[RFC weston] Views: Should we have a concept of "primary"?

[Bill Spitzak]

Having now understood what a "view" is, I don't think there should be a 
"primary view". This is because it looks like there can be zero views 
for a surface (ie if it is clipped by all the outputs or the compositor 
otherwise decides to hide it). So there will be instances when you 
cannot retrieve information.

The client is not supposed to know about views, so a transform should be 
stored in the surface. Whether or not that surface has a view using the 
same transform is up to the compositor.

Transient children are EXACTLY the same as subsurfaces, except the 
compositor is allowed to place other surfaces between the child and 
parent. Because of this a transient surface should probably create the 
same number of views as the parent has, since this is obviously correct 
for subsurfaces. Thumbnails may clip both transients and subsurfaces to 
the parent's area, though.

 From what little I have done with sync playback of video/audio, I think 
clients will want the frame callback for the "primary" output, no matter 
what output (including none) the surface is on. The buffer release is 
what will depend on the sync of each output.

Jason Ekstrand wrote:
> Hello all,
> Last night, I sent a set of patches to the list to split the geometry 
> information off of the weston_surface structure so that we can have a 
> single surface in multiple locations at the same time.  With this, 
> comes an interesting question: Should a surface have one view that is 
> denoted the "primary" or "default" view?  There are a number of places 
> where this comes up: