=== marcusto_ is now known as marcustomlinson | ||
duflu | tsdgeos: That setSurface(null) from the QML code... does that actually execute in the rendering thread or not? | 08:58 |
---|---|---|
duflu | mzanetti, ^ ? | 08:59 |
mzanetti | duflu, yes | 09:00 |
duflu | mzanetti: Damn. You sure? | 09:00 |
duflu | :) | 09:00 |
mzanetti | duflu, yes... I don't think there's any multithreading happening when inside QML | 09:02 |
mzanetti | duflu, communication between QML and C++ is happening through method invokations which Qt automatically decouples thread-wise | 09:03 |
tsdgeos | duflu: is the rendering thread the main thread? | 09:05 |
duflu | tsdgeos: I have no idea. I thought it was not... | 09:05 |
tsdgeos | in qt quick sense no | 09:06 |
duflu | tsdgeos: No?... http://doc.qt.io/qt-5/qtquick-visualcanvas-scenegraph.html#threaded-render-loop-threaded | 09:06 |
tsdgeos | no idea if some mir nomenclature is slipping in or not | 09:06 |
tsdgeos | since i've no idea about mir :D | 09:06 |
tsdgeos | duflu: so no, setSurface will be executed in the main thread, not in the rendering thread | 09:06 |
tsdgeos | there's very few things that run in the render thread | 09:08 |
tsdgeos | at least that we can touch | 09:08 |
tsdgeos | mzanetti: ↑↑↑↑↑ | 09:08 |
mzanetti | ermm... ok | 09:09 |
mzanetti | sorry.. | 09:09 |
mzanetti | right... with mir we're probably talking of different rendering threads... | 09:10 |
tsdgeos | i guess you mixed up the rendering thread and main thread names? | 09:10 |
tsdgeos | yeah E_TOO_MANY_THREADS :D | 09:10 |
mzanetti | to me the main thread is the rendering thread :D | 09:10 |
mzanetti | as it's the one where qt does the ui stuff | 09:10 |
tsdgeos | yeah, not in quick | 09:10 |
tsdgeos | there's a separate thread that does almost everything rendering wise | 09:11 |
=== alan_g is now known as alan_g|lunch | ||
=== karni-away is now known as karni | ||
=== alan_g|lunch is now known as alan_g | ||
dandrader | Saviq, now that mir 0.16 is out, can we get a silo with the mousePointer branches? | 12:59 |
Saviq | dandrader, I've already a silo for unity8 and qtmir, are mousePointer branches ready to land? | 12:59 |
dandrader | Saviq, yeah, the only thing holding them back was a mir bug and its fix came in mir 0.16 | 13:00 |
Saviq | dandrader, ok then, can you please let me know the MP links, I'll add to the silo and rebuild | 13:00 |
dandrader | Saviq, altough gerry left for holidays before being able to top approve them all because of this | 13:01 |
dandrader | Saviq, sure. just a sec | 13:01 |
dandrader | Saviq, https://code.launchpad.net/~dandrader/unity8/mousePointer/+merge/271608 | 13:01 |
dandrader | Saviq, https://code.launchpad.net/~dandrader/unity-api/mousePointer/+merge/271620 | 13:02 |
dandrader | Saviq, https://code.launchpad.net/~dandrader/qtmir/mousePointer/+merge/272027 | 13:02 |
Saviq | dandrader, update debian/control to say unity-application-9 where applicable | 13:05 |
Saviq | dandrader, I'll drop the bump-application-api branches then | 13:06 |
ltinkl | dandrader, hi, is there a branch that makes qtmir compile against mir 0.16? | 13:08 |
dandrader | ltinkl, yes and it should have landed in trunk | 13:08 |
ltinkl | dandrader, mir::events::make_event bails out due to deprecation warnings | 13:08 |
ltinkl | dandrader, I branched trunk a few moments ago and it still fails to build | 13:09 |
ltinkl | dandrader, mirbuffersgtexture.cpp also fails to compile (missing gl_bind_to_texture) | 13:10 |
dandrader | ltinkl, revision 378 ("Port gl_bind_to_texture to the new TextureSource interface. Update for Mir 0.16 release.") | 13:10 |
