/srv/irclogs.ubuntu.com/2013/10/14/#ubuntu-unity.txt

veebersscratch that,  it seems that stopping unity8 while any application is active seems to throw a spanner in the works00:21
=== Guest99228 is now known as amithkk
Saviqveebers, yeah, sometimes crash, sometimes hang?07:05
tvoss_Saviq, good morning07:05
Saviqtvoss_, yo07:05
tvoss_Saviq, hey, as you recently looked into it: do you know how to enable gdb pretty printers?07:06
Saviqtvoss_, nope07:06
veebersSaviq: yeah, I added code tp that MR to handle it (make sure OSK is gone when stopping unity8, and then start it again when needed)07:08
Saviqveebers, k thanks07:08
veebersSaviq: nw, do you know if anyone is working on a fix for: https://bugs.launchpad.net/mir/+bug/1238417 ?07:08
ubot5Ubuntu bug 1238417 in Mir "Unity does not process events from evdev device created before unity is restarted (autopilot tests)" [Critical,Confirmed]07:08
veebersSaviq: also fyi you'll see that MR is 'needs review'07:10
Saviqveebers, I don't think anyone is, since we have a workaround...07:13
tvoss_veebers, the clean fix would be to also fix https://bugs.launchpad.net/mir/+bug/123778407:15
ubot5Ubuntu bug 1237784 in Mir "Please move input detection to libudev" [Low,Triaged]07:15
tvoss_veebers, right now, mir uses the android input stack as is, which uses inotify to get notified about device node creation07:16
tvoss_veebers, while I haven't verified myself, I'm pretty sure that switching to libudev would also fix the autopilot event injection issue07:16
Saviqveebers, are you still "at work" to fix some issues with that MP or shall we take over?07:17
veebersSaviq: heh, no I'm just about to pop out to pick my partner up from work07:17
Saviqveebers, k o/07:18
veebersSaviq: ack re: the bug I mentioned. I will probabl be around a little bit later on07:18
veebersSaviq: also interested, what are the issues that you mentioned?07:18
Saviqveebers, just s/unity/unity8/ in the "you need to stop unity8"07:18
Saviqveebers, but am also wondering if we should stop it ourselves...07:19
veebersSaviq: right, both thomi and myself had that discussion too, I think it's safer in the long run to have the environment owner (the user sort that out)07:21
Saviqveebers, may very well be07:21
davidcalleGood morning all07:25
Saviqveebers, while still here, any reason to not use _patch_envirionment for QT_LOAD_TESTABILITY?07:26
Saviqdavidcalle o/07:26
veebersSaviq: oh um, that's still there from the transition from having a "are we using upstart to start unity", also it's guaranteed to be cleaned up in _cleanup_launching_upstart_unity07:30
Saviqveebers, right, will fix07:30
Saviqveebers, if someone set it before - we shouldn't be resetting it07:30
Saviqveebers, fyi, we need -g in cleanup, too07:30
didrocksSaviq: hey! I guess we have still a yummi unity8 crash with latest image, mind having a look?07:34
didrockshttp://jenkins.qa.ubuntu.com/job/saucy-touch_mir-mako-smoke-unity8-autopilot/21/artifact/clientlogs/_usr_bin_unity8.32011.crash/*view*/07:34
Saviqdidrocks, smokes truncate .crash files07:34
Saviqdidrocks, I won't get anything out of it07:35
didrocksSaviq: let's hope not, we need to get stable at some point :)07:35
Saviqdidrocks, they do07:35
Saviqdidrocks, a unity8 crash needs to be some 12MB at least07:35
didrocksargh07:35
Saviqdidrocks, either way07:35
Saviqdidrocks, we have fixes in queue for a few crashers07:35
didrocksany idea why they are truncated on this arch?07:35
didrocksok, any ETA?07:35
Saviqdidrocks, today07:36
didrocks(to know if we can handle + get the results today)07:36
Saviqdidrocks, utah doesn't wait for them to be completed, AFAICT07:36
didrocksSaviq: if you can get them merged before 10 UTC (so in 2h30), we can hope getting the results07:36
Saviqgreyback, ↑ the "waitforme" branch07:36
didrocksSaviq: ok, as soon as I have a crash on my phone, I'll fw it to you07:36
Saviqdidrocks, please apport-cli View, Keep it first07:36
didrocksSaviq: yep, I know about it for years ;)07:37
Saviqdidrocks, sorry :)07:37
didrocksno worry :p07:37
Saviqdidrocks, btw, smoke will also cough up only one .crash file per project07:37
didrocksyeah, that's how apport works…07:37
Saviqdidrocks, we should collect them and move away from /crash/ after every test07:38
didrocksagreed07:38
didrockswell, first, they have to fix the "wait for apport to finish"07:38
Saviq+107:38
nic-doffaySaviq, I don't suppose you ever got an answer to this question you asked? https://bugs.launchpad.net/unity8/+bug/115015108:38
ubot5Ubuntu bug 1150151 in Unity 8 "[SIDESTAGE] GU threshold for enabling sidestage should exclude 7" tablets" [Medium,Triaged]08:38
Saviqnic-doffay, right indeed08:38
Saviqnic-doffay, can you poke Oren about it/08:38
om26erbug 123957508:42
ubot5bug 1239575 in unity8 (Ubuntu) "left alignment of Music/Video tile' text looks bad" [Low,New] https://launchpad.net/bugs/123957508:42
nic-doffaySaviq, yep np.08:42
om26erCimi, ^ do you know if that was intentional ?08:42
Cimiom26er, intentional08:59
om26erCimi, looks bad doesn't it ?09:01
Cimiom26er, nope09:01
Cimiom26er, https://lh6.googleusercontent.com/DiFH5zJxUPhXES9dICX4ydlmQzpkKFMNeJlq3NsPhPIUAzbAH8XjX4YlZ0eii1w5csbAX2ZqnW9l6bjLSDzzd8Y1oCJf2qKzJYZ49lDeg5dPDbCiCEI-3eyi8Fs09:01
Cimiom26er, put covers09:01
Saviqmzanetti, can I ask you to https://code.launchpad.net/~unity-team/unity8/ap_launch_unity_with_upstart/+merge/19088609:02
mzanettiSaviq: ofc09:02
om26erCimi, doesn't look good to. In the current implementation there is still one bug. the text is overflowing a bit left09:02
om26er*to me09:02
Cimiom26er, it's prob because you're missing the cover09:03
Cimiit's grid aligned...09:03
om26ercover is supposed to appear inside the tile. so that should have no effect on the text alignment.09:05
SaviqCimi, om26er, maybe we just need a margin on the left09:07
Cimiom26er, I think there's another bug in that screenshot09:07
Saviqlike 0.5 GU09:07
CimiSaviq, 1du maybe09:08
Cimior 209:08
Cimi0.5 is way too much09:08
Cimiom26er, Linkin Park09:08
om26erCimi, lol. But I love them09:08
mzanettiSaviq: is data/unity8.conf supposed to be our upstart service file?09:08
SaviqCimi, remember that UShape "softens" the edges09:08
Saviqmzanetti, yes09:08
om26erthe later albums were not that good.. but still09:09
Saviqmzanetti, I need to fix, though09:09
mhr3Saviq, tsdgeos, we want to disable the category reordering, it seems to be causing more problems than it solves, comments?09:11
Saviqmhr3, yeah I think09:12
mhr3pstolowski, will you mp a revert branch?09:13
mhr3doesn't have to be revert...09:13
pstolowskimhr3, yes, i'm on it. it don't want to revert, just disable signal etc.09:16
