the image installed that we thought was supposed to be there (sanity check) [13:06] asac: then after making sure the device is really up, and that we really believe we have a functioning network, it calls utah, telling it all the places it needs to pull logs from, the test name to run, etc [13:07] asac: utah doesn't really do that much, it gives us a simple way of defining the tests and it gathers the logfiles and gives us the results in a consistent way that the dashboard knows how to interpret [13:07] ev: doanac: .. ok see above with plars. i believe we need to do somethin very soon to get rid of this problem (e.g. utah doing something, but others doing something else) [13:07] maybe the answer is: standardize on using utah rather than phablet-test-run, but maybe there is something easier [13:08] it must be super easy to use ... RUNTEST package-name -> does the right thing [13:08] anyway, think we talked about this at length [13:08] asac: do you have a specific example? we do intend to call phablet-test-run from utah rather than autopilot directly, maybe that will help, but if you have a specific problem it would help us to ensure that it gets addressed [13:08] so just hope we get a plan what to do [13:08] plars: that was the idea for month' [13:08] it didnt happen [13:08] hence we still have the problem [13:08] so maybe its the wrong idea [13:09] or we just dropped the ball or we had no time :) [13:10] asac: doanac has a branch that does most of that conversion already, but said he had some problem with it, we can try to take a look today and see if its something that can be resolved quickly. I do often try things with phablet-test-run at home and don't normally see any difference from what I get when I run things with our ci scripts [13:10] asac: but if you have a case where you're seeing that happen, I'd like to investigate why [13:11] plars: we had those cases yesterday [13:11] its always there [13:11] we still have to manually unlock etc. [13:11] i dont want to investigate more cases and fix/bandaid them [13:11] asac: can't believe the swipe left/right while scrolling up/down thing is already fix committed [13:12] i want a bullet proof solution that always does the same thing [13:12] that can be done without investigating more cases imo [13:12] just ensure there is no special code in utah [13:12] then we are done [13:12] asac: so this is about the unlocking before running a test? [13:12] its not a specific case [13:12] thostr_: click package scope is done, see the landing plan, it should be crossed [13:12] vila: I didn't have write access, hence the issue :p [13:12] utah should do nothing special but flashing and booting and then use something else for everything else [13:12] and this something else is what everyone else also uses [13:13] didrocks: ha crap, still remember which job encountered the issue ? [13:13] plars: anyway, just put it on the discussion agenda with ev doanac fginther and the rest of the team. i doubt that there can be any real doubts about the requirements. [13:13] vila: it was a publishing job, don't remember the stack (and focusing on the rest ;)) [13:13] asac: none of the code I just described was in utah, it is in our ci scripts. and we had to do it because it wasn't happening anywhere else. I thought it was generally agreed a long time ago that phablet-test-run was an equally bad place to do it, but until it gets done in autopilot, we have to have something that does it [13:13] didrocks: that includes also removal of old app lens scope? [13:13] sil2100: give me good news! what's up? [13:13] thostr_: yep [13:13] thostr_: this is already in image 92 [13:14] plars: ci script is for me in utah as well... just replace utah with "anything special we do in utah or our infrastructure" [13:14] plars: well, we discuss for month' ... i am just hear saying we must start deciding and doing [13:14] didrocks: thansk [13:14] all i want to ensure that the ball is carried on [13:14] feels for me this got stalled [13:14] asac: so we can either keep it in our scripts as it's working, or we can do the work to convert our scripts to run phablet-test-run (already in progress) and also move all the unlocking to phablet-test-run. But we'd still have to deal with the problem we dealt with yesterday of turning the display on early - probably before we get to phablet-test-run [13:14] didrocks: ack, let's try to collect that data next time. I may be able to find it differently but not right now. [13:16] didrocks, there is a mir merge back here that is not approved, is that expected? https://code.launchpad.net/~ps-jenkins [13:16] plars: what's preventing it getting done in autopilot? resources? [13:18] josepht: I thought that was the plan all along, but I wasn't involved in a lot of those discussions so I don't know the details. At some point it was supposed to be done there, then I heard it wasn't possible to do in autopilot right now though and we were back to square one. [13:18] fginther: I guess it's part of the issue I talk to vila about, you can remove it [13:19] ha an entry point may be [13:19] fginther: url ? [13:19] vila, https://code.launchpad.net/~ps-jenkins [13:20] fginther: sorry, I meant a recent job where bzr did hang [13:20] asac: also, I'm not sure if you saw my messages or heard me mention during the call yesterday - the new unlock method you mentioned using dbus is dependent on a package we don't yet install in the images: unity-greeter-session-broadcast [13:21] didrocks: hm, still fighting with psivaa to get sensful results for unity8, but I think I got some now! [13:21] vila, ah. was missing the context [13:21] plars: so we forgot to add that? [13:21] fginther: yeah, sorry, my bad [13:21] psivaa: how many failures you got? [13:22] plars: anyway. ignore that. we have to live for the next few days with what we have [13:22] sil2100: I will be interested in the whole story then :) [13:22] asac: no, we could add it but it's not supported by the images yet, do you know if there is still a plan to include unity-greeter-session-broadcast? [13:22] sil2100: so you are testing all services + unity8 (everything bu media-scanner) [13:22] plars: well, we wanted to add it 2-3 weeks ago [13:22] now its too late [13:22] not sure why noone asked that we needed it ... guess it just drowned in the noise :) [13:23] asac: well, we can unlock either way [13:23] vila, http://10.97.0.1:8080/view/cu2d/view/Saucy/view/Mir/job/cu2d-mir-saucy-3.0publish/50/console perhaps? [13:23] asac: I was just told that once this was in, it would be the preferred method. but it's still not in, so... [13:23] asac: how about this [13:23] plars: we wanted to use this unlock feature. i even stressed folks like mterry to get this feature for us [13:23] and then we didnt use it [13:23] popey: ping [13:23] no big problem, just definitely a fail [13:25] asac: sergiusens can we add an option to phablet-test-run to unlock the same way we are doing it now? Then we can direct people to use that option if they are running phablet-test-run by hand, and in the meantime it can be on or off while we get the patches into shape that convert our ci scripts to call phablet-test-run [13:25] sil2100: pong [13:25] plars: lets defer it until next week [13:25] its too late for all this. this topic was just dropped sometimes in the back [13:25] didrocks: so, it seems that with Mir enabled we cannot run unity8 tests normally - psivaa found a way of doing it, but it's error-prone [13:25] which is the part i want to prevent happening again [13:25] asac: wasn't there something else depending on unity-greeter-session-broadcast though? I was under the impression it was going in because it was needed - I don't think it was just for unlocking [13:25] I tried running other tests in the meantime which had 100% pass rate in smoketesting and all was fine [13:25] i dont know [13:26] popey: so, we found a way to run unity8 tests! [13:26] popey: http://pastebin.ubuntu.com/6222311/ [13:26] sil2100: how is UTAH running it then? [13:26] popey: psivaa found this method, I tried it and after a few tries it works in the end [13:26] asac: did you get answer on that from doanac? ^ [13:27] didrocks: no idea what they're doing, but for testing we're actually doing what UTAH is doing - we're installing utah locally and running the jenkins.sh for unity8-autopilot [13:27] sil2100: that's how the ci scripts work [13:27] sil2100: can you document that on the wiki? [13:27] sil2100: https://wiki.ubuntu.com/Touch/Testing#Testing_your_Ubuntu_Touch_Code_before_submission [13:28] add an unity8 section [13:28] with what to install and so on [13:28] asac: on deferring it to next week, do you want a meeting where we can flesh out some action items? [13:28] so that this doesn't get dropped on the floor again [13:28] Ok, psivaa if you don't mind I'll paste your solution there [13:28] sil2100: if you do the same things jenkins.sh does for setup (ensure the display is on), you don't get the same results as we started getting yesterday on unity8 tests by running phablet-test-run? [13:28] ev: +1 [13:29] sil2100: sure go ahead :) [13:29] fginther: nope, didrocks mentioned a push failing but the branch there is not the right one, don't worry, I only asked in case you were already on such a job [13:29] plars: I got 7 failures and 17 passes, so it's the same result as on smoketesting [13:30] sil2100: great! so services + unity8 are good to go in? [13:30] sil2100: ok, can test that [13:30] sil2100: ok, good... from what asac was saying I thought this was the issue he was talking about but maybe not [13:30] didrocks: btw. regarding mediascanner tests - we not including that? I thought lool said that after the analysis that thostr_ made, we'll be pushing what we have now [13:30] sil2100: everything but not mediascanner [13:31] So thumbnailer yes? [13:31] thostr_: ^ [13:31] I thought all is ok as we have in trunks - that it's ok, but just not fixing all issues [13:31] sil2100: didrocks: we said we'll push everthing for now as it's not really regressing [13:31] sil2100: did you confirm the non regression? [13:31] thostr_: the issues we're seeing have been there forever [13:33] didrocks: it's strictly better than what we have [13:33] sil2100: ok, if it doesn't