RE: [Axiom-developer] B#

[Bill Page]

On November 19, 2005 2:05 AM Hans Peter Würmli wrote:
> 
> In "How to Make AXIOM Into a Scratchpad" Jenks and Trager 
> describe "... a new user interface language ..." for Axiom,
> which they called B# ("B natural") in the article. If it ever
> had been implemented, it would offer a lot of the language
> elements at least I would have hoped for in the interpreter 
> interface for Axiom.
> 
> Does anybody know what ever happened to that project?
> 

http://portal.axiom-developer.org/refs/articles/axiom-scratchpad.pdf

The date on this paper, 1994, closely coincides with the time
when IBM decided to close Axiom as a research project and
negotiated with Numerical Algorithms Group to market Axiom as a
commercial product. It is easy to guess that B# might have been
a causality of this transition.

I agree with Peter that B# is probably what a lot of new users
of Axiom are expecting to find. I like most of the design ideas
for the B# language presented by Jenks and Trager in this paper
so I think implementing B# would be a great project to resurrect
for open source Axiom.

Also interesting in this article is mention of the project GAUSS
which was an attempt to provide an Axiom-like type system in Maple
GAUSS is contrasted with B# because Maple is essentially untyped
while B# was an attempt to provide an untyped user environment
within Axiom. As an active Maple developer at that time, I remember
briefly playing with GAUSS and not being particularly impressed.
My point of view has changed a lot since then. I am convinced of
the value of strong type system in Axiom but I find I often miss
the freedom of expression that the untyped environments of Maple,
Mathematical and Maxima provide. B# might very provide the kind
of bridge that Jenks and Trager envisaged in this paper.

Does anyone else have an interest in investigating the possible
implement of B# ?

Regards,
Bill Page.