=== _salem is now known as salem_
=== salem_ is now known as _salem
=== Mirv__ is now known as Mirv
=== ricotz_ is now known as ricotz
pete-woodsSaviq: hi. I have a silo with unity-scopes-shell in it ready to be landed (it does the RTM sync with settings, etc in it)07:51
pete-woodsI'm guessing you can predict the question. but are you happy with me landing it (and then you rebuild unity-scopes-shell in your silo)07:51
Saviqpete-woods, go for it07:58
pete-woodsSaviq: thanks! :)07:58
seb128Saviq, hey, r.e https://code.launchpad.net/~macslow/unity-notifications/fix-1354406/+merge/231005 ... do you have an url to a design document? I'm unsure people are along the same line on what are the requirements, from reading the comments08:07
Saviqseb128, just sent a comment there08:07
Saviqseb128, https://code.launchpad.net/~macslow/unity-notifications/fix-1354406/+merge/231005/comments/56314208:07
seb128Saviq, thanks08:08
MacSlowSaviq, seb128: I did not expect this particular issue to cause that much discussion08:09
seb128MacSlow, the behaviour seems wrong and likely to create user visible issues (like missing notifications)08:09
seb128did CI stop building amd64 debs?08:09
MacSlowseb128, well non snap-decision notifications were ok to be missed as far as I can remember... even on the phone08:10
seb128MacSlow, is that written down somewhere in a design spec?08:11
seb128MacSlow, I find it weird, we never did that on desktop, we always queued those08:12
larsuSaviq: this design document doesn't really think about a typical phone user...08:12
Saviqseb128, it never kept them around08:12
MacSlowseb128, larsu, Saviq: that spec has gone through too many design-hands in a too short time-period and several things got lost in between... thus the knowledge-base is very different in all involved "camps"08:13
larsuMacSlow: exactly. What I'm arguing is that the MR is clearly wrong, even though the design document says this is what we want. We should talk to design about it08:14
Saviqseb128, yes it is, see https://docs.google.com/a/canonical.com/document/d/1puQ9Z0yKqzsQ1VQ1OOBkxgp78iWGnAhAkFXWJFTWIrE/edit#heading=h.zf8gzpeasuio for example08:14
Saviqlarsu, seb128, MacSlow, remember that that's what we have the notification centre for, too, the missed notifications should end up there08:15
seb128Saviq, I'm not speaking about keeping, but about displaying them at least08:15
seb128Saviq, well, imho things shouldn't go missing from screen, expect if you opted out08:16
larsuSaviq: hm? What puts them there?08:16
Saviqlarsu, right now, the application, soon, the post office08:16
Saviqseb128, yeah, it's explicitly mentioned in that spec that they can "get lost" in the sense they will never be displayed as a bubble08:16
seb128Saviq, :-(08:17
seb128I'm glad we still have unity7 for desktop :p08:17
seb128we should talk to designers out of that08:17
seb128Saviq, MacSlow: do you know who is the design to talk to about those?08:18
MacSlowseb128, yes...08:18
* larsu hides08:18
Saviqnot any more I don't think08:18
MacSlowlarsu, no... mpt no longer does work on those...08:18
MacSlowlarsu, seb128: Daniela and Esti do notifictions on the phone now08:19
seb128MacSlow, what about "notifications with convergence in mind"08:19
seb128MacSlow, we don't design an OS that works on the phone but sucks on desktops08:19
* larsu thinks this doesn't even work on phones08:19
seb128larsu, well, you are less likely to be looking at your phone screen all day long, so I guess notification bubbles are less important there, as long as they are recorded in the notification log/indicator08:20
seb128but on my desktop I don't want to miss IRC pings just because designers decided it's fine to ignore a ping just because 2 people pinged me08:21
MacSlowseb128, the update an existing notification is still in place and available on the phone...08:21
SaviqMacSlow, did you manage to still reproduce the issue on the phone? with your example py?08:22
MacSlowseb128, but e.g. I've had to axe the "append"-feature on the phone... which was introduced on the desktop for e.g. IRC-ping use-cases08:22
seb128MacSlow, so, in which case do we dismiss notifications?08:22
MacSlowSaviq, not yet08:22
