=== Malsasa_ is now known as Malsasa [05:39] bregma: it seems even the newest 14.04 LTS compiz update is stopped :( http://people.canonical.com/~ubuntu-archive/phased-updates.html I remember the last time there was a big line of Problems, now after the blacklist adding there's just one [05:39] bregma: but you'll need to chat with release team on how to proceed if it's wanted to be spread [07:42] Saviq: tsdgeos: I've made a sidenote about https://codereview.qt-project.org/#/c/109756 - do you plan to "formally" ask me about landing it, or is the plan still to rather backport the bigger Provider part? [07:43] Mirv: no because we asked you to land the bigger async provider part (or did we not?) [07:43] we did [07:44] bug #1439263 [07:44] bug 1439263 in qtdeclarative-opensource-src (Ubuntu) "Backport QQuickAsyncImageProvider from upstream-dev" [Undecided,New] https://launchpad.net/bugs/1439263 [07:44] tsdgeos: you haven't really, the bug #1439263 mentions it should be backported and as a comment it's mentioned that that MP brings fix for http stalling without anything else, but there's no clear asking what to do [07:45] Mirv, what else do you want us to write? :D [07:46] Saviq: like, whether to take https://codereview.qt-project.org/#/c/108540/ or https://codereview.qt-project.org/#/c/109756 :) [07:46] since now both are mentioned in the same bug [07:46] Mirv, my bad wording [07:47] Mirv, because we don't have a separate bug for http stalling [07:47] Saviq: oh, so both would be needed? I somehow thought thet asyncprovider would solve the problem. [07:47] Mirv, updated the comment ;) [07:47] ok, so no both :) [07:47] Mirv, no :) [07:48] clear! I'll start a landing for that + the thumbnailer branch [07:50] Mirv: wait wait, the thumbnailer branch is not ready [07:50] tsdgeos: I know it's not ready, I'm not going to _land_ it [07:51] just rebasing/testing the qtdeclarative part [07:51] tsdgeos: perks of being on the landing team, silos for free! :) [07:51] k [08:02] weird, that yesterday's unity8 landing that the PPA decided to show complete qml-phone-shell changelog for... [08:02] if it happens again there must be some new LP bug [08:05] FYI https://codereview.qt-project.org/#/c/110910/ solved my Qt 5.5 blocker problem so I've been able to go forward with testing 5.5 alpha more (builds only, because of qtmir) [08:13] Mirv: what's with qtmir? crashes? doesn't build? [08:14] tsdgeos: I filed this a month ago https://bugs.launchpad.net/ubuntu/+source/qtmir/+bug/1437181 saviq looked it into a bit and I remember the change indicates it's not a simple build problem but requires some changes [08:14] Ubuntu bug 1437181 in qtmir (Ubuntu) "qtmir fails to build with Qt 5.5" [High,Triaged] [08:14] there's been a change to how the scenegraph profiler is to be used [08:17] right [08:17] at worst for testing you could just comment those lines [08:17] since it doesn't do anything else than ehlp profiling [08:17] Mirv: ↑↑↑ [08:21] tsdgeos: oh, ok, good info, thanks! === willcooke is now known as francooke === francooke is now known as JupiterWhistles === JupiterWhistles is now known as willcooke [08:37] tsdgeos: hmm, it's possible there's DBus activity in adding Ubuntu One account that gets worse with the 018 silo https://trello.com/c/doDBpNlq/1428-ubuntu-landing-018-qtbase-opensource-src-qtbase-opensource-src-gles-mirv - I saw that too, but I thought I reproduced that without the PPA but maybe it was different [08:38] Mirv: it's possible, no idea what the ubuntuone or accounts system does tbh [08:42] tsdgeos: I think the problem described is actual app installation. I did reproduce that adding U1 account was problematic, but I did not see problems with app installation possibly because I flashed clean + added new account [08:43] (and adding U1 account was problematic without the PPA too) [08:43] I don't know about what happens with app installation either === marcusto_ is now known as marcustomlinson [08:53] (asking on #ubuntu-touch) [09:06] MacSlow: you didn't run into that yesterday but we did a landing yesterday [09:06] so it makes sense :D [09:07] tsdgeos, yeah.. I guess so... just wish I didn't have to babysit that branch until the needed update for libunity-api-dev lands... so Jenkins can move on... and also provide an approve-stamp :) [09:08] tsdgeos, that branch is like a never-ending story for me :) [09:08] yaeh i can feel the pain of long branches [09:09] MacSlow: actually i think some of the merge issues come from the parent [09:09] so make sure you fix those first [09:09] tsdgeos, you mean lp:~unity-team/unity8/shellRotation [09:09] yes [09:10] tsdgeos, yes... debian/changelog, debian/control and tests/plugins/Unity/Launcher/CMakeLists.txt at least [09:22] MacSlow: well i wrote it in that branch MR :) [09:52] /food [10:26] Mirv: sorry for the delay: https://code.launchpad.net/~gerboland/qtmir/qt55-compat/+merge/257235 [10:27] tsdgeos: fancy giving