[futurebasic] Re: A cry in the dark for...(Curve Fitting)

Message: < previous - next > : Reply : Subscribe : Cleanse
Home   : March 2001 : Group Archive : Group : All Groups

From: lcs@...
Date: Fri, 9 Mar 2001 11:49:23 +0100 (MET)
%


Patient Cov!

LS>It looks to me like a hacker's quick try to do better than
>MacDraw or Illustrator.

RC>That is what I need, I was just using the freehand pen tool in Illustrator
>3.2 (ancient) and it rocks. I need that. Somebody. Help. Hurry.

It seems to me that MacDraw is maybe more user-friendly (to
non-experts).  It easily allows
addition and deletion of points.  Very convenient indeed.
This leads to beautifully controlled weeding out of
superfluous control points.

I miss a tool for cutting and joining paths. Am I blind?
(Yes.  Do it at the polygon level.)

The theory for MacDraw is neat.  The b'ezier endpoints are
the midpoints of the edges of the given piecewise linear
path (polygon).  The two b'ezier curves that abut at the
midpoint of an edge are there tangent to the edge, and the 
two control vectors point to the two adjacent vertices. 
Except at the two endpoints that are regarded as midpoints
of a prolonged edge. 

It seems most reasonable to realize the MacDraw 
path setup.

The Adobe Illustr. and MacDraw setups are both well behaved under
affine linear transformation.  Knuth&Hobby are only
invariant under rigid motions and scaling.

LS> PS.  I propose to eat the following honeyed 
> words I wrote:

RC> Hope that means something nice. :)

Oh yes the honey comes from down under:
David Cottrell <David.Cottrell@...>

"Let my words be as honey, for tomorrow 
       I will probably have to eat them."

Cheers

Larry S.


PS.  IF you reaally want to understand the the K&H method,
I can give a bit of motivation.  But the gory details are too 
much for me.