=== xnox is now known as doko_ === doko_ is now known as xnox === jono is now known as Guest92387 === duflu_ is now known as duflu === jalcine is now known as doge_ === doge_ is now known as jalcine === alan_g is now known as alan_g|afk === iahmad__ is now known as iahmad === jalcine is now known as doge_ === doge_ is now known as jalcine === alan_g|afk is now known as alan_g [08:59] sil2100, hey, did you see my comment on the bug: https://bugs.launchpad.net/unity8/+bug/1262743/comments/2 ? [08:59] Ubuntu bug 1262743 in Unity 8 "Latest lp:unity8 trunk autopilot shell.tests.test_notifications.EphemeralNotificationsTests.test_summary_only is flaky" [High,Invalid] [08:59] sil2100, filed a related bug on ci processes: https://bugs.launchpad.net/ubuntu-ci-services-itself/+bug/1262879 [08:59] Ubuntu bug 1262879 in Ubuntu CI Services "There should only be one, documented, way to run tests on devices" [Undecided,Confirmed] [09:00] Saviq: hi, just looking at it [09:00] Saviq: I wonder what happened that it failed that one time on my device - maybe unity8 couldn't properly got woken up by autopilot? [09:00] *get [09:01] sil2100, no, unity8 sometimes take quite some time to tear down everything [09:01] sil2100, in which time it might end up the screen suspends [09:01] sil2100, there's no way for unity8 to wake the screen [09:01] sil2100, when it's starting - I wrote all of that to ubuntu-phone ML [09:04] Saviq: ok, so what I'll do is run the test suite again and then try doing the release [09:11] sil2100, yes please, make sure to keep the screen on with the powerd-cli command [09:12] sil2100, FWIW this only happens in trunk for you due to https://code.launchpad.net/~nicolas-doffay/unity8/upstart-kill-fix/+merge/198931 [09:12] sil2100, before, upstart would kill unity8 after 5s [09:12] now it only times out after 30s [09:13] so before it would never take long enough to blank the screen (but also not long enough to collect a .crash, hence the override) [09:54] dednick, hey, can you mark affect the respective indicators that have incorrect positions: https://bugs.launchpad.net/ubuntu/+source/indicator-network/+bug/1262611 ? [09:54] Ubuntu bug 1262611 in unity8 (Ubuntu) "Network indicator hidden when there is no cellular signal" [Undecided,Confirmed] [09:57] Saviq: sure [09:57] dednick, positions come from indicators now, right? [09:58] Saviq: yep, service files if i remember correctly [09:58] dednick, yup, please comment on the bug what the values are now and where they need to be fixed - and mark unity8 Invalid afterwards [09:58] Saviq: ok [10:05] Cimi: did you run the calendar test suite yesterday? the entry is there, but i had the same results (might have just entered in incorrect box) [10:06] dednick, no calendars [10:06] Cimi: ok cool. [10:08] dednick, and battery of the tablet is dead :D [10:08] willl do in a bit [10:09] Cimi: ok, all mine failed. calendar doesnt seem to work at all with sidestage [10:50] Saviq, you probably got a bunch of emails about the json doc :) [10:50] thostr_, you might want to re-look at well [10:53] mhr3_, not yet, but will have a look [11:06] mhr3_, ah, so by EarlyPreview you actually mean with no round-trip between the UI and the scope [11:06] mhr3_, that's something I wasn't thinking about [11:06] Saviq, oh, that wasn't clear? [11:07] :/ [11:07] mhr3_, yeah, I always thought about a minimal round-trip, but now I get it - and yeah, probably nothing tragic happens to the JSON structure anyway [11:09] right === tvoss is now known as tvoss|lunch [11:37] sil2100, any luck? [11:39] Saviq: it's as you say, the tests seem to look fine - there's like one test that fails when the screen fades out [11:39] But rarely [11:42] sil2100, cool [11:43] Saviq: so I'll release it once cu2d says it's ok [11:43] have you guys tried greyback