/srv/irclogs.ubuntu.com/2014/03/13/#ubuntu-unity.txt

=== jasoncwarner___ is now known as jasoncwarner
=== jhodapp|afk is now known as jhodapp
=== jhodapp is now known as jhodapp|afk
mhr3mzanetti, btw your activate-appid branch landed09:58
mzanettimhr3: yeah, seen that. thank!09:58
mzanetti+s09:59
mhr3hope we didn't break clicks again :)09:59
mzanettimhr3: heh. no I fixed the reason why this was reverted the first time :=09:59
mzanettianyone happens to know a workaround for the "error during hwc set"? I'd need to test something when locking and waking up the device10:45
mhr3tsdgeos, what's the difference between new-scopes-cleanup and ..-clean-to-trunk?10:48
tsdgeosmhr3: cleanup has tests that pass and unneeded files removed10:48
mhr3ah, so it's a dep10:50
Saviqmzanetti, restarting unity8 while the screen is on should work11:01
mzanettiSaviq: hmm, no, doesn't seem to help. I use run_on_device. That crashes as soon as the screen blanks regardless of the state when it was started11:01
Saviqmzanetti, hum, other than reboot, dunno what else11:02
mzanettiSaviq: but I got it fixed by fiddling with the installed qml code and using upstart11:02
Saviqmzanetti, hmm I wonder if we need to tweak run_on_device for system compositor...11:03
mzanettistill this is a bit nasty in everydays work. Will this be tackled by the "who-talks-to-powerd" discussion?11:03
Saviqmzanetti, most probably, yeah11:03
Saviq/food11:04
tsdgeosCimi: ping11:21
Cimitsdgeos, ponbg11:21
tsdgeosCimi: if i send you some self compiled Qt 5.2 libs can you try if you can reproduce the crash with them?11:22
Cimitsdgeos, nope I can't11:22
Cimiof course I can silly!11:22
Cimitsdgeos, send them over11:22
Cimitsdgeos, but before you tell me the secrets of debugging such crashes11:23
tsdgeosi don't even know how i reached that conclusion :D11:23
tsdgeosso i compiled everything with ASAN11:23
Cimihah11:23
tsdgeosand ended up with something that resembles http://paste.ubuntu.com/7083801/11:23
tsdgeosand then i said something is going on with QQmlVME11:24
tsdgeosand then i did this patch http://paste.ubuntu.com/7084035/11:24
tsdgeosit's all a bit blurry tbh11:24
tsdgeosbut can't really explain mroe :D11:24
tsdgeosASAN = adding -fsanitize=address to all the compile lines of qt, unity8, etc11:24
tsdgeosCimi: sent11:32
=== jhodapp|afk is now known as jhodapp
=== MacSlow is now known as MacSlow|lunch
maxbArgh. The new unity lock screen is very pretty, but almost every time I unlock the screen, it relocks immediately once, effectively requiring my password twice every time I sit down at my computer. Is this a known bug?12:01
CimiTrevinho, ^12:03
mhr3Saviq, distro version of scopes-api now listening to the scope-ui-... signal12:03
Saviqmhr3, cool, reverting the temporary thing in new-scopes12:03
Cimitsdgeos, I think it fixes for me12:12
tsdgeos\o/12:13
paulliuSaviq: https://code.launchpad.net/~paulliu/unity8/zoomImage/+merge/20794112:34
paulliuSaviq: BTW, why jenkins failed?12:35
paulliuSaviq: I got 404 today when I tried to open the link.12:35
Saviqpaulliu, which link?12:35
paulliuSaviq: https://jenkins.qa.ubuntu.com/job/unity8-ci/2403/12:36
Saviqpaulliu, that's some really old job12:36
Saviqpaulliu, autopilot failures are known/fixed - i.e. the unicode ones are fixed, the process not found ones are Qt 5.0 crashes - fixed with 5.2 (to land today)12:37
Saviqpaulliu, here's the latest run for that branch https://jenkins.qa.ubuntu.com/job/unity8-ci/2468/12:37
paulliuSaviq: ok. I'll try to kick a rebuild.12:37
Saviqpaulliu, I can do that if you can't12:37
paulliuSaviq: ah.. please do.12:37
Saviqpaulliu, wait(300); is never a good thing12:38
Saviqpaulliu, is there something you can wait for? a signal, some property reaching some value?12:38
paulliuSaviq: Let me research that a bit. Actually something weird there.12:39
=== _salem is now known as salem_
rsalvetitsdgeos: mind checking bug 1291602?13:13
