=== duflu_ is now known as duflu === salem_ is now known as _salem [09:17] Saviq: i will be moving the qmltests for showing previews from tst_GenericScopeView to tst_DashContent since now we have that slide from left stuff, i think it's better suited in DashContent where the sliding actually happens, that ok with you? [09:18] tsdgeos, +1 [10:05] Hey, could anyone look at bug 1284536, it is blocking installer tests. [10:05] bug 1284536 in compiz (Ubuntu) "compiz crashed with SIGSEGV in two_way_long_needle()" [High,Confirmed] https://launchpad.net/bugs/1284536 [10:07] Trevinho, bregma: ^ [10:08] in fact it might be an xorg bug [10:08] jibel, it would help to have debug symbols for swrast_dri.so [10:08] jibel, you might want to try pinging mlankhorst about it (he's not on this channel) [10:09] seb128, okay, it is difficult to add debug symbol in this environment but I'll try. I'll ping mlankhorst when he is online. thanks [10:10] jibel, he should be online, he's based in .nl [10:10] he's just part of the xorg team and not hanging out on the unity channel [10:10] you can try e.g #ubuntu-desktop though [10:50] Saviq: you there? [10:51] tsdgeos, yeah [10:51] Saviq: a bit confused about something you added to qml/Dash/Previews/Preview.qml [10:51] tsdgeos, hit me [10:52] Saviq: bah, ignore me, can't read :D [10:52] sorted [10:52] somehow was confused into row.column and widgetColumnCount [10:53] thinking it was setting the value to itself [10:53] but then row.column is not the rows :D [10:53] but the columns [10:55] Saviq: mhr3_: so here i'm doing the fake unity scopes ng clases for the tests and i was wondering if we could somehow re-use the real classes instead of me having to copy&paste&adapt them and then having to follow the changes everything they happen [10:56] tsdgeos, well, we should abstract the classes to lp:unity-api, and base our mocks on that, only overriding what's needed [10:56] but that would mean installing the headers and some kind of C++ library we would use from the tests in unity8 [10:56] i mean for now it's fine since i hope it's api is not going to change much [10:56] tsdgeos, I'm not sure we should get anything more than just interfaces from the real plugin [10:57] but i just realized that the classes i did two weeks ago for scopes-ng-tests are now changed again :D [10:57] Saviq: rolenames [10:57] tsdgeos, yes, that too [10:57] tsdgeos, /me considers part of interface [10:57] ok :) [10:58] but yeah now i was having preview() return a PreviewModel and now it's returning a PreviewStack [10:58] i think that would be "enough-ish" if we could at least share this kind of things [10:58] everything that qml uses is an interface, and the idea was that the interface would be defined in unity-api [10:59] but it would be nice to have some tests magic that creates a model with predefined roles and content [10:59] yeah ;) [10:59] it's what i'm doing anyway [10:59] surely not only scopes test would benefit [10:59] create 4 scopes with n categories and k items inside [11:00] don't need more than that === _salem is now known as salem_ [11:49] xnox, btw, if you wanted to press for py32ap, you'll need to convince didrocks to let those in [11:50] didrocks knows what he is doing =) [11:50] Saviq: as long as py32ap does land, when it can, that's fine. just don't get it caught into another regression / stop-ship =) [12:08] Saviq, where did you get that "text" widget is supposed to have title? the spec doesn't specify that [12:08] Saviq, but anyway, i'm wondering whether we shouldn't make "Description" default there [12:26] mhr3_, from the visual [12:26] s [12:26] mhr3_, and we defaulted to empty, not sure "Description" would be common enough [12:26] mhr3_, especially since it needs to come pre-translated [12:27] Saviq, well, each scope we have has Description there [12:27] Saviq, an i18n was exactly why i thought that having that as default would be a good idea [12:28] mhr3_, problem is Description might not be as omnipotent in other languages === MacSlow is now known as MacSlow|lunch [12:33] Saviq: could you have a look https://code.launchpad.net/~andreas-pokorny/unity-mir/fix-1281075/+merge/207375 sometime.. will be back in a few hours [12:45] Saviq, can we do title-less / header-less categories yet? [12:45] mhr3_, no [12:46] mhr3_, would be purty simple if we had the JSON [12:46] what will blow up if i create a cat titled ""? :) [12:46] what do you mean, if we had json... don't we? [12:47] mhr3_, you want it to base off of empty title? [12:48] mhr3_, I thought it'd be explicit [12:48] Saviq, i meant as a temporary thing, agreed that it should be explicit [12:53] mhr3_, http://paste.ubuntu.com/6994212/ [12:58] Saviq, didn't you want