/srv/irclogs.ubuntu.com/2016/03/14/#ubuntu-unity.txt

=== Saviq_ is now known as Saviq
tsdgeosMirv: ping08:51
tsdgeospstolowski: rebuild silo!09:00
Mirvtsdgeos: pong09:02
tsdgeosMirv: how convincing do i have to be to get https://codereview.qt-project.org/#/c/151559/ into our packages?09:03
Mirvtsdgeos: filing a Launchpad bug would be enough09:10
pstolowskitsdgeos, 71?09:12
tsdgeospstolowski: filters, 54?09:13
tsdgeosMirv: ok, doing09:13
pstolowskitsdgeos, ok, I can, just remember Saviq has a new one with filters in it and plethora of other unity8 stuff (silo 41)09:23
tsdgeosah yeah09:23
tsdgeosi just was going to tell you that09:23
tsdgeosi was just reading his email :D09:23
tsdgeoshttps://requests.ci-train.ubuntu.com/#/ticket/110509:23
pstolowskitsdgeos, do you need silo 54 still for any testing?09:23
tsdgeosguess not09:23
tsdgeoscan anyone retrigger that silo or has to be mzanetti?09:23
mzanettiI'm here09:24
tsdgeosmzanetti: can you rebuild https://requests.ci-train.ubuntu.com/#/ticket/1105 ?09:24
pstolowskitsdgeos, i could but don't want to get in the way, not sure what's happening to that silo09:24
tsdgeosjust merged a few branches09:24
mzanettitsdgeos, fwiw, everyone that has approved lander status can. so you can do too09:24
mzanettibut no prob. I'll do09:24
pstolowskitsdgeos, will keep 54 around for a while just in case we need to test/fix something quickly09:24
tsdgeospstolowski: k09:24
tsdgeosmzanetti: k09:25
tsdgeosman Saviq goes away for a day and the CI already broke :D09:30
* tsdgeos retriggers all jobs09:30
mzanettihaha09:33
tsdgeosMirv: https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/155684209:35
ubot5Launchpad bug 1556842 in qtdeclarative-opensource-src (Ubuntu) "Backport patch so that forceLayout calls layout on more occasions" [Undecided,New]09:35
Mirvtsdgeos: thank you09:37
=== shuduo is now known as shuduo-afk
pstolowskimzanetti, please let me know when you guys sort out the issues Saviq mentioned in the email & silo 41 is ready for proper testing (i'd like to check filters once again in this silo)10:52
mzanettipstolowski, ack10:53
tsdgeosmzanetti: do you know who are we supposed to tell about all those errors on our CI? they seem like network unstability11:10
mzanettitsdgeos, if it's something in the surrounding network, I guess opening a ticket is way to go.11:11
mzanettihowever, there's a #jenaas irc channel11:11
mzanettitry asking there if others have the same issues11:11
mzanetti#jenkaas11:11
mzanettipstolowski, that dependency issue is not fixed yet, however, the silo is built and the code should be the final edition. you can test the scopes stuff12:12
pstolowskimzanetti, k thanks12:14
dandraderSo qtmir got infected again "799 tag(s) updated"12:26
* dandrader runs script on it12:26
=== alan_g is now known as alan_g|lunch
mzanettitsdgeos, pstolowski, is there anything from a user point of view to be tested in regard to filters?13:14
tsdgeosmzanetti: not *strictly* filters, but it comes with a new department navigation13:16
tsdgeosso the click scope navigation is different13:17
tsdgeoss/navigation/departmnet navigation/13:17
mzanettiright... departments vanished13:17
tsdgeosthey did not13:17
tsdgeosthey are on the serach popup13:17
mzanettiah13:17
mzanettiack. all clear13:17
tsdgeosmzanetti: filters there's a test scope that pstolowski has around that can be usde for testing13:18
mzanettitsdgeos, there seems to be sonething odd in the pressed highlight13:18
tsdgeosbut afaik there's no "real" scope using them yet13:18
mzanettitsdgeos, sent you a pic via telegram13:20
mzanettilol13:20
mzanettiPEBKAC13:20
mzanettior maybe not13:21
mzanettiI'm totally confused13:21
tsdgeosi'll flash and try to reproduce13:21
pstolowskimzanetti, yeah, as tsdgeos says. atm we should check if we haven't regressed. only my test scope uses filters13:24
