Closed Drakulix closed 2 months ago
The
toplevel_enter
event includes the rectangle when we get the event though, so I guess the panel can just do that calculation.
Yeah, that is what I thought about when adding that parameter. We could obviously also do a set_overlap_region
, but that feels weird and none of the already available regions fit this.
What we truly need would be something like set_window_geometry
for layer-shell, but I don't wanted to dive into that rabbit hole with a small extension protocol, since that probably would have implications for layout as well.
So if there is nothing wrong in your opinion with having the panel do the necessary calculations, I would just go on with this approach.
Yup, I don't have any problem with the panel doing the calculations.
Are there any cases where we anticipate two potentially-overlapping layer surfaces that both react to layer_enter
events? I can imagine that causing weird feedback loops.
Probably not an argument against this protocol (if there's no obvious design that prevents that) but it's something we should be aware of when using it.
Are there any cases where we anticipate two potentially-overlapping layer surfaces that both react to
layer_enter
events? I can imagine that causing weird feedback loops.
All use-cases we have right now should not cause any clients to resize/move their surfaces. So yeah, something to be aware of, but not an immediate issue.
This protocol is intended to give layer-surfaces some information about overlapping elements.
Use cases are:
Please confirm if this looks usable. The intention is for cosmic-panel to watch for
toplevel_enter/leave
events, while the cosmic-files would probably only track exclusive layer-shell surfaces on it's layer and higher up for it's grid.