dandrader | ltinkl, that's the one | 13:10 |
ltinkl | dandrader, hmm | 13:10 |
dandrader | ltinkl, it does update mir::events::make_event usage | 13:11 |
* ltinkl tries a merge | 13:11 | |
Saviq | ltinkl, it only landed an hour ago or so | 13:14 |
ltinkl | Saviq, ah right, that explains it | 13:14 |
dandrader | Saviq, they already do that (unity-shell-application=9) | 13:16 |
Saviq | dandrader, debian/control don't | 13:16 |
Saviq | dandrader, not for unity8 | 13:16 |
dandrader | Saviq, ah, checked CMakeLists.txt only | 13:17 |
Saviq | which is really the less important bit ;) | 13:17 |
mterry | ltinkl, heyo! you busy? I'm wondering about the error I emailed you about | 13:17 |
ltinkl | mterry, fixed, and rebuilt the silo | 13:17 |
dandrader | Saviq, it's not!! | 13:17 |
mterry | oh really? sweet, thanks! | 13:17 |
Saviq | dandrader, for us it is | 13:17 |
ltinkl | mterry, didn't you get my email about it? | 13:17 |
dandrader | Saviq, you can't compile if there's a missmatch in the CMake file | 13:18 |
Saviq | dandrader, exactly | 13:18 |
mterry | ltinkl, I didn't notice it... let me double check | 13:18 |
Saviq | dandrader, so you'll know there's a problem soon enough | 13:18 |
dandrader | Saviq, so the cmake is the most important bit | 13:18 |
mterry | ltinkl, no... and not in spam either :) | 13:19 |
ltinkl | mterry, weird, it's in my "sent" folder, nvm :) | 13:19 |
Saviq | dandrader, I disagree, because it's obvious when there's a problem there | 13:19 |
Saviq | dandrader, dependencies only show up as problematic in specific conditions | 13:19 |
Saviq | on top of that, Ubuntu is our only release target for now | 13:20 |
dandrader | Saviq, as a developer building the code, that's pretty effective/important. but anyway... | 13:20 |
mterry | ltinkl, huh. now I wonder what else I'm missing :) | 13:20 |
Saviq | dandrader, that's because you're never building the packages ;) | 13:20 |
Saviq | dandrader, I do | 13:20 |
dandrader | Saviq, yeah, building packages is so boring | 13:21 |
dandrader | :) | 13:21 |
dandrader | Saviq, so you mean "s/unity-application-impl-6/unity-application-impl-9" in debian/control of qtmir and unity8? | 13:21 |
Saviq | dandrader, fixing boottests is even more boring | 13:21 |
Saviq | dandrader, yes | 13:21 |
ltinkl | dandrader, package 'mir-renderer-gl-dev>=0.16' not found | 13:21 |
ltinkl | dandrader, what am I missing? :) | 13:22 |
Saviq | ltinkl, that's a new package | 13:22 |
Saviq | ltinkl, sudo apt-get build-dep qtmir | 13:22 |
dandrader | ltinkl, mir-renderer-gl-dev version 0.16 | 13:22 |
dandrader | :) | 13:22 |
ltinkl | obviously :) | 13:22 |
ltinkl | Saviq, build-dep didn't find anything new to install | 13:23 |
ltinkl | that's a broken dep imo | 13:23 |
dandrader | Saviq, done | 13:24 |
Saviq | ltinkl, or you didn't update your apt | 13:24 |
ltinkl | Saviq, I did | 13:24 |
Saviq | ltinkl, apt-cache showsrc qtmir | 13:24 |
dandrader | ltinkl, "sudo apt-get install mir-renderer-gl-dev"? | 13:24 |
Saviq | check Build-Depends | 13:24 |
ltinkl | dandrader, ye that works ofc | 13:25 |
Saviq | dandrader, don't ;) | 13:25 |
ltinkl | too late :) | 13:25 |
Saviq | ltinkl, anyway, wfm | 13:25 |
Saviq | dandrader, btw, "FIXME: Get this out of unity-api"? why do we put it there in the first place? | 13:27 |
dandrader | Saviq, because we dont have a "libqtmir-dev" | 13:28 |
dandrader | Saviq, we are using unity-api for that instead | 13:28 |
dandrader | Saviq, but first we have to move Application and ApplicationManager out of qtmir | 13:28 |