pstolowskis/it/I/09:16
mhr3yea, fine with me09:16
pstolowskimhr3, Saviq, tsdgeos https://bugs.launchpad.net/unity8/+bug/123958409:18
ubot5Ubuntu bug 1239584 in Unity 8 "Category reordering in Home should be disabled as it causes problems in UI" [Critical,Confirmed]09:18
tsdgeoswell09:19
tsdgeoswe ought to fix those empty categories :D09:19
Saviqpstolowski, osk misbehaving I believe is unrelated09:21
Saviqpstolowski, mhr3, tsdgeos so if we can fix the empty cats, that'd obviously be better09:21
mhr3Saviq, it surely would, but do you have someone who could look at it?09:22
Saviqmhr3, yeah, tsdgeos :)09:22
Saviqmhr3, should've asked first, thought it was past hope ;)09:23
pstolowski:)09:23
tsdgeosSaviq: well, i haven't had a look at it09:23
tsdgeosnot sure how do i repro tbh09:23
tsdgeospstolowski: just "search"?09:23
pstolowskimhr3, besides empty cats, there were some other glitches as well right?09:23
mhr3yea, the "shifting" of the override model09:24
mhr3having apps category twice there09:24
mhr3lots and lots of issue really09:24
mhr3tsdgeos, while i do believe you're awesome, i'm not sure if you'll be able to tackle all of that in the next couple of hours where we still can do fixes09:25
mhr3Saviq, that's why i suggested the revert instead ^09:26
tsdgeosmhr3: sure i'm not saying we should not rever it09:26
tsdgeosi'm just saying hiding them under the rug is not future proof09:26
tsdgeossince i don't even have a rug around here09:26
tsdgeos:D09:26
Saviqlol09:26
mhr3sure, i just want 13.10 to be usable :)09:26
mhr3and you know... without being able to see bugs after two searches09:27
mhr3should take at least dozen :P09:27
SaviqCimi, you're working on the music carousel overlay?09:29
SaviqCimi, and video carousel mount, for that matter?09:29
Saviq-mount, layout, too09:29
CimiSaviq, today yes09:30
SaviqCimi, ok09:30
CimiSaviq, as well with quitting termination mode09:30
SaviqCimi, start with the carousels please09:30
Cimiok09:30
CimiSaviq, is this really for me? https://bugs.launchpad.net/unity8/+bug/122622109:31
ubot5Ubuntu bug 1226221 in Unity 8 ""Recent searches" are not persistent and are not common between scopes" [Undecided,Triaged]09:31
Cimiiirc might be easy to store a property across scopes09:32
Cimibut how about the zeitgeist part?09:32
tsdgeosdidn't we already have a property across scopes for that?09:35
MacSlowSaviq, are ap tests meant to work as usual if executed on the device?09:37
SaviqMacSlow, kind of, yeah09:37
SaviqMacSlow, you need to make sure display is on09:37
SaviqMacSlow, easiest is to see that it's on, sudo stop powerd; stop unity809:37
SaviqMacSlow, from then on they should run ~fine, yes09:38
SaviqMacSlow, but we are fixing things around that atm09:38
MacSlowSaviq, http://pastebin.ubuntu.com/623520709:47
mzanettiSaviq: do you know what changed in the scope backend? Apparently I'm now getting a model that doesn't support get()09:48
mzanetti which breaks the previews09:48
SaviqMacSlow, maguro?09:49
SaviqMacSlow, try stopping maliit-server09:49
MacSlowSaviq, yes the GX09:49
MacSlowGN09:49
SaviqMacSlow, and make sure no app is running09:49
SaviqMacSlow, if all else fails - reboot09:49
Saviqreboot09:50
pstolowskitsdgeos, Saviq so ok to disable reordering?09:54
tsdgeospstolowski: sure09:54
Saviqpstolowski, tsdgeos yeah09:54
tsdgeosSaviq: i'm still having a look at the crashers on sigterm, my patch fixes the crash once everything has properly stated, but if we get a sigterm "while starting", we still crash a bit09:55
Saviqtsdgeos, yeah btw you know you opened a pandora's box? ;)09:55
tsdgeoswhy?09:56
Saviqtsdgeos, now that we're destroying everything correctly09:56
Saviqtsdgeos, we uncovered a bunch of bugs :)09:56
Saviqtsdgeos, 3 different crashes in notidy09:56
tsdgeoslike?09:56
tsdgeos:D09:57
tsdgeosi'm sorry :D09:57
Saviqtsdgeos, on exit, though - so it's not really bad09:58
Saviqtsdgeos, the ones at the bottom in https://bugs.launchpad.net/unity-notifications/09:58
Saviqtsdgeos, only other one I found was https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/123940909:59
ubot5Ubuntu bug 1239409 in unity8 (Ubuntu) "unity8 crashed with SIGSEGV in __libc_do_syscall()" [Undecided,New]09:59
tsdgeosSaviq: so that's unity8 crashing or notify-osd crashing?09:59
Saviqtsdgeos, there is no notify-ods09:59
tsdgeosah09:59
Saviqtsdgeos, it's unity8 in unity-notifications09:59
tsdgeosi see10:00
MacSlowSaviq, I can't stop/kill unity8 on the device...10:01
SaviqMacSlow, kill -9 in case10:01
MacSlowSaviq, ok... brute force seems to have worked... now let's hope the ap-test work10:04
om26erCimi, https://code.launchpad.net/~om26er/unity8/add_leftMargin_tile_text/+merge/19090510:04
tsdgeosMacSlow: you need to stop unity8 from the phablet user, not from root10:05
Cimiom26er, 0.2 gu is not processed iirc10:06
om26erCimi, its working10:06
tsdgeosproblem with 0.210:06
tsdgeosis that it may be a non integer in some places10:06
tsdgeoswhich will make people unhappy10:07
tsdgeos0.2 gu i mean :D10:07
MacSlowtsdgeos, ok... but I wanted to be sure I strike with all possible might ;)10:08
Cimiom26er, let's use 1du or 210:09
Cimiom26er, and remove your code from the running tile :)10:10
om26erCimi, yes. working on that10:10
Saviqmzanetti, pushed a fix to the upstart branch...10:13
Saviqmzanetti, will build packages for testing here10:13
nic-doffaySaviq, got any new bugs I can look at while some branches of mine are reviewed?10:13
mzanettiSaviq: ok... had to reflash my device anyways.10:13
Saviqnic-doffay, how's multi-selectors?10:14
Saviqnic-doffay, waiting for review?10:14
nic-doffaySaviq, yeah10:17
om26erCimi, what provides du ? api reference only have gu and dp.10:18
om26eris that from qml?10:18
Saviqnic-doffay, take https://bugs.launchpad.net/unity8/+bug/1127627 from dednick10:19
ubot5Ubuntu bug 1127627 in Unity 8 "[Shell] Search icon in menu bar should be removed when focus is not dash" [High,Triaged]10:19
Cimiom26er, du10:19
Cimiom26er, was my typo :)10:19
Cimisorry10:19
Cimidp10:19
Cimihah10:19
Cimimonday morning10:19
om26er\o/10:20
Cimiom26er, 1 du is 1 pixel10:20
Cimidp10:20
Cimidammit10:20
CimiDP10:20
Cimiso 1 or 2 max10:21
Cimimaybe 1 id fine10:21
Cimi*is10:21
om26erCimi, yeah, 2 looks fine here.10:21
Cimiwill look bad on desktop10:21
Saviqnic-doffay, then, bug #122622110:21
ubot5bug 1226221 in Unity 8 ""Recent searches" are not persistent and are not common between scopes" [Undecided,Triaged] https://launchpad.net/bugs/122622110:21
Cimion https://code.launchpad.net/~om26er/unity8/fix_1238837/+merge/19074310:24