ubot5bug 1291602 in qtdeclarative-opensource-src (Ubuntu) "com.ubuntu.developer.andrewsomething.stackbrowser_stackbrowser_0.2.2 causes qmlscene to segfault under Qt5.2" [Undecided,Confirmed] https://launchpad.net/bugs/129160213:13
Mirvtsdgeos: so with 5.2 we have that ^ + we might need to use qtwebkit 5.1.1 instead of the newer one because of OpenGL game problems13:27
mzanettiSaviq: Qt.inputMethod.hide() doesn't do anything. Is that intentional to not allow anything to interfere with it?13:38
mzanettior well, do you know something about this?13:38
Saviqmzanetti, sounds like a bug to me, check with bfiller13:39
mzanettiSaviq: ack, thanks13:39
tsdgeosrsalveti: Mirv: desktop too or phone only?13:43
rsalvetitsdgeos: only tested it with phone13:43
tsdgeosrsalveti: do you know how i get com.ubuntu.developer.andrewsomething.stackbrowser_stackbrowser_0.2.2 on the desktop?13:43
rsalvetitsdgeos: let me grab the click13:44
rsalvetitsdgeos: http://people.canonical.com/~rsalveti/com.ubuntu.developer.andrewsomething.stackbrowser.tar.bz213:45
tsdgeosrsalveti: do you know what provides U1db qml module?13:47
tsdgeosqtdeclarative5-u1db1.013:48
Saviqtsdgeos, yes13:49
Saviqtsdgeos, and you need qtdeclarative5-friends0.2, too13:49
Saviqtsdgeos, and yeah, crashed here on desktop13:49
tsdgeosyep, installing that one now13:49
tsdgeosyep, crash13:50
tsdgeosbinding loop for the model doesn't sound cool13:50
tsdgeosSaviq: so what's more urgent debug this crash or finish the testcase for the other crash i am fixing?13:51
tsdgeosi guess "both" :D13:51
Saviqtsdgeos, unity813:51
tsdgeosok13:51
=== MacSlow|lunch is now known as MacSlow
mhr3kgunn, fix your mic! :)14:11
kgunni know right....14:16
=== pete-woods1 is now known as pete-woods
Saviqkgunn, `rm ~/.config/pulse; pulseaudio -k`14:23
Saviqbtw, /me in vUDS session, might not make standup14:24
kgunnSaviq: ack14:29
mzanetti@unity: standup14:31
kgunnhang on...deleting my pulseaudio cfg's14:32
Saviqnice timing ;D14:43
mterrySaviq, just saw you join as I quit, yeah  :)14:43
kgunnhe planned it14:46
kgunndandrader: curious, can you list the branches & any instructions to see what you've done ? (per your description this morning with live previews)14:47
kgunnvery cool btw14:47
dandraderkgunn, here https://docs.google.com/a/canonical.com/document/d/1IiHBDIW_e0qnGt-po1D2z5HJKrNhBwh6pdILeEN2sgA/edit#heading=h.z8j86m70rm14:52
kgunndandrader: thanks...now instead of reading window management docs, i'll be completely distracted replicating this :P14:52
kgunntoo cool14:52
dandraderkgunn, what do you mean by "live previews"? I fear you might be having wrong expectations. :)  Actually, the code changes (cleaner arch., simpler unity8 code) are more exciting than what you see on the screen.14:58
=== gatox is now known as gatox_lunch
dandraderkgunn, the issues I listed there I would like Mir guys to help out with during the sprint. think that's feasible (will there be time for it)?15:04
kgunndandrader: yes!15:06
kgunnperfect15:06
mhr3_Saviq, hm, something wrong with cards in new-scopes-cleanup, searching in apps gives cards with no text15:24
mhr3_Saviq, and works fine with the old new-scopes branch15:25
Saviqmhr3_, checking15:25
CimiSaviq, might be tsdgeos fix for the tabs15:25
Cimiif he added it15:25
SaviqCimi, doesn't sound like it15:26
SaviqCimi, since tabs should only affect header, not cards themselves15:26
Cimiaaah ok15:27
Cimisorry misread15:27
CimiI meant this http://bazaar.launchpad.net/~unity-team/unity8/new-scopes-cleanup/revision/723#qml/Dash/DashContent.qml15:27
Cimifor reference15:27
Saviqmhr3_, terminate called after throwing an instance of 'kj::ExceptionImpl'15:40
Saviq  what():  src/capnp/layout.c++:1822: requirement not met: expected ref->kind() == WirePointer::STRUCT; Message contains non-struct pointer where struct pointer was expected.15:40
Saviqmhr3_, looks like you need to restart scope-registry from time to time ;P15:40
mhr3_Saviq, especially when we do major updates :)15:40
Saviqmhr3_, indeed15:40
Saviqmhr3_, not getting any scopes anyway, so probably not really important ;P15:41
* Saviq purges 5.215:41
mhr3_i'm running 5.2 as well, wfm15:41
