[11:59] <greyback> alan_g: are you aware of a bug like this lp:1633052
[11:59]  * alan_g looks
[12:01] <alan_g> greyback: I was not aware of that
[12:01] <greyback> far from critical anyway
[12:02]  * alan_g will try to investigate this afternoon
[14:27] <alan_g> greyback: I've not been able to reproduce (yet). Your 0.25 compatibility code is wrong. (See comments.) Going to try with Mir-0.24.0 next.
[15:48] <alan_g> greyback: https://code.launchpad.net/~alan-griffiths/miral/fix-1633052/+merge/308397
[15:50] <greyback> alan_g: that again? :)
[15:51] <alan_g> Mir stupidly has two structures to specify surfaces - one for create, one for modify
[15:51] <greyback> so I see. That's unfortunate
[15:52] <alan_g> I only fixed the "create" version and was too stupid to check "modify"
[15:53] <alan_g> It probably all comes back to trying not to break existing APIs
[15:54] <greyback> *nod*
[15:54] <alan_g> BTW your code is wrong https://bugs.launchpad.net/miral/+bug/1633052/comments/3
[15:54] <greyback> alan_g: yep, I've seen that
[15:54] <greyback> thanks
[15:54] <greyback> working on testing it with trunk mir, just to make sure
[15:55] <alan_g> Cool, I won't bother to do that
[15:58] <alan_g> If you're finding stuff like this then there must be a lot else working OK?
[16:01] <greyback> alan_g: well there's little funny things I cannot reliably reproduce. Sometimes when a surface is destroyed, it doesn't trigger a redraw to show the surface is gone
[16:01] <greyback> but making a testcase for it, I can't reproduce
[16:03] <greyback> I'm also seeing the way Qt does menus does make miral decide teh focused surface strangely, but Qt can jut work with it
[16:04] <alan_g> greyback: I've seen that occasionally with Xmir - but wasn't sure (yet) which was to blame.
[16:04] <greyback> in general, it's working well. Some crashiness when running on X, so am focusing more on actual hardware
[16:05] <greyback> I'm still getting drawing bugs with Qt, which I need to sort out