regress, please push what we listed then [13:33] didrocks: but we wont get the bug fix for a newly uncovered bug today [13:33] (bug is already in current image AIUI) [13:33] didrocks: not yet, but testing didn't reveal anything new at least [13:33] plars, yes we can, only reason it was never added was because we never defined where it was supposed to be [13:34] didrocks: right, we'll push it once popey ACKs ;) [13:34] plars, if you have something written up; get the MR and I'll test/happrove [13:34] lool: do you have the reference for it? [13:34] popey: hope you're already running the tests ^ ? ;) [13:34] sil2100: great! [13:34] * didrocks stare at popey [13:34] just setting up now [13:34] didrocks: sergiusens is preapring the daily release changes, but we will lose the .deb builds, no issue with that I guess? [13:34] thostr_: bug id for new bug? [13:34] (; [13:34] didrocks: I dont [13:34] sergiusens: well I thought the agreement was that it should go into autopilot but doanac would have more details about that. I was recently told that's not possible though. Let's talk to him when he gets in and make sure we have a sane plan for the time being. [13:37] lool: for mediascanner? https://bugs.launchpad.net/ubuntu/+source/mediascanner/+bug/1238599 will add another one, because there's yet another one :( [13:37] Ubuntu bug 1238599 in mediascanner (Ubuntu) "mediascanner hangs" [Undecided,Confirmed] [13:38] lool: no issue at all [13:38] thostr_: you think the fix will be a monday thing? [13:38] didrocks: i hope we could fix it still today [13:39] thostr_: ok, keep us posted [13:39] sil2100: do i need to kill powerd? [13:39] didrocks: it just takes a while to get the stack right now need to build with debug [13:39] thostr_: yep, I imagine, no worry [13:40] popey: no no [13:40] k [13:40] popey: just do what the pastebinit says [13:40] popey: of course, the -s needs to be different (or not provided) [13:40] Since it's the device of psivaa [13:41] popey: sometimes you need to try a few times before it works [13:41] sil2100: uhm, the paste isnt clear, is utah installed on device or laptop? [13:41] didrocks, sergiusens: smooch <3 [13:41] popey: it's all on host, not the device [13:41] k [13:41] I'm going to school to pick up my son, then I'll pop by shortly and call it a day [13:41] popey: since you do adb push ;) [13:41] ya [13:41] I'll clear it up on the wiki [13:41] didrocks, I'll ping you in an hour or so [13:42] I'll check a bit in the evening and over week-end [13:42] you guys keep rocking the boat cause frankly I'm baffled with the stuff landing; it's awesome [13:42] sergiusens: excellent, thanks man :) [13:43] LOL [13:46] sil2100: running [13:46] fginther: got a moment for a theory discussion ? [13:46] kgunn, sure [13:46] popey: s/running/passing/ [13:46] * didrocks updates popey's vocabulary :p [13:46] hah [13:47] optimistic as ever [13:47] heh ;) [13:47] fginther: so...not sure who the right person to talk to is...maybe didrocks has an opinion here as well....but [13:47] today is the perfect day [13:47] we have tons of stuff landing [13:47] everything is fine [13:47] We love Friday releases [13:47] as we are approaching freeze on touch...i assume we will need to branch [13:47] I didn't get a ping from the Mir's team yet [13:47] oh wait! [13:47] :) [13:47] didrocks: PING...now your day is complete [13:48] ;) [13:48] kgunn: not yet, we are all focusing on touch saucy right now [13:48] kgunn: we are discussing about how to handle T and upgrading today in fact [13:48] so let's sync that next week? [13:48] let's rock the Mir perf fix first I guess :) [13:49] didrocks: ok...surely we will branch....(worrying about mir & unity8 & unity-mir)....so just want to know...is it lp:~team/project-T...or lp:project/project-T [13:49] and i know you guys have some quid pro quos wrt that [13:50] kgunn: I like (as we don't have stable name) lp:~team/project-name/13.10, and so lp:project-name/13.10 is pointing to it [13:50] then lp:project-name will be 14.04 [13:50] the new trunk [13:50] kgunn: that way you keep all MP against 14.04 [13:52] Saviq: ^ eventually [13:52] cool that the desktop autopilots are back, happy to see green at cu2d as well [13:53] Mirv: yeah, it will be a nice way to end the week :) [13:53] kgunn, didrocks k [13:53] thanks jibel, stgraber and jdstrand :) [13:53] kgunn, glad to help [13:56] didrocks: updtaing the landing asks since I merged and built the SDK default thumbnail branch, and adding that qtubuntu-sensors is in the same landing # [13:57] Mirv: I changed the description + test, but good! thanks :) [13:57] * lool & [13:57] didrocks: yeah I noticed [13:59] my only problem is scheduling but I've now everything ready except test results, which I'm running starting now [13:59] Mirv: let's cross fingers! [14:00] popey: how's the testing going? [14:01] sil2100: still running I think.. [14:01] maybe not [14:01] the good thing about SDK is that the trunk is not scary thanks to the stopped automerging [14:01] 13:50:44.108 INFO testcase:554 - Killing process 4404 [14:01] thats the last output [14:02] sil2100: it's just sat there ☻ [14:02] asac: I've made you optional on the meeting, just in case you have further thoughts. If not, we'll take it from here. [14:02] popey: ok, so it means it hanged up ;) [14:02] popey: you need to kill it and re-run! [14:03] popey: psivaa said that tends to happen and re-runs help ;p [14:04] bah [14:05] balloons: ping [14:06] didrocks: it was tyhicks who did the heavy lifting on my team. I only tested and advocated [14:06] lool: can i upload ubuntuone-credentials with our fixes for touch to ubuntu now? [14:06] tyhicks: thanks then! :) [14:06] ev: which meeting? [14:06] didrocks: ui-toolkit AP still gets Killed with updated Unity8 (unity-mir not updated), some others not [14:06] jdstrand: testing is still a great assess. Thanks for the good news of the eow! :) [14:07] yw [14:07] balloons: do you know what's going on with the new event test failing on calendar under mir only? [14:07] Mirv: ok, try to reboot between tests then [14:07] asac: to resolve this phablet-test-run produces different results issue [14:07] Mirv: I meant, between each app test [14:07] tyhicks: fyi, testing is back to green ^ nice job :) [14:07] didrocks: it was the first test and got killed [14:08] I want to make sure we understand what needs to get done and then divvy up the work [14:09] Mirv: argh, not good :/ [14:09] didrocks: there are 60 tests there in ubuntuuitoolkit so a lot [14:09] hum, can be linked then [14:09] I'm getting a lot of 100% OK results from smaller apps [14:09] Mirv: then adb shell [14:09] top [14:10] to check unity8 mem size [14:12] balloons: also music app seems to have 2 failures with both mir and sf since image 90 - I think maybe it's related to https://bugs.launchpad.net/music-app/+bug/1235475 ? [14:12] Ubuntu bug 1235475 in Ubuntu Music App "Autopilot Tests fail with Upstream SDK" [Critical,New] [14:14] ev: yeah. i dont want to be part of that discussion again :) [14:14] :) [14:14] i will only make people including me feel uncomfortable and [14:14] simple rule is: if there is something that needs checking individual cases then we are on wrong track :) [14:14] e.g. why is something not working good in utah while people see better results in manual [14:15] balloons: oh wait, sheesh, I see the error there [14:16] balloons, did you get to the filemanager test issues while on click? [14:17] sergiusens, plars: reading backlog - you guys wondering what I know about screen unlock? [14:17] ERROR: FAIL: Not all tests passed [14:18] vila: jibel: I don't know what's happening there but seems stuck http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2523/console [14:19] nvidia finished fine which is nice [14:19] doanac: yes [14:20] didrocks: sil2100: when will #94 be available? [14:20] plars, sergiusens : i've never made any progress getting this fixed where it should be. [14:20] sergiusens: I think https://code.launchpad.net/~music-app-dev/music-app/expandable-and-more/+merge/189483 has the fix for the typo that is preventing music-app from passing, but it seems to be growing a large history quickly. I see you commented on it recently, do you think it's going to land soon? [14:20] thostr_: it will start building when rsalveti will finish the MM stuff (so will start building in ~2 hours) [14:20] plars, no I don't think it is [14:21] didrocks: ok, thanks [14:21] plars, and they broke my click integration as well [14:21] yw [14:21] sergiusens: do you think it's worth it for me to just propose a simple fix for the untits typo that's preventing those tests from working then? [14:21] sil2100: http://paste.ubuntu.com/6222590/ can you parse that gibberish? ☻ [14:21] popey: oh shit [14:21] popey: 22 failures, 2 passes [14:21] :O [14:21] plars, tests are completely broken for when it's a click app btw [14:21] What did you dooo?! [14:21] psivaa: ^ [14:21] sergiusens: :( [14:22] sergiusens: but it's not a click app yet, it would at least reduce the noise a bit [14:22] sil2100: followed your instructions ☻ [14:22] plars, I had a semi working MR for that and I was broken by a refactor :-/ [14:22] plars, well it has to be or it won't be on the image next week ;-) [14:23] popey: let's just agree your test results never existed, we like swiping dirt under our carpets ;D [14:29] popey: could you try again? One last time I would say [14:29] Since maybe it had issues with screen-unlocking or something [14:29] ok [14:29] will reboot, unlock, run tests [14:29] * fginther will be back in a bit [14:31] Mirv: rings no bell except it looks like nothing is happening inside the container [14:32] Mirv: the timeout is set to 330 (wasn't it 120 yesterday or was it somewhere else ??) [14:36] sil2100: seeing nothing on screen here [14:37] popey: press the power button just in case [14:37] i did [14:37] ah, now I see somethiong [14:38] ogra_: I retried ubuntu-touch-meta on amd64, it had chroot issues [14:39] sil2100: this just isn't working, sorry. [14:40] didrocks, ? [14:40] ogra_: it was happily waiting in proposed, I had a look and it failed on amd64 [14:40] oh [14:40] thanks [14:41] @ci can I get this merged? https://code.launchpad.net/~sergiusens/ubuntu-terminal-app/click_improvements/+merge/190686 [14:41] yw ;) [14:41] * ogra_ wonders why he didnt get a mail for the failure [14:41] balloons, psivaa: clock looks like it started failing on all tests today with both sf and mir, was there an update to the click package from yesterday? [14:43] plars: let me run that locally.. dont know if there is any updates [14:43] psivaa: you see any problems regarding releasing the unity8 stack? [14:43] psivaa: or is it relatively green light from you as well? [14:44] sil2100: my testing so far has not been worse than 93 alone. so i would say that's a green [14:44] plars, I believe it was updated a couple days ago, test-wise, but let's see [14:46] Ok, I'll finish one last test suite and release then [14:47] Mirv: do you think you will go through this sdk + qtubuntu-sensors today? [14:48] psivaa, balloons, asac: terminal app had a failed test on mako (focusing on that right now), but it went back green on the retry [14:49] balloons: weather app is down to one failure, not sure how much is just flakiness or if it really got better [14:49] I'm going to try that one locally rather than rerunning, lest I just end up making it worse [14:53] didrocks: continuing, yes, trying to pick pieces of needed test results while doing also other stuff. so ETA 1h or so. [14:53] Mirv: rocking! thanks a lot dude :) [14:53] plars, I haven't touched it.. but it seems that anything that enters text under Mir is having issues with Mir.. the entered text doesn't seem to line up with what AP is sending [14:53] Mirv: looking good until now? [14:55] plars: clock app tests failed locally as well [14:55] fginther, i really need this in https://code.launchpad.net/~sergiusens/ubuntu-terminal-app/click_improvements/+merge/190686 [14:56] @ci I get a strange failure in https://jenkins.qa.ubuntu.com/job/mir-saucy-amd64-ci/835/console [14:56] @ci [14:57] fginther, hmm, it's merged, nevermind :-) [14:58] * ogra_ grumbles watching the orange LED of death on mis mako [15:00] sil2100: done? we can publish everything? [15:00] slangasek: oh btw, you can remove your hint for rsalveti's gstreamer/android/hybris [15:02] didrocks, landing asks https://code.launchpad.net/~sergiusens/ubuntu/saucy/lxc-android-config/no_appmanager_process_on_mir/+merge/190692 [15:02] didrocks: I would say done, wanted psivaa to test one last thing for me - waiting for a sign from him [15:03] sergiusens: just saving memory or fixing a bug? [15:04] didrocks, I'll leave that for ricmm; but we might have unintended consequences if we run that for a while with mir [15:04] didrocks: dobey had an ubuntuone-credentials upload to fix some issues with account handling in the UI, is it ok to upload? I had postponed to at least after Mir stabilization [15:04] sergiusens: can you add a landing ask on the spreadsheet and update once ricmm has more info? [15:04] alf_: said line has 3 spaces between email and date instead of 2 for the other similar lines, didrocks ? typo ? [15:05] sergiusens: just to decide if it's today or if we handle that on Monday [15:05] didrocks, I can't, I don't have write access [15:05] rsalveti, ^^ [15:05] vila: hum? [15:05] content? [15:05] context* [15:05] didrocks: @ci I get a strange failure in https://jenkins.qa.ubuntu.com/job/mir-saucy-amd64-ci/835/console [15:05] lool: which landing ask? [15:05] didrocks: parsechangelog/debian: warning: debian/changelog(l11): badly formatted trailer line [15:05] vila: this is not upstream merger but fghinter, want me to have a look? [15:05] LINE: -- Ubuntu daily release Fri, 11 Oct 2013 06:59:30 +0000 [15:05] didrocks: it's mixed with other things from r.alsina [15:05] yeah, you need 2 spaces [15:05] vila: ^ [15:05] didrocks: but it's actually independent [15:06] didrocks: I think we landed the rest of the click stack [15:06] didrocks: annotation says you did the commit ? [15:06] ah no we didn't [15:06] lool: can we get a proper ask on the spreadsheet? [15:06] didrocks: ask is 158 [15:06] vila: oh that one (sorry, dealing with 100 stuff as you can see just on that channel), yeah, probably a typo when merging back [15:06] fixing it [15:06] didrocks: can that just be fixed (remove spurious space) by alf_ ? [15:06] lool: yeah, we need a new daily-release of unity-scope-click to get the other half landed [15:07] vila: yeah, please alf_ do it ;) [15:07] dobey: the click-scope and u1-credentials updates are completely independent yes? [15:07] didrocks: sure, no stone thrown, again, just trying to understand [15:07] lool: yes [15:07] dobey: which bzr rev? [15:07] dobey: we have latest bzr rev 76 in PPA: http://people.canonical.com/~platform/cu2d/results [15:07] lool: so, everything is listed in? [15:07] lool: r76 needs to be included in the unity-scope-click landing [15:07] lool: we have nobody free to handle it before robru comes [15:07] didrocks: yes, do you want us to update click scope too? [15:08] didrocks: probably dropping qtubuntu-sensors from agenda, I don't see/hear/feel any feedback on mako [15:08] lool: aren't you in week-end? :) [15:08] didrocks: I thought we were landing the app + click scopes today actually [15:08] Mirv: ok, keep them posted on that and please go on [15:08] didrocks: if there's someone who can do manual testing by updating qtubuntu-sensors, then that's welcome, I'll concentrate on the SDK [15:08] didrocks: Yeah, I said I check from time to time and I'm trying to answer pings blocking on me (when they shouldn't ;-) [15:08] didrocks: yeah pinged kalikiana but not sure if he's around anymore [15:08] lool: well, no, see the landing plan, we try to go as for as we can [15:08] lool: we had issues, but landed 40 packages [15:09] didrocks: looking [15:09] Mirv: ok, let me annotate the landing [15:09] didrocks: ok reportedly maybe workd on maguro, at least kalikiana says so. but nothing happens on mako - although if nothing means no problems, then I guess it could be ok if someone tests on maguro [15:10] Mirv: annotated [15:10] didrocks: I actually see scope-click made it in image #92 according to plan [15:11] ogra_: "Fix maliit shutdown crasher" which package was the stop in already? [15:11] lool: scope-click alone [15:11] lool: not the app [15:11] didrocks: Ok I see now that dobey has landed one last revision to scope-click (r76); I think I'll add that to the ubuntuone-credentials thing [15:12] didrocks, ubuntu-touch-session [15:12] lool: ralsina added the bug that rev fixes to the ask in the spreasheet, earlier [15:12] ogra_: 0.80, right? [15:13] lool, dobey: yes, I added it [15:13] ask 158 [15:13] didrocks, right [15:15] lool: you are going to help them in that landing? (you assigned dobey, I don't think he can release the click scope) [15:15] didrocks: I've added a new landing slot for scope-click last bzr revision + ubuntuone-credentials; I can help with it [15:15] didrocks: yeah, I'm adding myself now [15:15] lool: please image 95 [15:15] didrocks: ok, things are fine, let's publish \o/ [15:15] didrocks: I listed him as PoC; note that u1-credentials is not CI-ed [15:15] Mirv: some activity on http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2523/console suggests the job is not dead but encounter an issue with... unity ? [15:15] didrocks: Ok; when is image #94 happening roughly? [15:16] Mirv: /var/local/autopilot/autopilot.log: 15:06:50.040 ERROR __init__:64 - Unity doesn't appear to be running, exiting. [15:16] sil2100: please publish ALL of them! :) [15:16] lool: we'll start building in ~1h [15:16] Publish all stacks o/ [15:16] didrocks: Ok [15:16] ;) [15:16] didrocks: I can put a block hint in place and have him upload ubuntuone-credentials for me to test it in proposed? [15:16] lool: good enough to me, yeah [15:17] dobey: please upload ubuntuone-credentials [15:17] Mirv: could that be http://pad.lv/1238417 [15:17] Launchpad bug 1238417 in Mir "Unity does not process events from evdev device created before unity is restarted (autopilot tests)" [Critical,Confirmed] [15:17] ok [15:17] vila: ok. I'm running touch only testing now, so just reported that to you about a current example on desktop testing side [15:19] kalikiana will join here and is testing flash + qtubuntu-sensors from daily-ppa on maguro [15:19] Mirv: yeah, let's not block on that though [15:19] Mirv: ack, but who will be investigating that ? Doesn't look like a ci issue per se no ? We're far past the point were this could be considered related to the previous dbus/lxc/apparmor one right ? [15:20] didrocks: yeah, runing SDK tests all the time, plus booting [15:20] * didrocks crosses fingers even more [15:20] didrocks: in a car, to be exact ;) [15:20] not driving [15:20] Mirv: I'm stuck on "can't open /cache/recovery/ubuntu_command" :-( [15:20] Mirv: ahah, I was starting to get frigthened :p [15:20] oh wait it's doing something afterall [15:21] Mirv: we do want a photo now! [15:21] did anyone notice that we have SHELL=/system/bin/sh in each and every crash file ? [15:21] on utah [15:22] ogra_: popey: I think we should promote image 93 as well, did you try it a little bit? [15:22] hmm, locally too [15:22] didrocks: http://people.canonical.com/~tjyrinki/car/2013-10-11-182153.jpg [15:23] didrocks: packaging ACKs! http://paste.ubuntu.com/6222868/ <- I only browsed through them, still looking [15:23] didrocks: yes, not found any regesssions [15:23] Mirv: rock! :) [15:24] ;) [15:24] :) [15:24] sil2100: -usr/share/dbus-1/services/*MessagingApp*.service [15:24] this is expected? [15:24] http://10.97.0.1:8080/view/cu2d/view/Saucy/view/Phone/job/cu2d-phone-saucy-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_messaging-app_0.1+13.10.20131011-0ubuntu1.diff [15:25] this it's the same in dialer one [15:25] bfiller: can you enlight us? ^ [15:25] root@ubuntu-phablet:/# grep SHELL /etc/environment [15:25] SHELL=/system/bin/sh [15:25] ugh [15:26] so we should definitely fix this ^^^^ [15:26] asac, didrocks ^^^ [15:26] ogra_: indeed, do you mind tracking it? [15:26] or you want us to? [15:26] didrocks: I guess it is, as otherwise there would be missing files during package build - so I guess they switched that out [15:26] bfiller: ^ [15:26] didrocks, not at all, its a livecd-rootfs change i think [15:26] ogra_: thanks (and good catches! ;)) [15:27] didrocks, i wonder if that also influences the testing