Using that you could put relative ordering preferences of child surfaces in their MirSurfaceSpec instead. Which I think makes more sense.
(1) I think "order" is a more widely understood term than "arrangement" here. But also think the functions should be different in other ways than just name. See (0).
Instead, child surfaces could have an "int child_depth". Then their final Z-order is determined by (a) the depth of the parent surface within the main stack, plus (b) their relative int child_depth.
(0) I've proposed an attribute batching API that will probably land soon and might be helpful. Check it out: /code.launchpad .net/~vanvugt/ mir/rename- api/+merge/ 252864
https:/
Using that you could put relative ordering preferences of child surfaces in their MirSurfaceSpec instead. Which I think makes more sense.
(1) I think "order" is a more widely understood term than "arrangement" here. But also think the functions should be different in other ways than just name. See (0).
Instead, child surfaces could have an "int child_depth". Then their final Z-order is determined by (a) the depth of the parent surface within the main stack, plus (b) their relative int child_depth.