=== c74d is now known as Guest62930 === c74d3 is now known as c74d === chihchun_afk is now known as chihchun === c74d3 is now known as c74d === mibofra is now known as Guest6635 === chihchun is now known as chihchun_afk [11:53] Is mir packaging borked or is my system messed up? -> http://paste.ubuntu.com/11431212/ [11:54] In /usr/lib/x86_64-linux-gnu/mir I've only a client-platform subdir [11:55] something is borked [11:57] so you do not have mir-graphics-drivers-desktop installed? [11:59] anpok_, no === chihchun_afk is now known as chihchun [11:59] hm no I think thats ok.. the package that adds mir_demo_server / mir_proving_server requires neither the android driver nor the desktop/mesa drivers.. [12:05] I tried to remove all mir pacakges from my system, but for some reason it would bring along a whole lot of other packages such as xchat, update-manager, totem, etc :/ [12:06] I fail to see the connection between them [12:06] dandrader, likely because libgtk has a mir backend so link to libmirclient [12:06] trying to remove the lib removes gtk [12:06] which takes the gtk rdepends with it [12:07] seb128, sounds right [12:07] dandrader, why do you need to uninstall libmir? it doesn't force you to have the mir server installed [12:07] it's just a small lib [12:07] like you have libwayland and others installed for the same reason [12:08] seb128, just to get back to a clean slate and start over. I don't really need to remove it [12:08] k, I guess keep the lib then [12:08] yeah [12:08] or install --reinstall /vivid|wily [12:08] just found it curious that the system already relies on mir packages, even though it doesn't use them === mibofra is now known as Guest5770 [13:10] alf_, any more thoughts on: https://code.launchpad.net/~kdub/mir/surface-buffer-access/+merge/259049 ? === chihchun is now known as chihchun_afk [13:28] kdub: let me remember what was going on there... [13:30] alf_, basically, I just want to break ms::Surface : ms::SurfaceBufferAccess, because the streams provide the buffers === dandrader is now known as dandrader|afk [14:13] kdub: so, I guess I am ok with getting rid of the narrower interface at this point, and hopefully we can remove the with_most_recent_buffer() function in the future, when snapshotting is not needed anymore [14:15] alf_, I think it can even move to the private compositor::BufferStream inferface before the snapshotting is removed [14:15] once we root out the concept of the default_surface from the session [14:16] regarding daniels complaints.. have we discussed using buffer streams as-is for decoration purposes? [14:17] but I also do not see any problems to add that later [14:18] anpok_, in the context of ~kdub/mir/surface-buffer-access? or somewhere else [14:18] i was refering to yesterdays stuff.. um add-streams-to-surface [14:19] ah, so I think we're coming around on that one in general [14:20] but we have discussed it as a future possibility [14:21] ok [14:22] alf_, I think I'll collapse it then, seems everyone who has weighed in would be okay with that, thanks for the look [14:23] kdub: np === dandrader|afk is now known as dandrader === dandrader is now known as dandrader|paperw === dandrader|paperw is now known as dandrader|bbl === mibofra_ is now known as mibofra [17:04] vogons so for those keeping score on the qtubuntu/libmircommon snafu....silo 19 passed testing, but they just locked vivid [17:04] again... :-/ [17:05] :\ [17:05] so it'll land in wily and then robru said he'd create another silo to stick the pkgs for vivid [17:05] already passed qa as well...so should just be ok [17:05] just a little outta sync === davmor2 is now known as davmor2_hols === mibofra is now known as Guest52528 === Guest52528 is now known as mibofra === dandrader|bbl is now known as dandrader [19:42] racarr, running into an interesting issue. I cant seem to hold a key down with the new mir stuff? [19:42] ie. if i press and hold LEFT arrow, then press RIGHT UP DOWN, then release LEFT [19:42] all i get is LEFT down/up RIGHT down/up UP down/up DOWN down/up [19:43] same if i hold a, or b... it gets the release as soon as i press a new key [19:43] (or it possibly gets the release as soon as a press a key down) [19:45] racarr, the code that handles key events: http://paste.ubuntu.com/11438311/ [19:59] * bschaefer gets a timestamp of the events [19:59] bschaefer: processing [19:59] racarr, its a bit strange... i dont see a mir demo that really has key events [19:59] like moving an object around per say [19:59] MIR_CLIENT_INPUT_RECEIVER_REPORT=log or mir_demo_standalone_input_filter [19:59] uh [20:00] o that works as well [20:00] racarr, where does it log to? [20:00] stdout? [20:00] stdout yeah [20:00] ok [20:01] * bschaefer looks up code for LEFT/RIGHT [20:03] racarr, mir looks like its doing it right hmm [20:03] racarr, http://paste.ubuntu.com/11438557/ [20:03] holding a down while pressing b c d [20:04] bschaefer: I bet I know whats happening [20:04] action == mir_keyboard_action_repeat ;) [20:04] and you arne't handling it [20:04] so key_state is left as [20:04] 0 [20:05] which must be [20:05] SDL_RELEASED [20:05] ooo [20:05] racarr, i didnt have to handle this before :) [20:05] Tricky :( [20:05] hmm [20:05] Yeah... [20:05] well there was [20:05] repeat_count before [20:05] you should use switch [20:05] right, which i could ignore [20:05] for things like this [20:05] so the compiler will warn you [20:05] i should beable to just skip these events [20:05] IIRC [20:05] but [20:05] maybe... [20:05] lol [20:06] yeah this seems a little like a trap [20:06] (on our end) [20:06] as i think the client handles the repeat [20:06] ie. KEY IS DOWN KEEP MOVING [20:06] That's always an option yeah [20:06] and it waits for a release...but i should keep the count... [20:06] maybe... idk if sdl1.2 handles that [20:06] * bschaefer looks had headers [20:07] racarr, yeah sdl1.2 doesnt even have a repeat option for the key event [20:08] racarr, thanks for pointing that out! [20:08] bschaefer: np :) [20:08] Happy friday :D [20:08] racarr, ill have to poke you about sdl2 :) [20:08] racarr, and happy friday to you as well [20:08] for sdl2...i might have to keep track of the repeat count... [20:09] or rather...maybe it just asks if its repeating [20:09] * bschaefer will have to worry about that another time [20:09] Mm. yeah thats what most people seem to do (just bool is_repeat) [20:10] yeah ill have to double check that ... if thats the case that would be sooo much easier [20:10] * bschaefer assumes it is for now [20:17] confirmed fix my issue :), thanks again! [20:18] Yay :D [20:19] * bschaefer needs to stop playing etracer [20:23] bschaefer: you mean "testing SDL" [20:24] racarr, yes :) [20:52] hello soo unity8 desktop seems to be crashing in -proposed: [20:52] http://paste.ubuntu.com/11439219/ [21:19] bschaefer: hm i think you should have a graphics-mesa.so.2 instead [21:20] anpok, i see, installing now :) [21:20] it didnt remove so.1 soo hopefully it doesn't pick thatone [21:20] * bschaefer relogs [21:23] anpok, sweet thanks! Now it works [21:30] still curious why this is not picked by defaut [21:30] yeah that is an interesting point, as i've done many update/upgrades [21:30] since adding -proposed [23:05] racarr, sooo now another fun question... holding a key down works on a server i run my self [23:06] but if i run etracer on unity8 desktop hold right/left [23:06] does not work [23:06] err it works if i make my own server, but does not work when running through untiy8... idk if theres any differences between the layers there though...which i strange [23:06] * bschaefer adds a log [23:12] racarr, this time it does look like mirs fault [23:12] instead of sending repeat, its sending up [23:13] http://paste.ubuntu.com/11441144/ [23:13] racarr, maybe...im some how running the wrong version of the lib? (i was somehow running the wrong platform graphics lib...) [23:14] also the server im running my self is built from trunk mir [23:18] on the plus side etracer on unity8 desktop: http://i.imgur.com/TVS8bSr.jpg [23:18] (ignore bad phome camera) [23:52] bschaefer: and when you take an installed mir server? [23:53] anpok, if i use the one from mir_demos [23:53] it works fine [23:54] which should use the same libraries that unity8 desktop uses [23:58] hm for unity8 events get fed into qt and dispatched through the scene [23:58] and then converted back and sent to clients [23:59] anpok, i bet qt isn't handling the new repeat then [23:59] action [23:59] anpok, what package would that be in? [23:59] qtmir?