behavior in any way [15:27] (manual vs automated) [15:27] ogra_: influencing or not, it has to be fixed :) [15:27] heh, yeah. thats clear [15:27] ogra_: do you mind poking a little bit image 93 before promoting? [15:27] * ogra_ wouldnt be surprised if many of the unexplainable differences suddenly go away though [15:28] didrocks: don't need dbus services for those apps anymore [15:28] didrocks: they use qt.openUrl() now [15:28] ok great! [15:28] sil2100: please +1 [15:28] :D [15:28] thanks bfiller [15:28] np [15:29] Ok, publishing... hope I won't miss any components, but I made a list what to release [15:29] sil2100: we can recheck together [15:32] didrocks: ok, so I published: thumbnailer address-book-app address-book-service dialer-app history-service messaging-app telephony-service mediascanner unity8 unity-scope-mediascanner - did I miss anything? [15:33] didrocks, landing asks 187 ... can i move it over and upload (then it can make the next build, i think this is serious) [15:34] lool: ubuntuone-credentials 13.10-0ubuntu2 uploaded [15:37] ogra_: looking [15:37] ogra_: please add and upload now :) [15:38] great, thanks ! [15:39] dobey: yeah, see #ubutnu-release [15:39] sil2100: sounds the list is complete! [15:39] ogra_: thanks to you :) [15:39] I showed ubuntu-phone at the post office where they sell sim cards today [15:39] lool: not in there at the moment. did i miss something? [15:40] dobey: it's in unapproved, only seeded in touch, but somehow it's in an ubuntuone packageset, not sure what that meas [15:40] in terms of freeze [15:40] means [15:40] lool, and, did you sell it :) [15:41] lool: it's in the package set, but that's only an upload privileges matter afaik. the package itself is only in universe, so it really shouldn't be a problem to squeeze in with the freeze [15:41] dobey: Yeha that's whatI thought [15:41] dobey: waiting for it to come out, I might test it only later tonight if it's not out in 5 mn [15:42] ogra_: they were asking "but why do you need a prepaid if you're not going to charge it super often" etc. [15:42] heh [15:42] then I take the two nexus 4s out, android and ubuntu touch one [15:42] and start showing the UI [15:42] I looked a bit like a freak I'm afraid [15:42] they were like O.O [15:42] but it looked good! [15:42] hah, cool ! [15:45] ogra_: what needs fixing in the shell? [15:45] \o/ [15:45] asac, not in the shell [15:45] sil2100: all in proposed? [15:45] asac: done past [15:45] next topic :p [15:45] kk [15:45] asac, grep SHELL /etc/environment ... [15:45] nevermind [15:46] asac, SHELL=/system/bin/sh [15:46] pretty serious :) [15:46] didrocks: mediascanner is! Looking at others :) But it seems all are landing nicely [15:46] Sweetness [15:49] dobey: It's accepted now -- 17:43 -queuebot:#ubuntu-release- Unapproved: accepted ubuntuone-credentials [source] (saucy-proposed) [13.10-0ubuntu2] [15:50] dobey: but I'm out of time, so will come back to it later tonight [15:50] around 9pm my time or so [15:50] dobey: is there something specifically I should test outside of what I listed in the landing plan? [15:51] lool: just the 2 bugs that were in the ask, i think [15:52] sil2100: everything's in then? [15:52] dobey: Ok; thanks for the upload! [15:52] dobey: will ping you if I have issues or if it goes in [15:53] didrocks: just a status update this reboot + test + retest dance is quite slow, I've 15 of the AP:s ok (compared to image runs), gallery 20/22 pass (image has 100%, need to rerun), unity8 I'm right now running with one test at a time, weather app and rssreader app todo and then I'd need to somehow get uitoolkit's own tests run [15:53] lool: ok, thanks! [15:53] Mirv: ok! good luck :) (thanks for udpating) [15:54] ogra_, didrocks: if you're talking about the unity8 upstart socket change, the fix I uploaded was only half of the fix; we believe there are other uses of the socket in libupstart-app-launch which are contributing to the leak, tedg pushed a patch yesterday to fix one leak we identified but we haven't confirmed if that's actually what's causing the problem - I'm going to try to test on mako today to see for sure [15:54] didrocks: all is in -proposed! [15:54] didrocks: remove the hint> android hasn't been uploaded, does that mean this plan was dropped? [15:55] slangasek: @re socket: excellent, seems under control (at least we have progress) :) [15:55] hmmm [15:55] slangasek: hint: rsalveti detected an issue with Mir yesterday, he's testing the last fix and will upload, we want it in the image now [15:55] sil2100: nice! [15:55] Wait! [15:55] need to check something [15:57] didrocks: hmmm, I don't see address-book-service address-book-app in -proposed :| And in the queue as well [15:58] Even though PS Jenkins committed the release commits to trunks [15:58] sil2100, https://launchpad.net/ubuntu/saucy/+source/address-book-app/0.2+13.10.20131011-0ubuntu1 [15:58] sil2100, they are in there [15:59] sil2100, well, they moved to release [15:59] slangasek, and the udev filtering too [15:59] sil2100: there is a slight delay when they disappear when copying [15:59] slangasek, or does either of the other patches prevent us from doing this ? [15:59] sil2100: you will update the status on the spreadsheet while everything is transitionning? [15:59] didrocks: phew... since I didn't even see it in the "Latest upload:" piece [15:59] :) [16:00] So I got worried ;p [16:00] ogra_: there is already a patch in place to filter uevents on maguro. === jfunk is now known as jfunk-afk [16:00] and filtering events does nothing to fix the underlying memory issue [16:00] * lool leaves for a couple of hours or so [16:01] all: great landings today! [16:01] can't wait til next image [16:01] slangasek, there is no path that makes udevd not eat 10% of my CPU while the screen is on [16:01] slangasek, the filtering fix is in upstart-udev-bridge only [16:01] ogra_: that is a bug unrelated to what we're talking about [16:02] slangasek, thats udevd recieving the uevents [16:02] 60/s [16:02] yes, I know [16:02] (one for each vsync) [16:02] which has nothing at all to do with the bug we were talking about [16:03] that bug covers the uevent spam, no ? [16:03] robru: meeting? [16:03] ogra_: coming? [16:03] asac, yes [16:04] seems like the clock in the panel is still broken with this image [16:04] Mirv: any idea on the qtc problem? does it work for you to enable platform mode? [16:04] kalikiana: what's wrong with the clock in the panel? that should work... the one on the greeter is broken though [16:04] kalikiana: and I fixed the qtc problem [16:04] it's waiting for reviewers [16:05] kalikiana: https://code.launchpad.net/~mzanetti/qtcreator-plugin-ubuntu/fix-device_run_app-for-mir/+merge/190677 [16:05] kalikiana: I can't test now [16:05] mzanetti: the clock disappears from the panel, I saw the same bug with the latest released image on Wednesday [16:05] it briefly worked to "restart unity8" [16:06] kalikiana: oh... that one... haven't seen that yet. can you reproduce? or maybe check if you have a crash trace of indicator-datetime around? [16:06] kalikiana: but you'll happy to see this one: https://code.launchpad.net/~mzanetti/unity8/fix-greeter-time-update/+merge/190636 [16:06] +be [16:06] mzanetti: where would I find a trace? all I did was boot really [16:06] thostr_: still a hud crash here: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/93:20131011:20131010/4677/share-app-autopilot/ [16:06] kalikiana: /var/crash [16:09] hmmm nothing related to clock there, only the rss app (just installed it and switched in a quick smoke test) [16:09] asac: https://code.launchpad.net/~indicator-applet-developers/hud/fix-hud-crash-share-app/+merge/190711 [16:09] asac: we have I think 3 hud crashes in pipeline... so let's get those landed... (and merged before) [16:10] mzanetti: Saviq: hey, so seems for notes-ao there is unity8 crashing FIVE times ... http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/93:20131011:20131010/4677/notes-app-autopilot/ [16:10] for all others we just get one [16:10] mzanetti: ah, the mr is about the time getting out of sync, that would be lovely to see fixed, I felt quite insecure at the airpoirt with the delayed clock as it is… [16:10] (probably shut down crash is that one) [16:11] mzanetti: I'll reboot and see what happens wrt disappearing [16:11] kalikiana: hehe, +1 === dpm is now known as dpm-afk [16:17] didrocks: we'll need to postpone the multimedia stack a bit more, we found another crash with mir [16:18] didrocks: let you know once I have more updates, but feel free to spin a new image without it [16:19] didrocks, the oom stuff needs proper testing first, i just wanted to have it on the asks page already, dont move it to the plan yet [16:19] totally weird, cannot reproduce the clock issue anymore [16:19] ogra_: oh ok, let's not do it then [16:20] just keep it on asks [16:20] i'll update once it is ready [16:22] * Mirv learns that image tests do not run full suites at once either [16:22] rsalveti: ok, feel free to upload when you are ready [16:22] Mirv: same for you [16:22] we're going to kick an image soon [16:22] so will be on next for both of you [16:22] thostr_: ping! [16:23] thostr_: I seem to have found a problem ;/ [16:23] I'm so pissed off right now [16:23] Since I missed this and shit [16:23] :) [16:23] all will be good [16:23] didrocks: I'm really sorry, but we have a problem, and I might have to re-push mediascanner... [16:24] I knew it was too good to be true [16:24] * asac wonders if he should use the /clear command in irssi :) [16:24] thostr_, didrocks: so, it seems mediascanner is FTBFS on powerpc because of a failing unit test [16:24] ppc... nice one :) [16:24] didrocks: thanks [16:25] sil2100: hum, you didn't get that in the build? [16:25] rsalveti: thanks for keeping us posted :) [16:25] I'm pissed because I knew about this, ran re-build and FORGOT ABOUT THAT [16:25] ah ok [16:25] With hopes it was a 'flacky test' as always in ppc on mediascanner [16:25] at least you are honest :) ... thats a good thing! [16:25] well, no worry, let's hope thostr_ can fix it quickly [16:25] thostr_: https://launchpadlibrarian.net/153485520/buildlog_ubuntu-saucy-powerpc.mediascanner_0.3.93%2B13.10.20131011-0ubuntu1_FAILEDTOBUILD.txt.gz <- but