[00:31] <racarr> ok I need a break for a while...left some documentation ont he bug and assigned it to me
[00:32] <racarr> but should be back in an hour or so to continue
[01:52] <racarr> The program is stopped at a permanent breakpoint, but GDB does not know
[01:52] <racarr> how to step past a permanent breakpoint on this architecture.  Try using
[01:52] <racarr> a command like `return' or `jump' to continue execution.
[01:53] <racarr> what is this -.-
[02:14] <racarr> MirSurfaceManager::sessionCreatedSurface (this=0xa86cce68, surface=0xabe06b68) with surface name 'MaliitOnScreenKeyboard'
[02:14] <racarr> MirSurfaceManager::surfaceAttributeChanged surface = MaliitOnScreenKeyboard (this=0xa86cce68, surface=0xabe0adc4, attrib=0, value=6)
[02:15] <racarr> unfortunately we have implemented a new feature where surfaces change address
[02:15] <racarr> :(
[02:16] <duflu_> racarr: We did?!
[02:17] <duflu_> There must be a std::move or something
[02:17] <racarr> maaaaybe? I am more suspiscious of memory corruption
[02:17] <racarr> or something
[02:18] <duflu> Cosmic rays
[02:18] <racarr> lol very reproducible cosmic rays
[02:28] <racarr> nvm its abi
[02:38] <racarr> no it wasnt lol
[02:45] <duflu> Free karma: https://code.launchpad.net/~vanvugt/mir/frontend-server/+merge/207602
[03:23] <duflu> Spooky while the large planes fly low over my house to and from the southern Indian Ocean search area
[06:19] <duflu> Awesome. Now I'm involved in all the code reviews, I won't be doing much coding :)
[09:26] <duflu> alan_g: One of those links is invalid and the other seems to be pure Mir, no Q-anything !?
[09:27] <alan_g> duflu: it has been a while since I followed them. Let me dig...
[09:27] <duflu> dandrader, Saviq: Can you point me to the code of the Qt-Mir-Compositor ?
[09:28] <Saviq> duflu, https://code.launchpad.net/~unity-team/+junk/qpa-mirserver
[09:29] <duflu> Saviq: Cool thanks
[09:29] <Saviq> cheers
[10:05] <alf__> duflu: Renamed AncillaryBuffersConfig to GLConfig, if you get a chance please take look before EOD, so I can move forward with this
[10:06] <duflu> alf__: Technically passed EOD ;)
[10:06] <duflu> past
[11:48] <anpok> alf__: AlbertA: kdub: alan_g: should setting focus to a session also mean applying the display configuration of that session?
[11:49] <anpok> In the scenario I look at u-s-c sets focus to the greeter.. while display is off
[11:50] <anpok> as a reaction to the focus switch the mediating display changer picks a stored display configuration which has the power state on
[11:50] <anpok> and hence turns the display on
[11:51] <anpok> i tend to believe that usc should be aware that the display is off and should set the focus to the greeter, or someone along the lines should check if the focused session is hidden.. (but maybe it isnst.. have to check)
[11:52] <alf__> anpok: @focusing=>apply display config, yes that is the intention if the session has set a custom display config
[11:53] <anpok> and we dont want to delay that if display is off - or make an exception for the power state setting ..
[11:53] <anpok> in other words focus also implies - turn it on!
[11:55] <alf__> anpok: what will change the focus while the display is off?
[11:56] <anpok> lightdm+usc do that, in the split greeter rewor - if I understand it correctly they lightdm spawns another when display is turned off
[11:57] <anpok> as soon as the greeter is done with startup and has submitted its first frame to the session usc sets the focus, so that when display is turned on it is immediately visible..
[11:57] <anpok> *rework
[11:57] <anpok> -they
[11:57] <anpok> brb need to to get some food
[11:58] <anpok> i will see if usc is somehow aware of the display state
[12:18] <ice9> what are the advantages of mir over wayland or why Canonical decided to make its own display server?
[12:34] <anpok> i dont want to get into a debate, but since this is digital communication I have the urge to correct wrong statements.. wayland is an ipc system. You have to ask why mir did not use wayland as an ipc backend.
[12:35] <alf__> ice9: See bottom of https://wiki.ubuntu.com/Mir/Spec
[12:55] <anpok> alf__: another thought on that topic - is it valid to have the power mode as a member of the display configuration?
[12:58] <alf__> anpok: Go not to the elves for counsel, for the will say both yes and no :)
[12:59] <anpok> hehe
[13:00] <anpok> well the dwarves said I should shut up and just fix the bug and then continue digging..
[13:02] <alf__> anpok: Like the brightness issue we were talking about yesterday, power mode conceptually belongs in "display configuration" but at the same time its use case differs from normal display configuration (setting up the display resolutions etc)
[13:04] <anpok> yip
[13:04] <anpok> but differs mostly because we like it to change more frequently / aor daptively to other events
[13:04] <anpok> *"or a"
[13:07] <alf__> anpok: right, so we should either introduce a new interface to do so, or reuse the same interfaces (i.e. DisplayConfiguration structs) but more intelligently (e.g. the what-changed flags we were talking about yesterday).
