Re: 2.2.1 plans

[Arnaud Charlet <charlet ACT-Europe FR>]

> I'll gladly apply your first patch (to gdkevents-win32.c), if only
> Owen (or somebody) confirms that there isn't anything fundamentally

Thanks. Actually without my first patch, the current behavior is wrong
anyway, since gdk/win32 creates coordinates that are invalid (out of range),
so I am not sure why you would not want to apply it in any case.

All my first patch is doing is using the documented API for retrieve X and Y
coordinates from an event.

> wrong (from backend-independent GTK's point of view) in having
> negative coordinates. I mean, will the rest of GTK fall over in
> mysterious ways if windows have negative coordinates, mouse events
> occur at such, etc?

As I mentioned in the bug #, positioning of menus wrongly assumes positive
coordinates, but this should be fixable.

Since Windows (and possibly other systens) use negative coordinates when
using multiple monitors, I do hope that Gtk+ does not have the limitation you
are mentioning.

Arno