explicit? [12:58] although i'd be fine with this [12:59] mhr3_, lol, that's for the temporary thing ;) [12:59] mhr3_, problem with that is that empty title == non-expandable category [12:59] mhr3_, so yeah, I want it explicit [13:00] mhr3_, or wait, not with see more | see less === alan_g is now known as alan_g|lunch [13:05] Saviq, also true, so maybe it doesn't need to be explicit afterall [13:06] mhr3_, indeed [13:06] mhr3_, but should wait for see more / see less then [13:06] mhr3_, on that note [13:07] mhr3_, we need to think how do we communicate the intent to see more [13:07] mhr3_, and whether there's anything to see more [13:09] Saviq, is that 14.04 thing? [13:10] mhr3_, you tell me, from our PoV it's easy [13:10] for now we'll just give you result set and you can see more while you don't display everything [13:10] mhr3_, you mean just see more without querying the scope? [13:10] mhr3_, yeah, that we'll do [13:10] i think that's good enough for 14.04 [13:11] as for actual paging [13:11] need to think that through [13:11] mhr3_, ok [13:11] not exactly simple since scopes are stateless [13:11] mhr3_, sure [13:12] tsdgeos, so, how about scrapping LimitProxyModel? [13:12] tsdgeos, and relying on the renderers? it's actually not related to delegateRange at all, since we'll just make the *View yay-high, so the rest will just be out of the view, so not created [13:13] brb [13:47] mzanetti, needing your review https://code.launchpad.net/~dandrader/unity8/disableDDAWhileDragging/+merge/207965 [13:48] dandrader: ack [13:54] Saviq: might work === alan_g|lunch is now known as alan_g [14:22] dandrader: approved. looks good. There's a test failure. Seems flaky AP, I've retriggered the job [14:22] mzanetti, ok, thanks! === karni is now known as karni-food === karni-food is now known as karni [14:30] karni: that was quick :) [14:30] @unity: standup [14:30] kgunn, ↑ [14:30] mzanetti: hahahah well.. I decided changing on canonical irc was sufficient ;) === pete-woods is now known as pete-woods-lunch [14:39] Saviq: any ETA on the new scopes being ready to be merged? Sounded like its gonna be soon too [14:41] * mterry adds notes to doc [14:41] sorry for missing [14:41] In any case, I'll polish/add some tests for the right edge stuff today and tomorrow morning we can figure a plan on how to get stuff merged. [14:43] Saviq, ideas how to translate card background into mascot-only background in the preview? [14:44] Saviq, or well... should header widget support something like "mascot-background" attribute? [14:44] Saviq, i'm asking cause i just previewed TWC scope [14:45] mhr3, +1 [14:45] can I get a review here, please? [14:45] https://code.launchpad.net/~elopio/unity8/fake_app_from_toolkit/+merge/208002 [14:46] oh, no, wait, the scroll branch hasn't landed yet. [14:46] I saw it, I suppose I was confused. [14:46] mhr3, not sure what you're telling me? have an image? [14:48] Saviq, have working screenshoting in mir? :) [14:48] Saviq, just preview twc scope [14:48] mhr3, rm ~/.display-mir; reboot; lxc-console -nandroid -t0; screencap foo.png [14:48] mhr3, don't have mwc scopes on my devices any more :0 [14:49] Saviq, it's not mwc scope, it's regular server one [14:49] you'll get it on desktop too [14:49] mhr3, ah ok [14:49] oh... desktop [14:49] shouldn't be too hard to screenshot that [14:50] lol [14:50] mhr3, yeah, not here on desktop for some reason [14:51] Saviq, http://imgur.com/b4MCd7D [14:51] Saviq, missing blue background on the icon [14:51] mhr3, weeell, the icon is missing the background, not the mascot :P [14:52] Saviq, that's why i'm suggesting mascot-background [14:52] mhr3, comment on the spec please [14:52] ok [14:53] mhr3, but yeah, if we support card-background, it should map into the preview, too, probably [14:53] Saviq, right, i just don't want to map it completely, so that the entire header would be colorized [14:54] mhr3, no no, of course [14:54] mhr3, it'd be the whole preview (as the whole card is) ;) [14:54] heh [14:57] Saviq, btw who's going to respond to these comments now? :/ [15:07] mhr3, I am [15:07] Saviq, ah, cool, don't forget to reply to those comments about the table widget ;) [15:07] ;) [15:14] ok, now I'm ready for a review: [15:14] https://code.launchpad.net/~elopio/unity8/fake_app_from_toolkit/+merge/208002 [15:26] mzanetti, can you do ↑, I thought it could help with the click issue === charles_ is now known as charles === dandrader is now known as dandrader|afk === alan_g is now known as alan_g|tea [15:44] The diff has been truncated for viewing. [15:44] sorry :D [15:45] Saviq, is there a renderer that allows variable height cards? === alan_g|tea is now known as alan_g [15:51] Ran 60 tests in 138.775s [15:51] FAILED (failures=58) [15:52] autopilot is a bit broken in new-scopes :D === dandrader|afk is now known as dandrader [15:58] davidcalle, vertical journal, but it's not integrated yet [15:58] davidcalle, https://code.launchpad.net/~unity-team/unity8/new-scopes-vj-integration/+merge/201932 [15:58] Saviq, thanks [16:14] davmor2: mhr3: hi, i'm here [16:15] achiang, ETIMEDOUT [16:19] achiang: awesome I think it was ricmm that wanted to talk to you. I think we have a potential fix for the images performance issues tested with 10 apps open but ricmm can give you more details === pete-woods-lunch is now known as pete-woods [16:24] * tsdgeos hates autopilot tests, they are ultra broken in new-scopes :'( [16:28] elopio, can you help tsdgeos on the status of that ↑? [16:29] elopio, i.e. we should update the emulators for new-scopes, right? /me recalls a branch floating around, not sure... [16:29] Saviq: oh, didn't know someone he was trying to make them work, i was trying to get myself to convince to start doing that [16:29] Saviq: will you have a look at https://code.launchpad.net/~aacid/unity8/new-scopes-cleanup/+merge/207921 ? [16:30] it contains the removal of all files we don't need and makes qmluitests pass for me [16:31] tsdgeos, Saviq: I can help, that's why I'm here. [16:31] I'm in a meeting, will finish in 45 minutes. tsdgeos, will you still be around? [16:31] good stuff :) [16:31] wasn't planning to [16:32] but i can stay a bit longer [16:32] if you want to talk about it [16:48] tsdgeos: if you have to leave, send me an email with what you have done. So I'll continue on my evening, and we sync again tomorrow. [16:56] elopio: ok, will do that === jasoncwarner___ is now known as jasoncwarner [17:35] ah mzanetti forgot to let you know [17:35] mzanetti, had to revert the running apps aspect, it wasn't working well, some Changed signals were not emitted [17:36] I remember I had some issues like that with Qt.size at some point [17:54] Saviq: ok. no prob. seems to work well in my branches (I don't use QSize like that in there) [17:55] mzanetti, yeah, but you'll need to for rotation, won't you? [17:56] don't think so... as images come from the image provider directly. [17:56] mzanetti, right [17:56] mzanetti, but they'd have to be rotated in the provider then? [17:56] yep [17:56] mzanetti, ok that works === dandrader is now known as dandrader|lunch === robru is now known as robru-is-deathly === robru-is-deathly is now known as robru-is-dying [18:21] Saviq, something's super broken with unity on my device [18:21] mhr3, ? [18:21] Saviq, when i swipe left/right between scopes it doesn't work from most categories [18:21] ever seen that? [18:21] mhr3, no [18:22] mhr3, meaning like the renderer would take over the events? [18:22] Saviq, plus carousels don't scroll [18:22] mhr3, something got pretty broken your side [18:22] mhr3, that mwc? [18:22] yep === gatox_ultra is now known as gatox [18:23] Saviq, wanna see it? it's super weird [18:23] mhr3, /me flashes mwc [18:23] it was working fine for a long time [18:24] now after a couple of hours it's like this [18:24] mhr3, what device? [18:25] mako [18:26] Saviq, ah, got pattern - swipes/scrolling work fine from the first category, but not the rest [18:28] and maybe not really [18:29] wow.. [18:31] Saviq, maybe it's actually a MouseArea covering the bottom ~half of the screen that makes everything un-interactive [18:31] rings a bell? [18:31] mhr3, that sounds like OSK [18:31] mhr3, either maliit died [18:31] mhr3, or the connection between OSK and shell [18:32] let me check if it's alive [18:32] mhr3, still worth a bug [18:32] maliit-server? [18:32] that one's running [18:33] and osk does appear [18:33] and disappear [18:35] Saviq, and now started an app and can't get from it cause the edges don't work [18:35] eh, time for reboot [18:41] mhr3, yeah, there's a socket connection in unity8 to maliit-server that gives up after some time [18:41] mhr3, or to ubuntu-keyboard, actually [18:41] something of this sort must've failed === dandrader|lunch is now known as dandrader [19:33] Saviq, there's the OSKController item from unity-mir that puts a MouseArea to block events intended to the OSK from reaching items in unity8. So it might be that the OSKController thinks the OSK is up when it's not [19:33] that would cause this bug [19:34] @ that discussion you were having with mhr3 an hour ago :) [19:37] dandrader, yeah, that's what I meant [19:37] dandrader, maybe something didn't reach IRC? === veebers_ is now known as veebers === salem_ is now known as _salem === _salem is now known as salem_ === salem_ is now known as _salem === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader === bregma_ is now known as bregma