[07:30] veebers: ping [07:30] mzanetti: pong [07:30] veebers: hey. somethings wrong with my calendar [07:30] veebers: the meetings have been moved another hour earlier? [07:30] mzanetti: ah, no that's not your cal fault [07:31] mzanetti: it's now daylight savings here, and I must have made the original meeting with my local tz, not using UTC [07:31] so I'm an hour ahead of you than normal [07:31] oh... did you guys already switch daylight saving times [07:31] yeah, just yesterday [07:31] it another 3 weeks around here I think [07:32] Had I realized yesterday that it would affect this meeting I would have emailed, I didn't notice until a couple of hours ago though :_P [07:32] veebers: so, you still up to for a sync meeting or you already off? [07:33] mzanetti: I think a quick one now would be good. We can plan the next couple until we sync hours again too [07:33] ok [08:14] Saviq: hi [08:15] mzanetti, ih [08:15] mzanetti, veebers, sorry for not joining the sync lately... [08:16] Saviq: no worries. got it covered [08:16] (mostly) [08:16] Saviq: I've seen this one happening on the weekend and attached some debugging stuff: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1219871 [08:16] Ubuntu bug 1219871 in unity8 (Ubuntu) "when idle with screen blank, unity8 generates hundreds of context switches per second" [High,Confirmed] [08:16] mzanetti, oh, so you managed to repro? [08:17] Saviq: well, it happened. no idea still what's the trigger [08:17] Saviq: but there is one interesting thing [08:17] we have 14 threads, one of them is named "dconf worker" and according to gdb it's stack depth is 2300 and something [08:18] does that ring any bell for you? [08:18] mzanetti, well, gsettings-qt probablky [08:18] -k [08:18] mzanetti, but 2300? I'd imagine is a broken stack and a loop [08:18] probably [08:19] but still indicates dconf integration might be the issue [08:19] mzanetti, see Stacktrace.txt in https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1221492 [08:19] Ubuntu bug 1221492 in unity8 (Ubuntu) "unity8 crashed with SIGSEGV in _IO_vfprintf_internal()" [Medium,Incomplete] [08:19] Saviq: also I can confirm the futex thing as reported by cjwatson [08:20] strace prints wakes by futex like crazy when it happens [08:20] :/ [08:21] Saviq: my trace wasn't that useful as this one. but if they are indeed the same, this would be helpful I guess [08:21] mzanetti, I doubt they are [08:21] mzanetti, but just another example of a broken stack [08:22] probably not... seeing all that flickable stuff in there it doesn't look like a dconf worker thing [08:22] funny thing is, in my case I can see that broken stack depth info while its running [08:22] mzanetti, if you can get rid of Gsettings from Shell.qml [08:23] it doesn't crash. in my case [08:23] I don't think we're using anywhere else atm [08:23] and see if you can reproduce [08:23] launcher [08:23] but yeah, can remove it there too [08:23] but I'd need a way to reliably reproduce, yes [08:24] right, of course [08:32] Saviq: hmm... if there is a dynamic lib and the -dev package also ships a static lib for it, is there still a way to choose which one it should use? [08:35] mzanetti, yeah, should be [08:36] mzanetti, for cmake, you just need to point to it directly - i.e. include the .a in the linking [08:36] I think you can give the full path to select the static one. but iirc that was compiler specific. [08:36] right [08:36] or well, -lblah.a should work, too [08:36] yeah. I did that with qmake too already. but there was some caveat which I can't remember === pete-woods1 is now known as pete-woods === vila is now known as vila-afk-biab [09:03] mhr3, ping [09:03] pong === AlanChicken is now known as alanbell === alanbell is now known as AlanBell [09:27] mzanetti: hi. Can