http://bugs.winehq.org/show_bug.cgi?id=10142
--- Comment #34 from Alexandre Julliard julliard@winehq.org 2008-10-01 16:58:37 --- (In reply to comment #32)
Comment #10 says the exact opposite -- that wine is acting in the way it does to work around the behavior of other window managers; and that the behavior of wine is wrong but correcting it is not trivial.
My patch, which is rather trivial, is probably missing something obvious. But it works for my test case in compiz, fvwm2, metacity and kwin. I was rather hoping someone would comment on the proposed patch rather than getting in to a protracted circular argument about whether the (confirmed) bug is actually a bug.
Yes, it's a bug, it should be fixed, but that's not easy or it would have been done already. Your patch is not correct, what we really need is to keep track of the off-screen state and distinguish between window being moved off-screen for the first time and being moved after having been off-screened by the wm.