seb128MacSlow, get the "on the phone" out of your mind, that's supposed to be a convergent OS08:22
seb128you need to support the IRC usecase08:22
seb128you can't say "it's only a phone, screw that", it's going to bite us back08:23
MacSlowseb128, talk with Design not me .)08:24
larsuseb128: but if I have my screen on right now, I don't want to miss notifications08:24
seb128larsu, yeah, not sure what's the rational there08:24
seb128MacSlow, k, but please don't push those buggy changes in until we talk to design then08:25
seb128MacSlow, daniela and esti you say?08:25
MacSlowseb128, correct...08:29
seb128MacSlow, thanks08:30
MacSlowseb128, I've asked them to be on IRC daily, thus to have easier reach to engineering and vice versa08:30
seb128do you know if they are there/what nicknames they use?08:31
Saviqseb128, not around atm08:31
Saviqseb128, mup knows nicknames08:31
seb128Saviq, thanks08:33
MacSlowseb128, dferrai and ...08:38
MacSlowseb128, forgot Esti's nick...08:38
seb128MacSlow, don't worry, Saviq said mup has those08:38
MacSlowseb128, aht ehre you go...08:38
MacSlowSaviq, thx08:38
seb128Saviq, MacSlow: thanks08:39
MacSlowSaviq, larsu: so in order to unblock the train - for the moment - due to LP: #1354406, if I make multiple notifications from the same app (using the same app/sender-name) work with the backend... would be be happy?08:41
ubot5Launchpad bug 1354406 in unity8 (Ubuntu) "PIN-locked phone becomes unresponsive after simultaneous alarms" [High,In progress] https://launchpad.net/bugs/135440608:41
SaviqMacSlow, yes08:41
MacSlowSaviq, larsu: not saying we should not talk about this to Design and ask for more convergence-robust UX-design08:42
MacSlowSaviq, then I'll do that08:42
SaviqMacSlow, we need to fix the real failure, not prevent it from happening08:42
MacSlowSaviq, I'll be looking into the whole thing08:43
dednickSaviq: Fixed up the indicator branch now i think. Using qml ownership and seems to get deleted on GC.09:23
dednickSaviq: i was reading about the performance of gc in qml, and it seems to indicate that we should probably be scheduling some manual gc at times. Apparently it only gets done when memory gets scarse. Perhaps after turning off the screen we should schedule a gc.09:26
Saviqdednick, trueth09:29
seb128Saviq, can we get https://code.launchpad.net/~mterry/unity8/dismiss-old-pam-prompts/+merge/231363 in the next unity8 landing?09:44
Saviqseb128, we will for sure09:46
seb128Saviq, thanks09:46
CimiSaviq, is there a reason why you use header customisation as prerequisite for branches like card attributes?10:13
SaviqCimi, tried to get away from conflicts10:14
tsdgeoskarni: ping10:21
karnitsdgeos: pong10:21
tsdgeoskarni: i'm not sure i undersntad https://bugs.launchpad.net/unity8/+bug/131666010:21
ubot5Ubuntu bug 1316660 in Unity 8 "Scope preview should use SDK's Expandable*s" [Medium,Triaged]10:21
karnisure, 1 sec10:21
tsdgeoskarni: are you complaining about that the user needs to scroll the view after clicking the combo?10:21
=== Estilanda_ is now known as Estilanda
karnitsdgeos: I'm not complaining, I'm reporting :D After clicking the combo (which is (or used to be) a hack, not an SDK expandable [as Savi-q called it at the time]), if it overflows the bottom of the screen (say, there's many buttons collapsed), you need to scroll the whole screen to see them10:23
karnitsdgeos: i.e. imagine I have enough content that I have [ More ] [ foobar ] buttons at the bottom10:23
karniwhen I press the [ More ] button, part of the 3 (so, not so many) buttons overflows the bottom edge of the screen10:23
Saviqkarni, Expandable is unrelated to the combo button10:24
tsdgeoskarni: does overflow mean you can't get them? or it means the user has to scroll to them?10:24
karnitsdgeos: I don't mind what component is used, what I mean is the lack of "auto scroll" when they're shown10:24
Saviqkarni, but yeah, the fact that you want it to scroll magically as it expands, that's the Expandable10:24
Saviqtsdgeos, the latter10:24
karnitsdgeos: I need to scroll there down10:24
tsdgeoskarni: so you want autoscroll10:24
tsdgeoskarni: can you update the subject? i think it'd be more clear10:24
tsdgeossomething like "preview combo should autoscroll to make sure its child buttons are on screen"10:25
