No subject

deactivated in its WM_MDIACTIVATE handler, there is no effect at all. Only
the handler for the window that is becoming activated can, in turn, activate
a different window instead. This indicates that the second attempt in
Windows to deactivate won't fail, because it is not activating another
window on the second go.

Yes, this behavior is a bit odd and a regression test is warranted. I was
afraid someone would ask for one :) I'll see if I can find some time to work
on that this week.

Zach