Johan Malm

Will re-think this. Might use cmocka instead.

Ready for review. Not yet integrated with CI. It'd be nice to get this over the line + also potentially useful for the YAML patch. Cc @tokyo4j

Might this be a swapchain / scanout issue relating to the scene-graph implementation? Would you be able to test with this sway branch? https://github.com/Nefsen402/sway/tree/scene-graph

If you run “labwc -d 2> log.txt” do you get any swapchain related messages?

With the fullscreen alacritty + Firefox it feels direct scanout related. Do you get the same with Firefox starting in a non-maximised state?

@Consolatis Thanks for finding the river and sway-scene-graph issues and also homing in on the optimization relating to `WLR_SCENE_DISABLE_VISIBILITY`. Be nice to get to the bottom of this one, but...

Okay. I'll split without moving the key-handling thing into it. Yes, good idea about a private header.

Thanks. Nothing else is imminent. I've moved it to `0.7.3`

I haven't looked into this, but quick response nonetheless: - There are two protocols here. `wlr-output-power-management` (that wlopm uses) and `wlr-output-management` (that wlr-randr uses). - I think I'm right in...

What behaviour do we expect? What do openbox/mutter/etc do?