Code review comment for lp:~vanvugt/mir/hidden

I'm also of the opinion that it doesn't hurt to have 'hidden' as a surface state. Shells do not have to use it if they don't want to.

One problem that I noticed in the implementation is that set_hidden() gets called from within set_state(). Both of these notify the observers. Essentially we are reporting the same event twice.

I suggest we do away with 'hidden_set_to' since now that hidden is a state, it'll be reported as the others through set_state. However, that'd mean show() and hide() would no longer trigger 'hidden_set_to' to the observers. Not sure if that's something we want to keep. So something needs to be done about them as well.