Code review comment for lp:~mterry/mir/session-for-surface

We should be doing something different than passing the session ptr into the surface. This establishes a weird dependency circle, I think.

from 'bug report' "But I can't seem to get access to the session that a surface is being created for."
This is somewhat by design, at the moment, a design that probably needs rethinking to suit the requirements of the shell. I think we should meet on monday to chat about what exactly is needed.

(I think that we're exposing cracks in 1) the "depth id" stacking system that ms::SurfaceStack has that might need changing)