tsdgeosmzanetti: silo 41, right?13:24
mzanettitsdgeos, yes13:27
pstolowskiwhat is 'pressed highlight? i guess it's not ubuntu store highlight?13:32
mzanettipstolowski, http://imgur.com/BOGFJ2G13:33
pstolowskimzanetti, hmm how did you do that?13:34
mzanettipstolowski, press, drag a little and release13:35
mzanettitsdgeos, cimi, pstolowski: I believe this is a bug of the old ListItems. Have seen this in other places too13:35
mzanettiwe might want to consider moving to the new ListItem?13:36
pstolowskimzanetti, ok, got it. yeah, i don't think it's related to our changes13:36
tsdgeosi can't repro :/13:37
tsdgeosnow13:37
mzanettitsdgeos, big effort to move to the new ListItem?13:38
mzanettiI'm afraid the old ListItems won't ever get fixed any more13:38
mzanettiwhich one are you using? ListItems.Standard ?13:38
pstolowskimzanetti, tsdgeos if this is a lot of work then i suggest we log a bug and fix that separately13:39
pstolowskimzanetti, i found one minor issue with music scope, will try to fix that13:41
tsdgeosmzanetti: i object to blocking on that, that code is the same it was before this, so the bug already existed13:42
* tsdgeos reflashes without the silo to try to reproduce13:42
mzanettiodd, I can't repro on my dogfooding (rc-proposed) phone13:43
mzanettibut I figure the code of the actual list should not have changed13:44
mzanettiweird thing is, once I have an item in that state, I can't even click it any more13:44
mzanettiand not recover from it13:44
tsdgeosactually13:45
tsdgeosi think i changed from ListItem.Standard to ListItem13:45
mzanettiwell, I can recover by clicking the cancel button13:45
tsdgeosso it's the new one that regresses it seems13:45
mzanettioh really13:45
mzanettiwell, not denying that13:45
tsdgeosyep13:46
tsdgeossee https://code.launchpad.net/~aacid/unity8/optionselector_filter_visual_improvements/+merge/27713813:46
tsdgeosline 12813:46
tsdgeosmeh13:46
tsdgeosso can't object now13:46
tsdgeosshould not have trusted the sdk people :D13:46
tsdgeosor whoever convinced me to use the new stuff13:46
mzanettilol13:48
mzanettiwell, usually the new ListItem is quite a lot better than the old ones13:48
mzanettijust tried to create a standalone testcase13:48
mzanettiit doesn't seem to have a pressed highlight13:48
tsdgeosmzanetti: yeah no idea where that highlight comes from euther13:55
tsdgeosmzanetti: seems to be a default behaviour13:56
mzanettitsdgeos, yeah, but I just created a standalone example at doesn't do it...13:57
tsdgeosmzanetti: did you set a color: ?13:58
mzanettiah... it only does when a onClicked handler is assigned13:58
tsdgeosah wait i'm not either13:58
mzanettiand I can reproduce the bug13:59
mzanettiwith this: http://paste.ubuntu.com/15384113/13:59
mzanettiok... definitely an SDK issue then13:59
mzanettitsdgeos, we might want to try to use ListItems.Empty for now then as I don't believe the SDK will manage to do another landing before OTA-1014:00
mzanettilemme talk to zsombi14:00
mzanettioh, he's here in this channel14:00
zsombimzanetti: ?14:00
mzanettifound a bug in ListItem14:00
mzanettizsombi, this snippet: http://paste.ubuntu.com/15384113/14:01
mzanetti(the onClicked handler is the important bit)14:01
mzanettizsombi, when that's around, it will have some onPressed highlight14:01
mzanettithat gets stuck and makes the item not interactive any more14:01
mzanettireproducible with the above snippet14:01
zsombimzanetti: with touch or mouse?14:02
mzanettiboth14:02
=== alan_g|lunch is now known as alan_g
mzanettijust run the above in qmlscene, then start dragging the list with mouse/touch but only for a tiny bit and release it again14:02
mzanettieventually the pressed highlight will get stuck14:02
mzanettiit is considerably easier to trigger it with thouch14:03
