=== jono is now known as Guest55880 [02:53] Saviq: someone on G+ says they are experiencing the same slowdown on GNexus === hggdh_ is now known as hggdh [04:20] tsdgeos: qt 5.1.1 is more likely, although without not much testing done yet the only known blocker for 5.1 is the QPA plugin which is not done yet [05:58] morning [06:11] didrocks, hi. Can you tell me if a 0.9.10.0 release of Compiz is planned and imminent ? [06:12] MCR_: should be, we are transitionning Xorg 1.14 today, compiz is next [06:12] why? [06:12] because Sam mentioned something like this and I need to know if trunk is frozen... [06:12] but I have no details on it [06:13] MCR_: better to avoid changing it now [06:13] All I have is this: https://code.launchpad.net/~mc-return/compiz/compiz.merge-thumbnail-improvements/+merge/170548/comments/384428 [06:14] but no further info [06:14] Saviq, ping [06:14] I would prefer we stop evolved compiz for the time being [06:14] mlankhorst, good morning :) [06:14] until we land 0.9.10.0 [06:14] tvoss, pong [06:15] didrocks, do you have an ETA for that? -> a lot of MPs with finished reviews are waiting to land... [06:15] smspillaz, ^^ [06:15] morning [06:15] hi, btw (if you are here), Sam... [06:17] MCR_: as told, once xorg 1.14 lands [06:17] aha, prepare for troubles ;) [06:17] which is what mlankhorst and Mirv are coordinating [06:17] didrocks, https://code.launchpad.net/~unity-team/unity8/packaging-cleanup/+merge/172578 feels ready [06:18] MCR_: I don't handle releases [06:18] first tests with xorg-edgers (still 1.13) on R showed a lot of problems here :( [06:18] Saviq: yeah, +1! qtubuntu-android and libunity are all good :) [06:18] also the new fglrx is slow and buggy [06:19] didrocks, can you have a quick run through over debian/ there [06:19] didrocks, I'll try installing on devices and ask mzanetti to have a look through, too [06:19] didrocks, qu: who is responsible for Compiz in Ubuntu ? [06:19] Saviq: sure, in ~40 minutes good? I want to finish some google-mork for mirv [06:19] MCR_: sil2100 is the stack holder, but he's on holidays [06:20] didrocks, sure [06:20] didrocks, well -> then a freeze does not make sense... [06:20] MCR_: I know that andyrock is testing current trunk [06:20] ah, cool [06:20] MCR_: so don't make it unstable [06:20] I never make it unstable [06:20] didrocks: to be fair, its probably me who will make it unstable :) [06:21] but that's because I always take the hard stuff :) [06:21] 185 branches merged, none reverted is a good quote, no ? [06:21] smspillaz: heh :) [06:21] smspillaz, that is true (with the hard stuff) [06:22] I just can fix what I understand ;) [06:22] but I am learning fast ;) [06:22] smspillaz, we have quite a few branches which just need your gogo... [06:23] MCR_: don't really have time, I'm travelling the next few days [06:23] smspillaz, ok -> we'll wait then I guess :( [06:24] happy travellin' -> if you are near Vienna, Austria I buy you a drink... [06:27] my impression was we just have one Compiz release per cycle, at the end of the cycle... [06:34] Compiz + Unity7 + xmir + xorg 1.14 + proprietary drivers in Saucy -> I'm happy I'm not responsible for that stack... [06:52] I'm out, c ya. [07:55] tsdgeos, can you please have a look at the hud toolbar buttons with the latest ubuntu-ui-toolkit? [07:55] they look crap, right? [07:56] tsdgeos, yeah, not sure what changed [07:56] i think i saw them yesterday in the autopilot tests run [07:56] and thought "must be only this autopillot thing" [07:56] but then thought "but why?" [07:56] and then forgot :D [07:57] tsdgeos, fixed [07:58] tsdgeos, http://pastebin.ubuntu.com/5842753/ [07:58] that was fast :D [08:00] Saviq: you propose it? [08:01] tsdgeos, yeah [08:01] tsdgeos, we need to release (going into daily) [08:01] tsdgeos, but SDK crapped out on us ;) [08:01] he he [08:16] tsdgeos, https://code.launchpad.net/~saviq/unity8/abstract-button/+merge/172966 [08:16] tsdgeos, wth was the env var to print draw calls? [08:16] hmmm [08:17] i know how for qpainter, but not sure i have ever known the "qml scene" one [08:17] damn [08:25] Saviq: this are all the getenv that qtdeclarative has http://paste.kde.org/~tsdgeos/788726/ [08:25] might be QSG_RENDER_TIMING now [08:25] that is 5.1 though [08:26] let me branch to 5.0 [08:26] http://paste.kde.org/~tsdgeos/788738/ [08:26] 5.0 ones [08:27] QML_RENDERER_TIMING ¿ [08:28] mzanetti: ping [08:28] tsdgeos: hey [08:29] pos = qMax(qMin(bottomItem->position() - highlightRangeEnd, -maxExtent), -minExtent); [08:29] that's what gives the -38 the thing animates to [08:29] i think you are being tricked by the fact you do bottom up but the list does not and everyone gets confused [08:30] because otherwise not sure how bottomItem is the one causing you problems [08:32] tsdgeos: ? [08:32] not following [08:32] :D [08:32] the listview angle thing [08:32] for the launcher [08:32] tsdgeos: just set inverted to false [08:32] tsdgeos: the issue will be the same the other way round [08:33] ah sorry, i can't copy & paste [08:33] is not bottomItem [08:33] but topItem [08:33] >│1449 pos = qMax(qMin(topItem->position() - highlightRangeStart, -maxExtent), -minExtent); │ [08:34] where did you get this line? [08:34] gdb :D [08:34] http://paste.kde.org/~tsdgeos/788762/ [08:34] this is the problem [08:35] your topitem pos and the highlightRangeStart don't match [08:35] now, not sure where highlightRangeStart comes from [08:35] ok, it's preferredHighlightBegin [08:36] yeah... thats what I'm saying [08:37] the preferredHighlightBegin seems to be set, but not updated to the items [08:37] and on the first interaction with the mouse they are updated, which makes it jump to that semi-folded position [08:42] not sure i follow :D [08:43] oh man [08:43] https://jenkins.qa.ubuntu.com/job/unity-phablet-qmluitests-saucy/406/testReport/junit/%28root%29/ListViewWithPageHeaderTestSection/testInsertItemsBeforeValidIndex/ failed again :/ [08:43] mzanetti: can you detach me that vm so i try running the test there manually? [08:44] Saviq, thanks for the mail. [08:44] nic-doffay, cheers, makes sense? [08:46] Saviq, yep. Spoke to Florian about themes yesterday, so I'm busy implementing those and splitting up the component. Looks like I'll get to this after. [08:52] nic-doffay, cool [09:52] Saviq: ping [10:00] dednick, pong [10:01] Saviq: about the indicators-client. Did you mean on the desktop there was only datetime? [10:01] dednick, no, on device [10:02] Saviq: hm. no, that's not expected [10:02] dednick, the panel icons were there [10:02] dednick, but pages were empty (not even the "Empty!" labels) [10:02] Saviq: i c. i'll take a look [10:03] tsdgeos: this reproduces the issue: http://paste.ubuntu.com/5843019/ [10:03] Saviq: do you remember if the messaging/network icon showing the correct icon? [10:04] mzanetti: cool, that's so simple! [10:04] dednick, it seemed to, yes [10:05] dednick: Saviq: http://s-jenkins:8080/job/generic-mediumtests-runner-testing/2/ [10:06] dednick: does that work for you? [10:07] mzanetti: when build and run from local, yes. [10:07] dednick: ? [10:07] mzanetti, nice! [10:08] mzanetti: looks like it's trying to find it in the build folder, but doesnt ap use a installed package? [10:08] i ean ap o jenkins [10:08] urg [10:08] mzanetti: i mean autopilot on jenkins [10:08] dednick: yes [10:09] dednick: it installs *.deb from our repo and calls "autopilot run indicators_client" without a build dir [10:09] dednick: your tests would need to be prepared for that [10:09] mzanetti: yeah. i see it doesnt check for installed path === MacSlow is now known as MacSlow|lunch [11:43] Saviq: have good and bad news :D [11:51] Saviq: basically the render loop of 5.0 is "not that good" [11:51] and it's what leads to those random test failures in the "quite slow" CI VMs [11:52] i updated the CI VM to 5.1 and *boom* the failures are go [11:52] ne [11:59] tsdgeos: yay! thats quite good news I'd say === MacSlow|lunch is now known as MacSlow [12:00] tsdgeos: Do you think we should disable the failing tests until upgraded to 5.1? === alan_g is now known as alan_g|lunch [12:15] tsdgeos, yeah, I say that's good [12:16] tsdgeos, another reason to switch to 5.1 [12:17] didrocks, it's in, we can has daily release? :D [12:18] dednick, sorry about that... the packaging cleanup branch just went in [12:18] dednick, and there's quite some conflicts [12:18] dednick, ping me if you have issues merging [12:19] dednick, one important thing to do is to change the installation path to ${SHELL_PRIVATE_LIBDIR} [12:19] dednick, see https://code.launchpad.net/~unity-team/unity8/packaging-cleanup/+merge/172578 for the diffs [12:21] Saviq: queuing your request :p [12:22] didrocks, but! but... I've been in the queue for two weeks already! ;) [12:22] Saviq: hem, you do had stuff to do, don't tell you were waiting on me for 2 weeks :p [12:23] didrocks, of course I didn't, I was blocked on that! ;P [12:23] * Saviq biab, trying to fix the overheating laptop... [12:28] Saviq: ok, i'll take a look [12:30] didrocks / cyphermox: question - can you tell me where it is (a PPA, presumably?) that the releases of libusermetrics go? [12:30] pete-woods: it's going to distro [12:30] presumably the guys working on unity are able to get it from somewhere [12:30] oh, cool [12:31] * pete-woods is both pleased and scared :) [12:31] :) [12:40] pete-woods, i had some trouble compiling it, mostly because of missing #include [12:40] dednick: the multi-job capable mediumtests runner is live now. Please let me know when I can/should enable the indicators_client test suite [12:41] mzanetti: ok. will do [12:42] pete-woods, applied http://paste.ubuntu.com/5843395/ but it's a bit overzealous, run a few sed's on the tree :) [12:46] mhr3: are you compiling it on raring? [12:47] (i.e. backporting it) [12:47] mhr3: because if so you will need to build qdjango from source, too [12:48] using the source package from saucy [12:48] pete-woods, nope, i just grabbed the one from repos, worked fine [12:48] pete-woods, and yes, raring [12:49] mhr3: that's very strange, qdjango 0.2.6 is built against qt4 and would normally segfault if you mix it with a qt5 library [12:50] mhr3: it would compile happily, though, leading to a false sense of success [12:51] pete-woods, well, i didn't actually run the daemon, just wanted to compile unity8 and for that needed .pc file of usermetrics [12:51] mhr3: well I would advise if you actually want it to run, to also build the newer qdjango :) [12:54] pete-woods, unity8 runs, that's important ;) [12:54] mzanetti: well, the thing is that i am not really sure which are the failing tests [12:54] potentially it's all of them i'd say [12:55] since what happens is that sometimes i say "update" [12:55] and the "update callback" doesn't come back [12:56] tsdgeos: I for one would be ok with disabling all of them until we switch to 5.1. As long as we have tests ready to be enabled and know the reason why they fail for now... [12:57] mzanetti: not really sure what i prefer tbh [12:57] tsdgeos: you mean not switching to the new LVWPH until we have Qt5.1? [12:58] well we have lrady swithced :d [12:59] tsdgeos: whats the other option then? [13:00] keeping the tests randomly failing until 5.1? [13:00] ostrich-style [13:00] :D [13:01] lol [13:01] tbh don't know [13:01] tsdgeos: given that the switch to qt5.1 is still at least a month away, a -1 on that one from me === alan_g|lunch is now known as alan_g [13:02] mzanetti: sure [13:02] i'm just thinking what is better [13:02] revert LVWPH [13:02] or just comment the tests [13:02] well, it does improve things already [13:03] that's the question [13:03] does it? [13:03] ah, you're not sure on that one? [13:03] i mean, the failures we see on the tests [13:03] does this issue also happen with usage or only with tests? [13:03] we could see in real life [13:03] i don't see them [13:03] ever [13:03] and i think