[interfacekit] Re: about the BRect bug
- From: "Philippe Houdoin" <philippe.houdoin@xxxxxxx>
- To: interfacekit@xxxxxxxxxxxxx
- Date: Tue, 3 Dec 2002 10:16:47 +0100 (added by postmaster@wanadoo.fr)
Eric wrote:
> burton666@xxxxxxxxxxx wrote:
> > I don't think this is a bug. If a Rect
> > totally contains another, the two perimeters never intersects. So
it could be that
> > Intersects() means "the perimeter intersects...".
>
> According to the BeBook "Intersect() returns true if the BRect has
any
> area -- even a corner or part of a side -- in common with rect, and
> false if it doesn't." If "rect" is entirely contained within the
BRect,
> then the entire area of "rect" is "in common" and, therefore,
> intersecting. I'm still of the opinion that this is a bug.
I think that BeBook is bugged on this topic:
bool BRect::Contains(BRect rect) and bool BRect::Intersect(BRect rect)
looks like
way too much redundant for entirely contained "rect".
Sounds like rect=5Fa->Intersect(rect=5Fb) return true *only* if rect=5Fb
really intersect
rect=5Fa, that is some area is NOT inside rect=5Fa.
Not a valid implementation of geometric intersection concept,
however...
IIRC, it was an old BeBugs entry too, but was never fixed...
-Philippe
--
Fortune Cookie Says:
Life is like an analogy
Other related posts: