Code review comment for lp:~kdub/mir/RFC-surface-arrangements

> A multi-window client has no special control over the arrangements of its
> windows; I don't believe that's necessary for the nested-bypass case (or for
> any of the other use-cases raised), and I don't think it's a good idea to
> implement.

"...it seemed useful that advanced clients (such as a *multi-windowed client*, a OS virtualizer, or a nested mir server) need a way to specify the ordering of the surfaces that they have created."

I'm not insisting on the use-case but it is both plausible and mentioned in the description. (It is also one of the volatile parts of the window management specification - so I'm not yet prepared to ignore it.)