you help me to review this? https://code.launchpad.net/~paulliu/unity8/unity8-fixgenericpreview [09:27] mzanetti: https://code.launchpad.net/~paulliu/unity8/unity8-fixgenericpreview/+merge/187762 [09:28] paulliu, btw the prereq branch for the music preview was finally merged [09:28] so should be "fixable" now :) [09:28] mhr3: ok [09:28] Let me do it. [09:35] Cimi: ping [09:35] tsdgeos, pong [09:35] Cimi: wasn't the hud button supposed to stay around for a while? [09:36] tsdgeos, yeah [09:36] it doesn't anymore [09:36] tsdgeos, I have a branch and it's failing with jenkins [09:36] ah [09:36] i see [09:36] tsdgeos, autopilot on slow machines... [09:36] Cimi: ping [09:37] dednick, pong [09:37] Saviq: you there? [09:37] tsdgeos, yup [09:37] Saviq: do you know of any video i can use to test https://bugs.launchpad.net/unity-mir/+bug/1231125 ? [09:37] Ubuntu bug 1231125 in unity8 (Ubuntu) "Music and Videos do not launch from the dash" [Critical,Triaged] [09:37] all the videos i add to the Videos/ folder [09:37] Cimi: howdy. i've done some work on the ubuntu-settings-components branch. Would be great if you could take a look sometime. lp:~nick-dedekind/ubuntu-settings-components/pre-archive-fixes [09:37] fail to show on it [09:37] so if you have some file that you know it works [09:37] it'd be great [09:37] Cimi: https://code.launchpad.net/~nick-dedekind/ubuntu-settings-components/pre-archive-fixes/+merge/188283 [09:38] dednick, I'll do as soon as this jenkins will stop failing... [09:38] tsdgeos, http://bazaar.launchpad.net/~phablet-team/demo-assets/trunk/files/head:/videos/ should work, no? [09:38] Cimi: it's generally just restructuring so we can get it into archive at some point [09:39] didrocks: ping [09:39] Saviq: makes sense [09:39] * tsdgeos tries [09:40] dednick: pong [09:41] didrocks: hi. what does it take to get new packages into archive? [09:42] dednick: hum, is that the same request than thostr_? ;) [09:42] didrocks: um, don't know about his request. probably not. [09:42] didrocks: ubuntu-settings-components [09:42] dednick: get a landing ask describing the package and if it's needed in touch [09:42] i'm talking about the thumbnailer [09:43] so, it's a different one [09:43] ok, just similar requests and questions at the same time ;) [09:43] didrocks: ok, thanks [09:44] yw [09:45] meh [09:45] i get nothing in the Video shell [09:45] s/shell/scope [09:54] * tsdgeos tries to dist-upgrade to see if that helps at all with no videos === vila-afk-biab is now known as vila [10:14] Saviq: do you get local videos ? [10:14] i get nothing :-/ [10:14] tsdgeos, /me tries [10:15] not even in the pc [10:15] tsdgeos, pc is different scope than phone - phone uses unity-scope-mediascanner [10:16] tsdgeos, and yeah, I'm getting a carousel of local videos here [10:16] path being [10:16] ~phablet/Videos [10:16] ? [10:17] tsdgeos, /home/phablet/Videos yes [10:17] pfff [10:17] i get only the "online" part [10:17] tsdgeos, well, I just pushed them through MTP, but yeah [10:17] nothing local [10:17] tsdgeos, try searching and clearing the search? [10:18] did not help [10:20] interesting, mhr3, pstolowski, any idea for tsdgeos ↑ on how to get local videos to show up? [10:21] tsdgeos: codecs problem? mediascanner can only scan what's supported by gstreamer [10:21] Saviq, tsdgeos, we were just talking about it, seems codecs are broken, so if you didn't have something in the db already you won't get it there now [10:22] ¿? [10:22] so i can never get anything [10:23] it should show up once codecs are fixed [10:23] hopefully [10:23] mhr3, interesting, I get stuff... [10:23] jamesh would know [10:23] Saviq, depends on the codec [10:24] k [10:24] tsdgeos, let me push something for you [10:26] tsdgeos, still uploading, but