[12:49] <MacSlow> Does anybody know what package is needed for this issue http://pastebin.ubuntu.com/11767536 ?
[12:50] <MacSlow> I do have ubuntu-ui-toolkit-autopilot installed, so I wonder what else could be missing.
[14:47] <cimi> MacSlow, asked sdk?
[14:47] <MacSlow> cimi, already sorted
[15:48] <rhuddie> mzanetti, hello, I've updated this mp after your review, please do take another look: https://code.launchpad.net/~canonical-platform-qa/unity8/fix1306340-deprecate_emulators/+merge/259202
[15:53] <mzanetti> rhuddie, did you just really make our jenkins green? thanks so much!
[15:53] <rhuddie> mzanetti, i'm not sure if its a one off :)
[15:54] <mzanetti> rhuddie, yeah... maybe... there's still the issue that sometimes it doesn't connect, right?
[15:54] <mzanetti> I had a look list night but didn't come far with the time I had at hand
[15:54] <mzanetti> s/list/last/
[15:55] <rhuddie> mzanetti, I've had many troubles where it can't introspect unity8 process, just see some of the above jenkins
[15:55] <mzanetti> yep... that's the one I mean
[15:55] <rhuddie> yeah, that is still there
[17:46] <olli> mterry, dandrader|lunch, cimi, mzanetti, is it true that I can't use an env var in Qt/QML/JS without implementing a helper in C++?
[17:47] <mzanetti> olli, I think so, yes
[17:47] <mzanetti> unless someone came up with something in the meantime
[17:47] <olli> :)
[17:47] <mzanetti> olli, what env var are you interested in?
[17:47] <mzanetti> you should not need that normally
[17:48] <olli> mzanetti, SNAP_APP_DATA_PATH and friends
[17:48] <mzanetti> I see... no, quite sure haven't anything there...
[17:48] <mzanetti> actually
[17:48] <mzanetti> olli, QStandardPaths should have this
[17:49] <mzanetti> olthough I'm not sure if there's a QML element for that :D
[17:49] <mzanetti> let me search
[17:49] <olli> that's what I found http://stackoverflow.com/questions/16408691/get-home-and-or-username-in-qml
[17:50] <mzanetti> yeah, that would work... if you have a main.cpp you can even just set it as a context property
[17:50] <olli> hrg
[17:51] <olli> my run a .qml in snappy just got complicated by a order of magnitude
[17:51] <olli> +project
[17:51] <olli> mzanetti, thx for confirming
[18:00] <mzanetti> olli, hmm... makes me think of: is there Qt on snappy now?
[18:00] <anpok_> dandrader|lunch: ping - I want to add stuff to MirSurfaceItem
[18:13] <olli> mzanetti, so close
[18:13] <dandrader> anpok_, what stuff?
[18:14] <anpok_> dandrader: with 0.14 some of the nested bypass changes will land
[18:15] <anpok_> kdub added the ability to surfaces to contain multiple buffer streams
[18:16] <anpok_> that means a mir::scene::Surface will provide a vector of Renderables..
[18:16] <anpok_> I guess for now it would work if we just pull the first Renderable out..
[18:17] <kdub> yes, that is what should be done
[18:17]  * kdub suspects I broke something, but dont have the problem context
[18:17] <anpok_> but I was wondering if QQUickItem then needs multiple QSGTextureProvides..
[18:18] <anpok_> or if is fine when QSGTextureProvide::texture just yields one texture, while MirSurfaceItem::updatePaintNode adds multiple nodes..
[18:18] <anpok_> multiple image nodes to the scene graphi
[18:18] <anpok_> -i
[18:21] <dandrader> anpok_, btw, why does a surface (will) have multiple buffer streams?
[18:22] <kdub> to accommodate when a nested server wants to defer composition (specialized 2d operations, usally) to the host server
[18:22] <kdub> skip a composition step in the middle, for simple 2d scenes
[18:23] <anpok_> but not only for nesting... i.e. video player apps could stream the video in yuv surface while having a simple rgba overlay...
[18:23] <anpok_> so no composition is needed during playback
[18:23] <anpok_> in the client..
[18:23] <kdub> right, and other situations where rendering can be optimized
[18:25] <kdub> but the compelling first intended usage is to reduce latency in the nested case
[18:26] <anpok_> .. in which renderables[0] would be sufficient
[18:26] <dandrader> still don't understand it, but well...
[18:26] <anpok_> untill a fancy client comes along
[18:27] <kdub> dandrader, basically, to make it possible for u8 to use hwc overlays or bypass
[18:27] <kdub> even when u8 is running as a usc client
[18:29] <kdub> anpok_, does something need fixing?
[18:31] <anpok_> hm not in mir so far just some event header things .. smaller event_builder changes.. and now generate_renderables(...)[0]
[18:35] <kdub> anpok_, okay, let me know if I can help with the generate_renderables() stuff then