Hi,

> I've not been able to find it quickly (must have displaced it 
> somewhere with that beautiful proof for which the book margin
> was too small :-) I guess it might be on my notebook. Will look
> later today. Anyway, it was not a complete solution but mostly a
> proof of concept to test how fast rectangle lists for clipping
> regions on the BitBlt level could be when done completely in
Smalltalk.

That's what I'm interested in, too. Benchmarking the morphic redraw
cycle shows that there should be significant improvements possible by
keeping a list of invalidated regions in the canvas around. I was
suspecting that this may be in the range of 10-20+% but I never had the
time to get into it.

Cheers,
  - Andreas