https://chinstrap.canonical.com/~msawicz/test.avi [10:26] tsdgeos, 20s [10:29] Saviq: waht do i do with it? [10:30] tsdgeos, drop into ~phablet/Videos, search in dash videos, unsearch - see if it shows up [10:30] search anything, right? [10:30] tsdgeos, yeag [10:30] h [10:31] new ubuntu installation (friday was broken) [10:32] I just branched unity [10:32] ran ./build -s [10:32] then at ./build it complains of missing so many components [10:32] s/components/dependecies/ [10:33] what's wrong? [10:34] mzanetti, ping [10:34] Saviq: got something [10:34] but something i can't play :D [10:34] can you? [10:34] tsdgeos, isn't that what you needed? :D [10:34] well [10:35] it's supposed to work on SF [10:35] and it doesn't [10:35] nic-doffay: hey ho [10:35] soo no, that's not what i need [10:35] tsdgeos, build/install demo-assets-videos [10:35] tsdgeos, can get you packages in 5 if you want [10:36] if that's going to help [10:36] please [10:36] mzanetti, I'd like to use a mock scope in a test, got some info on that for me? [10:36] tsdgeos, will test first [10:36] actually I already have the packages [10:36] guys ^ [10:37] otherwise I install manually [10:37] nic-doffay: hmm... depends a bit on what exactly you want to test [10:39] http://paste.ubuntu.com/6175073/ [10:47] mzanetti, my search indicator branch which pops up with an indicator when a search is in progress. [10:51] mzanetti, you know why build doesn't install of of these? ^ [10:51] Cimi: because its broken [10:51] ok [10:51] Cimi: I think Saviq started to fix it already [10:51] I'm using build-dep [10:52] in the meanwhile [10:52] Cimi: yep. thats the workaround for now [10:52] Saviq: so where are the packages? [10:52] Cimi, just delete unity8-build-dep*.deb [10:52] tsdgeos, just pushed to my device, will get them to you in 5 [10:52] ah, ok, sorry [10:53] Saviq, what? is it required to remove the package? [10:53] nic-doffay: there is tests/mocks/Unity/fake_scope.cpp [10:53] nic-doffay: I think you can reuse that. [10:53] nic-doffay: just add your searchInProgress() stuff there too [10:54] Cimi, not required, but otherwise it might not install it (and that's the package that depends... on the dependencies) [10:54] Cimi, just delete and both ./build and ./run_on_device will work again [10:55] mzanetti, I need a qmluitest... [10:55] So using the mock scopes in a qmluitest to check the graphics. [10:56] nic-doffay: yeah... that's what it is [10:57] mzanetti, ok cool. [10:57] mzanetti, has it been used in a test before? [10:57] tsdgeos, actually - they're there in ppa:phablet-team, demo-assets-scopes and demo-assets-videos [10:57] tsdgeos, you might also need to fix the scopes dconf entry, will let you know in a sec [11:00] Saviq, qa broken? looks like qmluitests are not running [11:01] mhr3, looks like the vm nodes that run them are dead [11:02] mhr3, triggered a launch for them, let's see [11:02] thx === MacSlow is now known as MacSlow|lunch [11:04] nic-doffay: tst_ScopeView for example [11:08] anyone know what this is? http://paste.ubuntu.com/6175160/ [11:10] tsdgeos, ah shite, -scope-mock is outdated :/ [11:10] mhr3, that's a segfault during "qmlplugindump" [11:10] Saviq, why and how? :) [11:11] mhr3, it's a thing that generates .qmltypes files for QtCreator [11:11] mhr3, it's run automagically on plugins we generate [11:11] Saviq, hm, clean build fixed it... guess the abi change in unity-core [11:11] mhr3, you can do it yourself by exporting QML2_IMPORT_PATH and running "qmlplugindump Unity 0.1" or so [11:11] mhr3, yeah, very possible [11:17] Saviq: what does that mean? [11:17] tsdgeos, that means the demo assets scope won't work :/ [11:17] ok [11:17] debugging this is getting harder and harder [11:19] jamesh_, how does -scope-mediascanner activate the result upon "Play" in the