Re: [Axiom-developer] REDUCE on test.axiom-developer.org

[root]

Bob,

>I wonder if your last example on that page is correct...neither Maple
>nor Mathematica will do that integral.

The Trager/Bronstein Integration can be broken up into a set of cases.
This is one special case that was not implemented but should be integrable.

>I was thinking to separate your code into a different PageType.  i.e.
>the PageType specifies what this document is *written* in.  Separate
>from that is the presentation (i.e. png/mathml which should not be a
>page type -- as I have it now).  Is the new Reduce stuff in your darcs
>repository?

One use I expect to have for these pages is test cases. I'm working
on CATS (Computer Algebra Test Suite). The basic idea is to document
some mathematics and then implement the computation in a variety of
systems. The idea is to pool the tedious effort of developing a test
suite so that all the systems can benefit from the effort.

NIST funded a project years ago to do a classification of the various
types of numerical mathematics. There is no corresponding classification
scheme for symbolic mathematics. CATS could form the basis for a scheme
for classification of the mathematics and provide test cases for each
category.

Bill's idea of joining several systems in one page is an important tool
for this kind of work.

>But now we have two math backends...and it might be very useful to have
>both in the same document.  So should we define a
>StructuredText+LaTeX+Math pagetype where 'Math' refers to any symbolic
>backend?  It would also be useful to do a bit of stylesheet magic to
>differentiate between Axiom/Reduce/Maxima...  Perhaps some small text in
>the UR corner that says what language this is.

Tim