re-running didn't help this time [16:25] sil2100: feel free to republish mediascanner yourself once ready [16:26] I guess we just need to coordinate with ogra_ or sergiusens to kick an image build then [16:26] didrocks, what were you saying about signon earlier? do you need me for that still? [16:26] didrocks: maybe he cant easily figure/fix this? [16:26] (powerpc) [16:26] robru: no, cyphermox was the victim :p [16:26] ok [16:26] hmm can't add a gmail account anymore [16:27] didrocks, hey, so I can switch the terminal app to click now; the other two need some work (tests) [16:27] didrocks, want us to do that? [16:28] sergiusens: sure, just do it either right now (if less than 30 minutes to be ready) or wait for 94 build to start [16:28] sergiusens: maybe you will be the one kicking 94 btw :p [16:29] didrocks, one sec [16:29] didrocks: going to go buy fish so I can make lunch, bbl [16:30] sergiusens, we're waiting for livecd-rootfs 2.195 first [16:31] didrocks: yep, that's fine (next one), the UI Toolkit change alone is actually a pretty small one, the "fun" here is running all AP:s and such [16:32] Mirv: come on, you are doing that on a car! it's where you have the fun :) [16:32] is there some single reason why screen FPS drops to ~10 very often after running some time? some package to update? [16:32] Mirv: mir… [16:32] didrocks: hehe :) I left the car now, now eating tacos [16:32] Mirv, mako or maguro ? [16:33] ogra_: mako [16:33] Mirv: I have the same thing on mako... [16:33] * ogra_ has a dead mako today :/ [16:34] ogra_, didrocks https://code.launchpad.net/~sergiusens/ubuntu-seeds/terminal_click/+merge/190726 [16:34] I found that mimicking what image testing does is a good idea [16:34] Mirv: ;) [16:34] mako isn't a taco! [16:34] didrocks, does that need a landing entry ^^^ [16:35] sergiusens: approved [16:35] ogra_, that follows this http://paste.ubuntu.com/6223196/ [16:35] ogra_: it already has one [16:35] ah, great, so i can just merge and upload ? [16:35] #148 [16:35] ogra_: yes please, maybe remove the filemanager one [16:35] and duplicate in a new entry [16:35] or let's sergiusens add a new request for that one once ready [16:36] yes [16:36] didrocks, not sure if music app has an entry; but that and filemanager as click would remove the PPA [16:36] well, or update the existing one [16:36] now my mako rebooted after running certain amount of ubuntuuitoolkit tests, before which several tests run attempts were killed [16:36] sergiusens, so you want filemanager and terminal dropped ? [16:37] ogra_, what, dropped from where? seeds, list or what? :-) [16:37] yeah [16:37] touch seed [16:37] (and meta) [16:38] ogra_, not yet, it's under test still for click [16:38] oh, ok. so only terminal [16:38] ogra_: so, apart from this requests, all the others that are transitioning are marked in proposed [16:38] (and we need to check with sil2100) [16:38] ogra_, filemanager and music do some extensive monkey patching to $HOME which fails badly with upstart ;-) [16:38] and under confinement :-) [16:38] ouch ! [16:39] mzanetti: https://code.launchpad.net/~mzanetti/qtcreator-plugin-ubuntu/fix-device_run_app-for-mir/+merge/190677 [16:39] I tried it, but got some errors [16:39] see the comment [16:40] sergiusens, merged [16:40] robru: requests 200 and 201 for you then [16:41] didrocks, ok, will do [16:41] thanks! [16:41] robru: ensure that we have trunk built first for all of them, I think you will need to rebuild the stacks [16:42] ogra_, already sync too :-) http://people.canonical.com/~ubuntu-archive/click_packages/click_list [16:42] cool [16:42] you are faster than the metapackage :) [16:42] thats still updating here [16:45] @ci can we get this merged? https://code.launchpad.net/~sergiusens/ubuntu-weather-app/click_improvements/+merge/190484 [16:46] didrocks, everything is landed in trunk though, right? do i need to wait for some mps? [16:47] robru: I guess look at the trunks, normally, they should be there, but better to always double check TBH [16:50] sil2100: is everything going all right? how can we help you? [16:51] didrocks, ah, yes, all MPs in the ask are merged [16:52] robru: excellent, please kick some builds :) [16:58] * Mirv warmly recommends trying to land something that has a description "run all AP:s" :) [16:58] ;) [16:59] ui-toolkit itself OK [16:59] Mirv: so, you want to sneak into that image? [17:00] didrocks: I thought we don't care about powerpc any longer??? [17:03] didrocks: I want to still prove to myself that the same unity8 tests that passed in #93 are still possible to run. almost there. [17:03] thostr_: we do if it built before on it [17:03] Mirv: ok, if you have that working, feel free to push for 94 then [17:04] didrocks: ok, but new stuff like mediascanner doesn't need to build/run there,r ight [17:04] we're pending on sil2100/thostr_ anyway [17:04] ah, right, there was those delays [17:04] thostr_: it was building on powerpc, I don't think we want to remove it from now on because tests don't pass on it [17:04] didrocks: it's not even utilized on desktop [17:05] thostr_: you can go on #ubuntu-release and discuss about it with the release team [17:05] thostr_: it's clearly not my call [17:05] thostr_: I do agree with the rationale though, people know what I think about powerpc anyway ;) [17:06] It's removable if we remove unity-scope-mediascanner there as well [17:06] didrocks: :) [17:06] I don't especially mind doing it [17:06] cjwatson: we surely can do [17:06] Fortunately unity-scope-mediascanner Build-Depends: mediascanner so it won't reappear in an uninstallable state [17:06] cjwatson: so, then, I hardcode the list (including arm64) of archs on the binaries? [17:07] didrocks: No, just let it continue to fail to build [17:07] If it builds again it can come back, whatever [17:07] ah ok ;) [17:07] ok, great! [17:07] cjwatson: thanks for processing that one then [17:07] We should avoid hardcoding architectures wherever possible [17:07] yeah, I agree, cu2d already will skip those if last binary wasn't available on that arch [17:07] didrocks: ok, we got now all HUD crash fixes merged [17:08] thostr_: need a slot for latest HUD crash fixes? mind adding to landing ask? [17:08] thostr_: it's only the HUD right? [17:08] didrocks: yes, will do [17:08] off for a bit [17:08] (1h for now) [17:08] sil2100: FYI, cjwatson is going to ignore the powerpc failure and it's going to go to the release pocket [17:08] ogra_: ^ (if you are monitoring for building the image) [17:09] no, I'm not going to ignore the powerpc failure, I'm going to remove the powerpc binaries from saucy [17:09] cjwatson: yeah, abusing the language :) [17:09] which will have much the same effect but it's not the same thing :) [17:09] done now pending the next publisher cycle [17:09] right, I think I got the mechanic :) [17:10] thanks again [17:10] didrocks, on 93, dialer app still doesnt come up on incoming calls still cant save contacts ... so all normal :) seems good otherwise [17:11] ogra_: promoting then? :) [17:11] popey, around ? [17:12] * Mirv publishes ubuntu-ui-toolkit, after testing it a bit :P [17:12] didrocks: check #ubuntu-release ... we should be green now [17:12] ogra_: just cooking... [17:12] ogra_: wassup? [17:12] Mirv: just "a bit" ;) [17:12] Mirv: thanks man! [17:12] thostr_: yeah, we discussed there as well, thanks! [17:12] you're welcome. I hope the testing becomes 'slightly' easier very soon now :) [17:12] Mirv: me too :/ [17:13] popey, i want to promote 93/20131011 [17:13] ogra_: so ubuntu-ui-toolkit is publishing and mediascanner will transition, then image 94 is ready! [17:13] ogra_: letting you tracking those? [17:13] didrocks, and meta (for sergiusens change) [17:13] ogra_: I guess meta will be in first [17:13] ogra_: ok, 5 mins [17:13] that only went up now, i had some issues with updating it [17:14] but yeah, it will likely build faster [17:14] ogra_: ah ok, I thought you uploaded it a while ago :) [17:14] ok, those 3! [17:14] great [17:14] that way, you won't finish really late [17:14] now, I trust plars to see good results on Monday [17:14] :) [17:14] hah [17:14] no pressure :) [17:14] didrocks: I'm off on monday, but I'll be peeking in as much as I can [17:15] plars: no worry, I think you will have the image earlier than usual today :) [17:15] ogra_: no pressure at all! :-) [17:15] heh [17:16] ok, I think it's time for week-end! [17:16] enjoy [17:16] see you guys on Monday (or Tuesday for the US side ;)) [17:16] * Mirv presses enter on firefox form suggestion before selecting "force publication", apparently losing the information about available build in cu2d :S [17:17] didrocks: I guess I need to rebuild for it to have the information on what can be published again? [17:17] Mirv: yeah, it's the easiest, we can cheat, but then, it's getting into more troubles [17:17] didrocks: oh, right, good weekend to you too, I'm almost finished (though I was already) :) [17:17] yeah, rebuilding then, no new commits obviously [17:18] robru: added #202 for you [17:18] didrocks, ok [17:18] robru: when telling "running all AP tests", it's ensuring we don't regress compared to the dashboard itself [17:18] hmmm ... bug 1238696 ... what might be on this devs mind [17:18] bug 1238696 in Ubuntu Music App "untits typo is causing autopilot test failures" [Critical,Confirmed] https://launchpad.net/bugs/1238696 [17:18] robru: as we know some tests are failing [17:18] robru: and you need to restart after each app test (due to the memory leak we are discussing) [17:18] those sites are btw slightly annoying that listen for the Enter key press and the browser passes it when selecting a content from drop down list with that [17:18] Mirv: ok, I think you have to wait for an additional 25 minutes ;) [17:19] ogra_: ^ [17:19] Mirv: agreed… [17:19] didrocks: at this point, that's little :) [17:19] Mirv: heh, I think starting Tuesday, you'll be able to take short days :) [17:19] to compensate [17:19] thanks again for all the work Mirv! [17:19] didrocks: yes, because I'm on vacation Wed-Fri :D [17:19] plars, doanac need to change the terminal tests to click for the next run [17:19] sergiusens, doanac: ack [17:20] Mirv: let's ensure you finish early on Tuesday then ;) [17:20] no prob, this ui-toolkit