mzanettibut I've managed with mouse too14:03
zsombimzanetti: ok, so I have to start dragging it a bit...14:04
mzanettiyes14:05
mzanettizsombi, easier if you longpress and then drag a bit14:06
zsombimzanetti: ok, then the long press messes it in14:07
mzanettinot sure what the exact trigger is14:07
mzanettibut with the new scopes departments in silo 41 I am triggering all the time when trying to naviage through it14:07
=== dandrader_ is now known as dandrader|afk
tsdgeosmzanetti: so do i go back to listitem.standard ?14:12
mzanettitsdgeos, this fixes it for me: http://paste.ubuntu.com/15384174/14:13
mzanettitsdgeos, maybe add a fixme. lemme report a bug for the sdk14:14
mzanettizsombi, tsdgeos: https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/155697114:17
ubot5Launchpad bug 1556971 in ubuntu-ui-toolkit (Ubuntu) "ListItem pressed highlight gets stuck when scrolling" [Undecided,New]14:17
tsdgeosmzanetti: is it ok if i make it a new branch on top?14:17
mzanettiyeah, I guess14:19
* mzanetti contiues with silo testing14:20
mzanettitsdgeos, please test that patch too... not sure if it breaks other things or so...14:20
tsdgeosyes14:20
=== dandrader|afk is now known as dandrader
tsdgeosmzanetti: https://code.launchpad.net/~aacid/unity8/listitemworkaround/+merge/28893314:38
mzanettitsdgeos, thanks14:50
dandradermzanetti, similar to your findings on mock issues: https://code.launchpad.net/~dandrader/unity8/duplicateMirSurfaceSignals/+merge/28893815:18
mzanettidandrader, heh, yeah... those hit in the darkest moments15:19
=== maxb_ is now known as maxb
mzanettipstolowski, hey, would you confirm that the OSK is a bit odd with silo 41?15:26
mzanettiI frequently hit the key above the one I try to press15:26
mzanettihmm... went away after a reboot now.. but it was hitting me on 2 boots already15:27
mzanettihave it again15:28
mzanettidafuq15:28
pstolowskimzanetti, doh, you're right. doesn't happen too often15:28
pstolowskibut i just reproduced15:28
dandradermterry, did "make tryTutorial" ever worked?15:40
mterrydandrader, I think so?15:40
dandradermterry, crashes with a mysterious error now. no idea what's going on...15:41
mterrydandrader, in trunk or tutorial-redesign?15:42
dandradermterry, trunk. hold a sec, something likely wrong with my build. am rebuilding from scratch15:42
mterrydandrader, you see a crash in trunk or tutorial-redesign?15:50
dandradermterry, there's something wrong... "make tryShell" is also crashing with trunk. Can you reproduce it?15:51
mterrydandrader, yeah...  "ASSERT failure in QList<T>::operator[]: "index out of range", file /usr/include/x86_64-linux-gnu/qt5/QtCore/qlist.h, line 51815:51
mterryAborted15:51
mterry"15:51
dandradermterry, yes, that's what I get15:52
dandraderso something broke a number of tryFoo targets in the latest release I suppose15:52
mzanettipstolowski, I'm afraid I can't repro that issue with trunk15:55
pstolowskitsdgeos, ^15:55
mzanettipstolowski, but I've really no clue atm which branch could be the culprit15:55
dandradermterry, got a backtrace. there's MirSurfaceItem::setSurface in it. weill keep digging15:55
tsdgeosdandrader: mterry: that's fixed by me15:55
tsdgeosdandrader: mterry: https://code.launchpad.net/~aacid/unity8/makeMakeTryOrientedShellWork15:55
dandradertsdgeos, ah, great! will try that15:56
pstolowskimzanetti, we have various 'kbdLayout' MPs in that silo..?15:56
mzanettipstolowski, they should only affect the physical keyboard15:56
mzanetti"should"15:57
tsdgeospstolowski: mzanetti: yeah osk weird, sure it's silo 41?15:57
pstolowskiyeah.. i don't see have filters could affect osk15:57
pstolowski* how15:57
* tsdgeos reflashes withous silo 4115:57
mzanettitsdgeos, so far I haven't managed to repro with trunk15:57
dandradertsdgeos, eek. huge prereq15:57
* dandrader cherry-picks15:58