tsdgeosor something10:25
tsdgeoswhich is frankly what we want10:25
karnitsdgeos: I'd be happy to. I actually updated it initially after Savi-q's suggestion, because my title sucked lol :D10:25
* karni fixes10:25
karniI suppose the intention was the title would contain a possible suggestion for the fix :)10:26
tsdgeosbecause with that description i am not sure it's such a problem sure, the user has to scroll, but sill can do most of his stuff10:26
tsdgeosit's not like things explore :D10:26
karnitsdgeos: yes, they don't explore :) it's just an UI annoyance. and since it's in previews (and scopes are important to Phone), I guess it would be a 'nice to have' category10:27
karni*they don't explode10:27
karnihaha, I repeated the same typo10:27
CimiSaviq, https://code.launchpad.net/~saviq/unity8/fix-empty-attributes/+merge/23107610:42
dandraderMacSlow, https://code.launchpad.net/~macslow/unity8/fix-1348092/+merge/228090/comments/56320610:44
dandraderMacSlow, is it a preexisting unstable AP test? could it be because jenkins is missing lp:~macslow/unity-notifications/fix-1348092 ?10:44
dandradermzanetti, greyback, so what's next for the lifecycle branches?10:46
mzanettidandrader: I'm charging my Nexus 10 to give it a run there, but if that's good I'll approve10:46
greybackdandrader: mzanetti was doing most of the review, so if he's happy, I'm happy10:47
tsdgeoslarsu: can confirm i'm correct in https://bugs.launchpad.net/ubuntu/+source/indicator-messages/+bug/1359152 ?10:47
ubot5Ubuntu bug 1359152 in indicator-messages (Ubuntu) "indicator-messages has no .pot nor .po files" [Undecided,New]10:47
* mzanetti hopes 5% battery is enough to flash the N1010:48
greybackSaviq: https://code.launchpad.net/~gerboland/unity-api/surfaceSizerCallback/+merge/230270 - I tried your Binding idea, fails as QJSValue not a QVariant. Can convert, think it worth it?10:51
tsdgeosneed to reboot router 75% packet loss11:01
=== pete-woods is now known as pete-woods|devic
=== pete-woods|devic is now known as pete-woods|away
mzanettiSaviq: will we get the brightness slider in the indicators back?11:08
mzanettior is that gone for good?11:08
=== MacSlow is now known as MacSlow|lunch
Saviqmzanetti, didn't know it was gone11:17
mzanettiSaviq: I miss it really hard11:18
mzanettiSaviq: the auto doesn't cut it11:18
mzanettiits too bright at night, too dark in daylight sun11:18
mzanettiI'm not even sure if it changes automatically any more11:19
mzanettiSaviq: any outcome for the inactive app thing yesterday?11:27
mzanettiSaviq: otherwise I'd say we get daniel's branch landed as is for now and can pimp visuals later11:28
Saviqmzanetti, inactive meaning getting resumed?11:29
Saviqmzanetti, https://code.launchpad.net/~dandrader/unity8/lifecycle/+merge/230090/comments/56284611:30
mzanettiah ok. thanks11:30
mzanettiok... I vote for a new branch, dandrader, your opinion? ^11:30
dandradermzanetti, me too. already replied to that comment yesterday11:35
mzanettidandrader: found a failing test11:35
dandradermzanetti, https://code.launchpad.net/~dandrader/unity8/lifecycle/+merge/230090/comments/56291111:36
tsdgeosMirv: Saviq: we had a ppa with "newer" qt, right?11:36
dandraderdang it11:36
mzanettiyeah, just seen the comment11:36
Saviqtsdgeos, if anywhere... https://launchpad.net/~canonical-qt5-edgers/+archive/qt5-beta211:38
larsutsdgeos: indicator-messages has quite a few translated strings in it. Translations come from launchpad. No idea how it works exactly (but seb128 might)11:56
seb128larsu, tsdgeos: I was just looking at that11:57
seb128https://translations.launchpad.net/ubuntu/utopic/+source/indicator-messages has strings but not updated ones11:57
seb128let me debug/fix it11:57
=== alan_g is now known as alan_g|lunch
=== MacSlow|lunch is now known as MacSlow
dandraderSaviq, should all qmltests pass fine with trunk?12:05
dandraderI'm getting loads of failures in "make testCard"12:05
=== _salem is now known as salem_
=== Malsasa_ is now known as Malsasa
dandradermzanetti, ^^^ ?12:22
Mirvtsdgeos: Saviq: bookmark https://wiki.ubuntu.com/Touch/QtTesting12:23
Mirvbut yes, beta2 atm12:23
MirvI think I need to delete unity-scopes-shell + api temporarily from there since trunks had '+rtm' version numbers for a moment so now new builds won't get automatically built12:30
