/srv/irclogs.ubuntu.com/2015/04/15/#ubuntu-unity.txt

=== marcusto_ is now known as marcustomlinson
=== spy is now known as Guest59833
tsdgeosMirv: https://codereview.qt-project.org/#/c/110325/ has been updated, but not with any crucial bits i think, may leave the rebuild for final in case we accept it08:15
* Mirv context switches from qtdeclarative to qtbase08:17
Mirvtsdgeos: ok.08:17
tsdgeosMirv: sorry :D qtdeclarative one for regexp was merged upstream and Saviq and me checked yesterday it seems to fix the issue we were having with it08:18
Mirvtsdgeos: yes, it's in 004 now, I'd be happy to hear results from testing the PPA specifically (the armhf has also finished building with ~test1 version, even though the final build is still ongoing)08:20
tsdgeosok08:24
tsdgeosMirv: seems to be fine for me09:04
Mirvthanks!09:16
Saviqdednick, diff here looks busted https://code.launchpad.net/~nick-dedekind/unity8/desktop-app-focus/+merge/25628711:04
tsdgeosCimi: how's the review of https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1442085 going?11:10
ubot5Ubuntu bug 1442085 in unity8 (Ubuntu) "Overlapping labels in table widgets on previews" [High,In progress]11:10
tsdgeosMacSlow: https://code.launchpad.net/~aacid/unity8/pot_file top approve?11:11
Cimitsdgeos, I was wondering if we needed to fix qt 5.5/5.4.2 too there11:11
tsdgeosCimi: we do11:11
tsdgeosCimi: https://codereview.qt-project.org/#/c/110348/11:11
Cimitsdgeos, so it was a regression in 5.4.2 ?11:12
tsdgeosno clue tbh11:12
tsdgeosmay be a 5.4.2 regression11:12
tsdgeosor that somehow our 5.4 packages have something that make this work11:12
tsdgeosdidn't feel like building 5.4.1 from upstream and checking11:12
SaviqMacSlow, crash of maliit-server and unity8-dash likely when they try and launch to unity8 that crashed on startup, your "process not found" error, too, I've not seen that hapen though11:12
Cimitsdgeos, I don't know if we do 5.4.2 first ot 5.5, but if we do 5.4.2 we should have tests not to fail... a fix in the tests is not possible?11:13
MacSlowtsdgeos, generally yes... but why does autopilot/DBus crap out again on Jenkins?11:13
SaviqMacSlow, you should merge ~unity-team shellRotation branch into yours btw, there were quite a bit of conflicts11:13
tsdgeosMacSlow: good question :/11:13
tsdgeosCimi: if we do 5.4.2 the tests will fail which is what they should be doing, no?11:14
MacSlowSaviq, I thought I did that yesterday... will double-check11:14
SaviqMacSlow, didn't know you did, that's fine then11:14
SaviqMacSlow, I'll tweak demo-stuff recipe configs so that it builds your branch instead11:15
MacSlowtsdgeos, looking at the log it fails with createPlatformOpenGLContext... some issue with Ubuntu-L.ttf ... ?11:15
SaviqMacSlow, tsdgeos, the jenkins failure seems to be a timeout11:15
MacSlowSaviq, I still want to check...11:15
Cimitsdgeos, ah, so the visual itself is broken in 5.4.211:15
MacSlowSaviq, tsdgeos: in that case I'm good to top-approve tsdgeos branch... I can hardly imagine the pot-file fixes to affect that :)11:15
Cimitsdgeos, I thought visuals were fine, tests were failing11:16
tsdgeosCimi: yes11:16
Cimigood then11:16
SaviqMacSlow, indeed11:16
dednickSaviq: yup. i know. fixing11:16
MacSlowtsdgeos, done11:16
MacSlowSaviq, doh... didn't merge with unity-team shellRotation branch... *sigh*11:17
MacSlowSaviq, so forget my alarm-bells for the moment :)11:18
tsdgeosCimi: i.e. if we update to 5.4.2 there can be two things happening, 1 the tests break because the visuals break and then we need to patch 5.4.2 with my upstream patch (if accepted) or we do actually have some other patch in some place that fixes it too and then we're fine :D11:18
SaviqMacSlow, there were alarm bells? ;)11:18
MacSlowSaviq, maybe I used teh wrong term :)11:19
Saviqtsdgeos, I think our last release must've caused the timeouts in Jenkins11:20
tsdgeosmay be :/11:20
Saviqtsdgeos, this is the first instance that took 2h40 http://s-jenkins.ubuntu-ci:8080/job/unity8-ci/5534/console11:20
tsdgeosthere's more than one thing failing11:20
tsdgeosthing -> job11:20
Saviqor actually11:20
Saviqthat's 4d21h ago11:21
Saviqso before the last release11:21
tsdgeoswhere do you check how much it took11:21
tsdgeos?11:21
Saviqhttp://s-jenkins.ubuntu-ci:8080/job/unity8-ci/5543/11:22
Saviqtop right11:22
Saviqor https://jenkins.qa.ubuntu.com/job/unity8-ci/5546/ for the public instance11:22