debs on the nexus10? [11:43] i'm getting a diying unity8 too [11:43] nexus4 was fine [11:44] but nexus10 unity8 keeps restarting [11:45] tsdgeos, not restarting here, works fine? [11:45] Saviq: extra ? in your sentence? [11:46] tsdgeos, yeah, probably ;) [11:46] tsdgeos, it does not restart here [11:46] tsdgeos, although what mzanetti reported happens - unity8 tests don't work (but they don't on mako, either) [11:46] well i get unity8 on screen on the nexus4 [11:46] nexus10 all i have is blackness [11:47] tsdgeos: with AP too? [11:47] mzanetti: why would i run autopilot if untiy8 doesn't even start [11:47] tsdgeos: it works fine when running it manually. But with AP it fails [11:47] for me at least [11:47] on Nexus 4 [11:47] well, both. I see the same on 4 and 10 [11:48] the question was rather if it works on Nexus 4 with AP for you [11:49] is qa vpn down? [11:49] dednick: can't connect either [11:50] mzanetti: sure, nexus4 is all fine [11:50] hm. i think all jenkins is down [11:50] oh, no [11:50] dednick, your branch failed due to otto failure, but couldn't restart [11:50] Saviq: otto? [11:50] dednick, the testrunner [11:51] Saviq: ah, yeah. I was trying to restart, but cant get to it.; [11:51] dednick, otto is the project that actually runs the tests in s-jenkins and daily build [11:51] dednick, yup, me neither [11:52] tsdgeos, greyback's sidestage packages worked for me on nexus 10 as well [11:52] :/ [11:52] dandrader: image 79 ? [11:53] tsdgeos, but I did have a terminal on with "powerd-cli display bright" during the tests to make sure I would see what's happening throughout [11:53] my unity8 is crashing constantly [11:53] that is not my problem [11:53] didn't work for me... AP doesn't work neither on nexus 4 nor nexus 10 for me [11:53] hmm once Nexus10 sleeps here, I can't wake it up :/ [11:53] tsdgeos, the image number I used I put on that spreadsheet [11:53] starting it manually works fine on both for me [11:54] hey, ricmm was working on fixing bugs with his sidestage stuff, here's his latest packages: http://people.canonical.com/~ricmm/sidestage-debs.tar.gz [11:59] tsdgeos, I'm on 79 and works fine [11:59] that is weird :_/ [12:00] * tsdgeos reflashes and reinstalls the debs [12:00] tsdgeos: i think mine is crashing pretty often as well. Have you got the sidestage mods installed? [12:00] tsdgeos, hmm maybe it crashed just now... [12:01] * Saviq clears /var/crash [12:01] dednick: mods? [12:01] and maybe that's why I can't resurrect it when it goes to sleep [12:01] tsdgeos, he meant the .debs [12:01] tsdgeos: the packages for sidestage we are testing with [12:01] dednick: yes, with greyback's debs crashes a lot [12:01] otherwise it's fine [12:02] tsdgeos, check out ricmm's debs above ↑↑ [12:02] Saviq: how much above? :D [12:02] ah [12:02] got them [12:02] tsdgeos, not a lot ;) [12:03] lol [12:03] anyone else getting upside-down screenshots from time to time? [12:04] in the side stage, when swiping through the launcher, I'm getting upside-down screenshots sometimes ;) [12:06] crash [12:06] tsdgeos, so yeah, it's pretty crashy, not as much as yours is, probably ;) [12:08] greyback, you going through the issues on mako? I'll go on manta then [12:09] Saviq: yep [12:09] greyback, gallery was manta first [12:09] greyback, should be mako? [12:09] Saviq: oh darn, fixing [12:10] greyback, right-edge switch, do you see a flicker when the apps get into place (so on focus change?) [12:11] greyback, think it's the same as mzanetti's flicker you discussed earlier this week? [12:11] Saviq: no actually [12:11] greyback, I do, on manta, will report separate [12:12] Saviq: I do get flicker if i activate running app from launcher [12:12] greyback, how about upside-down screenshots? ;) [12:12] Saviq: you're holding it wrong :D [12:13] greyback, no, I