Saviqdandrader, passes here?12:42
Saviqdandrader, and jenkins isn't too fussy about them either?12:42
dandraderSaviq, what's wrong with my desktop environment then....12:42
Saviqdandrader, can you paste the output?12:43
dandraderSaviq, hmm, I do build unity8 on a separate directory, not inside a ./build subdir. that might be it12:44
dandraderlet me try building it inside a ./build subdir...12:44
Saviqdandrader, yeah, just go ./build.sh12:44
sil2100Saviq: hey! How are the reviews of the two blocker merges going? ;)12:45
mzanettidandrader: passes here too12:45
Saviqsil2100, one's almost there (datetime), solution for the other is being worked on, the original solution did not meet expectations ;)12:46
sil2100Awwww :<12:46
sil2100Saviq: ok, good to hear, we're waiting for those with anticipation12:47
sil2100Saviq: is there a possibility for a fix today still?12:47
dandraderSaviq, yeah, that's exactly the problem. A bug in the Card tests I say having to expect that it's built in a source subdir12:47
Saviqsil2100, unlikely to land today still (especially seeing how I'm running for the airport in a mo, unless kgunn takes that on)12:47
sil2100I might poke him later on then, we would love getting rid of TRAINCON asap12:48
mzanettibregma: hey, I heard you're working on high dpi support12:54
bregmamzanetti, yes, and it works for me12:55
mzanettiit works for me to!12:55
mzanettiits awesome!12:55
mzanettistill a few things here and there12:55
mzanettilike count emblems on the launcher etc are still tiny (I reported a bug in LP already for that)12:56
bregmamzanetti, yes, there's a delayed patch in the queue for that already12:56
mzanettibregma: but also for instance the mail status icons in thunderbird are so tiny that you can't really see what they are12:56
mzanettior the thread expand/collapse triangle seems hardcoded small too12:57
bregmamzanetti, there are plenty of poorly written applications out there that need fixing upstream12:57
mzanettiyeah, I agree12:57
bregmafirefox makes my life miserable12:58
mzanettibregma: oh, taht works fine for me12:58
mzanettibregma: install the nosquint plugin12:58
mzanettibregma: and in about:config set devPixelsPerPx to 1.512:58
mzanettibregma: but yeah, I think we should automatically install/set those things if detected12:59
bregmait's just the chrome that's wrong, the content is OK12:59
mzanettibregma: yeah, the about:config thing fixes the crome12:59
bregmabut like Gnome, it's broken by design on multiple heterogenous displays13:00
mzanettibregma: hmm... for me that is fixed by hardware (macbook pro)13:01
mzanettibregma: seems the graphic chip scales it so it looks the same as on the highdpi screen13:01
mzanettibregma: but anyways, my question: does it make sense if I collect such things as the thunderbird issues in Launchpad?13:01
mzanettior will they just be ignored and waited for upstream to fix things13:02
bregmawell, they should be filed as bugs against project in Launchpad so they can be tracked and fixed, but there isn't really a good central project to file them against13:03
mzanettiso what would you suggest I should do?13:04
mzanettibregma: ah right, one more thing... the mouse is tiny when above unity elements. its fine when inside other windows. you don't happen to know a fix for that?13:07
bregmamzanetti, I'm thinking the best way to track cross-package HDPI bugs would be a metaproject, since it gives the best reporting ability, but I'll want to consult other folks and get back on that13:08
mzanettibregma: cool, works for me13:08
bregmaas for the cursor size problem, I would suggest filing a bug against Unity, we'll have someone look into it (I believe we did some work on the cursor not too long ago)13:09
mzanettiack, will do. thanks bregma!13:10
Saviqgreyback, you saw Project ERROR: lttng-ust development package not found13:12
Saviqgreyback, the gles sync needs you to update deps13:13
Saviqkgunn, I'm flying home, won't be there until late, if you wanna land greyback's silo 16 and dednick's fix, they would need a re-review that I won't be able to do13:15