tsdgeosi have how we haven't been able to convince QA people to install the timestamp extension :/11:22
tsdgeoss/have/hate11:22
Saviqtsdgeos, vote https://trello.com/c/f1V9Y5Gj/213-timestamps-on-all-ci-logs up if you can ;)11:23
tsdgeosso we went from 1:41 to 2:40 ?11:24
tsdgeosno voting power11:24
Saviqtsdgeos, well, it times out at 2:4011:24
Saviqtsdgeos, I mean the AP runner times out after 120 mins11:25
Saviqhttp://s-jenkins.ubuntu-ci:8080/job/generic-deb-autopilot-runner-vivid-mako/1929/console11:25
tsdgeosright11:25
Saviqtsdgeos, I'm calling that's actually bug #142100911:26
ubot5bug 1421009 in qtbase-opensource-src (Ubuntu) "unity8 sometimes hangs on boot" [Critical,In progress] https://launchpad.net/bugs/142100911:26
tsdgeosmay be11:26
SaviqAP complains about no reply on dbus straight after launching unity811:26
Saviqmust be that11:26
tsdgeosthat or the other bug i fixed in ap11:27
tsdgeosthink it's unreleased yet11:27
Saviqtsdgeos, it is https://launchpad.net/ubuntu/+source/autopilot/1.5.0+15.04.20150408-0ubuntu111:27
tsdgeoshttps://code.launchpad.net/~aacid/autopilot/dbus_search_no_seen_connections/+merge/25410911:27
Saviqyeah that's there since last week11:28
Cimitsdgeos, why that change in widgetData assignment after oncompleted?11:29
tsdgeosCimi: because is whattriggets the bug11:30
Cimitsdgeos, ah so if the data changes after the component was created the layout gets confused11:31
CimiI get that11:31
tsdgeosSaviq: those autopilot releases are confusing11:31
MacSlowSaviq, last time I merged with unity-team/shellRotation was on Friday... and I'm missing the merge with trunk.11:31
MacSlowSaviq, so doing that now11:32
Saviqtsdgeos, I agree, train really helped with obviousness of what's released and what's not11:33
=== MacSlow is now known as MacSlow|lunch
=== alan_g is now known as alan_g|lunch
dandradermzanetti, is https://code.launchpad.net/~dandrader/unity8/ddaImprovements/+merge/255896 ok for you now?12:22
dandradertsdgeos, you need to review  https://code.launchpad.net/~dandrader/unity8/ddaImprovements/+merge/255896  again as I added a bunch of new stuff.12:24
dandradertsdgeos, mostly API grooming, moving code around12:24
mzanettidandrader, from an API point of view, I think yes12:33
dandradermzanetti, can I get a review from you then? Not the full code review, as tsdgeos was doing that. could be review a type like "API and manual test"12:35
tsdgeosdandrader: i can have a look yes12:36
mzanettiah ok. tsdgeos, you already ok with the code etc?12:36
mzanettiI'll give it a test too then12:36
tsdgeosmzanetti: i was fine with the "old" code12:36
tsdgeosnot sure how much it has changed12:36
mzanettiright... I don't think it change a whole lot, but yeah. let's give it a check12:36
mzanettidandrader, looks like a lot of jenkins failures12:39
mzanettiall related to dragging12:39
dandradermzanetti, you mean autopilot?12:40
mzanettidandrader, qmltests12:41
mzanettidandrader, https://jenkins.qa.ubuntu.com/job/unity-phablet-qmluitests-vivid/701/12:41
dandradermzanetti, ah, found them now12:41
tsdgeosah yeah i think that's why i didn't review last time i went over it12:41
SaviqCimi, when you're free, bug #1330959 would be a nice fix now the shape can do it12:50
ubot5bug 1330959 in unity8 (Ubuntu) "Art in cards with backgrounds are Ubuntu-shaped" [High,Triaged] https://launchpad.net/bugs/133095912:50
mzanettidandrader, interestingly the right edge is "smooth as silk" indeed, but the launcher seems jumpy now :)12:50
dandraderoh boy12:51
mzanettistill good enough I guess12:51
mzanettiwouldn't block on it I think12:51
mzanettiit's really only when you intentionally try to get the slowest possible speed that is still recognized as a drag12:52
tsdgeoselopio: do you think you're getting https://code.launchpad.net/~canonical-platform-qa/ubuntu-ui-toolkit/fix1401517-overwrite_swipe_borders/+merge/248986 in soon?12:53
tsdgeosasking to know how "urgent" is to review https://code.launchpad.net/~canonical-platform-qa/unity8/click_item_with_swipe/+merge/245051 that seems to depend on it12:53
CimiSaviq, I was having a look at the recency in the spread, but I can do this easily probably12:55
SaviqCimi, as you wish12:56
mzanettidandrader, do we still have something like dragArea.status == Undecided?12:56
mzanettidandrader, I think that's the issue. the launcher should be visible during that state too12:56
mzanettidandrader, also from an API point of view I guess having a property "pressed" in addition to the "dragging" one would make sense I think12:58