battle was sort of "battle against AP in mir-by-default" sort of learning experience [17:20] indeed ;) [17:20] * didrocks really waves good evening now [17:20] see you guys! [17:20] ogra_: ok, back [17:20] ttfn didrocks [17:20] didrocks: yep, bye! [17:20] popey, great [17:20] bye popey, Mirv! [17:21] popey, publishing 93/20131011 [17:21] cool [17:21] writing mail [17:22] thx [17:26] sent [17:26] \o/ [17:29] nice [17:29] was that 3 images today? [17:29] :) [17:29] 2 iirc [17:30] ok.. still pretty good for a leisure friday :) [17:30] 92 and 93 [17:30] and lots new landing asks [17:32] so someone pinged me somewhere and now i lost the tab :/ ... sorry if it was yiour ping :) [17:33] heh [17:33] if it is important they will call again [17:34] is jenkins going bonkers? or just being manually frobbed to restart at the moment? [17:35] dobey, i think Mirv restarted some builds, not sure thats related [17:35] ogra_: no, just sdk stack / ui-toolkit, shouldn't be seen outside that [17:35] ah, k [17:36] * ogra_ doesn know exactly nothing about our CI infra :) [17:36] ogra_: please learn more... especially how to retry jobs (lool knows by now) [17:36] plars and siva will be gone for next week mostly :) [17:36] deep diving image testing for desktop [17:37] k, i'll hook up to lool [17:37] (next week though) [17:37] all a bit messed up, but we keep the job flowing! [17:37] yeah... ogra_ close this week :) [17:37] once 94 is building [17:37] great work here. 93 is awesome [17:37] and we will land more stuff until almost release day [17:37] if we can :) [17:37] we can [17:38] well, stuff that we own [17:38] we just continyue to process as much as we can [17:38] not everybody will get in, yes, but we can still move small hills :) [17:38] inidicators, NM etc are all ruled out [17:38] not necessarily [17:38] those can also still go in :) [17:38] with a lot of begging, yeah [17:38] just a bit annoying and requires strict control what [17:38] asac: well, not gone [17:38] drowning though [17:38] its not begging. its applying strict pre-landing-rules :) [17:38] :) [17:38] similar to ours [17:39] asac, they involve begging and bribing since ubuntu exists [17:39] plars: we shouldnt put you on the hook to own too many critical things. [17:39] alone :) [17:39] thats not good [17:39] its just that the bribes changed over the years :) [17:39] no i think its all a known process now [17:39] asac, see /topic in #ubuntu-release :) [17:40] * ogra_ points to " we accept payment in cash, check or beer" [17:40] ogra_, Mirv: i see a build was aborted, and the message posted on the merge proposal has the private IP links rather than the jenkins.qa.ubuntu.com links, for the build logs [17:40] ogra_: yeah :) [17:40] nice one [17:40] i saw that before but forgot [17:40] (used to be chocolate and fuzzy teddybears) [17:41] they got more greedy [17:41] :) [17:41] more realisitic :) [17:41] heh [17:41] i have beer and cash, but no fuzzy teddybears === gatox is now known as gatox_lunch [17:41] see, it used to be quite some effort in the past :) [17:42] so to sum up it might indeed reflect an increase in corruptability of the release team :)... one could make a study on this i guess [17:42] haha [17:44] cjwatson, wonderful mail to ubuntu-users ! [17:51] allrighty, ui-toolkit really released this time [18:05] dobey: if it's 'sdk' stack ('ubuntu-ui-toolkit' package), then it's ok and handled, and otherwise it's something completely else I don't know about [18:08] dobey: and the only merge proposal I've handled is https://code.launchpad.net/~ps-jenkins/ubuntu-ui-toolkit/latestsnapshot-0.1.46+13.10.20131011.2-0ubuntu1/+merge/190735 which I merged manually (since auto-merger is disable for ubuntu-ui-toolkit) [18:09] Mirv: no, it's this one: https://code.launchpad.net/~mikemc/ubuntuone-credentials/add-qml-unit-tests/+merge/189426 [18:11] dobey: yeah, no idea / haven't been there [18:13] fginther: ^^ any ideas? [18:15] dobey, the node that was running the job went offline during the build [18:16] fginther: will it auto-retry? is that why the URL has the IP instead of the public jenkins domain? [18:16] dobey, no, but i did restart it [18:17] dobey, yes, that's why the link is there [18:18] dobey: fginther: actually I'm interested in that if by any chance it affects the -credentials autopkg tests that ubuntu-ui-toolkit needs to have run to get migrated to -release http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html [18:19] just noticed that [18:20] but yeah right I'm really going to sleep now bye -> [18:20] fginther: right, the links being there make sense. but it using the private IP for the log links instead of jenkins.qa.ubuntu.com doesn't [18:20] dobey, the job actually runs on the private IP, the public jenkins is basically just a read-only instance [18:21] the job can't be restarted from the public instance [18:21] fginther: yes, but it normally puts the public jenkins URL for the logs and only the private IP for the restart link [18:21] fginther: so one doesn't have to connect to a VPN to see the logs [18:21] however, in this case it didn't do that [18:22] dobey, ahh... the other IPs were caused by a misconfigured script after the jenkins service move. those are now being replaced [18:22] this job was ran before the fix was in place [18:24] fginther: oh, because i've seen failure messages elsewhere that had the public address after the move. (after I brought it up on the list in reply to the move e-mail that is). so it seemed weird to me that this one didn't use the public address. [18:25] dobey, the rebuild is done, the current comment has the correct urls [18:30] thanks [18:34] sergiusens, doanac: terminal app is good to go from our end [18:39] plars, thats good, since it already went :) [18:40] ogra_: we don't have a new image yet though [18:40] plars, nope, but meta is in the archive [18:40] ogra_: speaking of which... what's the best way to find out what changed in the click packages for a new image? [18:41] plars, the next image shoul dhave them listed in the .changes [18:41] ogra_: oh awesome! [18:41] i uploaded a change that attaches them to the manifest (hope it works as expected :) ) === gatox_lunch is now known as gatox [19:00] ogra_: Liam has some curious ideas sometimes ... [19:01] heh, yeah [19:01] he's a good guy though [19:02] ogra_: heya [19:02] ogra_: what kind of jobs would you want to retry? [19:02] ogra_: I got all my ci / autolanding training from Francis, who was a great teacher for the 15mn it took to get me there :-) [19:03] ogra_: but you'll need permissions on jenkins too, and it's also becoming less necessary once the projects are back to autolanding enabled [19:03] ogra_: there are 2 jenkinses with autolanding [19:04] ogra_: coreapps, in our canonistack cloud, and internal canonical one for other projects [19:04] I'll DCC send you my notes because that seems like a good 90s way of sending a file [19:11] ogra_, lool: running unity8-autopilot on my phone, I get 12 failures (as opposed to 7 listed in smokeng), so i guess we can't release this unity-mir fix then? [19:13] lool, nne for tonight, asac meant that in general [19:14] *none [19:14] ogra_, lool: I'm not very familiar with unity8's autopilot generally http://paste.ubuntu.com/6223834/ [19:14] lool, i'm just waiting for mediascanner and ui-toolkit to be landed, will kick off an image and be gone for the night ... [19:20] robru: which image are you on? [19:20] robru: I dont think it's the way to run the tests though [19:20] lool, latest stable, but I installed unity8, unity-mir, ubuntu-keyboard from daily PPA [19:20] robru: https://wiki.ubuntu.com/Touch/Testing#Testing_your_Ubuntu_Touch_Code_before_submission [19:20] lool, how should i run the tests then? [19:21] robru: you want latest *proposed* image though [19:21] robru: see wiki for how to run the tests [19:21] ogra_: ok; these are tested though? [19:21] lool, those instructions will just install latest stable [19:21] ogra_: Yeah I understood it wasn't urgent; I've sent you an email with them [19:23] robru: I'm speaking of the way to run autopilot [19:23] phablet-test-run -n unity8 [19:23] robru: this is the official run rune [19:24] ok. how do i install latest dev image? [19:24] robru: seriously? :-) [19:24] robru: --channel devel-proposed [19:24] lool, ok [19:24] lool, well, these were tested abnd are the last two bits to go into #94 [19:25] * ogra_ goes back to the TV [19:25] will be back in 1h and kick the build [19:27] ogra_: great; thanks for confirming [20:06] lool, you'll be happy to hear https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1238637 is Fix Committed already [20:06] Ubuntu bug 1238637 in unity-mir "Tapping on OSK does nothing once unity8 is restarted" [High,Fix committed] [20:07] Saviq: great [20:08] Saviq: did I have a personal link to this bug? or is it particularly important? the one I personally saw recently seemed to be around mir connection with slighlty different symptoms [20:08] lool, it was in the "important bugs" document [20:08] lool, https://docs.google.com/a/canonical.com/document/d/1niO7JEyGEcs8W1h_DsAsAnAUMAOCHh3od0NARwGf3N8/edit# [20:09] ah the Julien list [20:09] "The Julien list" [20:09] :-) [20:09] Saviq: it also feels like it could help u8 tests [20:09] Saviq: cause I saw another one which was similar but different [20:10] Saviq: I changed the unity8 upstart job to rm the mir socket on startup [20:10] just to see if that'd help [20:10] Saviq: but the results was a bunch of apps such as osk talking to an old socket [20:10] which actually made we wonder if we should have a mir socket management more like dbus's [20:10] Saviq: the symptom was different in that typing would just crash the keyboard; it'd come back but crash again [20:11] Saviq: anyway; good news! [20:11] Saviq: you guys never stop! :-) [20:11] lool, actually we have everything in that should help u8 tests - except for any crashers we've missed [20:11] Saviq: using the phone I feel like we're lowering the bar on bug reports worth reporting because the most offending crashers are so quickly going away [20:12] I think we'll be jobless in a week at this rate [20:12] lool, HAHAHAH [20:12] isn't that the point? you'll relax on a beach then [20:12] yes, because bugs will never