Saviq | dandrader, qtmir bit has no ACK from anyone, but has a NeedsFixing form ltinkl, not ready to land in my book, is it... should it wait until Monday? | 13:29 |
dandrader | Saviq, as I said, gerry left before being able to give the final ack. could you put it in a silo at least? I'm sure gerry will be happy to bless it on Monday | 13:29 |
Saviq | dandrader, ok, separate silo then | 13:30 |
dandrader | could you mir-renderer-gl-dev\ltinkl | 13:30 |
dandrader | bleh! | 13:30 |
Saviq | ;d | 13:30 |
dandrader | ltinkl, could you update your mousePointer comment please? | 13:31 |
Saviq | dandrader, btw, on the implementation side, so we're drawing and compositing the cursor ourselves? so no hardware layer for it? | 13:31 |
dandrader | Saviq, right | 13:31 |
ltinkl | dandrader, sure | 13:32 |
Saviq | shame :/, we need to think of using HWC asap | 13:32 |
dandrader | Saviq, don't find it so crucial. if your compositor is stuttering, having a fluid mouse hovering over the mess won't save you anyway. | 13:34 |
ltinkl | Saviq, dandrader: comment on qtmir/mousePointer updated | 13:34 |
Saviq | dandrader, maybe it'd stop stuttering if you composited the cursor in HWC ;) | 13:35 |
Saviq | dandrader, but also, battery life, all kinds of things | 13:35 |
dandrader | Saviq, yeah, I'm not advocating *against* it | 13:35 |
Saviq | I just mean we degraded in that sense | 13:36 |
Saviq | because it was hwc'd when u-s-c was drawing it | 13:36 |
Saviq | dandrader, https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-015 | 13:36 |
Saviq | building unity-api first | 13:36 |
dandrader | Saviq, thanks | 13:36 |
dandrader | ltinkl, you can open cursor files with gimp, btw. it's pretty nice | 13:37 |
ltinkl | dandrader, ye I did try, it's fine; I just got scared by the xterm cursor name initially :) | 13:38 |
dandrader | ltinkl, yes, it's misleading | 13:38 |
dandrader | ltinkl, "sudo apt-get build-dep qtmir" did bring the new mir-renderer-gl-dev dependency. but I'm on wily | 14:15 |
ltinkl | dandrader, not on vivid | 14:15 |
dandrader | ltinkl, I think the overlay ppa in vivid doesn't include source packages or something, so you don't get updated build-deps info from it | 14:15 |
dandrader | ltinkl, I've seem this kind of issue before in vivid+overlay | 14:16 |
dandrader | ltinkl, can't trust build-dep too much there | 14:16 |
=== dandrader is now known as dandrader|afl | ||
=== dandrader|afl is now known as dandrader|afk | ||
Saviq | dandrader|afk, unity8 failed to build in the silo, please have a look https://launchpadlibrarian.net/218825620/buildlog_ubuntu-wily-amd64.unity8_8.11%2B15.10.20150925-0ubuntu1_BUILDING.txt.gz | 15:29 |
Saviq | https://requests.ci-train.ubuntu.com/#/ticket/420 | 15:29 |
Saviq | o/ | 15:29 |
=== dandrader|afk is now known as dandrader | ||
dandrader | Saviq, fixed. I had it fixed before but it must have been lost in some local branch... | 16:29 |
dandrader | kgunn, do you have the permissions to trigger a rebuild of unity8 in silo 015? https://requests.ci-train.ubuntu.com/#/ticket/420 | 16:30 |
dandrader | kgunn, just pushed a fix for it | 16:30 |
kgunn | dandrader: lemme ee | 16:31 |
kgunn | dandrader: weird that you don't ? | 16:31 |
dandrader | kgunn, I don't have landing rights | 16:32 |
kgunn | dandrader: ok, just kicked | 16:33 |
dandrader | kgunn, thanks | 16:34 |
=== alan_g is now known as alan_g|EOW | ||
=== dandrader is now known as dandrader|afk | ||
=== dandrader|afk is now known as dandrader |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!