Saviqkgunn, if you'd want to sort that out, there's also some other top-acked branches of unity8 that we could land13:17
kgunnSaviq: yack13:20
kgunn...or ack13:20
kgunnor both13:20
kgunnaltho i think greyback's unity8 mp has addressed the needs fixings13:20
kgunnSaviq: and tsdgeos you each had one https://code.launchpad.net/~gerboland/unity8/initialSurfaceGeometry/+merge/23049013:20
kgunnSaviq: ;) i got ya...will get as many approved as possible and rebuild13:20
=== dandrader is now known as dandrader|afk
Saviqk, I'm off o/13:22
kgunnsafe travels13:22
sil2100kgunn: sooo! Could you make sure the work on the 3 blockers we have is prioritized above anything else?13:22
mzanettibregma: hah! just figured the about:config devPixelsPerPx trick works in thunderbird too!13:23
kgunnsil2100: yes! those are #1....the 2 date/time indicator ones have an mp i'm trying to get reviewed/approved13:23
kgunnthe dual notification one needs a little more time13:26
kgunntsdgeos: could you review https://code.launchpad.net/~nick-dedekind/unity8/lp1328646/+merge/231335 as well ?13:26
tsdgeoskgunn: sure13:26
MacSlowSaviq, current trunk (unity8, unity-notifications) does correctly handle multiple notifications (any type) from the same app... I've yet to run that same test on my N4... latest image-update did fail and I've to reflash it.13:27
sil2100kgunn: thanks!13:38
sil2100kgunn: we're really pressured by the need of a new promoted image13:38
tsdgeosmterry: are you sim pin guy?13:48
mterrytsdgeos, mzanetti mostly, but maybe me as well13:49
tsdgeosmterry: any clue what may be causing this? https://chinstrap.canonical.com/~aacid/sim.mp413:49
tsdgeosit happens say 1 out of 4 times i write the PIN wrong13:49
mterrytsdgeos, that black screen thing?13:50
tsdgeosmterry: yep13:50
mterrytsdgeos, mzanetti would know -- something about how we receive notifications or something?13:50
mterryI remember hearing that in the past13:50
tsdgeosmterry: mzanetti: is there a bug about it?13:50
mterrytsdgeos, I don't know13:52
=== pete-woods|away is now known as pete-woods
* mzanetti reads13:55
mzanettitsdgeos: well, that's the notifications thing13:56
mzanettitsdgeos: it closes the screen, and if its wrong it reopens it13:57
tsdgeosmzanetti: eh?13:57
mzanettitsdgeos: I think Wellark is supposed to fix this13:57
tsdgeoswhich notification13:57
tsdgeosthere's no notification :D13:57
mzanettitsdgeos: the SIM pin entry is a notification13:57
mzanettia snap decision13:57
tsdgeosbut why black screen13:57
tsdgeosand moreover with a text on the top left13:57
tsdgeosis that the notification being slowly built up?13:58
mzanettitsdgeos: I guess a bug in the notifications code... it paints the default background color for a bit13:58
mzanettiMacSlow: ^ ^13:58
mzanettifix it :P13:58
mzanettitsdgeos: anyways, the proper thing is to not close the notification until we know the status13:58
mzanettiinstead of closing and reopening. however, I don't know what the ETA on that is13:59
MacSlowmzanetti, tsdgeos: I don't think I can do much about this... right now at least with on my back 135440614:02
=== dandrader|afk is now known as dandrader
mzanettiIs there a way to stop files from opening every time I plug the phone?14:09
tsdgeosdednick: curious about the change in qml/Panel/Indicators/IndicatorBase.qml in https://code.launchpad.net/~nick-dedekind/unity8/lp1328646/+merge/23133514:12
tsdgeosdednick: is it just stylistic? or has real impact?14:13
dednicktsdgeos: impact14:13
tsdgeosdednick: interesting, what's the difference between the alias and the property?14:14
dednicktsdgeos: oh right. sorry, i thinking you were talking about the mp in general. That change was regarding Saviqs comment "There's a problem with using types like this in properties... Those objects get created on parent object creation, just to be replaced with the bound one and destroyed"14:15
=== alan_g|lunch is now known as alan_g
dednicktsdgeos: but that was for the UnityMenuModel property14:15
tsdgeosdednick: ah right14:16
dednicktsdgeos: but not entirely convinced on that point.14:16
tsdgeosdednick: it'd be created twice14:17
tsdgeosone for the property and one for the right hand side and one for the left hand side14:17
tsdgeosQML "bug"14:17
dednicktsdgeos: i c. but isn't it a "pointer" ?14:17
