[metapost] re: MetaPost 0.891 announcement

[Larry Siebenmann]

Hi Taco,

Your answers are helpful, but there are still a few
patches of fog in my view.

Concerning 5:

 > 5. The EPS output no longer contains spaces
 > within PostScript strings

Which processors were troubled by spaces in 
those PostScript strings; and why?

Concerning 6:

 > 6. The turningnumber was sometimes wrong in unexpected ways. The new
 >    implementation is still sometimes wrong (when there are strange path
 >    segments involved), but in a much more predictable way: the new code
 >    always draws straight lines between the knot points, and calculates
 >    the turningnumber based on that path instead of the actual path. The
 >    effect is that cusps and loops within segments are now completely
 >    ignored. A more thorough fix of turningnumber is planned for the
 >    next release.

[[ Knotty remark: Knuth does not use the term "knot point" in mfbook
(?!).  But one can talk of the "integral points"  or "integral time
points" of a path and be understood by all users of mp and mf.]]

I am not keen on introducing systematic functionality in any public
mp version that is ab initio known to be inappropriate and impermanent.

A more orthodox alternative is to introduce the new primitive

              "secantturningnumber"

for what you proposed for the interim "turningnumber"
and maintain it forever as a useful sanity check on the 'optimum'
turningnumber that has not yet been decided and debugged.

I have railed often against ignoring the inevitable instabilities
of turningnumber. They should be signaled somehow.  But how?
Example: Consider the mirror image 'almost perfect' circles

             (100,100){left}..(100,200)..cycle;
             (100,100){right}..(100,200)..cycle;

With your current proposal, what does mp say about their respective
turningnumber's (ie secantturningnumber's).


Cheers

Laurent S.