just can't rotate it that fast [12:15] tsdgeos, crashes in the application plugin here, for you? [12:16] qtubuntu somewhere [12:17] already reflashed to try ric's packages in a clean environment [12:19] greyback, how about general jitter when doing transitions? [12:19] greyback, it looks like frames are delivered in wrong order or something... [12:20] greyback, whenever I launch apps or switch between them, the screenshots jump around like crazy [12:21] Saviq: aside from white flickers where screenshots not loaded yet, most animations are smooth here. I'm on image 77 though (yesterdays). I need to update to today's [12:21] ric's packages also make it die [12:22] tsdgeos: backtrace? New Mir release was made, maybe API/ABI break [12:23] greyback: http://paste.ubuntu.com/6605309/ [12:23] greyback: but it is weird, since it worked on the Nexus4 :-S [12:23] tsdgeos: ok, looks like qtubuntu fail for tablet [12:24] greyback: want me to try to compile something manually? or? [12:25] tsdgeos: I'll be ~30 minutes before I can reproduce. If you'd like to dig, be my guest. [12:26] tsdgeos: https://code.launchpad.net/~ricmm/qtubuntu/papi-setdimensions/+merge/198498 & https://code.launchpad.net/~ricmm/platform-api/set-dimensions are the branches [12:34] greyback, tsdgeos new mir not there in image 79 yet anyway [12:34] the look of http://s-jenkins.ubuntu-ci:8080/job/unity8-ci/ is quite sad -/ [12:34] i am totally puzzled as why it works for Saviq and not for me :-( [12:34] tsdgeos, it's pretty crashy here, too, so... [12:34] dudes, we got a unity8 release, thanks sil2100! [12:34] ah ok [12:35] * Saviq awaits the 100% for unity8 on maguro finally :) [12:35] \o/ (finally ;p) [12:35] Yeah sil2100! Why did that take so long, huh?! [12:35] Oh, right, that's me [12:35] tsdgeos, otto seems to get locked again... http://s-jenkins.ubuntu-ci:8080/job/autopilot-testrunner-otto-trusty/1673/artifact/results/autopilot/artifacts/unity8.shell.tests.test_hud.TestHud.test_show_hud_appears%20%28Desktop%20Nexus%2010%29.ogv [12:36] :/ [12:36] tsdgeos, not so for the last three runs, though, so hopefully an isolated issue [12:44] greyback: will we get a possibility to know when an application actually starts painting after startup? [12:44] mzanetti: when we use the scenegraph work, yes. [12:45] ok. great. [12:49] greyback: lunch calling, didn't have time to compile anything, back layer [12:50] ok === alan_g is now known as alan_g|lunch [13:22] Saviq: just testing manta, I see what you mean about jitteriness [13:22] greyback, yeah, it's rather bad [13:22] yep === alan_g|lunch is now known as alan_g [14:02] greyback: any luck? did you repro the crash? [14:03] tsdgeos: what were you doing to make it crash? I've not crashed it yet on tablet [14:03] (no I lie, I did once, but failed to do it again) [14:04] greyback: install, reboot -> done [14:05] tsdgeos: nope, not getting that [14:05] * tsdgeos shakes fist [14:05] tsdgeos: I instlled image 79, installed ricmm's packages, it's ok [14:05] same i did [14:05] greyback: you did reboot, right? [14:06] tsdgeos: yes [14:06] * tsdgeos tries for the fourth time [14:07] Saviq, gauge! [14:07] mhr3_, comments? ;) [14:10] mhr3_, aah rating? [14:11] * Saviq probably connects "gauge" to "12-gauge" most... not sure that's a good connection to have in the scopes system ;) [14:12] Saviq, lol, i thought you're asking me to add it as a comment [14:12] anyway, i did :) [14:12] had to google what 12-gauge is [14:15] mhr3_, now you now ;) [14:16] yea, at least it's not as bad as last time when i was googling something about one cup [14:18] mhr3_, MAN, you don't mention that [14:18] you DO NOT MENTION THAT [14:19] * Saviq goes to find gasoline to pour on his eyes [14:20] * mzanetti is curios now [14:20] greyback: insatlling all of