=== zoktar_ is now known as zoktar === xnox_ is now known as xnox === Pici is now known as Guest56659 === jamesh__ is now known as jamesh === jhodapp is now known as jhodapp|afk === duflu__ is now known as duflu [06:11] Can someone perhaps kick off a build please? https://code.launchpad.net/~unity-team/+recipe/unity8-daily-newscopes [06:11] times out for me (known lp bug) [08:39] lol my unity7 launcher doesn't open totally [08:39] only something like 80% or something [08:40] tsdgeos: probably because it's not fully awake yet :) [08:41] yeah [08:41] no [08:41] i got some hickup while it was expanding [08:41] and got upset [08:41] had to restart unity [08:41] interesting… I remember we had that issue in the far far past [08:41] when there was a race due to xorg barriers and unity [08:42] the fix was in xorg to send the right offset [08:42] but as we moved to a new one and the barrier has changed, I bet that additional code has been dropped [09:00] mhr3: https://bugs.launchpad.net/unity-scopes-shell/+bug/1279439 ? [09:00] Launchpad bug 1279439 in Unity 8 "Icons for non-click apps are shown empty" [Undecided,New] [09:05] tsdgeos, good question, how do we want to handle that? [09:05] tsdgeos, should we make the scopes use the image://theme/ uris? [09:06] that would fix it [09:06] depends on what you say that result is [09:06] but then we're committing to those uris [09:06] if you return image://theme then making a PHP based shell is harder [09:07] because they have to replace all that [09:07] but if it is "standard" it's ok [09:07] mhr3: i'd do it at the unity-scopes-shell level [09:07] which is a "translate scopes to shell" layer [09:07] so then we can do a unity-scopes-php-shell project that does the appropiate translation [09:08] tsdgeos, but the scope is already tying itself to the system by using theme icons [09:08] sure [09:08] it just boils down if you want to commit to "image://theme" officially or not [09:08] indeed [09:08] but i'd prefer art to always be a uri [09:09] so we don't need crazy if (this_looks_like_a_uri) ... else ... [09:16] mhr3: your team's call, i don't care :D [09:27] what is the difference between launching applications with upstart-app-launch and /usr/bin/dialer-app --desktop_file=path-to-/dialer-app.desktop? [09:32] tsdgeos, does http://paste.ubuntu.com/6924594/ look ok? [09:32] (the plugin will finally expose the per-column models) [09:33] mhr3: hmmm, don't know [09:33] mhr3: what does get(0) do? [09:34] tsdgeos, gets the first item from the model [09:34] what is a "per-column" model? [09:35] just writing this into mp [09:35] Split PreviewModel into 3 separate models - PreviewStack which holds a bunch of PreviewModels (for nested previews, currently always holds just one), and PreviewModels contain per-column PreviewWidgetModels, which in turn contain the actual widgets. [09:36] tsdgeos, if shell tells the PreviewStack to use two columns (as it would on the tablet), each PreviewModel will expose two PreviewWidgetModels [09:37] i am still confused [09:37] you wrote 3 separate models [09:38] but i don't understand what that means [09:38] tsdgeos, quick hangout? [09:38] sure [09:40] tsdgeos, looks like i'm calling you, but it didn't give me a link [09:40] :D [09:40] you should see it if you log onto plus [09:42] you calling my canonical or the gmail account? [09:42] an the gmail one [09:48] back in a sec [09:48] my unity has died === MacSlow_ is now known as MacSlow [09:57] hey guys, would someone please explain to me how to get an app to go full screen over the dock and the menu bar with _NET_WM_STATE_FULLSCREEN ? [09:57] The only I manage to get my window on top of those is with override_redirect = True, but that has lots of other problems... [09:57] tsdgeos, https://code.launchpad.net/~mhr3/unity-scopes-shell/preview-split/+merge/206130 [09:58] pstolowski, could you take a look too? has the layout bits and other stuff [09:59] woot [09:59] that's big [10:01] anyone got a clue why Unity always shows the dock on top of my 'fullscreen' app ? [10:04] mhr3, sure [10:05] pstolowski, will need some api additions, to be able to do the nested previews :) [10:05] but not today [10:05] nor this week [10:06] mhr3, yeah, I was about to say this === mirco is now known as MacSlow [10:34] mzanetti: can you do https://code.launchpad.net/~aacid/unity8/lvwph_resize_item_bug/+merge/206140 ? [10:35] tsdgeos: ok. in a bit [10:39] mhr3: so waht do we do with the icon theme thing? [10:39] tsdgeos, commented on the bug, i want uris [10:39] ah, didn't see the comment [10:39] mhr3: so close as invalid? [10:39] please [10:52] mhr3: i tried turning the onlinemusic to a carousel, but then i can't make it return few results, so can't test the carousel -> grid transformation :/ [10:52] tsdgeos, "nonordinary" [10:53] ok [10:53] that works, tx [10:54] macSlow, macSlow_: hey, how are you? [10:54] macSlow, macSlow_: do you want to review https://code.launchpad.net/~larsu/notify-osd/always-use-workarea/+merge/206145 ? [10:55] hey seb128... busy debugging some branch-failure on jenkins [10:55] macSlow, macSlow_: I've tested it in quite some configs, seems to work fine, let me know if you want to review/test it as well or if I should just approve it [10:55] seb128, I can look into that branch today too np [10:55] MacSlow: thanks [11:08] mhr3: https://code.launchpad.net/~aacid/unity8/new-scopes-carousel-as-grid/+merge/206148 [11:08] mhr3: btw i did add the progress stuff [11:08] to new-scopes [11:21] tsdgeos, hmm, won't that break the way from grid to carousel? [11:21] mhr3: no [11:21] mhr3: it is inside CardFilterGrid.qml [11:21] it's always a Grid if you're there [11:22] tsdgeos, consider that the model has 3items after first bunch of results get back from the back, 2seconds later another 15arrive [11:22] backend* [11:22] yes [11:22] that will not be a CardFilterGrid [11:22] can't be [11:22] wheter that works or not [11:22] i don't know [11:22] but it can't be breaking it [11:23] after the 3items it would be FilterGrid [11:23] and then you modify the layout, so it won't even try to be carousel [11:23] that's how i see it [11:23] you see it wrong :) [11:23] I forgot [11:23] what does the ternary var || "something else" we use imply? [11:23] tsdgeos, ^ [11:24] which conditions on var? [11:24] mhr3: getRenderer in GenericScopeView will say "oh wait the number of results has changed so let's use a CardCarousel instead of a CardFilterGrid" [11:24] undefined/null what? [11:24] Cimi: example? [11:24] tsdgeos, so you're saying the var t = ...; makes a deep copy/ [11:24] ? [11:25] widgetData["review-label"] || "Review" [11:25] mhr3: yes [11:25] tsdgeos, in which cases it chooses "Review" [11:26] mhr3: or maybe not [11:26] * tsdgeos scratches head [11:27] Cimi, same as it's use widgetData[review-label] if ( widgetData["review-label"] != undefined && ditto != null ) else use the default you specify [11:27] mhr3, sure? [11:27] mhr3, good then [11:27] yes [11:27] mhr3, I just wanted to be sure of which checks [11:28] mhr3, cause I had issues with assigning null [11:28] although i'm not completely sure how it works with nulls and empty string and the like [11:29] js secrets [11:33] tsdgeos, so, did you make up your mind? :) [11:33] mhr3: so it does modify the original [11:33] but i'm not so sure it's a problem [11:33] but i'll have a different solution anyway just in case [11:34] well... on one hand it forbids the grid -> carousel transition [11:34] on the other maybe we don't want it [11:34] would be disruptive [11:35] f5 [11:35] ahhhhhhhh [11:35] damn [11:35] pushed to wrong place [11:38] mhr3: now https://code.launchpad.net/~aacid/unity8/new-scopes-carousel-as-grid/+merge/206155 [11:38] this is safer [11:40] tsdgeos, have you had bad experiences with qFuzzyCompare? Is it that broken or am I doing something wrong? http://paste.ubuntu.com/6921245/ [11:40] he he [11:40] dandrader: i know it is some level of broken when comparing against 0 [11:40] but to be honest never encountered such a case myself [11:41] tsdgeos, now I'm just doing QVERIFY(qAbs(actual - expected) > 0.001) [11:41] tsdgeos, property string categoryLayout: template["category-layout"] || "" [11:41] ? :) [11:42] mhr3: that will fail if template is null [11:42] dandrader, you might want < instead :) [11:42] mhr3: well not fail but spit out a warning [11:43] mhr3, the "id" for rating-input [11:43] mhr3, what is it used for? [11:43] have enough warnings already, don't want to add more :D [11:43] how shall I use it? [11:43] tsdgeos, ok [11:43] Cimi, which id exactly? [11:44] mhr3, there's an id in the json spec [11:44] maybe tsdgeos knows [11:44] looks like similar to