preview? and any idea why would it work for music and not for videos? [11:21] Saviq, it's public holiday in au today [11:21] mhr3, you mean yesterday? [11:21] Saviq, on australian monday :) [11:22] Saviq, anything missing in https://code.launchpad.net/~mzanetti/unity8/switching-previews/+merge/187327 ? [11:22] mhr3, yeah, a review [11:22] design? jouni was looking at it [11:22] gave it +1 afaik [11:23] mhr3, no, code [11:23] mhr3, and no support for carousel, for that matter [11:24] i see [11:24] Saviq: carousel support is in by now [11:25] mzanetti, oh is it? didn't see it in the commits? you pushed? [11:25] * mzanetti checks [11:26] Saviq: yep. rev 367 [11:26] mzanetti, k thanks [11:26] * Saviq needs to see an eye doctor [11:27] Saviq: there is one minor glitch I'm aware of... the positioning of the first/last item is a little bit off. not sure how critical that is. seems something in the carousel [11:28] mzanetti, yeah, I was suspecting there will be something tricky there with the carousel [11:28] mzanetti, it's conflicting again, btw [11:29] will merge [11:31] /food [11:33] mzanetti: there? [11:34] tsdgeos: yes sir [11:35] let me try first, sorry the ping was early :D === alan_g is now known as alan_g|lunch [12:07] mzanetti, trying to load the scope like this: http://pastebin.ubuntu.com/6175332/ [12:07] it's null however, any idea what I'm missing? === MacSlow|lunch is now known as MacSlow [12:11] nic-doffay: I think you just want one scope, not all scopes (not the "s") [12:11] note [12:12] but I haven't really worked with this myself. [12:12] perhaps one of the scope guys can give better advice? ^ [12:12] sil2100: mornin/afternoon [12:13] sil2100: is autopilot working now ? (in general...not wrt mir) [12:21] sil2100, are you landing the thumbnailer pkg into universe? [12:21] sil2100, any eta on that? cause we need people to review the sdk branch that integrates it, and it'll be much simpler if they can actually build/test it [12:26] mhr3: I will be doing that, let's say that I'll prepare things for it today - but I guess it will be in universe not sooner than tomorrow morning [12:26] sil2100, is there anything missing other than the qa bits? === _salem is now known as salem_ [12:42] mzanetti, https://code.launchpad.net/~paulliu/unity8/unity8-fixgenericpreview/+merge/187762/comments/430455 === thostr_ is now known as thostr_brb [12:43] mhr3: no... its actually mostly one level below, in DashPreview.qml [12:44] ok [12:46] mhr3: yeah, it is one level below. So other previews have to be adepted for that changes. [12:47] mhr3: The description is not so accurate right now. [12:47] paulliu, ok [12:48] nic-doffay, bzr fail, it can't merge your branch with scope-isactive without conflicts :( [12:50] mhr3, yeah I noticed. [12:50] mhr3, what do you propose? [12:51] nic-doffay, one of them will need to be fixed [12:51] depends which one lands sooner [12:51] nic-doffay, so perhaps land your first, and i'll then fix isactive === alan_g|lunch is now known as alan_g [12:52] otherwise you can change the owner of your branch to unity-team, and i'll fix that one [12:57] mhr3, reckon mine will land as soon as I'm done with this test. [12:59] nic-doffay, would be easier if it did [13:03] greyback: a run_on_device with unity8 code and mir gives me [13:04] terminate called after throwing an instance of 'boost::exception_detail::clone_impl >' [13:04] what(): Could not unblank display [13:04] any idea? [13:04] tsdgeos: happens as powerd blanks screen. Press button on side, then try again === boiko_ is now known as boiko [13:09] greyback: not really helping [13:09] let me reboot [13:10] tsdgeos: https://bugs.launchpad.net/unity-mir/+bug/1233155 [13:10] Ubuntu bug 1233155 in unity-mir "Mir fails due to "could not unblank