these, right? http://paste.ubuntu.com/6605904/ [14:20] greyback: that's the sha1sum [14:20] mzanetti, don't be, you'll regret it [14:21] mzanetti, DON'T [14:21] heheh, I remember that one cup craziness [14:21] this doesn't help, Saviq :D [14:21] tsdgeos: http://pastebin.ubuntu.com/6605912/ [14:21] mzanetti, really, not before christmas [14:21] lol [14:21] Saviq, think it'll soon claim another victim [14:21] mzanetti, leave it for new years [14:21] ok. remind me if I forget [14:22] mzanetti, just keep your webcam recording you as you find out what's that about [14:22] will make a good video [14:22] ;) [14:23] there's probably more "guy watches two girls one cup for the first time"-type video results when you google for it these days than the actual video ;) [14:23] s/guy/someone/ [14:23] uh oh [14:23] now he knows what to google for [14:23] hmm... [14:24] it does sound like I should not use the canonical managed google account for it [14:24] leave it after work hours at least [14:24] nsfw [14:24] ok [14:25] nsfwaa [14:25] as in "not safe for watching at all" [14:25] :D [14:27] nope [14:27] nothing [14:27] * kgunn really wonders what all this one cup thing is [14:27] somethins wrong with my nexus10 [14:28] * kgunn heeds advice he sees not look [14:28] or what? [14:28] * tsdgeos confused [14:32] tsdgeos: it's a mystery to me, sorry [14:37] kgunn, do not. [14:37] :) [14:37] taking the advice [14:37] woohoo....AP tests running on sidedstage mako for me now! [14:37] Saviq, https://code.launchpad.net/~mhr3/unity-scopes-shell/json-overrides/+merge/199508 pls? [14:38] guessing i just did somethign stupid [14:38] Saviq, you mentioned some packages, but qtubuntu is not packaged [14:38] hmmm...i missed standup huh [14:38] mzanetti, ^ or maybe you want to see some scopes code again :) [14:38] Saviq, you have a special link for it? === dandrader is now known as dandrader|afk [14:38] mzanetti, this time with 100% fewer threads :) [14:38] Cimi: http://people.canonical.com/~ricmm/sidestage-debs.tar.gz [14:39] mhr3_: :D [14:39] Saviq: also noticed that apps that are supposed to be running in sidestage, actually dont under autopilot. Is that expected? [14:39] thx [14:39] mhr3_: yeah. I guess I can do that one [14:39] dednick, should be fine with the packages above [14:39] mzanetti, thanks [14:39] ah, new ones.. [14:40] dednick, new ones this morning, yeah [14:40] Saviq: oh, no, i already have those [14:41] dednick, hmm so they run in sidestage manually, but not under ap? [14:41] and now it works [14:41] * tsdgeos is even more confused [14:41] Saviq: yeah [14:41] dednick, that sounds like we're not reading the hint from the .desktop file in that case [14:42] Saviq: greyback: so i did the phablet-flash with --boostrap instead of the regular one and now it's working, no clue why that would mean anything different [14:42] tsdgeos, interesting indeed [14:42] dednick, since they're not launched via upstart, maybe we're not taking the stage hint into account [14:42] greyback, WDYT ↑? [14:42] greyback, apps not going to sidestage under autopilot? [14:43] Saviq: hm, let me check what happens when i start with cmdline [14:44] Saviq: yeah, that's it [14:44] dednick, add bug to http://pad.ubuntu.com/qP0HD1BUn4 please [14:46] there's the stage_hint command line switch. But ricmm added to platform-api code to read the APP_ID env var, and from there open the desktop file and read the stage. So APP_ID needs to be set [14:46] this is excellent work guys... [14:49] does terminal app for for anyone? [14:49] s/for for/work for/ [14:49] greyback: looks like starting it with stage_hint=SideStage is doing "something", but not the correct thing :) it get's the size right, but the it's on the left. [14:50] dednick: Cimi ...i'll do the mako runs for the tests you have [14:50] i don't think either of you