Cimithere's a mouse area inside an item10:24
Ciminevermind, API helped10:25
om26erCimi, so this branch is good to go now ? :)10:26
Cimiom26er, think so10:26
Cimiom26er, I wrote the same code friday :)10:26
Cimiom26er, didn't realise you were doing it too10:26
om26erCimi, oops, I came here to ping you to make sure you were not already working but couldn't catch you. I guess you were offline already10:27
Cimiom26er, actually not10:27
Cimiom26er, (I already fixed it then I went EOD)10:27
Cimiom26er, btw10:27
Cimiom26er, you need to write a new test or extend the current one10:28
Cimiom26er, because you're no longer testing app termination now10:28
om26erCimi, if your branch have a test, feel free to propose yours10:28
Cimiom26er, you should now check that clicking the icon closes the app10:28
om26erCimi, right, I made a comment about that https://code.launchpad.net/~om26er/unity8/fix_1238837/+merge/190743/comments/43814210:28
Cimiom26er, so yes, exactly10:29
tsdgeoswhat?=¿10:29
tsdgeos../../../../../src/modules/Unity/Application/ubuntukeyboardinfo.cpp: In member function ‘void UbuntuKeyboardInfo::tryConnectingToServer()’:10:29
tsdgeos../../../../../src/modules/Unity/Application/ubuntukeyboardinfo.cpp:61:32: error: ‘class QString’ has no member named ‘isEmtpy’10:29
tsdgeos:D10:29
tsdgeosah10:29
Cimiom26er, in RunningApplicationTile add 'objectName: "closeIcon"'10:30
tsdgeosa q_assert10:30
Cimiom26er, then you find it in the test10:30
om26erCimi, how do I run these tests? to me it says Unity.Test is not installed10:30
Cimiom26er, ./build10:30
Cimi….10:30
Cimiom26er, cd builddir10:31
Cimiom26er, make testRunningApplicationsGrid10:31
nic-doffayCimi, are you planning on dealing with this soon or can I take it? https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/122622110:31
ubot5Ubuntu bug 1226221 in Unity 8 ""Recent searches" are not persistent and are not common between scopes" [Undecided,Triaged]10:31
Ciminic-doffay, if you are doing nothing, take it10:31
Ciminic-doffay, should work with a shared property across scopes10:32
Ciminic-doffay, or if you can use zeitgeist or something else10:33
Ciminic-doffay, you can push and pull recent searches10:33
Ciminic-doffay, maybe mhr3 can help you10:33
Cimiom26er, you want to do it or I can do it if you want10:34
om26erCimi, I am working on it (trying)10:35
nic-doffaymhr3, mind giving me a shout about this?10:35
mhr3nic-doffay, you'd need to store it somewhere, but we don't have zg on the device, and it won't be for 13.1010:36
Cimidbus? :D10:39
Cimigsettings10:39
CimiI meant10:39
nic-doffaySaviq, you mentioned this could possibly be a regression ^10:41
Saviqnic-doffay, the fact that they're not common between scopes is a regression that should be relatively easy tacklable10:42
Saviqnic-doffay, the fact that it's not persistent - not so much - we should probably store it in U1Db for syncing10:42
nic-doffaySaviq, can this be done now?10:44
Saviqnic-doffay, "this"?10:45
Saviqnic-doffay, U1Db? http://developer.ubuntu.com/api/devel/ubuntu-13.10/qml/u1db-qt5/overview.html10:45
nic-doffaySaviq, just wondering about your approach in comparison to mhr3 's10:47
mhr3nic-doffay, "my" approach is just saying that it needs to be stored somewhere, i'm fine with somewhere being u1db, plus if it's supposed to be shared between all scopes, it can be purely qml patch afaict10:51
nic-doffaymhr3, right then.10:52
nic-doffaymhr3, where are the recent searches currently?10:53
Saviqnic-doffay, read, please, read10:53
mhr3nic-doffay, in qml :)10:53
nic-doffaymhr3, Saviq I meant what source files.10:57
Saviqnic-doffay, in qml source files10:58
Saviqnic-doffay, you've been around it for the past month10:58
Saviqnic-doffay, PageHeader10:59
=== MacSlow is now known as MacSlow|lunch
Cimiom26er, update?11:04
om26erCimi, Just pushed a test11:04
om26erCimi, https://code.launchpad.net/~om26er/unity8/fix_1238837/+merge/19074311:04
Cimiom26er, remove empty line11:06
Cimibefore bracket11:06
Cimiseems fine here11:06
Cimibravo11:06
Cimiom26er, rename the other test11:06
Cimitest_clickTile maybe11:07
CimitoTerminateApp doesn't make sense anymore since we drop the functionality11:08
om26erCimi, test_clickTileNotClose sounds fine ?11:08
om26ersince we are making sure clicking it does not close. the name of the test should also indicate that11:09
Cimiom26er, sounds better11:09
Cimiom26er, yeah yours is better11:09
mhr3Saviq, is it possible to create a map for ListElement?11:09
Saviqmhr3, .get() should work11:10
mhr3not what i meant, a sec11:10
mhr3Saviq, something like http://paste.ubuntu.com/6235442/11:11
mhr3but something that actually works :)11:11
Saviqmhr3, ah no11:11
mhr3:/11:12
Saviqmhr3, a list, yes - an object, no11:12
CimiSaviq, u1db?11:18
CimiSaviq, that way will sync across devices11:19
SaviqCimi, yes, that's the plan11:19
CimiSaviq, I would not do this11:19
CimiSaviq, searches depends on local content11:19
SaviqCimi, that's why I put ubuntu-ux on the bug11:19
CimiSaviq, it's not like searching google where it makes sense to share your search history11:20
SaviqCimi, anyway, storing in U1Db does not automagically equal syncing11:20
CimiSaviq, I know but it's not required11:20
SaviqCimi, and dconf isn't the right place to store it11:20
CimiSaviq, I'd go for something local11:20
SaviqCimi, U1Db *is* local11:20
Saviquntil you sync it11:20
=== hikiko is now known as hikiko|lunch
pstolowskimhr3, Saviq https://code.launchpad.net/~unity-team/unity8/disable-reordering/+merge/19091811:35
Saviqpstolowski, FIXME++ ?11:36
pstolowskik11:36
mhr3eh, acked already11:37
mhr3switched back11:37
Saviqmhr3, good we don't have autolanding ;)11:38
mhr3oh?11:38
mhr3good to know11:38
mzanettiSaviq: https://code.launchpad.net/~mzanetti/unity8/fix-lazyimage-initialSize/+merge/19093011:49
Saviqmzanetti, thanks11:50
pete-woodstsdgeos: hi, we have a report of the proposed HUD update breaking the unity8 ap tests, I want to check I'm running the tests correctly myself11:51
tsdgeospete-woods: hi11:51
tsdgeossure11:51
tsdgeoswhatcha doing11:51
tsdgeos?11:51
pete-woodstsdgeos: on the device I'm stopping unity811:52
pete-woodsthen running autopilot run unity8.shell.tests.test_hud.TestHud.test_show_hud_appears11:52
tsdgeosyou need to be insde tests/autopilot11:53
tsdgeosor tests/whatever11:53
tsdgeosfor that to work11:53
pete-woodstsdgeos: that's from the installed unity8-autopilot package11:53
tsdgeosah11:53
tsdgeosok11:53
tsdgeosthen no, you don't need to be anywhere11:53
=== tvoss_ is now known as tvoss|dentist
tsdgeosok, and it's not working?11:53
pete-woodsthe HUD gui isn't appearing11:53