it a quick look^^ [10:28] greyback: so dropped the profiling? [10:28] it's really not useful? [10:28] tsdgeos: yeah, it always printed 0. binding a texture is a really cheap [10:28] k [10:28] let me compile it in case it fails somewhere else [10:45] greyback: all good [10:45] tsdgeos: thanks [11:13] Saviq, i figured a lot of the stuff I was pestering you about was on https://unity.ubuntu.com/getinvolved/development/unity/ Would be great to have this page in the topic [11:14] sidi_, unity.ubuntu.com is there in the topic, and there's a "Get Involved" link there... [11:15] well, that's fair enough === MacSlow is now known as MacSlow|lunch [11:30] greyback: thanks! not the biggest priority anyway, I'm just happy I'm early in doing these build tests myself this time. [11:31] and now when I have time at some point I can even try to see if the phone runs with 5.5! maybe next week. [11:31] and for the first time my build tests are truly covering all non-required rebuilds too, and I already found about 5 such FTBFS:s [11:36] MacSlow|lunch: yeah QObject was removed from some big global include which makes some stuff not compile [11:36] should be trivial to get back into compiling [11:38] tsdgeos, you meant Mirv ↑? [11:38] i did :D [11:38] i.e. http://cgit.freedesktop.org/telepathy/telepathy-qt/commit/?id=494a955128a193382a698ee3d71071d8ff8388e9 === Malsasa_ is now known as Malsasa [11:41] :) [11:42] tsdgeos: right, I hinted towards your adding of it to unity8 in a couple of the bug reports that seemed clearly to be similar to that [11:42] then there's also some possibly more tricky bugs, but on the other hand it's Alpha we're talking about.. [11:43] in a parallel universe when Qt releases on time Beta was released March 12th, but sadly in this universe we're still at Alpha [11:43] Mirv: yeah for sure there's a crasher in one of unity8 tests [11:43] but let's wait for the beta before complinign more [11:43] (https://wiki.qt.io/Qt-5.5-release) === Malsasa_ is now known as Malsasa === MacSlow|lunch is now known as MacSlow === marcust__ is now known as marcustomlinson_ === marcustomlinson_ is now known as marcustomlinson [13:22] mzanetti, I can has a review of the last commit in https://code.launchpad.net/~unity-team/unity8/homeKey/+merge/257254 [13:23] sure [13:31] Saviq, haven't tested it yet, but reading through it, it seems a long-press of the button would not bring you to home any more [13:31] intentional? [13:31] mzanetti, yes [13:31] mzanetti, it just built in silo 25 if you want to test [13:31] ah cool, yes, I will [13:31] Saviq, but why not having it working on longpress? [13:32] i.e. starting the timer if not running on release [13:32] mzanetti, because by then you started a lazy bottom swipe [13:33] mzanetti, ah I know what you mean [13:33] Saviq, not that I want that. just thinking out loud if it wouldn't make sense [13:34] mzanetti, but I think we'll leave long-press for a different use case, we only care about tap today, and I wouldn't like to introduce a long press people might get used to [13:34] Saviq, fair enough [13:34] mzanetti, you'll need to boot the kernel from https://chinstrap.canonical.com/~okubik/tangxi/boot.img [13:34] mzanetti, (fastboot boot boot.img) [13:34] Saviq, hinting the bottom edge onPress would be cool :) [13:35] mzanetti, apps could do that, we're delivering the events to them anyway [13:35] oh. yeah, would need to be an app-thing [13:35] mzanetti, but we'd need someone from design to actually look at this at all ;P [13:35] firt [13:35] *first [13:36] sure [13:36] om26er, updated silo 25, it will only interpret taps now [13:40] Saviq, great, will get to it now [13:50] mzanetti, oh and also, in this case we likely need to be a bit more granular between tap and long press because of the bottom edge (i.e. tap must be short, and long press acts immediately as soon as its threshold is reached) [13:50] vs. tap being any press+release that's shorter than long press [13:51] yeah... [13:52] Saviq, tested. feels good [13:53] mzanetti, tx, before, a very lazy bottom swipe would trigger dash, can you confirm that's not the case still? like that you really can't see anything unexpected happen (I can't) [13:53] confirmed [13:54] Saviq, I just notice that verly slow bottom edges work in apps but not in the dash :) [13:54] (yes, I know why - just didn't realize it so far) [13:54] ;) [13:54] we got a card for that :D [13:54] yep [13:54] Saviq, btw, new calculator app asked me on an ETA for that [13:55] Saviq, they have a problem that the bottom line of the the buttons is partially obscured by the bottom edge drag are [13:55] a [13:55] right [13:56] I guess it's probably worth for them coming up with a workaround for now [13:56] might be a while until it really ends up in the SDK [13:56] and in a stable image for that matter [13:56] in a framework etc. [13:56] yeah [13:57] popey, ^ [13:57] wat? [13:57] calculator app, bottom