regress, or new bugs never introduced ever again [20:12] good one [20:12] lool, but don't worry, plenty of features still coming - with new bugs! [20:12] I'll stay on the beach and never upgrade ever again [20:13] * lool takes the 3310 out of the drawer [20:13] hehe [20:13] ogra_: src:ubuntu-ui-toolkit in release pocket === jfunk-afk is now known as jfunk [20:14] ogra_: and src:mediascanner [20:14] ogra_: were there other mediascanner related packages you were tracking? [20:15] sergiusens: Ah your mp to add a manifest to music-app is gone?! [20:16] jfunk: can you please do something about the pass rate of touch_mir going up; it's about to beat SF pass rates [20:16] Saviq: we're indeed getting unity8 crashes in tests [20:16] Saviq: e.g. http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/93:20131011:20131010/4677/dialer-app-autopilot/ [20:17] http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/93:20131011:20131010/4677/gallery-app-autopilot/ [20:17] lool, there were a slew of fixes today [20:17] lool, or are coming [20:17] as in are proposed - but most are committed [20:17] Saviq: Cool; note that we have to count the shots next week [20:17] lool, I'd like to see a run after we release current unity8 and unity-mir trunk [20:18] Saviq: likely stopping on wed, so realistically one landing on monday and perhaps a last minute landing on tuesday [20:18] + https://code.launchpad.net/~saviq/unity8/workaround-lp1238645/+merge/190724 for good measure [20:18] lool, yeah, I imagine [20:18] Saviq: curious why you need the inictl there [20:19] lool, although it'd be good to release over the weekend, so we can see the real stuff on Monday [20:19] lool, in case maliit crashes [20:19] Saviq: just "start maliit-server" doesn't work? [20:19] lool, ah, ogra_ ↑ he wants a slap [20:19] hmm? [20:19] lool, I'm not entirely sure why, but "initctl" I imagine is more reliable / future-proof [20:19] lool, "stop" "start" are pretty generic [20:20] we use them everywhere too :-) [20:20] lool, and the "correct" way is to feed them through initctl apparently [20:20] ogra_, ↑↑ he wants more slaps [20:20] ;) [20:21] lool, anyway - once we (re)start unity8 for testing with upstart [20:21] lool, that's going to go away [20:21] lool, 'cause upstart will make sure maliit is running for us === dpm-afk is now known as dpm [20:21] Saviq: the other thing is that *theoritically* maliit-server should restart if unity8 restart [20:21] lool, there's already a branch, just not complete and I wanted to see the results asap [20:21] lool, yeah, that happens with upstart [20:22] ah but we don't launch it with upstart? [20:22] lool, but atm we're directly launching unity8 in the tests [20:22] I see [20:22] lool, there's a branch [20:22] lool, it's mostly the same for all the apps' tests - need to transition over to upstart [20:22] Saviq: I see robru has landed unity8 in archive says the spreadsheet [20:22] but TBH I'd rather see a python binding somewhere instead of calling subprocess everywhere [20:22] so you'll get a run later tonight [20:22] lool, not the latest one - nor unity-mir - several commits behind [20:23] lool, it's in the big ass MP now [20:23] sergiusens: haha [20:23] sergiusens, added super() btw [20:23] sergiusens: yeah I've seen this one [20:23] sergiusens: "last week of development of all missing features in one branch" mp? [20:23] sergiusens, sorry for being an ass on this today - thanks for safe-guarding [20:24] robru: FYI INPROPOSED != INARCHIVE [20:24] ogra_: there's an unity8 in proposed which is probalby worth waiting for before kicking the build [20:24] lool, oh, sorry [20:24] robru: so tests passed for you? [20:24] lool, ogra, yes, unity8 confirmed to have many fixes [20:24] robru: how many tests passed? [20:25] lool, only 3 failures, very impressed [20:27] lool, ah, indeed unity-mir just published [20:28] yay, everything's getting pushed through, next run will be a good run [20:32] lool, btw, the .crash files on smokeng seem to be truncated [20:32] lool, a .crash for unity8 needs to be ~14MB [20:33] Saviq: ah that's a very good one [20:33] Saviq: I bet the tests stops, the crash file is captured before being done [20:33] Saviq: there are two bugs here [20:33] lool, yup, probably [20:34] Saviq: a) apport is writing straight to .crash (instead of renaming it once written) b) infra should wait for settle and grab the crash [20:34] lool, yup [20:34] Saviq: good catch; not sure where we file this, but worth pinging plars / doanac on [20:34] plars, doanac: ^ looks like we capture crash files before they are done writing [20:34] plars, doanac: Perhaps looking at their permissions (000 when being written IIRC) is enough to detect? [20:35] plars: or sleep 5 or something, but that seems costly [20:35] lool, plars, doanac, it'd be good if it did preprocessing, too (the thing apport-cli "View", "Keep" does) [20:35] this way we could avoid having to collect locally before retracing [20:35] lool, plars, doanac - one more, it should actually pick them up per-test-run [20:36] lool, plars, doanac as now we only get one .crash where 5 tests error out due to crashing [20:36] hey guys. i was just walking out the door for a bit. i'll have to re-read this transcript when I get back [20:36] doanac, no worries [20:36] doanac: no hurry, important improvement to make, or to log to do soon, but not blocking anything [20:36] lool, do you want to drive #94 ? [20:37] assuming unity8 wil still take a while [20:37] ogra_: wont take much, but can do [20:37] ogra_: whatever you prefer [20:37] ogra_: not sure I want to take away the pleasure of kicking an almost crash free build ;-) [20:37] well, i wanted to kick it off now ... and dont want to do an all nighter [20:38] ogra_: ok will kick in a bit; unity stuff stillin proposed [20:39] lool, the manifest shuld have click: lines too now [20:39] i guess the first changelog will look a bit odd [20:40] ogra_: oh that's cool, thanks [20:40] well, i only tested locally, hopefully it all works :) [20:41] what's the worst that could happen? fire starts in build farm? [20:41] :-P [20:41] ogra_: have a good night / we [20:41] you too ! [20:41] and stay away from kbd ... do something with your kid ;) [20:41] typing on my kid feels weird [20:42] I have this popey trend of doing a star wars miniset each saturday [20:42] i think cjwatson does that at times :) [20:42] awful habbit [20:42] heh [20:42] worst thing that happened to me is that they are opening a lego store some 2 km from my house [20:42] fortunately the ceiling broke or something and they delayed the opening by a couple of months [20:43] so I can still work a bit [20:43] lol [20:47] lol [20:47] robru: was pushing unity-notifications intentional? [20:47] lool, uhhhh, i just pushed the stack [20:48] robru: Was that the landing subject or was it just the unity8 source package? [20:49] lool, the landing plan was unity8, unity-mir, and ubuntu-keyboard, and hud. I forgot to check the stacks and just pushed whole stacks instead of individual packages. you can drop unity-notifications if it's a problem, i don't care about it personally [20:49] (and i'm still working on hud) [20:49] haha lool ☻ [20:50] robru: the thing is I cant drop it [20:50] mfisch, are you the powerd-cli guy? [20:50] it's going in release pocket now [20:50] lool, by 'drop' it I meant get the release team not to approve it. [20:50] fginther: I can help, but not officially anymore, whats up? [20:50] robru: it's already past that [20:50] lool, does it break anything? i don't really understand the problem [20:50] it's already approved for entering release pocket [20:50] mfisch, can it be used to query the device for the screen state (i.e. is it on?) [20:50] robru: I dont know whether it breaks anything, that's the problem! [20:51] lool, well let me install it and run the tests again i guess. [20:51] robru: that's be helpful; thanks [20:51] mfisch, I see a 'list' but don't know how to interpret [20:51] fginther: I dont think so, it can only request a state [20:51] fginther: let me fire up my phone, 1 sec [20:51] robru: the changes are around wifi password UI changes in the snap dialog, and around support for action less snap decisions albeit I dont know how to trigger [20:51] perhaps volume button [20:52] which might be useful for indicator-sound [20:52] lool, i saw some wifi password stuff pop up during the unity8-autopilot. [20:52] fginther: list is just what people have requested for the state [20:53] fginther: there's no way to ask it what the status is currently [20:53] mfisch, bummer [20:53] you can tell what the state should be from the list [20:53] and so powerd has told the display what state it wants based on those [20:53] but in theory the display could not match I suppose [20:54] mfisch, I think I'm seeing that. I ask the screen to be on, and it's not :-( [20:54] fginther: you should enable debug mode for powerd and restart it [20:54] fginther: directions are in the upstart job itself [20:55] mfisch, I shall try that, thanks [20:56] robru: as a bug or just as a running test? [20:56] robru: haha there was another landing for ubuntu-keyboard + unity-mir [20:56] lool, as a running test. what i was trying to say is that unity8-autopilot should help confirm that unity-notifications works [20:58] Saviq: ubuntu-settings-components entered landing asks many moons ago, I never knew what it was for and it sat there; is it something we still want? [20:58] robru: ack [20:58] lool, probably too late for 13.10 - obviously since it's not there yet - we're not using it [20:58] lool: the release team would like this: http://paste.ubuntu.com/6224151/ [20:58] lool, but it's components common between indicators and settings app, more or less [20:59] lool: it is a trivial one character shange to the /usr/sbin/aa-status script to use python3 instead of python [21:00] that will get python 2 off the image. why this wasn't adjusted before I don't know-- we wrote it to be bilingual in the first place [21:00] lool, wtf?? http://paste.ubuntu.com/6224218/ but the package is clearly in the PPA, which i have enabled [21:00] robru: apt-cache showsrc unity-notifications [21:00] we even have debian/control use python3 already. simple ommission [21:00] robru: to see the list of binary