pete-woodsit gets to the point of showing the funny symbol in the middle of the screen11:54
pete-woodsbut then waits there until timeout11:54
tsdgeosthat is weird11:54
tsdgeosi tried them here last week and it worked fine11:54
tsdgeospete-woods: what's the autopilot backtrace?11:55
Saviq/away11:55
pete-woodstsdgeos: do you mean this part? http://pastebin.ubuntu.com/6235587/11:55
tsdgeosyeah that11:56
tsdgeoshave to run for lunch sorry, will try to see what's wrong later11:56
tsdgeosis that ok?11:56
om26erjamesh, hello11:56
pete-woodstsdgeos: this bug is classified as super-urget, but yeah, you may as well get some lunch11:57
pete-woodstsdgeos: some further info, this test fails for me regardless of HUD version, so I don't think that it's the backend at fault11:57
CimiSaviq, I go and merge omen's branch? reviewed11:58
Cimi*omer11:59
* Cimi pushed12:02
=== MacSlow|lunch is now known as MacSlow
=== _salem is now known as salem_
=== dednick is now known as dednick|lunch
=== hikiko|lunch is now known as hikiko
=== jhodapp|afk is now known as jhodapp
=== alan_g is now known as alan_g|lunch
tsdgeospete-woods: back12:53
tsdgeospete-woods: ok, let me try to see if i can repro12:53
pete-woodstsdgeos: I'm suspicious that this test might just not work under Mir?12:54
tsdgeospete-woods: does it pass on the pc but not on the device?12:54
tsdgeospete-woods: it works fine in mir, i tried12:54
tsdgeosa few days ago i mean12:54
tsdgeosnot now12:54
pete-woodsokay12:54
tsdgeospete-woods: works for me :_S12:57
tsdgeoshttp://paste.ubuntu.com/6235761/12:57
pete-woodstsdgeos: how are you setting up the test? are you stopping unity8 first?12:57
pete-woodswhat am I doing differently12:58
tsdgeossure12:58
tsdgeosstop unity812:58
mzanettiSaviq: do you have some packages for your branch?12:58
tsdgeosand then run then run the autopilot line you said12:58
mzanettiSaviq: I can't install the ones from jenkins for some reason12:58
tsdgeospete-woods: nothing i see different12:58
=== tvoss|dentist is now known as tvoss
mhr3mzanetti, try this lp:~mhr3/unity8/route-overridden-previews13:02
mhr3i'm not completely sure it works :)13:02
mzanettimhr3: building13:03
mzanettimhr3: hmm... just merging it into the switching-previews branch doesn't really fix it13:03
mzanettimhr3: well, there is now an empty image13:04
mzanettimhr3: before there wasn't even that one13:04
mhr3mzanetti, that's better, no? :)13:04
mzanettiits one step closer, yes13:04
mzanettinot good enough tho :P13:04
mhr3well it theoretically should work on the phone13:04
mzanettiah ok13:04
mzanettilemme try13:04
mhr3you don't have click scope nor the apps on the desktop13:04
mzanettimhr3: hmm... how come I have the suggested apps section here?13:05
mhr3it's from software center, not clicks13:05
mzanettiah wait...13:05
pete-woodstsdgeos: have you got time to try running those tests again with the proposed version of HUD?13:05
mzanettimhr3: well, in the apps scope I do see the preview for the camera-app. it doesn't work with the camera-app in home13:06
mhr3:/13:06
tsdgeospete-woods: what do i need to do? install a few hud packages? or?13:06
=== alan_g|lunch is now known as alan_g
pete-woodstsdgeos: it's the newest hud build from the proposed PPA13:08
pete-woodsyou should only need to install the hud package itself, as nothing else has changed13:08
Saviqmzanetti, yeah, it conflicts with the ubuntu-touch-session13:09
mzanettiSaviq: already removed that one... still some strange issue13:10
Saviqmzanetti, like?13:10
=== dednick|lunch is now known as dednick
mzanettiSaviq: some circular dependency I think. the locak packages required version ...pkg0saucyX something while that pulled in something riquiring ...ubuntu013:11
Saviqmzanetti, huh, will know in 513:11
tsdgeospete-woods: do you have a link for the .deb package? i don't want to add the proposed ppa to my phone since i'm in the middle of debugging something and took me all day to get here13:12
pete-woodstsdgeos: I don't actually know where the PPA is hosted, mail of deb file good enough?13:15
tsdgeospete-woods: yep13:18
MacSlowdoes it make sense to use strace on running an autopilot test?13:19
MacSlowor is there a specific python-counterpart for it?13:20
MacSlowhm... guess strace it good enough...13:20
tsdgeospete-woods: standup now, will try it just after13:30
tsdgeossorry for the delay13:30
pete-woodstsdgeos: np, at least I have someone who the tests originally pass for now :)13:30
mzanettiSaviq: dednick: standup13:32
didrocksSaviq: do you know who is working on https://bugs.launchpad.net/ubuntu/+source/maliit-framework/+bug/1233988? (I reopened downstream as well)13:35
ubot5Ubuntu bug 1233988 in platform-api (Ubuntu) "With Mir enabled: platform-api apps crash with SIGABRT in __gnu_cxx::__verbose_terminate_handler(), thrown from mir::client::DisplayConfiguration::copy_to_client()" [High,Confirmed]13:35
Saviqdidrocks, alf13:35
didrocksSaviq: ok, thanks13:35
dandraderCimi, what's your launchpad id?13:39
Cimidandrader, what you think?13:39
dandradertried cimi, andra, cimitan13:39
dandraderCimi, coulnd't find it. just tell me13:39
Cimidandrader, cimi13:40
dandraderCimi, weird: No items matched "cimi".13:40
dandraderCimi, well, could you please reassign https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1228756 to yourself then :)13:40
ubot5Ubuntu bug 1228756 in unity8 (Ubuntu) "After closing a recent app via the dash, unable to launch recent apps from dash" [Medium,New]13:40
Cimidandrader, https://launchpad.net/~cimi :-\13:41
nic-doffaySaviq, any idea which revision the common scope regression is a part of? I've been trying to find it...13:53
nic-doffayHaven't found anything of relevance yet.13:53
Saviqnic-doffay, no, no it happened quite some time ago13:56
Saviqdednick, can you take a look at bug #1239708?13:56
ubot5bug 1239708 in unity8 (Ubuntu) "Welcome screen clock missing" [Undecided,New] https://launchpad.net/bugs/123970813:56
dednickSaviq: yeah13:57
tsdgeospete-woods: still works here13:58
pete-woodstsdgeos: okay, so it sounds like that was a red herring then - thanks for checking it for me!13:58
tsdgeospete-woods: no worries13:58
tsdgeospete-woods: where's the bug btw, why assigned to you?13:59
pete-woodstsdgeos: https://bugs.launchpad.net/ubuntu/+source/hud/+bug/123901614:00
ubot5Ubuntu bug 1239016 in hud (Ubuntu) "latest hud totally busted" [Critical,Incomplete]14:00
=== dandrader is now known as dandrader|afk
pete-woodsbecause pete-woods = general dogsbody who fixes random bugs on project's he's not really on14:00
tsdgeos:D14:00
tsdgeosmake sense14:00