tsdgeosdandrader: the prereq has landed ;)15:58
mzanettidandrader, that prereq has landed15:58
mzanettitsdgeos, can you repro the weird OSK too?15:58
tsdgeosmzanetti: yes15:59
mzanettidarn15:59
tsdgeosa gave me q15:59
mzanettior well...15:59
mzanettibetter easy to repro15:59
tsdgeoson places it seemed to me should give me a15:59
mzanettiyep15:59
dandradertsdgeos, mzanetti right :)15:59
tsdgeosmzanetti: we should add https://code.launchpad.net/~aacid/unity8/makeMakeTryOrientedShellWork/+merge/288743 to https://requests.ci-train.ubuntu.com/#/ticket/110515:59
mzanettiok16:00
mzanettiso... now we need to figure what's wrong with the OSK tho16:00
mzanettiltinkl, ^16:02
mzanettiyou might have an idea?16:02
mzanettidon't see how the physical kbd layout would affect it either, but atm I'm still in the dark16:02
tsdgeosyep defenitely seems silo 41 related16:03
mzanettiI wonder if the shell_chrome stuff might do something... it seems to move surfaces between fullscreen and not16:03
pstolowskiwow this silo is massive16:03
mzanettipstolowski, no... 64 was massive... this is just biggish16:03
pstolowski:D16:03
tsdgeosi'm going to random point to the shell-chrome branches :D16:03
* pstolowski too scared to ask what was in silo 6416:03
mzanettipstolowski, 60+ branches16:04
pstolowskiphew16:04
tsdgeosmzanetti: can we create  a second silo and add group of features until we found otu what's the one that seems to cause it?16:05
ltinklmzanetti, OSK broken in that silo? pff but the keymapSwitching stuff has (almost) definitely nothing to do with it16:05
mzanettitsdgeos, you mean a bisecting silo?16:06
tsdgeosmzanetti: yeah16:06
mzanettiyeah, I guess we could... not sure if that's really faster tho16:06
mzanettidednick, hey, you did the most work of the shell_chrome branches, right=16:07
dednickmzanetti: yes16:07
dednickright=true16:07
mzanettidednick, could you imagine them having affect on the OSK surface? it seems input is a few grid units lower than it should be16:08
mzanettidednick, i.e. pressing "a" gives us "q"16:08
mzanettithose branches mess with fullscreen. and the offset seems pretty much the panel height16:08
tsdgeosinteresting point, it only happens the second time the OSK shows16:09
tsdgeosthe first time it's fine16:09
tsdgeosi.e. adb reboot -> trigger OSK -> good -> hide and show OSK -> bad16:10
dednickmzanetti: um, it could do maybe.16:10
dednickmzanetti: the panel hint hack.16:10
* mterry is now on irccloud, hopefully this is more reliable16:11
dednickmzanetti: is there a silo for it?16:12
mzanettidednick, 4116:12
mzanettithis totally explains why the heck I have not been able to enter 2 identical passwords testing the OOBE before :D16:12
dednick:)16:12
dednickdidnt seem to be a problem at mwc though... :/16:13
dednickthat i heard about anyway16:13
mzanettidednick, I'm not saying it is those branches... atm I'm still in the dark and those look like the most suspicious ones to me...16:13
dednickmzanetti: phone mode?16:14
dednickstaged16:14
mzanettidednick, yeah, tested on nexus416:14
mzanettimterry, ltinkl, Mirv could fix the 8.13 bump in the silo. should we or are you relying on the fact that it's bumped to 8.13 already?16:16
mterrymzanetti: hrm, "could fix?"  -- I bumped in that branch because we moved files, so needed a version for debian/control16:18
tsdgeosmzanetti: dandrader: happens on arale here too fwiw16:19
ltinklmterry, what version are we reyling now on with the breaks/replaces?16:19
mzanettimterry, so you manually bump it to 8.13 somewhere now?16:19
dandradertsdgeos, what?16:20
ltinklmzanetti, to 8.1216:20
tsdgeosdandrader: meh, that was for dednick16:20
tsdgeosmzanetti: dednick: happens on arale here too fwiw16:20
mzanettiltinkl, mterry, right, but the silo builds 8.13, not 8.1216:20