have n4s [14:50] kgunn, thanks [14:50] kgunn: thanks [14:51] dednick: another bug, please report :) === dandrader|afk is now known as dandrader [14:53] mhr3_: fails to build with http://paste.ubuntu.com/6606060 [14:54] heh... nice pastebin number [14:54] mzanetti, old libunity-api? [14:54] mzanetti, update && upgrade [14:54] mhr3_: on it, yeah [14:55] out for a bit. bbl [14:55] unity8 is crashing for me as well on manta... === dednick is now known as dednick|lunch [14:56] dandrader: all the time or just sometimes? [14:57] tsdgeos, I left the device alone for some minutes, when I came back to use it unity8 was gone. restarting it doesn't help. screen remains black even with "powerd-cli display on bright" [14:58] dandrader: adb shell and ps -A | grep unity8 [14:58] is it different pids all the time? [14:58] dandrader: recent crash file for unity8 in /var/crash/ ? [15:00] hmm, pulseaudio using 9% of CPU on tablet after fresh boot, wonder why [15:00] greyback, no [15:00] shame [15:00] is it normal (or unheard of) to have an upstart app launch crash is screen blanks right before an AP test runs ? [15:00] just checking if the crash i just saw is junk (expected) [15:01] kgunn: I'd argue that any crash is not normal :) [15:01] but that's news to me [15:01] greyback: most people run the tests correctly ;) [15:02] dammit powerd, stop reducing my default brightness to sod-all [15:05] well [15:05] the sidestage packages are broken [15:05] i'm gonna say that crash was crap [15:05] not really sure makes sense testing more :D [15:06] start calculator, works, close calculator, start calculator, doesn't work, close calcular, start calculator, doesn't work, close calcular, everything blows up [15:06] tsdgeos: assume you mean on n10 ? [15:06] yep [15:07] tsdgeos: yep, am seeing similar, it's fragile [15:07] greyback: think we have enough bugs to halt...then we can repeat this effort in the new year with a new branch ? [15:08] or is it useful to complete at least the AP testing [15:08] kgunn: yep, I think we can stop. ricmm & I will start fixing the obvious problems. Since app launch & stop isn't reliable, AP will just give us lots of false fails [15:09] Saviq, did you have time to rebase the scope tool on trunk? [15:09] ok... dandrader dednick|lunch Cimi Saviq ^ [15:09] Saviq, if not i can do it [15:09] mhr3_, no, didn't get to it - please do [15:09] ok [15:09] greyback: you and ricmm need any help ?...or is 2 cooks already crowded ? [15:10] Saviq, k, got anything unpushed for the branch? [15:10] if so, pls push [15:10] mhr3_, don't think so, let me check [15:10] ricmm what you think? I think 2 is enough [15:10] now I got 55 failures [15:10] :-\ [15:10] mhr3_, ready for you [15:10] Saviq, thx, on it [15:11] * ricmm readsback [15:11] kgunn: 2 is enough [15:12] kgunn: a handful of these bugs are not new bugs, these are just bugs that came from MWC, so not really regressions [15:12] others are simple to squash [15:12] 2 is good, we might have something cool in a few hours [15:13] ricmm: love to hear that [15:20] mhr3_: added one comment. let me know if I should approve or you want to change === dandrader is now known as dandrader|lunch [15:33] * mzanetti -> away for an hour [15:36] mzanetti, pushed [15:40] Saviq, https://code.launchpad.net/~unity-team/unity8/unity-scope-tool/+merge/199831 [15:42] mhr3_, cheers === dandrader|lunch is now known as dandrader === dednick|lunch is now known as dednick [16:17] can anyone start the music app on manta with sidestage? using the application scope, not a song. [16:18] dednick, me [16:18] Cimi: hm, it's working now after a reboot === greyback is now known as greyback|food [16:22] dednick, how are tests going for you? [16:22] dednick, for me it's like random [16:22] once 5 errors, once 11, once 20 [16:23] I dunno if I do