tsdgeosit' even "better"14:18
tsdgeosif you do14:18
tsdgeos property RootActionState rootActionState: null14:18
tsdgeosit still creates one and then assigns null14:18
tsdgeosto the pointer :D14:18
tsdgeosneeds some work14:18
dednicktsdgeos: eek14:19
tsdgeosdednick: code looks good, i'll test it in the phone now14:21
mzanettimterry: reviewed this: https://code.launchpad.net/~mterry/unity8/sim-unlock-on-boot/+merge/23101514:23
mterrymzanetti, OK14:25
MacSlowSaviq, not sure if you seen it... provided an small update on LP: #1354406 https://code.launchpad.net/~macslow/unity-notifications/fix-1354406/+merge/231005/comments/56330314:41
ubot5Launchpad bug 1354406 in unity8 (Ubuntu) "PIN-locked phone becomes unresponsive after simultaneous alarms" [High,In progress] https://launchpad.net/bugs/135440614:41
=== dandrader is now known as dandrader|lunch
=== salem_ is now known as _salem
mzanettimterry: https://code.launchpad.net/~mterry/unity8/sim-unlock-on-boot/+merge/231015/comments/56334715:40
mterrymzanetti, :(15:41
mterryI tried it...15:41
* mterry tests more15:41
mzanettimterry: well, let me know if you need logs or something15:41
mzanettimterry: I installed the packages from jenkins and rebooted15:41
tsdgeosmterry: ping15:41
mterrymzanetti, and you have a locked SIM I suppose.  I was using the phonesim package to fake it15:41
mterrymzanetti, oh do you have the connectivity branch installed?15:42
mterrymzanetti, linked in the description15:42
mterrytsdgeos, yo15:42
tsdgeosmterry: the ""Incorrect passcode. Try again." thing comes from ubuntu-system-settings, right?15:42
mterrytsdgeos, yeah if you're switching passwords15:42
tsdgeosdednick: didn't work15:43
tsdgeosmy clock is still in the past15:43
tsdgeosthough it did update after a while15:43
mzanettimterry: that explains things... stupid me...15:44
dednicktsdgeos: hm. that's weird.15:44
dednicktsdgeos: perhaps i fixed one thing but missed something else15:44
mterrymzanetti, I'm testing your suggested change of instantiating QDBusInterface on the fly.  I'm concerned with how that interacts with an asyncCall, hence the test15:45
mzanettiah... I see15:45
tsdgeosdednick: let me try to get a video15:45
=== gatox is now known as gatox_lunch
dednicktsdgeos: did it happen immediately, or take awhile?15:47
tsdgeosto update?15:47
tsdgeossay a few seconds15:47
dednicktsdgeos: to stop syncing15:47
tsdgeosi was able to clearly see the old time15:47
tsdgeoscame here, wrote it was wrong15:47
tsdgeoswhen i looked at the phone again was fine15:48
tsdgeosdednick: oh just after a reboot15:48
dednicktsdgeos: ah. well the sync only happens once every few seconds, so it's possible the labels are getting out of date15:49
dednickmight not be updating if screen is off15:49
tsdgeosdednick: did take something like 15 secs to recover15:50
tsdgeosdednick: can we make that trigger on screen unlock?15:51
tsdgeosit defeats the only purpose of a mobile phone15:51
tsdgeosthat is to replace my watch15:51
tsdgeosif it can't tell the time correctly :D15:51
tsdgeosdednick: or has nothing to do with this thing?15:51
MacSlowSaviq, kgunn: I'm a bit closer to the cause of the "blocking/invisible" snap-decisions... it's some opacity-animation-related issue... certainly the frontend... not the backend15:51
dednicktsdgeos: true :)15:51
kgunnAlbertA2 greyback ....so i lost track, but with the flashing dash on app start...were you guys thinking there was an errant alpha15:52
MacSlowSaviq, kgunn: disabling the opacity-animation "solves" the issue... so that's our fallback plan at least :)15:52
AlbertA2kgunn: yeah, the first frame from a client15:52
kgunnwondering if it might relate to MacSlow 's issue here ^15:52
AlbertA2is blank for some reason15:52
AlbertA2which in a nested case, since transparency is enabled15:53
AlbertA2will also have transparent alpha15:53
MacSlowSaviq, kgunn: still trying to figure out why the opacity-animation can get stuck at opacity=0 (leading to invisible nsap-decisions)15:53
greybackkgunn: I don't think the two issues are related15:54
kgunnMacSlow: can you go ahead and gen up an MP with disabling opacity-animation in order to unblock....and we'll keep you working on the bug, but without all the pressure :)15:55