mterrymzanetti: I manually bumped in https://code.launchpad.net/~unity-team/unity8/sessionIndicatorForDevices/+merge/28894016:20
mzanettimterry, still 8.12 :D16:21
mzanettimterry, the silo would give us 8.1316:21
mterrymzanetti: oh sorry.  I meant 8.1216:21
mterrymzanetti: did anyone bump to 8.13?16:21
mzanettiso I ask Mirv to clear the packages and we do a clean rebuild. that would give us 8.1216:21
mterrymzanetti: sure?  You mean clear the silo of packages?  my changes shouldn't need that, but ok16:22
mzanettimterry, the OOBE branches did have a 8.13 but it was reverted. however, PPAs reject lower versions and silos seem to be clever enough to just bump it on their own in that case16:22
mterryah16:22
mterrymzanetti: got it.  So no, not relying on 8.1316:22
mterrymzanetti: only relying on 8.1216:22
ltinklmterry, should be still fine right to rely on << 8.12?16:24
ltinkleven if we "skip" it16:24
mzanettitsdgeos, bisecting silo 35 building16:24
mzanettitsdgeos, so far only with the chrome branches16:25
mterryltinkl: yeah16:25
dednickmzanetti: dash is all white on 41 :(16:26
mterrymzanetti: you had started reviewing https://code.launchpad.net/~unity-team/unity8/tutorial-redesign/+merge/288839 back in the day.  Do you want to finish that, or shall I pass that off to dednick?16:27
Mirvmzanetti: so sounds like "go ahead"?16:27
Mirvto remove unity8 packages temporarily from the silo and rebuild in half an hour to see what happens16:27
tsdgeosdednick: what?16:28
dednicktsdgeos: hang on, i'm cleaning device and will check again.16:29
dednickmight have had some of my code on there.16:29
dednickcould have had some occlusion stuff still on it.16:29
Mirvmzanetti: I did that, so please try rebuilding unity8 in 041 in around 30 mins to give LP a bit of time to update whatever it might want to update. I also removed it from train's packages (it will come back since there are MP:s for it). if that doesn't help, then only silo switch would help.16:35
mzanettimterry, if dednick has the time, please16:36
mzanettiMirv, ack16:36
=== francisco is now known as Guest95587
=== pstolowski is now known as pstolowski|bbl
=== dandrader is now known as dandrader|afk
mzanettipstolowski|bbl, about the fix for filters you mentioned, did you push that somewhere? what do I need to rebuild?16:45
mterrydednick: ^ got time for tutorial-redesign when done with geonames?16:50
dednickmterry: i've got to look at problem mzanetti is seeing with osk, but after that yes16:51
mterrydednick: cool thanks16:51
mzanettidednick, silo41 is not installable atm... we had to delete the unity8 packages. but will be there in a bit. I can upload them somewhere to if you want to test now16:52
mzanettior not... apt cache is cleared16:52
dednickmzanetti: seemed to install for me.16:53
mzanettireally16:53
dednickmzanetti: and can confirm shift.16:53
mzanettiok... you might installed it before we nuked the packages then. great16:53
dednickperhaps16:53
mzanettidednick, ok. cool16:53
mterryltinkl: do you have review cycles?  https://code.launchpad.net/~mterry/phablet-tools/tutorial-redesign/+merge/277764 is a relatively small one in our silo that isn't approved yet17:12
ltinklmterry, sure, can do after I've finished the keymapSwitching stuff17:13
mterrymzanetti: tsgeos snuck lp:~aacid/unity8/listitemworkaround into silo 41.  Was that expected?17:18
mterrymzanetti: just want to make sure if I should review it or not17:18
mterrygreyback: and same for lp:~nick-dedekind/qtubuntu/shell_chrome -- looks like it just needs final stamp17:23
greybackyep17:23
mterryoh and u8 too.  OK, you must know all this  :)17:24
mterrygreyback: hey are your eyes fully restored now?17:24
mzanettimterry, yeah, it fixes an issue that I found in silo testing17:24
mterrymzanetti: ok cool, will review then17:25
mzanettimterry, read the linked bug report, it should be clear from that17:25
