Re: [Axiom-mail] blog post

[daly]

This appears to be a conference stressing teaching mathematics.
(http://atcm.mathandtech.org/)

I would appreciate it if you'd point out that THE key focus of the
Axiom project is to make it possible to understand how to read, write,
and modify computational mathematics.

The whole point is to try to keep Axiom alive and growing beyond the
lifetime of the original authors. Computational Mathematics is the
result of the collision of mathematics and computers. Axiom is the
"Newton's Notebook" of that collision.

In general, Axiom was created as a piece of research software used
only by fellow researchers, usually to complete their thesis work.
Thus, the algorithms are world class. But the documentation isn't.

It is important to document the algorithms. The code is NOT sufficient
since there are design decisions and tradeoffs that have nothing to do
with the algorithms. You can see this clearly in work by Jack Dongarra
(http://netlib.org/utk/people/JackDongarra/etemplates/book.html) where
he raises all kinds of implementation issues like stability and
convergence. In Axiom's case there are even more issues, some of which
are addressed by the new proof machinery being created.

I will grant you that this is a very long term goal which will take a
while to accomplish. Longer now, sadly. The Axiom community used to
have 568 unique contributors. The forks ended that. Now it seems there
are only a few people left. Nevertheless, it is vital that Axiom
becomes easier to understand, maintain, and modify so it can help the
next generation of Computational Mathematicians.

Ideally I would like Axiom to be THE software used as the platform
for teaching computational mathematicians.

Tim