Hi All,

A bit of rivalry of different approaches is a good thing. I like 
to keep an open mind and judge these issues on a case-by case 
basis. Nobody's prejudices deserve to be law.  Not even Paul 
Bourke's, who is a leading virtuoso**  *both* in computer graphics and
and in Mac programming, toolbox included.

Bob Covington <t88@...> remarks

 > LINETO can be rewritten in FB 3 to be twice as fast or better 
 > using a custom routine, for example. It is too slow for realtime 
 > 3D use.

Hmm.  I didn't know that.  Does that use PPC assembler?

Statz' neat 5 lines of toolbox code can indeed replace
Franklin's neat 10 lines of C (or FB). Both are very clear.
But what is the speed ratio?

Joe Wilkins opines

 > Apple has done a very thorough job, and most of
 > it is faster, now that we have access to PPC,
 > than any C code anyone can come up with, so why
 > not spend all of this energy exploring the
 > enormous potential of the Toolbox, and leave the
 > "convoluted" C thinking where it belongs - on
 > PCs.

C comes from unix and the 1970's.  So forget PCs. Let's say
linux to be more relevant; linux is in effect open source
unix.  The big argument for classical C like Franklin's code
snippet is that will be good 25 years hence -- as it was 25
years ago. I know of no distinctly better way of salting away
a neat algorithm for future use.  (Except maybe plain English;
I say maybe since translation to and from English is
problematic.)

That is why I want two-way access C <==> FB.

Bob Covington <t88@...> adds:

 > I have posted a Bezier curve demo and source at
 > http://www.mindspring.com/~artlythere/fb.html

At it Future Baby programmers!--
B'ezier curves, informally known as French curves, are
essential for programming good hooters.

I'm only half kidding.  They were really introduced for
French autobody design --- by French engineers B'ezier and 
de Casteljau.

               Cheers

                    Larry Siebenmann

** PS. Check out Paul C. Bourke's web site 
(via any search engine). Truly sensational.