mzanettimterry: \o/15:55
kgunngreyback: what's your line of thinking as to why that wouldn't be related ? (sorry..curious)15:56
mterrymzanetti, worked this time?  sweet15:57
mzanettimterry: yes. still can't do emergency calls though15:57
MacSlowkgunn, was about to suggest that now that I added another comment on the initial bug... https://code.launchpad.net/~macslow/unity-notifications/fix-1354406/+merge/231005/comments/56335615:57
mzanettimterry: but I guess not on your plate15:57
mterrymzanetti, I thought I fixed that15:58
mzanettimterry: well, says no network15:58
mzanettimterry: yes, it opens the dialer15:58
mterrymzanetti, oh ah15:58
greybackkgunn: sure. The flashing dash issue is due to the first frame (or 2) of an application apparently being not be drawn in - and therefore fully transparent - which lets the dash shine through underneath briefly, until app draws a frame which is opaque15:58
mzanettimterry: but the dialer says no network15:58
mterrymzanetti, yeah I'm throwing my hands up on that front  :)15:58
mzanettimterry: granted15:58
MacSlowkgunn, still want to do some more tests on the N4 to further verify the validity of the unblocking work-around.15:58
kgunnMacSlow: please do15:58
greybackkgunn: something is odd in Mir, as I've QtComp designed to only show a client surface when it has drawn to it. So something confused somewhere15:58
mzanettimterry: but...15:59
kgunngreyback: i was just thinking in the notification case, its likely to draw 1(or2) frames...then nothing else....so it'd be invisible, just like the first few frames of an app15:59
mzanettimterry: I entered the pin wrong 3 times... still asks me for the pin15:59
* kgunn grabs snack16:00
mzanettiwell, still not your issue, I realize16:00
greybackkgunn: snap decision is drawn by unity8/compositor itself. It does not go through mir's client buffer codepath16:00
greybackkgunn: so I expect it to be a QML bug16:00
anpoki try to debug a mesa crash with unity8-dash - what do I have to do to mimic unity8 and launch unity8-dash myself?16:03
=== Malsasa_ is now known as Malsasa
MacSlowSaviq, kgunn: the workaround does really unblock us... the two-alarms-at-the-same time issue no longer happens... MP is incoming16:09
Wellarkmzanetti: I will fix it once I get there16:09
mzanettiWellark: which one?16:10
Wellarkthe sim dialog acting weird16:11
WellarkI will probably encounter it many times when I add dual sim stuff to the pin unlock dialog16:11
Wellarkand that has to happen before we can land automatic sim unlocking from the greeter16:11
mzanettiWellark: why that?16:12
mzanettiWellark: we're landing that about now16:12
mzanettiright... because otherwise the user won't know which sim card16:13
kgunnanpok: hmmm, wondering if dash as app might help there... mzanetti any ideas for anpok ? like could he hack something to bring up unity8 shell...but not the dash, so he can launch it manually16:14
mzanettisure :)16:14
Wellarkmzanetti: you can't land it before indicator-network lands it's private dbus API to do that16:14
kgunngreyback: ok, that makes sense16:14
mzanettianpok: drop /usr/share/upstart/sessions/unity8-dash16:15
mzanettianpok: or well, just drop the like "start on unity8 started"16:15
mzanettianpok: then you can start/stop it manually16:15
Wellarkmzanetti: do you have this in a silo already?16:15
Wellarkif, so, please postpone to wait on the dualsim sim unlock logic16:16
mzanettiWellark: no, not yet... but I'll approve the unity8 related branch in a minute and then it'll be in the next silo16:16
mzanettikgunn: ^16:16
mzanettimterry: ^16:16
kgunnmzanetti: ack on the "wait for dual sim backend to get sorted"16:17
kgunnWellark: please hit me up when you're ready16:17
greybackanpok: http://pastebin.ubuntu.com/8099103/ - on the phone, that is the command I use to manually launch unity8. I expect nearly the same will work on desktop. Then you should be able to launch dash, setting the MIR_SOCKET correctly16:17
mterrymzanetti, sorry still testing (I had to reflash my device)16:17
greybackanpok: oh, or what mzanetti said :)16:18
mzanettimterry: no worries... seems you got some more time anyways16:18
