Stacking subsurface siblings

[Pekka Paalanen]

On Tue, 16 Jun 2015 23:42:43 +0800
Jonas Ådahl <jadahl at gmail.com> wrote:

> On Tue, Jun 16, 2015 at 08:02:52AM -0700, Jasper St. Pierre wrote:
> > I was not aware you could stack subsurfaces under a parent surface at
> > all. Is this intended protocol behavior? The fact that you might be
> > able to do that at all in Weston might be a bug.
> 
> From wl_subsurface.place_above:
> "... The reference surface must be one of the sibling surfaces, or the
> parent surface."
> 
> place_below is semantically equivalent except the stacking order. To me
> this reads as it is explicitly allowed to stack a subsurface below a
> parent surface.
> 
> Now how that should work if the parent surface stacks itself above some
> other surface is undefined I assume (whether the subsurface that is
> stacked below its parent surface should be below or above the surface
> the parent surface is stacked above, but I assume it subsurface tree
> should be considered tightly coupled and be stacked together, if we'd
> want to define that behaviour.

Yeah, think of it as a tree. A surface with its immediate children is
tightly coupled. No ancestor surface can get between a surface and its
siblings+parent.


Thanks,
pq