progress? [11:44] Cimi, it's the regular widgetId, every widget has that [11:45] Cimi: mhr3: this one "id": null, // mandatory here! [11:45] ? [11:45] tsdgeos, yup [11:45] i have no idea why it says "mandatory to null" [11:45] but yes it's the thing we store in widgetId [11:45] just pass it up in the triggered signal [11:45] up/down/somedirection [11:46] if we are using triggered [11:46] if not, ignore i'd say [11:46] we are [11:51] mhr3, what if you set visible to rating but required review? [11:51] it won't work :D [11:52] what shall we do in those cases? [11:52] Cimi, indeed, and the scope author needs to fix it [11:52] ok [11:53] tsdgeos, so what about the background-color thing? [11:53] mhr3: we need a new json for that? [11:53] tsdgeos, i'd treat it the same way we do title and all the other components [11:54] sure [11:54] i meant a new property [11:54] sorry [11:54] you mean role in the model? [12:03] mzanetti, greyback: I've added more fixes to https://code.launchpad.net/~dandrader/unity8/rotatedDDA/+merge/205845 [12:03] dandrader: ok... will take a bit before I come this again tho [12:03] mzanetti, greyback so you might like to ta a look at those new commits before top approving once again [12:03] mzanetti, np [12:04] dandrader: this I confirm works correctly === zequence_ is now known as zequence [12:22] hey sil2100, didrocks we're down to 9 failures in Unity CI, time for a distro landing I think [12:22] \o/ [12:22] Ouuu yeaaa [12:22] didrocks: ^ should we use CITrain for that as well, or can we simply release through cu2d? [12:22] waow :) [12:22] sil2100: I guess we should use CI Train (if you are happy handling it) [12:22] that would be the right time to switch it [12:23] didrocks: ok ;) [12:23] sil2100: I think nothing else is on your plate, right now, isn't it? you have time for that? [12:24] (apart from the remaining AP messaging-app) [12:24] didrocks: I'm switching it off in cu2d in a moment [12:24] excellent! [12:24] sil2100: I guess it's not only unity, but nux, compiz… and the whole world [12:24] check with bregma I guess :) [12:24] mhr3: arg [12:24] bregma: excellent work, 9 failures, we never had that :p [12:25] mhr3: yes i mean role in the model [12:25] bregma: what do you want us to release? unity + nux + compiz? libunity has been already released if anything ;) [12:25] didrocks, sil2100, yes the whole stack [12:25] mhr3: can you reproduce the first tab title not showing until you swipe back and forth? [12:25] * bregma is squirming in excitement [12:26] ahah :) [12:26] tsdgeos, happens with scope tool pretty much all the time [12:26] sil2100: maybe confirm with bregma the whole source package list? [12:26] mhr3: ok, i'll give you a patch for it and tell me if it fixes it [12:26] k [12:27] sil2100, far as I know it's just compiz, nux, and unity everythiong else has been landing [12:27] mhr3: https://code.launchpad.net/~aacid/unity8/fix-new-scopes-tab-name-missing/+merge/206162 [12:27] bregma: I'll set you as a lander for those ;) [12:28] Feel invited to the CITrain bootcamp on Monday! [12:28] mhr3: give it minute, still not pushed [12:29] tsdgeos, tests are taking time.... [12:29] tsdgeos, I'm testing all conditions [12:29] with rating/reviews [12:29] Cimi: good! [12:29] tsdgeos, not much code but trying it simply [12:29] mhr3: now, https://code.launchpad.net/~aacid/unity8/fix-new-scopes-tab-name-missing/+merge/206162 :D [12:30] tsdgeos, fixes it [12:30] mhr3: ok, so i'd say let's merge it, since somebody said people want tabs to die there anyway, no need to investigate much more [12:31] tsdgeos, ack, +1ed, pls merge [12:31] mhr3: done [12:32] tsdgeos, pushing also the previewmodel splitup fix === MacSlow is now known as MacSlow|lunch [12:33] mhr3: do we have the new unity-scopes-shell too? [12:33] otherwise it'll break too, no? [12:33] not yet, going to build it in the ppa [12:33] yea [12:33] ok [12:35] bregma: I should use lp:compiz/0.9.11 for compiz, right? [12:35] mhr3: let's spec the background color thing if you want, that's the last feature i think as missing (besides billions of fixes here and there) [12:35] tsdgeos, and the other billion for next week :) [12:35] sure [12:37] sil2100, yes, compiz 0.9.11 [12:41] bregma, landing coming?! [12:42] bregma, good morning ;-) [12:42] I'm preparing a landing slot ;) [12:42] \o/ [12:45] didrocks, bregma: slot is being prepared, I'll kick the build for it and deal with it after lunch ;) [12:45] great! [12:46] bregma, sil2100, didrocks: I'm going to test from the ppa once it's built [12:46] * seb128 wants the new version [12:47] seb128: bregma: sil2100: I guess we still want all AP tests being run once on the final .debs as well [12:48] Cimi: mhr3: any chance you guys can quick review https://code.launchpad.net/~aacid/unity8/new-scopes-header-fix/+merge/205987 ? It's just a quick fix for something that was commited and not throughly reviewed (my fault) [12:49] Yep ;) [12:49] looks good to me === _salem is now known as salem_ [12:50] mhr3: actaully give me a sec, i think it's not totally correct [12:50] mhr3: or it is... [12:50] mhr3: can a preview header ever be overlayed? i guess no, right? [12:51] tsdgeos, don't think so [12:51] mhr3: ok, then it's fine [12:51] mhr3: if you can approve we'll save a few warnings when showing previews [12:51] lunch! [12:51] that is a good idea [12:54] mhr3, tsdgeos what should happen on typos? [12:54] Cimi, what do you mean? [12:54] widgetData["visibility"] = "reviewefe" [12:54] instead "review" [12:54] fallback to default [12:54] ok [12:55] it's clearly specified that it supports 3 values [13:00] tsdgeos, if I change visibility of a parent, does children get different visibility [13:00] propagates [13:00] ? [13:01] yes [13:01] qt doc === jhodapp|afk is now known as jhodapp === dandrader is now known as dandrader|afk === MacSlow|lunch is now known as MacSlow === dandrader|afk is now known as dandrader [13:55] tsdgeos, when widget data changes [13:56] tsdgeos, what should happen to the review input? [13:56] keep the same values? [13:58] Cimi: "rating-input" ? [13:59] mmm [13:59] no when one changes the json [13:59] tsdgeos, and the user did a review before or so [14:00] tsdgeos, just like when you press submit, what should happen? [14:00] it's tricky [14:00] in the meanwhile, I'll try to finish this so you can review for now [14:00] Cimi: i don't think the json would change tbh [14:08] again with signalSpy drunk [14:08] seeing 2 instead 1 [14:08] mzanetti, help :( [14:08] Cimi: :D [14:09] Cimi: I didn't have any issues with the SignalSpy lately [14:09] I have it [14:09] I cannot see any bugs here [14:09] I have a clear onTriggered: console.log("triggered"); [14:09] ONE console line [14:10] signalSpy says 2 [14:10] not following. need to see code [14:10] tsdgeos: ping. I need your help to land a branch that's needed by autopilot. [14:10] Cimi: can you paste the snipped and perhaps the output too? [14:10] elopio: what's up? [14:11] mzanetti, it's not easy, I'll paste [14:12] mzanetti, http://paste.ubuntu.com/6925651/ [14:12] hah, I told you [14:15] elopio: i can help you...that way tsdgeos can do real work [14:20] kgunn, any new on getting mir 0.1.5 in ubuntu? [14:20] dandrader: you can't make this stuff up.... [14:20] so its blocked on xorg rebuild [14:20] b/c xorg is curently failing to build for [14:20] (wait for it) [14:21] ppc64el [14:21] sorry, they distracted me. [14:21] kgunn: https://code.launchpad.net/~elopio/unity8/no_uinput_side-effects/+merge/203876 [14:21] who knows why SignalSpy goes to the Oktoberfest in mid February?? [14:21] ...builds for everything else...so we're help up on powerpc [14:21] :-/ [14:21] kgunn: this branch will let us refactor code in autopilot to start removing the ugly things we are doing in unity8. [14:21] mzanetti, the output is [14:21] elopio: ack [14:22] tsdgeos: you want the preview_progress branch as well ? [14:23] kgunn: sure [14:23] kgunn: and if you give mzanetti 10 min, i'm sure he'll have another one done [14:23] mzanetti: how is the lvwph review going? [14:23] tsdgeos: sure..makes sense..will wait [14:23] oh my...lvwph [14:23] do we ever stop fiddling with lvwph :) [14:23] tsdgeos: looks good to me. still figuring what's wrong with that branch on jenkins [14:24] mzanetti: ah yeah, somehow jenkins exploded on it [14:24] oh god [14:24] mzanetti, I understood why [14:24] kgunn: when desing redesigns the dash next time maybe :D [14:24] mzanetti, we can talk it later on mumble [14:24] Cimi: ok [14:29] elopio: you're gonna get someone to review tho right? [14:30] kgunn: to review this branch? That's why I pinged tsdgeos. [14:30] ah...sorry, i thot it was already reviewed/approved.... [14:31] elopio: so that branch says you need a new autopilot, does that autopilot exist already or i need a new one [14:31] aaaah standup [14:31] tsdgeos: ahh, I didn't update the description. [14:32] I'll do it. Now it's compatible with current autopilot, and with my new branch that will refactor uinput,. [14:41] MacSlow: :D did you buy a new Mic? [14:41] mzanetti, no... is the sound quality better? [14:42] MacSlow: better than ever before [14:42] mzanetti, I mainly messed with mumble-settings [14:42] MacSlow: as in, you sound like a human now, not any more like a terminator robot [14:42] mzanetti, damn ;) [14:42] tsdgeos: I updated the description, and added the checklist. I keep forgetting about that, sorry. [14:44] Cimi: https://qt.gitorious.org/qt/qtdeclarative/source/586ebf6c55ed095c2627185d90153ccfdc51c3d2:src/imports/testlib/SignalSpy.qml [14:46] Cimi: you really made me curious how this can happen [14:48] mzanetti: so sorting out ppa for rightedge this mornin' and merging your right-edge-2 [14:48] branch with ~unity-team/unity8/new-scopes [14:49] conflicts, just wondering if you need to [14:49] rebase on trunk [14:49] ok [14:49] kgunn: might be, yeah [14:49] kgunn: will do now [14:49] ta [14:50] elopio: so with that branch it will work now and in the future autopilot? [14:51] tsdgeos: yes. In a really ugly way using try catch for future compatibility, but as soon as autopilot lands I'll get back and clean that, and the nasty calls to private _fingers. [14:52] elopio: looks reasonable [14:52] kgunn: merged & pushed === redj_ is now known as redj [14:53] elopio: i'll re-run the CI, ok? [14:55] mzanetti: thanks! [14:55] anyone knows if karni will be around today? [14:56] tsdgeos:oh i think he's in california right ? [14:56] yeah but he has helping $SOMEONE that was having medical problems [14:56] not sure how that ended up [14:57] tsdgeos: ok. The error is weird, on os.killpg(self._notify_proc.pid, signal.SIGTERM) it seems. [14:57] completely unrelated. [14:57] garg [14:57] now autopilot fails in my branch too [14:58] MacSlow: did you guys break notifications again? [14:58] MacSlow: see https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/2817/artifact/results/autopilot/artifacts/unity8.shell.tests.test_notifications.InteractiveNotificationBase.test_sd_incoming_call%20%28Desktop%20Nexus%204%29.ogv and https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/2817/testReport/junit/unity8.shell.tests.test_notifications/InteractiveNotificationBase/test_sd_incoming_call_Desktop_Nexus_4_/ [14:59] tsdgeos, certainly not as I've not managed to get my branch in yet [14:59] MacSlow: no change in unity-notifications that may cause that? [15:00] no [15:00] no release since january [15:00] tsdgeos, that has also not chagned [15:00] * tsdgeos tries to reproduce locally [15:04] obviously works locally [15:04] how not [15:05] i'm reading this [15:05] def _tidy_up_script_process(self): [15:05] if self._notify_proc is not None and self._notify_proc.poll() is None: [15:05] logger.error("Notification process wasn't killed, killing now.") [15:05] os.killpg(self._notify_proc.pid, signal.SIGTERM) [15:06] MacSlow: can it happen that the thing disappears in bettwen we check if it's there and we try to kill it ↑↑↑ ? [15:06] ah this is not yours [15:08] mzanetti: can you try running unity8.shell.tests.test_notifications.InteractiveNotificationBase.test_sd_incoming_call there and see if it works for you? [15:09] tsdgeos: on desktop? [15:09] tsdgeos, I just wish I could remove all AP-tests and somehow do everything with qml [15:09] mzanetti: yeah [15:09] on it [15:11] tsdgeos: is this what you get? StateNotFoundError: State not found for class ''. [15:11] mzanetti: nope [15:12] mzanetti: it works for me, and the error is different [15:12] mzanetti: you need to make isntall and then from the builddir [15:12] I did that [15:12] PYTHONPATH=../tests/autopilot autopilot unity8.shell.tests.test_notifications.InteractiveNotificationBase.test_sd_incoming_call [15:13] tsdgeos: same error [15:13] that is weird [15:13] pyhton is so helpless reporting errors sometimes [15:13] mzanetti: can you paste the whole error? [15:13] tsdgeos: http://paste.ubuntu.com/6925924 [15:14] weird [15:14] tsdgeos: seems unity8 is not starting [15:14] yeah [15:14] which is most likely an issue with my setup [15:14]