dandradermzanetti, did you make a PPA or a handy set of .debs to try out your right-edge-2 work? Yesterday I accidentally triggered the system image update in the N4 where I had built the right-edge-2 stuff15:47
dandraderrebuilding it all again might be a pain...15:47
mzanettidandrader: https://code.launchpad.net/~mzanetti/unity8/right-edge-2/+merge/20479815:47
mzanettidandrader: the description holds a link to a chinstrap folder15:47
dandradermzanetti, great! thanks15:48
mzanettiyw15:48
=== gatox_lunch is now known as gatox
=== dandrader is now known as dandrader|lunch
smbI notice some weird behaviour related to dual-monitor boot on a netbook of mine. Is this something known? So when I boot with an external monitor connected performance is quite bad (i945GME), like using llvm-pipe and compiz draws a lot of cpu (20% even when doing nothing). Booting without external monitor and plugging it in later is performing as expected.16:43
=== dandrader|lunch is now known as dandrader
Saviqtsdgeos, do you have a patch against 5.2 that we could push into our packages already?17:11
tsdgeosSaviq: for the unity thing sure17:12
Saviqtsdgeos, let's17:12
tsdgeosSaviq: http://paste.ubuntu.com/7084035/17:12
SaviqMirv, ↑17:12
tsdgeosthe 5.3 version lives at https://codereview.qt-project.org/#change,8089317:12
tsdgeosawaiting review, i'm not ultra sure if that one is correct since stuff changed in 5.317:13
tsdgeosbut the 5.2 one should be pretty correct17:13
Saviqtsdgeos, thanks17:13
dandrader mzanetti. flashed a device, installed these packages, and now I get this when trying to run unity8: http://paste.ubuntu.com/7085841/17:25
mzanettidandrader: hmm... something might have landed today and I haven't merged it yet17:26
mhr3_Saviq, btw this is what i see in -cleanup + click http://imgur.com/29cprfP17:27
mhr3_not sure if you were able to get it to run in the end or not17:27
Saviqmhr3_, can you paste some of the messages from console? the "undefined to int" seems most interesting17:28
Saviqmhr3_, and no, didn't, UDS took me over17:28
mhr3_here you go http://paste.ubuntu.com/7085862/17:29
Saviqthis is crazy https://www.youtube.com/watch?v=VRgFG90fmHI :D17:30
mhr3_whaaat17:30
mhr3_lol17:30
Saviqmhr3_, ok, nothing obvious, will look tomorrow morning17:31
mhr3_k, me out17:31
mzanettidandrader: so design wanted to do the "autoCompleteGesture" in case the user did a one way gesture, but they did not want the speed to have any impact17:35
mzanettido you think the EdgeDragEvaluator is any good for that?17:35
mzanettitried to use it and struggling a bit making it to do exactly that17:35
dandradermzanetti, nah, leave it out17:37
Saviqmzanetti, check out the youtube above ↑, you guys there around the Alps are crazy :P17:38
Saviqmust be oxygen deprivation or something17:38
mzanettiSaviq: yeah, just seen it. hehe, yeah. I could imagine some of my school mates playing this :D17:38
dandradermzanetti, although I do think it should follow the same autocomplete vs. rollback logic that the panel uses, for consistency17:38
tsdgeosSaviq: you around? have 5.0 installed somewhere?17:39
Saviqtsdgeos, yes17:39
Saviqtsdgeos, desktop or device?17:39
tsdgeosSaviq: desktop preferably17:39
Saviqtsdgeos, have both17:39
tsdgeosSaviq: ok, let me send you a tar17:39
tsdgeosSaviq: sent17:41
mzanettidandrader: fair point... it is a bit different tho... while on the right edge you want to flick the app over the whole screen, that's not really what you want to do for the launcher as that's a "short" gesture17:41
mzanettidandrader: I do agree with the general thing you're saying tho17:41
Saviqtsdgeos, no, does not crash17:42
Saviq<Unknown File>: SyntaxError: Unexpected end of input17:42
tsdgeosoh really?17:42
tsdgeosthat's weird17:42
tsdgeosSaviq: http://paste.ubuntu.com/7085942/ ?17:42
mzanettidandrader: does this work for you? https://code.launchpad.net/~mzanetti/unity8/right-edge-2/+merge/204798/comments/49700817:43
Saviqtsdgeos, yup17:43
tsdgeosSaviq: running over stackbrowser.qml, right?17:43