nic-doffaySaviq, check out my comment here too: https://bugs.launchpad.net/unity8/+bug/116925814:02
ubot5Ubuntu bug 1169258 in Ubuntu UX "There is no ListItem implementing the expansion behaviour" [High,In progress]14:02
Saviqnic-doffay, the ExpansionContainer would just have to *be* a Flickable or a ListView, is all14:03
dednicklarsu: hi14:04
nic-doffaySaviq, about that common search, how could I make SearchHistoryModel in PageHeader common to all scopes? My assumption would be that it was already?14:07
Saviqnic-doffay, it would work then, wouldn't it14:11
Saviqnic-doffay, there's a PageHeader per-scope14:11
Saviqnic-doffay, so you need a SearchHistoryModel in Dash.qml or Shell.qml, and pass it to PageHeaders14:12
mhr3Saviq, does run_on_device stop the upstart job or am i supposed to do that myself?14:13
Saviqmhr3, it does14:14
mhr3and it works just fine with mir, right?14:14
* mhr3 is bored while it's building14:14
Saviqmhr3, yes14:14
mhr3thx14:15
=== dandrader|afk is now known as dandrader
davmor2guys quick question I see nux on unity8 should it be there?14:27
mhr3mzanetti, finally compiled the branch on the phone and it works there14:28
mhr3minus the webapps, those give error14:28
tsdgeosdavmor2: where?14:33
=== alan_g is now known as alan_g|lunch
davmor2tsdgeos: ii  libnux-4.0-0                                          4.0.3+13.10.20131011-0ubuntu1                 armhf        Visual rendering toolkit for real-time applications - shared lib14:35
davmor2ii  libnux-4.0-common                                     4.0.3+13.10.20131011-0ubuntu1                 armhf        Visual rendering toolkit for real-time applications - common files14:35
tsdgeos ah14:35
tsdgeosyou mean it's installed14:35
tsdgeosi thought you meant in the unity8 cod14:36
tsdgeose14:36
tsdgeosi think libunity uses it for some non visual thing14:36
davmor2tsdgeos: yeah sorry,  I'm assuming it's pulled in by something but didn't think that unity8 used it so wondered why it was being pulled in14:36
davmor2tsdgeos: ah thanks, so it should still be on the image then?  That was the important bit to confirm14:38
tsdgeosi'd say so yes14:38
tsdgeosnot terribly sure14:38
tsdgeosmhr3: pstolowski: maybe you guys know more?14:38
mhr3yes, it's needed for the non-ui things14:39
mhr3since unity-core links to it14:40
dakirahi. is this the right channel to query about scopes?14:40
davidcalledakira, yes14:41
dakiradavidcalle: hey! I was the one who reported the torrent lens not working. What are the necessary changes to get it working again. I've got some time to spend. What changed in 13.10 that breaks the old lens?14:43
dakiraI mean scope ;)14:43
=== alan_g|lunch is now known as alan_g
dakiradavidcalle_:  got my last message?14:44
davidcalle_dakira, no, screen froze :)14:45
=== davidcalle_ is now known as davidcalle
dakiradavidcalle: I was the one who reported the torrent lens not working. What are the necessary changes to get it working again. I've got some time to spend. What changed in 13.10 that breaks the old lens?14:45
pstolowskitsdgeos, unitycore uses logger and proprerty classes from nux14:46
pstolowskidavmor2, ^14:46
davidcalledakira, oh right. It needs to be adapted to the new framework. We are still missing a new Python tutorial on that, but if you want to give it a shot, you can use https://launchpad.net/unity-scope-soundcloud as an example.14:47
davmor2pstolowski: Great thanks, I was just double checking that it should be on the image, it seems it should and I'm happy I don't have to write a bung now :)14:47
davmor2bug even14:47
dakiradavidcalle: I'll have a look, thanks. Do I have to change both, the scope and the lens?14:48
pstolowskidavmor2, yw14:48
Saviqtsdgeos, so, you were saying that we're crashing / hanging in join()?14:49
Saviqtsdgeos, i.e. bug #1239409 ?14:49
ubot5bug 1239409 in unity8 (Ubuntu) "unity8 crashed with SIGSEGV in __libc_do_syscall()" [Undecided,New] https://launchpad.net/bugs/123940914:49
tsdgeosSaviq: sometimes yes14:50
tsdgeoswith that backtrace can't really say "sure"14:50
davidcalledakira, it's a bit different now, there are no more lenses, just scopes. If you look in you /usr/share/unity/scopes you will see folders that act as categories for scopes (that's what is replacing lenses). So, you would need to updates the scope and put its .scope file in one of its category or create a new one if none fits.14:50
davidcallein your*14:50
Saviqtsdgeos, I just uploaded a (hopefully fully) symbolic .crash of that14:51
tsdgeoslet me see14:51
dakiradavidcalle: so categories are basically the new lenses. any documentation on how to create one?14:52
Saviqtsdgeos, https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1239409/comments/514:52
ubot5Ubuntu bug 1239409 in unity8 (Ubuntu) "unity8 crashed with SIGSEGV in __libc_do_syscall()" [Undecided,New]14:52
dakiradavidcalle: and looking at that folder I was wondering on where to find the skimlinks scope. Is it hardcoded somewhere else?14:53
davidcalledakira, look at an existing scope file  (eg. graphics.scope) in /usr/share/unity/scopes, that's all you need.14:53
dakiraDaviey: thanks!14:54
davidcalledakira, it's on the scopes server https://productsearch.ubuntu.com/smartscopes/v1/remote-scopes14:54
dakiradavidcalle: oops.. thanks14:54
tsdgeosSaviq: but was it crahing in the thread that did the join() or in another thread?14:56
Saviqtsdgeos, it didn't crash14:57
Saviqtsdgeos, hung14:57
tsdgeosah14:57
tsdgeoswas stuck14:57
tsdgeossure14:57
dandradergreyback, would you have time to review that? https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/123648914:57
ubot5Ubuntu bug 1236489 in unity-mir (Ubuntu) "In MIR landscape mode keyboard only works on left side" [High,In progress]14:57
tsdgeosif you control+c in some specific moment14:57
Saviqtsdgeos, I forced a crash with kill -SIGSEGV14:57
tsdgeosyou can make it hang14:57
tsdgeosworking on that14:58
tsdgeosi know what's wrong14:58
Saviqtsdgeos, ah awesome14:58
greybackdandrader: sure14:58
tsdgeosnot sure how to fix it "properly" yet14:58
dandradergreyback, thanks!14:58
Saviqtsdgeos, happens almost every time on stopping unity8 during autopilot here...14:58
tsdgeosok14:58
tsdgeosthen it's not what i say14:58
tsdgeosSaviq: can you thread aplly all bt?14:59
tsdgeosapply14:59
tsdgeosi'm interested in what's going on with the thead that has startShell in it15:00
Saviqtsdgeos, let me get it to hang15:00
Davieydakira: no problem?15:00
Saviqtsdgeos, http://paste.ubuntu.com/6236270/15:05
Saviqtsdgeos, let me know which symbols you'd like15:05
tsdgeosSaviq: get the glib2 ones and the qtbase ones if possible15:06
tsdgeosSaviq: unity-mir won't hurt either15:06