dandradermzanetti, no there's no "dragArea.status == Undecided" in the public API12:59
dandradermzanetti, but now touchX, touchY etc will only move once the gesture is recognized12:59
dandradermzanetti, and touchX,touchY,etc will move in a smooth way, instead of jumping to the actual touch position13:00
mzanettiright.. wonder why the launcher appears too late then13:00
dandradermzanetti, wider drag threshold most likelly13:01
mzanettidandrader, also line 2820 in the diff makes me suspicious a bit13:01
mzanettidandrader, so the idea from design was to show the launcher's shadow immediately when you touch, but only bring in the launcher when actually dragging13:02
mzanettidandrader, so having a "pressed" property that is true when a touch is there, but no gesture recognized would get us that I think.13:02
mzanettilike: pressed && !dragging == Undecided13:03
dandradermzanetti, ok13:07
=== alan_g|lunch is now known as alan_g
=== MacSlow|lunch is now known as MacSlow
=== dandrader is now known as dandrader|afk
elopiotsdgeos: not getting it too soon. I've put it back to work in progress.13:59
tsdgeosok13:59
=== dandrader|afk is now known as dandrader
tsdgeosMirv: tried the new qt with the patch that was supposed to fix kuniqueapplication, it doesn't14:55
Saviqdednick, desktop-app-focus diff still looks very much b0rked, is it based on something other than trunk?15:07
Saviqor is it supposed to be almost 2kloc15:07
Saviq?15:07
dednickSaviq: i've added mocking for Utils plugin15:08
dednickand it's almost 1.2k ;)15:09
Saviqdednick, ah, please be more descriptive in commit msg :)15:09
tsdgeosdednick: you almost got me in https://code.launchpad.net/~nick-dedekind/unity8/1436982.message-freeze/+merge/255702 ;) https://code.launchpad.net/~nick-dedekind/unity8/1436982.message-freeze/+merge/255702/comments/63819215:14
dednicktsdgeos: : woops :)15:15
dednickcopy paste is my enemy15:16
dednickbut couldn't live without it15:16
dednickguess i should add tests for those...15:18
tsdgeosdednick: also i'm kind of confused as to why https://code.launchpad.net/~nick-dedekind/unity8/1436982.message-freeze/+merge/255702 fixes the said bug15:28
tsdgeoswhich property is being updated on pressing clera all?15:28
dednicktsdgeos: i believe it's a property on the org.freedesktop.Accounts.User interface. There is only a "Changed" signal, so any property that is altered causes the maybeChanged slot to be called.15:30
dednicktsdgeos: i'm guessing it's XHasMessages15:30
dednickit's a bit of a shite interface.15:31
dednickwell, in the sense that it doesn't tell you what's changed.15:32
tsdgeosyeah15:32
tsdgeosi can see15:32
Saviqdednick, tsdgeos, looks like what mzanetti was fighting on the launcher?15:32
kapiteinedSaviq: did you see i was able to reproduce the screen rotate lock bug?15:33
mzanettierm, wat?15:33
Saviqmzanetti, lag in launcher15:33
mzanettiah15:33
dednickSaviq: could be. i think he put an async call in there15:33
Saviqmzanetti, due to AS15:33
Saviqkapiteined, not yet15:33
mzanettiyeah... changing one of the builtin properties causes *everything* to update15:33
kapiteinedok, take your time, it was just a heads  up.15:34
dednickmaybe worth looking into fixing underlying issue (ie Changed(propertyName)). assuming it's safe to do that.15:34
mzanettidednick, I had a chat with mterry about that... proper fix would probably be to patch accountsservice15:35
mzanettiand yes, we're not exactly sure what else relies on the general changed() signal. but sure, we could just not use that ourselves15:35
mzanettidednick, where do you see the issue?15:36
dednickmzanetti: clearing or adding a message in indicators15:36
dednickcalls the maybeChanged, which refreshes background as an async call.15:37
dednickwhich can take ages for some reason15:37
dednick*as an sync call.15:37
mzanettidednick, yeah, exactly what I had in launcher drag'n'drop. this is what I did: http://bazaar.launchpad.net/~unity-team/unity8/trunk/revision/167815:38
dednickmzanetti: right. i've changed the getUserProperty to async for updates.15:40
Mirvtsdgeos: ok :(16:27
=== dandrader_ is now known as dandrader|lunch
=== dandrader|lunch is now known as dandrader
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
=== Zhenech_ is now known as Zhenech
=== \b is now known as benonsoftware
=== tmpRAOF is now known as RAOF

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