packages it produces [21:00] Saviq: ah QML components, got it [21:00] anyway, yes I tested aa-status with python3 on mako and it works [21:01] Saviq: should we drop it from asks then? [21:01] plars, lool was there an sdk change? [21:01] jdstrand: please upload now if you can [21:01] lool, yes, please [21:01] lool, oh, ok [21:01] plus, aa-status is run as part of the security smoketests [21:01] sergiusens: there was an ui-toolkit change, and there are some qt changes too [21:01] sergiusens: what's up? [21:01] lool: ack, thanks [21:01] plars, lool this is what I get with the clock app: creating surface at (0, 58) with size (720, 1222) with title 'Window 1'creating surface at (0, 58) with size (720, 1222) with title 'Window 1'This plugin does not support propagateSizeHints() [21:02] that's clock app [21:07] rsalveti: ofono isn't ci-ed right? Do we need to merge https://code.launchpad.net/~rsalveti/ofono/apn_mnc_passwd_fixes/+merge/190424 by hand? [21:07] rsalveti: Or I can kick autolander [21:07] oh wow the release is behind... [21:07] sergiusens: I read it was broken indeed [21:07] Saviq: which one? [21:07] lool: hm, thought cyphermox fixed that [21:07] lool, unity8 [21:07] lool, 435 released as r449... [21:08] Saviq: Hmm changelog says 449 [21:09] lool, it's not from a clock change though [21:09] lool, Releasing 7.82+13.10.20131011.1-0ubuntu1 (revision 435 from lp:unity8) [21:09] lool, there's a difference between where the release *changelog* is merged in, and where it actually fits [21:09] lool, according to popey this is the build that broke it http://people.canonical.com/~j-lallement/touch/changes/20131011.html [21:09] Saviq: this merge is crash, but the uploaded changelog says snapshot of r449 [21:09] *crack [21:10] robru: what happened to the unity8 changelog? [21:10] lool, oh ok [21:10] lool, where? [21:10] lool, somebody must have done a manual release to distro, because there was a changelog in distro that wasnt' in trunk. jenkins gets fussy about that, so i had to merge it back into trunk despite being quite stale [21:10] lool, 44 [21:10] [ Ubuntu daily release ] [21:10] [21:10] 45 [21:10] * Automatic snapshot from revision 435 [21:10] Saviq: https://launchpad.net/ubuntu/+source/unity8/7.82+13.10.20131011.2-0ubuntu1 [21:10] lool, ah so that's not merged to trunk yet [21:11] Saviq: I dont think it will though [21:11] Saviq: let me try [21:11] *idea* daily release should be tagging revs on release [21:12] lool: uploaded. #ubuntu-release pinged [21:12] Saviq, it does, but some genius turned off autolanding, so now we have all these changelog MPs sitting around getting stale [21:13] Saviq: https://bugs.launchpad.net/mir/+bug/1238417 shouldn't say fix committed or should it? [21:13] Ubuntu bug 1238417 in Mir "Unity does not process events from evdev device created before unity is restarted (autopilot tests)" [Critical,Fix committed] [21:13] lool, nope [21:14] robru, right... [21:15] Saviq, robru: Kicked autolanding of unity8 changelog mp [21:15] lool, thanks [21:15] lool, kicked? how? it was already approved. what did you do [21:17] lool, also, tests are completely fucked here. won't even run. doesn't seem like unity-notifications fault though (for that I would expect simple test failures). no idea wtf is happening [21:18] Saviq: https://bugs.launchpad.net/ubuntu/+source/indicator-sound/+bug/1236292 should be fix commited in unity8? [21:18] Ubuntu bug 1236292 in Unity 8 "Using volume buttons on phone triggers spurious popup" [High,In progress] [21:18] robru: I actually ran the autolanding job [21:18] robru: since as you noted it's disabled [21:18] robru: did you reboot? [21:18] lool, released, even [21:18] lool, marked so [21:18] lool, can it be enabled? because it's gonna need to get run for every package that we release, ever [21:18] lool, yes, rebooted many times. [21:19] robru: it's been reenabled on packages with green testsuites [21:20] robru: so what did you change since it was mostly passing? [21:20] lool, uh, installed hud I guess. [21:20] robru: maybe try downgrading it then? [21:20] robru: did you install all hud packages or just hud itself? [21:20] lool, just hud i think [21:21] hmm actually it's a single package it seems [21:21] robru: try downgrading to saucy? [21:21] lool, ok, gonna try it [21:22] Hmm not it's NOT a single package [21:22] ah right typo here [21:22] robru: You need to install ALL binary packages [21:22] robru: (that were prevously installed) [21:22] lool, ok, trying that too [21:22] robru: That's crticailly important for testing the right thinsg [21:25] lool, do you make sense of this? https://bugs.launchpad.net/ubuntu-clock-app/+bug/1238798/comments/5 [21:25] Ubuntu bug 1238798 in Ubuntu Clock App "Clock app doesn't work on mir on maguro" [Critical,Triaged] [21:26] sergiusens: The hud warnings, do you think they are serious? [21:26] Could not determine application identifier. HUD will not work properly. [21:26] Provide your application identifier in $APP_ID environment variable. [21:27] sergiusens: if you want to start it, use upstart-app-launch-tools or start application-click APP_ID=com.ubuntu.clock_xyz [21:27] not sure of exact vars [21:27] lool, running through upstart just makes the app blank [21:28] sergiusens: check .cache/upstart/application-click-com.ubuntu.clock_xyz.log? [21:28] sergiusens: oh dconf [21:28] sergiusens: dconf wont work, will it? [21:28] sergiusens: dmesg|grep DEN? [21:29] lool, https://bugs.launchpad.net/ubuntu-clock-app/+bug/1238798/comments/6 [21:29] Ubuntu bug 1238798 in Ubuntu Clock App "Clock app doesn't work on mir on maguro" [Critical,Triaged] [21:29] lool, also, I'm not running it under confinement [21:30] lool, dconf works for certain cases; jdstrand can confirm [21:30] clock has the 'calendar' policy which lets it talk to eds [21:30] or at least was working until today [21:32] lool, we need a release of upstart-app-launcher for bug #1229551 [21:32] bug 1229551 in unity8 (Ubuntu) "Facebook launcher doesn't work on Home scope" [Medium,Fix released] https://launchpad.net/bugs/1229551 [21:32] -er [21:32] eeerm [21:32] url-dispatcher, that is [21:33] dobey: somehow I managed to create an empty email address account again; keyboard was crashing, and I ended up with literally 2 single quotes ('') in email address and then coulnd't enter password (keyboard was crashing); fixed keyboard that I had specifically forgotten to upgrade; empty account was there and broke install [21:33] dobey: but I guess that's not a regression [21:33] lool, shall I add to ask or is it there already? [21:33] Saviq: it was released earlier today [21:34] lool, https://launchpad.net/ubuntu/+source/url-dispatcher [21:34] 1003 [21:34] Saviq: r76 is what we have [21:34] of u-al [21:34] lool, sorry - not u-al that's fine [21:34] u-disp [21:35] that one is riskier but it's needed for stuff [21:35] I thought we had queued it for today [21:36] Saviq: Ok, I know about that change and talked to didrocks about landing it to allow more clickifcations [21:37] Saviq: if it's useful for other things, could you mention which? [21:37] Saviq: I thought it was only useful for opening urls with click packages [21:37] do you have appid:// use cases? [21:37] lool, yeah, although temporary - we have a set of "frequent" apps in dash home [21:37] lool, those that are clicks would need hardcoding their version without appid:// support [21:38] lool, so tricky [21:38] Saviq: aha ok [21:38] Saviq: Yeah, +1 [21:39] lool, so my system was completely busted, I went to sf to get a bt [21:42] sergiusens: erf ok [21:43] Saviq: I've updated the ask [21:43] lool, thanks [21:43] Saviq: will let didier pick it up monday [21:43] dobey: rejecting the click scope update -- list of installed apps only updated on install, not on removals; unity8 crashed on second removal [21:46] dobey: https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1238980 [21:46] Ubuntu bug 1238980 in unity-scope-click (Ubuntu) "Doens't update list of installed apps on removal and sometimes crashes unity" [Undecided,New] [21:48] robru: where are you with unity-notifications and hud testing? [21:49] lool, still not working after updating all of hud; just reflashed to latest dev image; just installed unity8 but not hud; just about to start tests to see if notifications are the source of the breakage [21:54] robru: I'll be building an image in a few; wont wait for hud landing; unity-notifications went into archive with unity-mir anyway, so will kick it with image 94 [21:55] robru: please confirm/infirm the AP regression with unity8 APs and report as one bug against unity8+unity-notifications and add the bug id to the spreadsheet [21:55] lool, ok. right now i tried as best as possible to recreate the env that only had 3 failures, and the tests seem to be running ok. when that's done i'll install just unity-notifications and then see how badly broken it is; that will determine whether it needs to be reverted or not [21:55] ok [22:00] rsalveti: lool: what is it that you thought I had fixed? [22:00] cyphermox: nothing specifically, was just going to test the attach bug fix for ofono [22:00] wanted to make sure I was using the latest nm in the pipeline === dpm is now known as dpm-afk [22:02] rsalveti: there is a bug fix for NM that also helps with the Attached thing, but that's not uploaded... it's not something you'll hit every time though, it's just another bug that you can hit sometimes after reconnecting [22:02] cyphermox: got it, have the bug number for that? [22:34] robru: I've scheduled a build for 2 hours from now unless another fix lands [22:35] @ci am I the only one? The server at 10.97.0.26 is taking too long to respond. [22:36] sergiusens: wfm [22:36] sergiusens: but isn't clock-app on the canonistack one? [22:36] lool, yeah, but the click build is here ;-) [22:36] sergiusens: I can load e.g. http://10.97.0.26:8080/view/Apps%20Team/job/unity-mir-autolanding/ [22:36] over VPM [22:36] VPN [22:37] yeah, vpn reconnect worked [22:37] seems my vpn connection went stale [22:44] lool, great news, unity-notifications didn't break unity8-autopilot. [22:45] lool, still working on hud though === dpm-afk is now known as dpm [22:47] robru: ok [22:53] * lool & [23:02] LOOL AND WHAT?? LOOL AND WHAT?!?!?!