Saviqtsdgeos, they're there15:06
tsdgeosare they?15:07
tsdgeosweird15:07
tsdgeosthread 1 should be calling join() from unity-mir15:07
tsdgeosand all we see is a lock() :-S15:07
dednick_larsu: can you take a look when you have a minute? https://code.launchpad.net/~nick-dedekind/qmenumodel/lp1239394/+merge/19096515:09
greybackdandrader: one comment https://code.launchpad.net/~dandrader/ubuntu-keyboard/osk_rotation_lp1236489/+merge/19094615:11
Saviqtsdgeos, http://paste.ubuntu.com/6236299/15:13
tsdgeosSaviq: there are the other 7 threads?15:14
Saviqtsdgeos, they must've disappeared since :D15:14
Saviqtsdgeos, let me re-hang it15:14
tsdgeosSaviq: it's weird, thread 1 should always be there15:14
Saviqtsdgeos, crap, need to reboot phone, display b0rked15:15
Saviqtsdgeos, http://paste.ubuntu.com/6236344/15:25
Saviqtsdgeos, had to manually bt threads < 8 - must be gdb crapped out after memory corruption15:25
tsdgeosbrr15:25
mhr3mzanetti, works fine for me - https://code.launchpad.net/~mhr3/unity8/route-overridden-previews/+merge/19098415:26
tsdgeosright15:26
tsdgeosso thread 1 is in the join15:26
tsdgeosand there is no thread with startShell in the stacktrace15:26
tsdgeosthat is weird15:26
tsdgeosif the thread ended why didn't it return after join?15:27
tsdgeosstartShell is the thread that actually does qApp->exec15:27
tsdgeosSaviq: this is in an autopilot run you say, right?15:27
tsdgeoslooks very much like the lock i'm having15:27
Saviqtsdgeos, yeah, first test is enough15:28
tsdgeosbut i can only get there if i kill it very early15:28
tsdgeosone would think autopilot lets it live a bit more15:28
tsdgeoslet me push the "fix" i have15:28
tsdgeosand let's see if that also fixes this case15:28
tsdgeosdon't think so15:28
tsdgeosbut who knows15:28
Saviqtsdgeos, k15:28
tsdgeosSaviq: try lp:~aacid/unity-mir/crashes_stop_starting_up15:31
tsdgeosSaviq: which is the autopilot test you are running exactly, just for me to try the same15:31
Saviqtsdgeos, just the first one... but now I'm running through upstart it doesn't hang anymore... ugh15:32
tsdgeosSaviq: unity8.application_lifecycle.tests.test_application_lifecycle.ApplicationLifecycleTests.test_app_moves_from_unfocused_to_focused ?15:32
Saviqtsdgeos, no, one of the mock ones15:33
tsdgeosi have no mock ones :S15:33
* Saviq stops15:33
Saviqtsdgeos, all the others are mock ones ;)15:33
Saviqtsdgeos, except for the lifecycle ones15:33
tsdgeoswhat do you mean mock?15:34
tsdgeoshow is the hud one mock?15:34
Saviqtsdgeos, ApplicationManager is using a mock15:38
Saviqtsdgeos, it's not really launching apps15:38
Saviqtsdgeos, autopilot run unity8.shell.tests.test_hud.TestHud.test_hide_hud_click15:38
tsdgeosok15:38
Saviqtsdgeos, that gets stuck here with unity-mir + waitforme15:39
tsdgeoshow often?15:39
Saviqtsdgeos, more than 60% I'd say15:46
tsdgeosSaviq: ok yeah it hangs here too, nothing to do with what i was fixing15:46
Saviqgood thing upstart doesn't care15:46
* greyback breathes sigh of relief, his GNexus not bricked15:50
tsdgeosshould i see qdebugs i write in main.cpp in autopilot runs?15:51
tsdgeosor any clue how i see them?15:51
tsdgeosnoone?15:55
dednick_tsdgeos: i think you might need to redirect std out.15:58
dednick_or std err. cant remember15:59
dednick_2>&1 i think15:59
tsdgeosdidn't work :'(16:00
* tsdgeos resorts to more manual debugging than debugs16:01
dednick_hm. i think ap eats everything. cant remember how to tell it not to.16:01
dednick_perhaps a command line option16:01
dednick_tsdgeos: ^16:01
tsdgeosthere's -vv16:02
tsdgeosbut it's just verbose for autopilot itself16:02
tsdgeosnot for the underlying app16:02
dednick_tsdgeos: give me a minute. i'm sure there is something16:02
larsudednick_: I commented on the patch16:02
dednick_larsu: thanks16:02
davmor2charles_: ping16:06
charles_davmor2: pong16:06
=== chrisccoulson_ is now known as chrisccoulson
davmor2charles_: I've noticed something on testing the headset on saucy (desktop) in bluetooth settings bluetooth say it is turned off, but the indicator says it is on and setting up the headset looks to of worked16:08
dednick_tsdgeos: hm. maybe not. but i'm sure that ap tests during ci used to log qt messages16:09
tsdgeosyeah i had that impression too16:09
tsdgeosbut..16:09
davmor2charles_: http://ubuntuone.com/2IE73Sj7W0XF1gecpYywOe and http://ubuntuone.com/4BBBPVXq8Sekr2kVrGi2SB16:11
=== charles_ is now known as charles
tsdgeosSaviq: i think i know what's happening16:14
Saviqtsdgeos, young man, it's a quarter past your EOD, what are you still doing here!? ;)16:14
Saviqtsdgeos, but cool!16:15
tsdgeosSaviq: i slept a bit more this morning ;)16:15
charlesdavmor2: could you repeat the system-settings test while running bt-monitor on the same system in a terminal?16:15
tsdgeosSaviq: from what i can see, the qml view doesn't like being deleted when there's stuff moving16:15
Cimiin the running application grid16:15
tsdgeosthat's the reason every time i debug it16:16
tsdgeosand put breakpoints16:16
CimiI'd like to be able to tap anywhere on the screen to dismiss the termination mode16:16
tsdgeosit always works :D16:16
CimiI'm looking for a solution that doesn't su*k16:16
davmor2charles: sorry for saucy desktop or on the phone?16:16
tsdgeosbut autopilot usually "stops" while e.g. the hud thing is closing16:16
Cimicoding wise...16:16
tsdgeosand doesn't like that16:16
tsdgeosno idea why though16:16
Cimibecause inversemousearea on the tile is limited to the grid, and not the whole listviewwithpageheader16:17
tsdgeosand no idea how to fix it either :D16:17
charlesdavmor2: whichever system you're running system-settings on & the scan isn't finding your headset16:17
charlesdavmor2: bt-monitor is provided by bluez-tools16:17
tsdgeosyep can totally repro without autopilot \o/16:19
Saviqtsdgeos, make sense16:19
tsdgeosSaviq: so what i do to repro16:19
Saviqtsdgeos, we can make autopilot wait-for-it16:19
tsdgeosstart unity816:19
tsdgeos./run16:20
tsdgeosand then open the indicators16:20
tsdgeoswhile it's scrolling down16:20
tsdgeoskillall unity8 in a shell16:20
tsdgeosboom16:20
tsdgeosgets stuck16:20
tsdgeosif i let it finish16:20
tsdgeosworks16:20
davmor2charles: will do16:22
Cimiguys, in the running application grid I'd like to be able to tap anywhere on the screen to dismiss the termination mode16:24
Cimibut inversemousearea is restricted to the list of running tiles due to the hierarchy of the files16:25
CimiI am looking for a solution that doesn't look bad16:25
Cimicoding wise16:25
Cimiyou have ideas?16:25
CimiSaviq, tsdgeos ^ ?16:31