Saviqtsdgeos, yes17:43
tsdgeosok17:43
tsdgeossomething defintely broke then17:43
tsdgeosthis crashes in 5.2 and 5.317:43
dandradermzanetti, yes17:44
mzanettiok, cool17:44
Saviqugh, jslint is not happy with that file17:44
tsdgeosSaviq: which of the 3?17:45
Saviqtsdgeos, the .js oen17:45
Saviqone17:45
tsdgeosSaviq: still shouldn't crash17:45
Saviqtsdgeos, well, it doesn't17:45
tsdgeosit does :D17:45
Saviqtsdgeos, on 5.0 I mean17:45
Saviqtsdgeos, srue17:45
dandraderSaviq, interesting video, btw17:47
Saviqdandrader, indeed :D17:47
mzanettiits unbelievable ho fast they can kick that ball, isn't it?17:48
jetsaredimanyone alive in here?18:05
jetsaredimtrying to figure out how to get the notification area to only show up on one screen of a dual screen config18:05
jetsaredims/screen/monitor18:05
Saviqtsdgeos, yeah, no crash, and actually no error any more18:24
Saviqjetsaredim, best file a bug18:24
tsdgeosSaviq: simplified even more to http://paste.ubuntu.com/7086183/ + .js file18:24
Saviqtsdgeos, yup18:25
tsdgeoswill bug simon tomorrow18:25
dandraderwhere does MathUtils come from?18:49
Saviqdandrader, it's ours I think18:56
Saviqdandrader, or UITK actually18:56
Saviq./modules/Ubuntu/Components/mathUtils.js18:56
dandraderSaviq, right18:57
dandraderit's weirs that you have to "impot Ubuntu.Components" to get MathUtils...18:57
dandraderweird18:57
dandradermzanetti, can't build your branch: http://paste.ubuntu.com/7086394/19:08
dandradermzanetti, unity-scopes-shell/activate-appid-2, I mean19:09
mzanettidandrader: that was merged yesterday night19:09
mzanettidandrader: so if you flashed today you might not need that one any more19:09
Saviqdandrader, that branch is in trunk already19:09
dandraderah, good. will build trunk then19:09
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
Saviqdandrader, if it's in trunk, means it's in packages, so no need to build :)20:26
dandraderSaviq, I thought so too, but built it anyway just in case. :)20:26
Saviqdandrader, since CI train, everything that's below the changelog commit (like the top of https://code.launchpad.net/~unity-team/unity-scopes-shell/trunk) is in the package, unless it got reverted, when the package version will have .is.the.actual.version.in.trunk20:28
CimiTrevinho, ping21:03
bschaeferCimi, Trevinho is on vacation for once :)21:08
bschaeferhe'll be back on Monday21:08
Cimibschaefer, he'll reply :P21:08
bschaeferCimi, he will, hes bad at vacations haha21:08
Cimibschaefer, I just wanted to propose a small fix, don't show LIM when resizing a window21:10
Cimibschaefer, you can resize from topleft corner and see LIM appearing/disappearing, looks like a glitch21:10
* bschaefer attempts21:11
Cimisame thing when moving21:11
Cimilike, LIM shouldn't change state when moving or resizing21:11
bschaeferCimi, hmm my LIM seems fine when doing that21:11
Cimibschaefer, move faster21:11
bschaeferCimi, maybe hes fixed it in trunk unity...or compiz21:12
bschaeferCimi, as everything seems to be staying in place and no glitching21:12
Cimibschaefer, I haven't looked at the code, but I believe LIM is shown on entering mouse event21:12
Cimiand hidden on leaving21:12
bschaeferCimi, oo there it is21:13
Cimiso if you move faster than the resize, and your mouse enters21:13
bschaeferCimi, i was looking at the window buttons for some reason haha21:13
Cimiyou show and hide21:13
Cimiah ok :)21:13
bschaeferCimi, should be a pretty easy fix when dragging21:13
bschaeferto hide21:13
CimiI believe they're both easy fix21:13
bschaeferCimi, sorry got kicked off the Internet, not sure if you saw what i said last... but ill get a bug up for Trevinho21:20
bschaeferand assigned to him :)21:20
bschaeferCimi, if you want to pay attention to the bug :): https://bugs.launchpad.net/ubuntu/+source/unity/+bug/129222921:27
ubot5Ubuntu bug 1292229 in unity (Ubuntu) "Decor flashes between LIM and title when moving window or resizing" [Medium,Triaged]21:27
=== salem_ is now known as _salem

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