On 2/17/07, Richard Fateman <fateman at cs.berkeley.edu> wrote:
> Putting the wxmaxima stuff in a hash table might be a way to get it off the
> property list so that other programs more intensively using the property
> list would not have to shuffle through wxmaxima properties.
(1) I've yet to see any evidence that the length of the property
list makes a measurable difference in Maxima run time.
(2) WxMaxima properties are appended by wxMaxima at run time.
Those properties have no effect when some other user inteface is running.
> I'm not entirely convinced that running the test suite is a good test of
> whether the move-to-front heuristic for the simplifier is worth doing or not.
Feel free to present some other kind of evidence, then.
> The length of the plist for %sin in wxmaxima is 46. To say that it is just
> as fast to get item 1 as item 46 is clearly false.
It's very far from clear that the speed difference, if any, has a measurable
effect on Maxima. However, I'm willing to be convinced by some
supporting evidence.
Robert