[13:52] <alan_g> Hmm trying to build unity-mir against devel:
[13:52] <alan_g> Linking CXX executable unity-mir-test-app
[13:52] <alan_g> /home/alan/local/lib/x86_64-linux-gnu/libmirserver.so: undefined reference to `mir::load_library(std::string const&)
[13:52] <alan_g> Where did mir::load_library go now?
[18:59] <dansuf> Hi, could you look at this log http://pastebin.com/tSnra0F7 ? I'm porting my device to ubuntu-touch and I'm stuck at black screen.
[19:02] <dansuf> also, in var/log/lightdm/unity-system-compositor.log
[19:02] <dansuf> terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<std::runtime_error> >'
[19:02] <dansuf>   what():  error binding buffer to texture
[19:03] <anpok> dansuf: dont run away
[19:03] <anpok> kdub: he is back ^
[19:04] <dansuf> anpok: I don't understand
[19:04] <anpok> dansuf: last time kdub wanted to help you but you dropped out of IRC
[19:05] <kdub> the log says EGL_BAD_CONTEXT, not sure what might be going on
[19:05] <kdub> i do see some errors around not being able to prod various files too
[19:06] <kdub> also, why is it reporting resolution 320x480, but the build keys look to be mako?
[19:08] <dansuf> kdub, it's not mako, I just used w-flo's adreno binary drivers for adreno because he wrote it solved some problems
[19:08] <dansuf> kdub, w-flo was porting touch to vision
[19:09] <dansuf> kdub, he made a commit in his repository where he updated the drivers and wrote which issues it solved so I copied them
[19:09] <dansuf> kdub, and now i can run more tests
[19:10] <dansuf> kdub, test_hwcomposer doesnt seem to fail now
[19:10] <kdub> how about mir_demo_standalone_render_to_fb?
[19:10] <kdub> you should get an animation
[19:10] <dansuf> kdub, I don't see this program in my path
[19:11] <dansuf> Maybe I have too old version
[19:12] <dansuf> I think I have mir 1.6 or sth like that
[19:12] <kdub> mir-demos package
[19:12] <kdub> may as well try the new stuff, we're fixing it all the time :)
[19:12] <dansuf> kdub, i have to leave for 10-20 minutes
[19:13] <dansuf> but i'll be back..
[19:24] <dansuf> I'm back
[19:28] <racarr> terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<std::runtime_error> >' what():  /usr/lib/arm-linux-gnueabihf/libmirplatformgraphics.so: undefined symbol: _ZNK5boost15program_options29value_semantic_codecvt_helperIcE5parseERNS_3anyERKSt6vectorISsSaISsEEb
[19:28] <racarr> mer?
[19:29] <racarr> thats a great symbol name
[19:31] <dansuf> kdub, mir-demos are installable through apt-get?
[19:31] <kdub> yep
[19:31] <dansuf> haha, and the internet isn't working on my device
[19:34] <dansuf> I'll try to chroot into ubuntu on adnroid and in that way install demos, should work
[19:46] <racarr> hmm if I ld_preload boost options
[19:47] <racarr> (why isnt mirplatformgraphics linked against it? not sure, also not sure why this doesnt present otherwise/ust started)
[19:47] <racarr>   what():  /usr/lib/arm-linux-gnueabihf/libmirplatformgraphics.so: undefined symbol: _ZN3mir7options13off_opt_valueE
[19:48] <kdub> i thought i saw a review about this
[19:51] <racarr> https://code.launchpad.net/~alan-griffiths/mir/add-mirplatform-to-dependencies/+merge/213077 maybe related
[21:23] <dansuf> kdub, I just installed the newest daily touch and now logcat gives me more info http://pastebin.com/xwYz3BYP (I might have forgotten to make some changes I did to the previous version which could cause this more extended log)
[21:23] <dansuf> I will now try to install demos (it can take some time)
[21:52] <dansuf> kdub, I've installed demos and I get the animation
[21:53] <kdub> excellent
[21:53] <dansuf> it is moving image with a green border
[21:53] <dansuf> and lots of "mdp not donfigured" in logcat because i diasbled it
[21:54] <kdub> dansuf, of course, but even at that, its a good step
[21:54] <kdub> the next step would be try running some of the server/client binaries together
[21:55] <kdub> like mir_demo_server_shell (in one terminal) and (mir_demo_client_egltriangle) in another
[21:57] <dansuf> http://pastebin.com/4awGHdir
[21:58] <dansuf> kdub, after running triangles, i'll try the others now
[21:58] <racarr> RAOF: Ping?
[21:59] <kdub> dansuf, so it looks like the server is having trouble using the buffers that the client is sending to it
[21:59] <kdub> does dmesg look suspicious?
[22:01] <dansuf> kdub, nothing interesting
[22:05] <kdub> hmm, getting the display going is usually the tougher part
[22:06] <kdub> probably something having to do with the egl errors that were in the logcat
[22:06] <dansuf> kdub, surfaceflinger also fails and gives egl bad context
[22:11] <dansuf> kdub, thanks for trying to help me, I'll do more research and I'll maybe write to people developing device-specific part
[22:12] <kdub> dansuf, yw, good luck!
[22:12] <kdub> once you get sf past that problem, mir will probably be past that problem too
[23:27] <RAOF> racarr: Yo!
[23:28] <RAOF> racarr: I'm not sick today! Yay!