edge [13:57] kk [13:57] you better add some workaround for now [13:57] ok [13:57] thanks [13:58] popey, it'll be new API in the SDK, so new framework etc. [13:58] ok [13:58] popey, or you guys can do what browser did [13:58] popey, just copy the code from us [13:58] would get us some out-in-the-wild feedback [14:01] what did the browser do? [14:02] popey, there's some classes that you could copy from the unity8 codebase so you'd have that special edge drag area available in calc [14:02] browser did that [14:12] ok [14:26] mzanetti: greyback: Saviq: i did a first take on https://code.launchpad.net/~aacid/unity8/suspend_screenshoting/+merge/257119 that afaics does everything that it should do, but some review would be nice :) [14:26] tsdgeos, coolz [14:26] tsdgeos: nice. I can look at it later on this evening [14:27] tsdgeos, one thing I'd ask for is some minimal docs [14:27] doxygen style and all [14:28] Saviq: for sessionscreenshoter.h ? [14:28] josharenson, latest jenkins for the greeter mode branch looks OK, right? [14:28] still 4 errors, but maybe not related? [14:28] mterry, we only expect 3 errors these days [14:29] Saviq, humph [14:29] Saviq, ok good to know [14:29] bug #1446846 [14:29] bug 1446846 in unity8 (Ubuntu) "Three consistent autopilot test failures" [High,Triaged] https://launchpad.net/bugs/1446846 [14:29] Saviq, thanks [14:29] (and complete FAILURE due to bug #1421009) [14:29] bug 1421009 in qtbase-opensource-src (Ubuntu) "unity8 sometimes hangs on boot" [Critical,In progress] https://launchpad.net/bugs/1421009 [14:29] mterry: my eyes are barely open, ill go take a look [14:29] josharenson, unity8.shell.tests.test_notifications.InteractiveNotificationBase.test_modal_sd_without_greeter is the new failure with your branch [14:30] mterry: ah that sounds real :-/ [14:31] mterry: humm it passes locally... === Malsasa_ is now known as Malsasa [14:40] popey, if you decide to take the code, we have a few unlanded branches that will take the bottom edge bits closer to what will likely end up in the SDK, so please just talk to us when you're there [14:54] Saviq, so you're saying that a long press should not show home? [15:00] Saviq, it might be that the only way to "fix" this use case is adding even more delay to the button response [15:00] Saviq, not sure this is really a bug [15:00] Saviq, it's like trying to do a very lazy edge swipe. it won't work [15:02] dandrader, already fixed [15:02] dandrader, http://bazaar.launchpad.net/~unity-team/unity8/homeKey/revision/1738 [15:05] dandrader, and don't you have a public holiday today? [15:05] Saviq, hmm... the light won't match activation though. but if people are happy [15:05] dandrader, it really behaves better [15:05] Saviq, it's a holiday in Rio only, and we get only national holidays, right? [15:05] dandrader, tsdgeos disagrees ;) [15:06] Saviq, hmm... tsdgeos tell me more :) [15:06] dandrader, /me dunno, don't have any non-national ones [15:06] dandrader, I think the deal is that you observe all the holidays that are observed by others in your area [15:06] well i've been doing "barcelona holidays" since the beginning [15:07] tsdgeos, ok, so we have a precedent! :D [15:09] this week was unique. national holiday on Tuesday and local holiday on Thursday. local universities (those lazy bastards!) took the whole week off. My wife teaches in one of them. :) [15:10] mterry, you should move to Rio ^^^ [15:10] dandrader, teachers have it like that ;) [15:12] tsdgeos, btw, screenshoter surprisingly small, /me likes ;) [15:12] yeah [15:13] tsdgeos, what's the transition between screenshot and app right now? we were considering desaturating the screenshot so it's clear the app is not ready (if it's not in greyscale to begin with ;D) [15:14] dandrader, agreed! [15:14] Saviq: it's "pixelated" since we only show an image with width/2 height/2 resolution [15:14] it's not horrible as an indicator of "this is not the real app" [15:15] could be better too [15:16] on that note, I think we need to stop doing the /2, darken+desaturation will likely work better [15:17] + ActivityIndicator [15:18] Saviq: different MR though? [15:18] tsdgeos, sure, could be [15:18] pushed some docu, it's not great, but gives a basic grasp of what the thing does [15:19] tsdgeos, btw, I'm starting to fall to "SessionGrabber", we already have a "ScreenGrabber" [15:19] s/to/for/ [15:19] yeah [15:19] that's the thing that actually saves screenshots for the user [15:19] yup [15:20] and it's grabToImage, too [15:20] hmmm not really [15:20] i mean kind of [15:20] QImage screenshot = main_window->grabWindow(); [15:21] I mean on QQuickItem? [15:21] http://doc.qt.io/qt-5/qquickitem.html#grabToImage [15:21] * tsdgeos got lost [15:21] tsdgeos, nvm, I'll write stuff on the MR [15:21] :) === alecu is now known as alecu-lunch === Malsasa_ is now known as Malsasa === alecu-lunch is now known as alecu === Malsasa_ is now known as Malsasa