something wrong [16:23] Cimi: it's a struggle [16:24] sigh.... i want to go home now :( [16:25] dandrader: if you're in monday you can have a look at https://code.launchpad.net/~aacid/unity8/horizontalJournal/+merge/199834 [16:25] dednick, you don't work from home? [16:26] Cimi: I am home [16:26] tsdgeos, I won't, but I can get started on it today [16:26] dednick, so why yu want to go home? :) [16:26] Cimi: :) because I'm tired of fighting with tests! [16:27] dednick, it's one of the easiest days... you run your command echoing to a log file,then you use grep :P [16:45] Saviq: do you know where to find the logs of an app trying to be started by unity8? [16:46] dednick, ~/.cache/upstart/application-*.log [16:47] Saviq: just says "Error" [16:47] Saviq: it's using upstart right? [16:47] where are the upstart logs? :) [16:47] dednick, yes, upstart-app-launch to be exact [16:48] dednick, they *are* the upstart logs [16:48] dednick, i.e. if anything's output by the job, it'd be there [16:48] Saviq: yeah, but isn't there a log for the upstart daemon? [16:48] as in the thing calling upstart-app-launch [16:49] dednick, not sure [16:49] dednick, maybe http://upstart.ubuntu.com/cookbook/ has some info === tvoss|lunch is now known as tvoss|dinner [16:55] ok, guys, happy holidays, see you next year! [16:56] mzanetti, https://code.launchpad.net/~mhr3/unity-scopes-shell/scopes-list-delay-envvar/+merge/199840 [16:56] quick workaround for the frickin qmlplugindump issue ^ [16:58] lol, wtf is that?! [16:59] dednick, that's qmlplugindump being broken pile of ... [17:00] ahha, because it creates the objects? [17:00] yea, and doesn't delete them [17:00] yet it does unload the module on exit [17:00] very clever [17:01] scratch that, if it did it on exit, it'd be fine, it unloads it after it's done with the introspection === greyback|food is now known as greyback [17:04] dednick, feel free to approve it since you've looked at it :) [17:04] mhr3_: lol, eww. don't want my name next to that! [17:04] mhr3_: should just remove plugindump until it's fixed. [17:05] we all know noone will fix it for this usecase :P [17:05] you could! [17:05] first i'd have to convince them there's a problem [17:06] don't want a fix next year, want it now! :P [17:06] mhr3_: right, and until that time, remove it! :) [17:06] but then you wouldn't have introspection in qtcreator [17:06] that'd be terrible [17:07] dednick, and the best part - it allows me to run the tests with lower delay, so they run 900ms faster!!! [17:07] mhr3_: is the load always supposed to be async? [17:07] and that's worth it! :P [17:07] yes [17:08] mhr3_: and presumably you mean 90ms faster? [17:08] dednick, no, i mean 900ms [17:09] or 95 rather. you changed from 100 to 5 [17:09] dednick, each test is standalone and recreates the whole thing [17:09] mhr3_: oh, you mean in total [17:09] ok [17:10] marginal! :) [17:10] right?! [17:10] mhr3_: uuu, does adding a static to clas change ABI? [17:11] dednick, doesn't matter, it's a plugin, not a library [17:11] mhr3_: is scopes.h not an interface? [17:11] no, it just defines the Scopes object for qml [17:12] but fwiw i don't think static would change the abi [17:12] mhr3_: ok. good enough for me then. except that it's pretty gross. :) [17:13] dednick, meh, you need to close one eye from time to time :) [17:13] rather ;) [17:13] approved [17:13] thx [17:15] and with that i bid farewell, enjoy holidays everyone! [17:23] dednick: thanks ... was eating [17:24] mhr3_: is this that qmldump issue? https://jenkins.qa.ubuntu.com/job/unity-scopes-shell-trusty-i386-ci/34/console [17:25] mhr3_: doesn't like... seems a real issue somewhere === sil2100 is now known as KURWA === KURWA is now known as sil2100 [18:23] * greyback eow, will be here Monday === jhodapp is now known as jhodapp|afk