From: cd_smith@ou.edu
  To  : ggi-develop@eskimo.com
  Date: Mon, 26 Jul 1999 16:05:04 -0500 (CDT)

Re: Freeze in _ggi_clip2d()

On Mon, 26 Jul 1999 I wrote:
> I'll track this down more in a sec [...]

It looks like the problem was a numerical overflow in the calculations in
_ggi_clip2d().  So it could probably be reasonably blamed on me.  But if
you'd be interested, I could look at checking for this possibility and
doing the clipping in a different way to avoid it.  It would probably be
slower, so there'd be a fast path clip with the current algorithm and a
slow path if an overflow would happen.  Interested?  In my case, I could
probably just neglect to draw the lines that result in overflow without
much problem, but that's definitely not a realistic solution for the
common case.

So what about it?  Should I write a slower large value version of the 2D
clipper?

Chris Smith <cd_smith@ou.edu>