it's just because the "old" scheduler in 5.0 has issues with slow cpus [13:04] but you know how threads work [13:04] whats the symptoms? a visual glitch or a complete lockup or crash? [13:04] the fact that you don't see the problem doesn't mean it's not there [13:04] the symptom is that a redraw of the scene is not triggered when it should [13:04] probably on real life it's not so bad [13:04] since you'll have another one soon [13:04] but on tests [13:05] where i do only 2 things [13:05] it's very "obvious" [13:05] so a visual glitch... and most likely even a small one [13:05] since sometimes the 2nd thing doesn't trigger the update [13:05] on real life you'd have something else coming over and updating and you probably wouldn't notice anything [13:05] my vote goes to keeping LVWPH and temporarily disable tests [13:06] but probably Saviq is the one to decide ^ [13:07] * dandrader jumps in the middle of the conversation [13:07] Saviq: so we now have a separate plugin and modules folder, but they end up going into the same folder anyway. sense? [13:07] mzanetti, you mean temporarily disabling autopilot test or qmluitests? [13:07] dandrader: qmltests for LVWPH only [13:07] ah, ok === _salem is now known as salem_ [13:27] can you connect to mumble guys? [13:28] * Saviq reads backlog [13:28] * tsdgeos can't [13:28] * MacSlow can't connect to mumble-server anymore [13:28] I can't. [13:28] dednick, yes [13:28] dednick, they end up on QML2_IMPORT_PATH [13:28] dednick, we could just as well split it up, but there's probably no reason to [13:29] dednick, or we could merge all of them into "qml" in our tree [13:29] i'm in! [13:29] Saviq: yeah, i wasn't seeing the purpose of the split in the tree [13:29] seeing as they're merged in the tests/mocks as well. [13:29] s/QML2_IMPORT_PATH/SHELL_PRIVATE_QML_IMPORT_PATH/ [13:31] nic-doffay, standup [13:38] Saviq trying to get mumble working... [13:41] pete-woods: yeah, you should have had a release at least yesterday [13:42] pete-woods: and I'm going to go check it out for today [13:43] rsalveti: I think we're missing a chewie release for the changes I made yesterday and renato's device-switch branch [13:44] * cyphermox flashes today's image [13:47] Saviq: i dont think there is anything except the plugin folder at the moment. The module folder only exists in indicators, so i could just move it back to plugins fairly easily. [13:48] dednick, we should rename it to "qml" probably [13:48] dednick, to just be consistent with the installation path [13:48] Saviq: ok [13:50] Cimi, https://code.launchpad.net/~nicolas-doffay/+junk/option-selectors [13:51] nic-doffay, thx [13:57] nic-doffay, what are you working on? [13:57] nic-doffay, a new component? [13:58] Cimi, http://design.ubuntu.com/apps/building-blocks/option-selector [13:58] got it [14:01] Cimi, yeah OptionSelector [14:02] dednick, why do you reckon it's using fakes on ./run? [14:03] Saviq: sorry, i think it was just my branch. [14:03] dednick, other than fake apps and LightDM, which is on purpose? [14:03] dednick, k [14:04] Saviq: forgot to add the modules path to the import overrides. === tvoss is now known as tvoss|errands [14:16] Saviq: first unity8 daily release completed! [14:19] cyphermox: thanks, it was just that someone was asking where they could get hold of one of the libraries, and I looked in all the PPAs, but couldn't see it. hadn't realised it was going to distro! [14:24] didrocks, !!! [14:25] didrocks, so now's the time to read through https://wiki.ubuntu.com/DailyRelease/FAQ right? ;) [14:25] Saviq: I guess it is time :) [14:26] Saviq: the integration tests are not running, tell me if/once they are ready so that I can add them [14:26] didrocks, it's still in my browser-tab-queue [14:26] (for unity-api, unity8 and the whole stack) [14:26] great! [14:26] didrocks, there are some integration tests, we could have them enabled, really [14:26] Saviq: as you wish, just tell me :) [14:26] pete-woods: np [14:26] didrocks, we'll be growing them soon [14:26] didrocks, but enabling them now is a good idea as any [14:26] greyback gone? [14:27] Saviq: if you prefer to wait, that's fine as long as it's landing to a PPA [14:27] didrocks, I don't see any advantage in waiting [14:28] Saviq: ok, so installing unity8-autopilot and running "unity8" autopilot target, anything else? === alan_g is now known as alan_g|tea [14:28] didrocks, nothing else now [14:29] mzanetti, lol looks like I terribly broke the calendar somehow [14:30] Cimi: indeed [14:30] mzanetti, I swear it was working before publishing the branch [14:30] ! [14:30] :) [14:31] mzanetti, it's onCurrentDateChanged [14:31] mzanetti, it conflicts with onCurrentIndexChanged [14:31] mzanetti, because they both do stuff... [14:33] Cimi: why do you reset the model anyways onCurrentDateChanged? [14:33] Cimi: shouldn't you just set the currentIndex in there? [14:33] Cimi: unless its outside of min/may. In that case print a error message and do nothing [14:33] s/may/max/ [14:35] didrocks, ah wait, we still depend on indicators from https://launchpad.net/~phablet-team/+archive/desktop-deps/ [14:36] didrocks, that dep will be gone next week at the latest [14:36] Saviq: you have conditional code building? [14:36] didrocks, no, it's a runtime dep [14:36] didrocks, it's actually a recommends [14:36] ah, anyway, it's in the next ppa, so not as important as if it was in distro :) [14:37] didrocks, but there is a dep for indicators-client [14:37] didrocks, so the integration tests will fail without that [14:37] mzanetti, yes yes [14:37] didrocks, so we might wait for that before enabling them [14:37] mzanetti, it should change if no element has that month [14:38] mzanetti, I need to rewrite that crap [14:38] Cimi: ok. let me know if you need something or are ready for another review [14:38] Saviq: no worry, it's a 10 minutes thing to enable it, just ping me :) [14:38] didrocks, will do [14:38] mzanetti, I thought code worked [14:39] Cimi: yeah... guess what happened to me with the launcher [14:39] mzanetti, then my plan was to use for cycles to fetch inside the current model and append elements instead clearing all [14:39] on a second review [14:39] mzanetti, at this point I just rewrite all of them into a 'bigga review' [14:39] Cimi: ok [14:47] didrocks, shall powerpc be disabled for ppa:ubuntu-unity/next? nothing builds due to missing Qt? [14:48] Saviq: no, we keep the ppa with the same configuration than distro [14:48] Saviq: it just dep-wait, which is fine [14:48] didrocks, ok [14:48] (I know, the cross is a little bit puzzling) [14:48] but daily releases deals with that [14:48] didrocks, ok [14:48] if one day, we have powerpc, it will wait on it before copying [14:49] as long as it dep-wait, it will treat that as non existant :p [14:57] cyphermox: yeah, didn't create a release because of an annoying bug when using any switch in the network-indicator [14:58] it just goes to 'Empty!' [14:58] ah? [14:58] so if you disable wifi, for example, you can't enable it again [14:58] the switch works fine here [14:58] not related with your change [14:58] actually it could be [14:58] can someone have a look at Merge Req for Bug #1196878 and #1196800 and #1196822 [14:58] bug 1196878 in Unity Videos Lens "Port Unity Lens Video to libzeitgeist2" [Undecided,Confirmed] https://launchpad.net/bugs/1196878 [14:58] bug 1196822 in unity-lens-applications "Port unity-lens-applications to libzeitgeist2" [Undecided,Confirmed] https://launchpad.net/bugs/1196822 [14:58] bug 1196800 in unity-lens-files "Port Unity Files Lens to libzeitgeist2" [Undecided,Confirmed] https://launchpad.net/bugs/1196800 [14:58] rsalveti: I'll test this some more [14:58] has everything landed? [14:59] I mean, in trunk? [14:59] cyphermox: bug 1195787 [14:59] bug 1195787 in touch-preview-images "Network indicator shows empty list after associating" [Medium,Confirmed] https://launchpad.net/bugs/1195787 [14:59] same happens when disabling/enabling wifi [14:59] yeah [14:59] cyphermox: also, noticed that nm is unable to auto connect with my sim card, seems it's trying it too early [14:59] I think it's still the same issue [14:59] ah, ok [15:00] can you ship me logs [15:00] sure, let me just do our stand up first [15:00] like, file a bug for that and add syslog so I can see what happens? [15:02] mzanetti, tsdgeos, greyback, dandrader who wants to try (and ideally has the time today still) and strip the shell (or build something minimal) that shows this bug https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1197801 ? [15:02] Launchpad bug 1197801 in Ubuntu UI Toolkit "Unity8 is hogging the CPU when idle" [Critical,In progress] [15:03] Saviq: do we have any clue of what's causing it? [15:03] tsdgeos, we've narrowed it down to http://bazaar.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/trunk/revision/570 [15:04] tsdgeos, Kaleo_ is on it, but he's on raring, so can't get the shell to run [15:04] :/ [15:07] rsalveti: I think today's image also didn't take the latest NM [15:07] Saviq, well, I can try. have 75% of my work day ahead of me still [15:07] Saviq: i'm a bit busy with mzanetti's listview snap bug, if someone else can take it, it'd be great [15:08] dandrader, I'll try and get as close as possible (my current plan is to just let him use -f) [15:08] I'm a bit busy too, but can try later on this evening if I have time [15:09] Saviq, erm, so you will do it as well? does it make sense to have 2 persons working on the very same thing? [15:09] dandrader, I mean that I'll get as close as possible [15:09] dandrader, and let you take over in 20 mins or so [15:09] Saviq, ah, ok. let me know when you're done [15:09] dandrader, actually I might've just finished :d [15:09] that was fast! [15:10] dandrader, need to see on a clean(er) machine === salem_ is now known as _salem === _salem is now known as salem_ === salem_ is now known as _salem === _salem is now known as salem_ === alan_g|tea is now known as alan_g [15:40] cyphermox: hm, that's weird [15:41] so I'm not dreaming? [15:41] I do upgrade and get it though [15:41] * rsalveti checking [15:41] network-manager 0.9.8.0-0ubuntu13phablet1 [15:41] seems to be the latest [15:42] yeah [15:42] somehow the image build didn't catch it yet [15:42] perhaps it was just a tad too late to get published [15:43] well, that's from the manifest [15:43] http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/saucy-preinstalled-touch-armhf.manifest [15:43] oh [15:43] then I didn't get the right image [15:43] but we'll respin another image in a few anyway [15:43] ah [15:43] I didn't get the right image, that's why [15:43] I had 03.1 [15:44] ok, indeed indicator-network gets empty when you deselect a wireless item [15:47] Saviq: hey, how long do you think it'll take to fix bug 1197801? [15:47] bug 1197801 in Ubuntu UI Toolkit "Unity8 is hogging the CPU when idle" [Critical,In progress] https://launchpad.net/bugs/1197801 [15:47] just trying to figure out if we should wait it to be fixed before triggering a new image [15:47] rsalveti, probably not [15:47] rsalveti, Kaleo_ is looking into it [15:48] rsalveti, I'm preparing a trimmed-down shell for him to reproduce on raring [15:48] right [15:48] rsalveti, and he'll revert if failed [15:48] Kaleo_: just update to saucy, lazy boy :P [15:48] rsalveti: no [15:48] rsalveti: we support app developers [15:48] rsalveti: app developers don't use saucy [15:49] Kaleo_: I know, but I like to bug you [15:49] rsalveti: in that case; come here [15:49] rsalveti: we are waiting for you :)$ [15:49] Kaleo_: indeed! [15:49] and now, lunch [15:49] see you soon [15:50] yeah, let me eat something as well === boiko_ is now known as boiko === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader === alan_g is now known as alan_g|life [17:12] mzanetti, I kinda (maybe) fixed it [17:12] mzanetti, but tomorrow I'll continue working on the for things... [17:12]