tedgSaviq, Can you send your indicator-network.log file on bug 123899016:32
ubot5bug 1238990 in Network Menu ""Unlock SIM" gets stuck if snap decision failed" [High,Triaged] https://launchpad.net/bugs/123899016:32
tsdgeosCimi: hmmm nothing really smart, all i can think of is using the "global" shell variable and anchor to it16:34
davmor2charles: Hmm odd now it shows up http://paste.ubuntu.com/6236655/16:34
charlesdavmor2: it's like taking your car to the mechanic :)16:34
charlesdavmor2: do you see any pattern between when the headset is recognized & when it isn't?16:35
Cimitsdgeos, which is not nice16:35
tsdgeosagreed16:36
davmor2charles: now I have the system in RW to enable me to install bluez-tools, let me set it back to RO and see if it still works16:37
charlesok16:37
davmor2charles: I have noticed that on desktop and phone that I don't think it is actually using the headset as such but that will be a different issue all together,  Android on the same device used it no issues so I'll dig into that when we have more time16:39
Cimitsdgeos, and also does not work16:40
Cimitsdgeos, Cannot anchor to an item that isn't a parent or sibling.16:41
tsdgeosright16:41
Cimitsdgeos, which means the only solution is to add code in the scope list view16:41
Cimiand enable ad hoc16:41
Cimiquite ugly16:41
Cimiit's ugly because we don't use it anywhere else apart here16:42
tsdgeoswell there's more solutions, btu sure that's one16:42
davmor2charles: hmm okay set back to RO but it is still listing my headset is there a way I can get rid of it's listing?16:42
tsdgeosanother would be create a parentless element that fills all the screen16:43
tsdgeosbut it's not much nicer16:43
Cimiouch16:43
Cimididn't know you could do this16:43
tsdgeosand at the moment i don't remember how parentless elemnts are crated :D16:43
om26ermhr3, hey! is there a separate package for apps scope ?16:43
Cimihah16:43
tsdgeosCimi: it's how you do "dialogs" for example16:43
Cimitsdgeos, well, maybe can work here16:43
charlesdavmor2: no. iirc there's a ticket open for adding unpairing, but the design for it didn't get completed for 13.10 :/16:44
tsdgeosthink about it16:44
tsdgeosneed to go and do fun stuff i mean, taxes16:44
tsdgeostalk to you tomorrow :D16:44
davmor2charles: no worries I'll reflash it add the bluez-tools and then make it RO and try again16:44
charlesdavmor2: you can do it from the command line with "bluez-test-device remove $address"16:45
charlesdavmor2: ok16:45
mhr3om26er, unity-scope-click16:45
Cimimm that is weird16:49
CimiSaviq, inversemousearea should automatically be full screen?16:49
CimiSaviq, or within parent?16:49
SaviqCimi, by default - root element16:50
SaviqCimi, within parent wouldn't make much sense, would it16:50
Saviqtedg, sure, can do16:51
CimiSaviq, mmm so it doesn't really work for me16:51
Saviqtedg, I had to break the notification to get there though ;)16:51
CimiSaviq, looks like it's limited to the component I am adding it16:51
Cimior other mouse area collides16:51
tedgSaviq, I'm confused, is it when it doesn't return anything at all, or when it closes?16:51
Saviqtedg, my use case was that I had the new network indicator, but old unity816:52
Saviqtedg, so didn't get the pinpad16:52
Saviqtedg, restarted unity816:52
Saviqtedg, so the notification was gone, but indicator-network was still greyed out16:53
Saviqtedg, really minor use case ;)16:53
Saviqtedg, basically, we'd have to make it crash now16:53
Saviqtedg, since we'll be deleting / closing all the notifications on unity8 shutdown soon-ish16:54
Saviq/away16:54
tedgSaviq, Okay, so I'm going to move that one down to medium.16:56
tedg:-)16:56
Saviqtedg, at most ;)16:56
tvossSaviq, test suite runs for me with the osk executables I gave you17:02
tvossSaviq, can you cross-verify?17:02
=== alan_g is now known as alan_g|EOD
davmor2charles: right in RO mode now http://paste.ubuntu.com/6236816/ and it's failing to connect again17:16
davmor2charles: daft question is there a RW rule for where bt stores it's config for devices?17:18
charlesdavmor2: that's a good question. I'll look at the bluez code17:19
davmor2charles: I'll reboot into RW mode and see if it connects again17:21
* greyback going home17:23
davmor2charles: so in RW mode it finds it straight away17:24
hashkenIn Unity, how can I mark a window so that it does not get minimized when I do "Show Desktop".17:29
hashkenI am using XPad for Desktop notes and when I do "Show desktop", even the XPad windows are getting minimized. I want to prevent this.17:30
Saviqtvoss, even without maliit, that is?17:32
om26erhttps://code.launchpad.net/~om26er/unity8/search_indicator_dash_only/+merge/19101217:33
om26erany takers ?17:34
Saviqtvoss, thing is... if I start unity8, stop maliit-server, and type through autopilot.input.Keyboard(), it works fine17:35
Saviqtvoss, and since we're starting maliit for unity8 now, I can't reproduce the issue at all - maybe it's when it never launches?17:36
=== dandrader is now known as dandrader|afk
om26erCimi, can you review that? ^ its a one liner17:39
Cimiom26er, yep17:39
om26erthanks17:41
tvossSaviq, might well be17:43
tvossSaviq, so issue solved?17:43
Saviqtvoss, well, avoided I'd say ;)17:50
=== dandrader|afk is now known as dandrade
=== dandrade is now known as dandrader
tvossSaviq, true ;)18:00
=== thomi_ is now known as thomi
thomimorning19:23
Saviqhey thomi20:09
Saviqveebers, ping20:09
thomihi Saviq20:15
veebersSaviq: pong20:36
Saviqveebers, hey, so I'd say lp:~unity-team/unity8/ap_launch_unity_with_upstart is ready if you could give it your last spin20:37
Saviqveebers, one important thing there is that I stopped using initctl for LD and QML2 paths20:37
Saviqveebers, so as to not leave them around in case of interruptions20:38
Saviqveebers, after all, we were setting them globally for the session, when you can just pass them to initctl start for this particular run20:38
Saviqveebers, I left QT_LOAD_TESTABILITY so that launched apps run under that20:38
veebersSaviq: ah nice thought, although if they are re-set using an addCleanup they should always be reset20:39
veebersSaviq: ah ok, makes sense then20:39
Saviqveebers, well, you can always kill autopilot enough to not call the cleanups ;)20:39
veebersSaviq: fair enough20:39
Saviqthis is just safer and doesn't affect the whole system unnecessarily20:39
* Saviq looks through the chlog20:39
Saviqveebers, dropped the maliit handling completely - it wasn't doing us any good (not any more than upstart is already)20:40
veebersSaviq: so is there anything left to do for that branch?20:40
Saviqveebers, AFAICT just testing - worked for me on both device and desktop20:40