display"" [Undecided,New] === hikiko is now known as hikiko|lunch [13:12] hmmm [13:12] i'm gettint unity8 to 100% use the CPU :-S [13:13] tsdgeos: is there much dbus traffic? [13:13] mzanetti: I also commented. If you have any ideas please follows up. https://code.launchpad.net/~paulliu/unity8/unity8-fixgenericpreview/+merge/187762 [13:14] greyback: nope [13:14] greyback: let me reboot in SF and see if i see it there too === jhodapp|afk is now known as jhodapp [13:29] MacSlow, nothing major, but https://code.launchpad.net/~macslow/unity8/extended-snap-decisions-part1/+merge/187312/comments/430493 [13:29] greyback: now i got http://paste.ubuntu.com/6175584/ :-/ [13:29] this is painful [13:29] i can't start anything [13:30] ah wait i can "continue" that in gdb?¿?¿ [13:30] yes i can [13:30] weird :D [13:30] tsdgeos: let's check. You reboot your phone, does unity come up (in mir mode) [13:30] yep [13:30] coming... just changed my password... [13:30] greyback: i am running it now [13:31] just had to conitnue that sigill [13:31] don't ask me why [13:31] but it's taking 100% cpu [13:31] tsdgeos: yes here too [13:31] ok [13:31] at least i'm not the only one :D [13:33] tsdgeos: I think the sigill is normal. The openssl will detect some acceleration instructions to optimize the encryption. [13:33] tsdgeos: so it "tests" the CPU by illegal instructions. [13:33] paulliu: i see [13:40] Saviq: ping! switching-previews :P [13:41] will get ther ;) [13:44] * greyback really needs to check the public hols :) [13:45] mhr3, ok so... vms are dead :/ [13:45] fginther, ↑ [13:46] mzanetti, so I've added the scope it's cool now, however when I trigger a search the searchInProgress property never changes to true which means I probably haven't done something else. [13:46] greyback: sorry... it's in fact thursday [13:46] greyback, Google calendar can give them to you [13:46] mzanetti: yep, just saw that. [13:47] greyback, https://www.google.com/calendar/embed?src=german@holiday.calendar.google.com [13:47] Saviq: thanks. I wasn't bothered to track that down :) [13:48] greyback: what's the branch saviq wants me to use? [13:48] tsdgeos, https://code.launchpad.net/~gerboland/unity-mir/use-upstart-app-launch2 [13:48] greyback, right? Albert should use it to track stuff not being activated on scope action? [13:49] Saviq: yes [13:50] waht i'm getting atm is [13:51] ApplicationManager REJECTED connection from app with pid 3324 as no desktop_file_hint specified [13:51] when i try to launch the music playet [13:51] -t+r [13:51] since a recent unity/compiz update on saucy, launcher icons are no longer highlighted when an app is running, is that by design, or is it a known bug? [13:51] let's see if that helps at all [13:55] Saviq, investigating [13:55] fginther, thanks [13:55] mdeslaur, they get a pip on the right, that was always the design? [13:56] mdeslaur, what else do you mean by "highlight"? [13:56] Saviq: the icon would change background color IIRC [13:56] Saviq: one sec, let me try a raring vm [13:56] mdeslaur, nope, don't think it should [13:57] mdeslaur, it does when you alt+f1 for keyboard navigation [13:57] Saviq: yep, confirmed...raring changes background color when you open an application [13:57] mdeslaur, and then it gets a glow when it's urgent or some such [13:57] let me try precise [13:58] mdeslaur, ah that, works for me [13:58] mdeslaur, you mean all your icons have the same background? [13:58] yeah, whether they are active or not [13:58] mdeslaur, firefox is orange here, thunderbird is blue etc. [13:58] Saviq: hrm, let me update again and reboot [13:59] mdeslaur, but yeah, bug, if they're not for you [13:59] Saviq: ok, thanks [13:59] mdeslaur, do you get the pips at all? [13:59] mdeslaur, maybe it's a bamf issue? [13:59] yeah, I get proper pips [13:59] ok [13:59] just not background colors [14:00] mzanetti, mind taking a look at tst_PageHeader for me and check what I'm missing? lp:~nicolas-doffay/unity8/search-indicator [14:01] It's the test_search_indicator function in that test accompanied by the scope. [14:01] It doesn't appear to begin searching even though the scope is bound to the pageHeader. [14:02] nic-doffay, no [14:02] nic-doffay, you're not supposed to type anything [14:02] nic-doffay, you're supposed to *set* searchInProgress to true [14:02] nic-doffay, and then test that PageHeader behaves as expected [14:02] Saviq: you got that or is the question still valid for me? [14:02] mzanetti, got htis [14:03] nic-doffay, you're not supposed to test the mock Scope object (which is what you're trying to do now) [14:03] Saviq, right. Is that simply because it's been included in another test? [14:03] nic-doffay, you're supposed to *use* the mock Scope object to instrument testing of the PageHeader [14:03] nic-doffay, no relation at all [14:04] nic-doffay, the mock Scope object needs a (if it doesn't yet) searchInProgress property [14:04] nic-doffay, you then set that property to true [14:04] nic-doffay, and check that the activity indicator is visible and running, and that the indicator is not [14:04] s/indicator/icon/ [14:10] mzanetti, can you trigger a rebuild? [14:10] mzanetti, http://s-jenkins:8080/job/unity8-ci/1148/rebuild [14:10] Cimi: I can give you permissions to do that [14:10] mzanetti, I have but I don't remember where all my stuff is [14:10] mzanetti, keys boa bla [14:10] *bla bla [14:11] dude [14:11] I format the system [14:11] today [14:11] as said on the standup [14:11] mzanetti, I have a backup but on another laptop in the office [14:12] hmm... is it only me who finds this message from jenkins weird? "Started by remote host 127.0.0.1" === hikiko|lunch is now known as hikiko [14:12] Cimi: done [14:12] mzanetti, thx. it will fail but packages are handy [14:13] thomi: ping [14:13] greyback: lets discuss here [14:13] ricmm: agreed [14:14] thomi: what exactly are the AP requirements? when doesi t need geometry? [14:14] does it need geometry *before* the shell is started? [14:14] ricmm: doubt thomi awake. Note: mzanetti tells me veebers will be working on a unity8 plugin for autopilot soon [14:14] ok [14:14] so we just need a band-aid [14:14] yea we need to match current functionality [14:15] but we need to know that first ;) [14:15] ricmm: yep [14:16] mhr3, that property is difficult to test being readonly... [14:17] veebers: ping? [14:18] greyback: so I see it being used to create the input device, which probably happens before anything is run [14:18] in which case we have a problem [14:20] Saviq: reboot fix it, sorry for the noise [14:20] s/fix/fixed/ [14:20] mdeslaur, well, still a bug, if you can reproduce [14:20] yeah, I'll keep an eye on it [14:24] greyback: https://bugs.launchpad.net/unity-mir/+bug/1231125 is fixed for me with your branch [14:24] Ubuntu bug 1231125 in unity8 (Ubuntu) "Music and Videos do not launch from the dash" [Critical,Triaged] [14:24] want me to assign it to you and link the branch? [14:25] tsdgeos: yes please [14:26] come on [14:26] kgunn, bug #1193099 is not gonna happen for v1 [14:26] bug 1193099 in Unity 8 "Unity: thumbnails for running apps are not cached" [High,Triaged] https://launchpad.net/bugs/1193099 [14:26] music playing stops when the screen darkens?¿?¿?¿ [14:26] who do i complain about that? [14:27] kgunn, we need to do it while reworking the whole thing to support splash screens [14:27] Saviq: maybe there are 2 bugs there... [14:27] Saviq, the issue that we talked a few hours ago, aka slowness. You need to make sure you have the USB cable disconnected when you put your phone for a few minutes [14:27]