[10:35] <duflu> Glorious 1-frame lag compositing. I'll try to get it down to zero before Christmas :)
[10:39] <alan_g> duflu: client of nested?
[10:40] <duflu> alan_g: No nesting :)
[10:40] <duflu> Just talking about system compositors
[10:40] <alan_g> pity
[10:40] <duflu> alan_g: Next week or so will get the compositor lag from 2 to 1 frame. Then will think about zero
[10:42] <alan_g> Sounds like a plan
[10:42] <duflu> And regretfully work around intel bugs
[10:43] <duflu> They on guarantee rendering will finish if you glFinish() once or swap buffers _twice_.
[10:43] <duflu> Crazy
[10:43] <duflu> *They only
[10:44] <alan_g> Have they accepted it is a bug?
[10:44] <duflu> alan_g: Haven't reported anything. Only got the confidence that it is a driver bug this arvo
[10:44] <alan_g> ack
[10:45]  * duflu -> weekend
[10:45] <alan_g> Have a good one!
[13:40] <greyback_> alan_g|lunch: do trust sessions need an ordinary app in them? Or could a trust helper start its own trust session and that's enough. I'm thinking of future when OSK will be a trusted helper of shell
[14:03] <alan_g> greyback_: I think they do: PromptSessionClientAPI.when_application_pid_is_invalid_starting_a_prompt_session_fails
[14:13] <greyback_> alan_g: ok, that might be something we'll need to look at later.