veebersSaviq: hmm ok, the reason I introduced it was I guess mainly that trying to stop unity8 while the kb was up caused issues20:41
veebersSaviq: cool, I'll flash my device to latest and give it a spin20:41
Saviqveebers, that's handled by upstart - maliit is stopped before unity8 is20:41
Saviqveebers, i.e. maliit has "stop on stopping unity8"20:41
Saviqveebers, you will see maliit crashes due to maliit starting too early to connect to Mir20:42
Saviqveebers, but that's a different topic - and it's not critical20:42
Saviqveebers, I forgot to verify on sflinger if you could20:42
Saviqand I will do maguro as well20:42
Saviqveebers, jenkins is at least happy on mediumtests-saucy, not so much on mediumtests-touch yet20:43
Saviqbut there's a slew of fixes in queue that will hopefully help that20:44
Saviqfginther, is there anything we're doing on mediumtests-touch to keep the display alive during mediumtests?20:44
fgintherSaviq, before the test is started, we do "nohup powerd-cli active&"20:45
fgintherSaviq, I'm not sure how reliable it is20:45
fgintherSaviq, there's no way to query powerd of the actual state20:45
Saviqfginther, yeah so it's not reliable enough it seems :/20:47
Saviqfginther, https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/2477/testReport/junit/unity8.shell.tests.test_notifications/EphemeralNotificationsTests/test_urgency_order_Native_Device_/20:47
Saviqveebers, I'm still seeing the keyboard tests fail on jenkins, but locally it seems to work fine - will do a bit more testing locally20:47
Saviqfginther, I wonder if it's too late20:47
veebersSaviq: ack, thanks20:47
Saviqfginther, well, it did light the display up here for me on mako...20:48
fgintherSaviq, the timing is fragile, if it's not down while the display is on, it won't work20:49
Saviqfginther, although the help says "This does not modify the screen state." there20:49
CimiSaviq, tried an inversemousearea filling the application thumbnail tile20:49
CimiSaviq, onPressed (and clicked) are detected only in the frame within the application grid (category header excluded), idea? bug?20:49
SaviqCimi, isolate please20:50
SaviqCimi, try with a simple MainView application20:50
Saviqfginther, maybe "powerd-cli display on bright" would be better20:51
Saviqfginther, yeah that seems reliable20:52
fgintherSaviq, I can try, I've been experimenting for the last few days to improve the test runner20:52
Saviqfginther, from a blanked display - after running that I can start unity8 no problem20:52
fgintherSaviq, hmm20:53
fgintherSaviq, before running the unity8 autopilot tests, does unity8 need to be stopped20:56
fginther?20:56
Saviqfginther, yes20:56
Saviqfginther, while rebuilding them to use upstart we thought of stopping it, but we decided to bail out instead20:56
Saviqfginther, saying that you should stop it yourself20:56
Saviqfginther, felt safer20:56
fgintherSaviq, "stop unity8" is all that is needed, correct?20:57
Saviqfginther, "initctl stop unity8" is preferred for explicity reasons, but yes20:57
fgintherSaviq, what about unlocking the screen, does that even apply?20:57
Saviqfginther, for unity8 it does not20:57
Saviqfginther, I mean for the unity8 suite20:58
fgintherSaviq, thanks, (and I was referring to unity8-autopilot as well)20:58
Saviqfginther, and even for apps you shouldn't need that temporarily20:58
Saviqas we'll unlock when an app is launched behind the greeter20:58
fgintherSaviq, yeah!20:59
Saviqbut we'll be locking this down at some point soon to plug that security hole ;)20:59
fgintheroh well20:59
Saviqfginther, but at that point we should have all apps transitioned to unity820:59
Saviqs/unity8/upstart/20:59
Saviqand then in the suite you'd restart unity8 under testability20:59
Saviqand use unity8's emulator to unlock21:00
Saviqso no more blindness21:00
fgintherSaviq, I like this plan21:00
Saviqfginther, that was the plan all along ;)21:00
Saviqfginther, just not executed yet :)21:00
Saviqwe should get rid of two of {otto,utah,phablet-test-run}...21:06
=== jhodapp is now known as jhodapp|afk
Saviqmhr3, still here? got a fix for blurred switching previews if you want to test it out21:58
Saviqmzanetti, breaker :P22:01
mzanettiSaviq: huh?22:02
Saviqmzanetti, for sourceSize 0 is a special value, different than undefind22:02
Saviqundefined22:02
mzanettiawww man22:02
Saviqalthough I agree that's broken22:02
Saviqmzanetti, pushed a fix to mhr3's issue with blurred images in previews22:03
mzanettiSaviq: cool, thanks22:03
veebersSaviq: finally finished running all those tests on my desktop and phone and they all pass fine for me22:07
veebersSaviq: pitty about the CI failures :-\22:07
Saviqveebers, yeah, we'll get there - awesome22:08
Saviqveebers, approve?22:08
Saviqmzanetti, if you want - you could approve too :)22:09
Saviqhttps://code.launchpad.net/~unity-team/unity8/ap_launch_unity_with_upstart/+merge/190886 I mean22:09
veebersSaviq: I've comment-approved now22:09
veebersSaviq: did I see you mention earlier that you've had issues turning on the screen after starting unity8 again?22:17
Saviqveebers, yeah, the bug's still there22:24
Saviqveebers, on mako at least, screen needs to be on to start unity822:24
Saviqveebers, otherwise you get "could not unblank display" and SIGABRT on startup22:25
veebersSaviq: ah ok, it turned on after a while. I was able to start unity alright (as far as I could tell) it's just that pushing the power button didn't turn on the screen22:25
Saviqveebers, push once, wait22:25
Saviqveebers, and monitor ~/.cache/upstart/unity8.log22:25
veebersbut I'm impatient ;-) I just mash at the key until something happens22:25
Saviqveebers, upstart will respawn it, as it will be SIGABRTing all the time22:25
veebersSaviq: ah right, thanks I'll do that next time22:26
Saviqmergin' that shit!22:26
=== salem_ is now known as _salem
veebersSaviq: \o/22:29
Saviqveebers, yeah, I'm much happier with that setup :)22:29
Saviqveebers, now we need to switch apps over to upstart22:29
Saviqveebers, and we can start unlocking unity8 while knowing what happens - and not blindly instead22:30
veebersSaviq: awsome, thanks for finishing it off22:30
veebersSaviq: that would be _awesome_ :-)22:30
Saviqveebers, we'll get there, mark my words! ;)22:30
veebers^_^22:30
Saviqmzanetti, dammit, confused switching-previews again ;(22:31
mzanettiSaviq: ack. will fix tomorrow22:40
mhr3Saviq, i'll trust you :)22:40
Saviqmhr3, yeah, managed to break it ;P22:40
mhr3lol22:40
Saviqmhr3, well - not my fix, just switching-previews ;)22:42

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!