greybackmterry: things a little blurry still - text not sharp for example. As day goes on, eyes tire more so it gets worse. But am still able to get most things done17:25
mterrymzanetti: only 11 unapproved branches by my count now17:26
mzanettihaha17:26
mzanettimterry, well, most of them have been looked at, tested well before17:26
mterrygreyback: good!  Do you love not needing glasses so far?17:26
mterrymzanetti: yeah some for sure have17:27
mzanettimterry, like all the OOBE, I would just skim over them to see if lukas fixed the things I mentioned in the last review17:27
greybackmterry: yeah, it's a small thing, but it is nice17:27
mterrygreyback: I can feel my eyes slipping.  I'm guessing I'll eventually need glasses / get some laser work done myself17:28
mzanettiso far everyone I've been talking to has said that the laser thing has been the best thing ever17:29
greybackI'll not be that encouraging, there's nothing wrong with wearing glasses. They can make you look more dignified/sophisticated17:31
mterrygreyback: I don't know if my face could stand being any more dignified17:31
greybackmterry: indeed. They'll be carving it into a mountain fairly soon17:32
greybackyou Americans like doing that17:32
mterrygreyback: I've known Mt Rushmore as a done-thing all my life.  But how fricking nuts of an idea that must have been originally.  Shocked they went through with it17:33
greybackyeah. Have always found it a bit much17:33
greybackbut then, why not17:33
greybackthey did it in lord of the rings universe at lot too17:34
mterrygreyback: because it's a beautiful mountain as is!  :)17:34
greybackwill give people thousands of years from now puzzlement17:34
dednickmzanetti: i've fixed that bug with osk. not sure why it was working previously and not now though. It was something specific i added to get it not to do that same issue!17:38
dednickbut seems to have backfired now.17:38
mzanettidednick, oh, cool!17:38
mzanettiI was still prepping the bisecting silo to confirm it's really those branches17:38
mzanettiI guess I can throw that away then17:39
mzanettidednick, which packages to rebuild?17:39
dednickmzanetti: qtubuntu17:39
mzanettita17:39
* mzanetti does17:39
dednickthere must have been a change elsewhere in the mwc code which required it... :/17:39
dednickso it'll probably come back!17:40
mzanettiltinkl, mterry, 8.13 it'll be... not going to completely destroy and re-setup the silo17:45
mzanettiltinkl, mterry, your code should still work with that, right?17:46
mterryyeah17:46
mterrymzanetti: ^17:46
mzanettiok... unity8, indicator-session and qtubuntu builds on the way...17:47
mzanettiactually. Mirv's hack helped!17:51
=== jhodapp is now known as jhodapp|errand
mterryltinkl: indicator-session built in silo, nice18:04
=== pstolowski|bbl is now known as pstolowski
pstolowskimzanetti, still there?19:00
=== jhodapp|errand is now known as jhodapp
=== dandrader|afk is now known as dandrader
mterryMirv: you did something to clear out the 8.13 version of unity8 from the silo 41 ppa?19:17
mterryIt still seems to be in the ppa (though isn't showing up in LP)19:17
mterryspecifically, only for unity8-common and unity8-schemas packages19:18
mterryltinkl: ^ u8 in the silo isn't even building unity8-common and unity8-schemas...19:20
ltinklmterry, no idea why is that happening...19:20
mterryltinkl: ah... because those are built by the amd64 build19:20
mterryltinkl: which failed to upload because 8.13 packages are still somehow in the silo?19:20
mterryMirv: so I guess back to you -- do you know why the 8.13 packages are still in the silo?  What did you do besides press "delete packages" in the LP ui?19:22
mzanettipstolowski, hey19:42
=== boiko_ is now known as boiko

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