Wellarkkgunn: I will. if I remember correctly there is some hooking up to do on the unity8 side as well, so finishing the backend logic will also involve a unity8 MP16:18
Wellarklet's put them all to the same silo then16:18
MacSlowSaviq, kgunn: here's the unblocking MP https://code.launchpad.net/~macslow/unity-notifications/fix-1354406/+merge/231005/comments/56337516:20
kgunnmzanetti: hey, can you review MacSlow's workaround branch ? ^ i'll test16:24
kgunnits pretty simple16:24
kgunncode change that is16:25
mzanettikgunn: didn't test it, but reading the code and the bug description I'm having a hard time to believe this is the cause16:27
mzanettibut I might be wrong...16:27
kgunnmzanetti: i won't approve until i test16:27
kgunnand verify16:27
=== _salem is now known as salem_
mzanettikgunn: added some comments https://code.launchpad.net/~macslow/unity8/disable-opacity-animation-1354406-workaround/+merge/23158816:34
=== gatox_lunch is now known as gatox
=== salem_ is now known as _salem
=== dandrader|lunch is now known as dandrader
=== alan_g is now known as alan_g|EOD
mterrymzanetti, OK updated branch to create object on the fly17:15
mzanettimterry: cool17:16
kgunngrabbing lunch, bbiab17:24
=== _salem is now known as salem_
elopioSaviq, mzanetti: how can I collect more information out of a QML test that segfaults during build on Jenkins? It passes nicely on my machine17:42
mzanettielopio: which one?17:44
elopiomzanetti: https://code.launchpad.net/~canonical-platform-qa/dialer-app/qmltests1/+merge/23041217:44
mzanettielopio: run it in xvfb17:44
mzanettilet me find the command17:44
elopiomzanetti: I'm running it on xvfb.17:44
mzanettihmm, ok17:44
mzanettithat sucks then17:45
* mzanetti reads logs17:45
elopioI have no idea is how to identify the cause for the Segmentation fault.17:45
* mzanetti branches phone-app17:46
mzanettielopio: what's with this? file:///tmp/buildd/dialer-app-0.1+14.10.20140806bzr224pkg0utopic156/src/qml/HistoryPage/HistoryPage.qml:22:1: module "Ubuntu.History" is not installed17:50
mzanettistill the segfault is odd17:51
elopiomzanetti: I uninstalled qtdeclarative5-ubuntu-history0.1 and the tests still pass here. How did you get that error?17:54
mzanettielopio: its in the jenkins log17:54
mzanettielopio: fails for other imports here17:54
elopiomzanetti: you will need to install all the build deps I added on that branch.17:55
=== salem_ is now known as _salem
mzanettielopio: op... now I see the same log as in jenkins, but no segfault17:59
elopioyes, that's what's puzzling. I changed the run to show more info, now it shows the same as here so what's printed is not the cause of the segfault.18:00
mzanettielopio: hm... I guess you can always log into that jenkins machine and try there18:02
mzanettibut this is odd18:03
mzanettishouldn't happen18:03
mzanettiits also regardless of the arch...18:03
elopioI'll ask for help from CI.18:04
elopiomzanetti: I've just noticed on my machine I'm not getting tp-qt 0.9.3 WARN: Account filtering requires AccountManager to be ready18:04
elopiodo you get that message?18:05
mzanettino, I dont18:05
* mzanetti kills accounts-daemon18:05
mzanettielopio: hmm... maybe something with telephaty-qt or so18:11
mzanettielopio: try asking boiko about this warning18:11
elopiomzanetti: I will. Thanks.18:11
=== _salem is now known as salem_
mhall119bregma: ping18:53
=== dandrader is now known as dandrader|afl
=== dandrader|afl is now known as dandrader|afk
bregmamhall119, poing19:02
mterrymzanetti, when you get a chance, could you approve the sim-unlock-on-boot branch?19:57
=== salem_ is now known as _salem
=== fginther` is now known as fginther
Saviqelopio, loop it21:17
Saviqelopio, run it in a loop, when it crashes you should get a .crash file as usual21:17
Saviqelopio, you should be able to get it (.crash) out of jenkins, too21:17
kgunn_Saviq: is there any problem if i run unity8 ap with a pin code set ?21:29
kgunn_never thot about it before...21:29
Saviqkgunn_, it should use a dummy backend21:30
Saviqkgunn_, so no21:30
Saviqkgunn_, same as when you ./run.sh, it uses a dummy auth backend that won't ask you for passwords21:31
kgunn_hooray! ap tests pass22:10
=== _salem is now known as salem_
=== dandrader|afk is now known as dandrader

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