RE: [Axiom-mail] RE: AXIOM / TeXmacs problems (bugs?)

[Bill Page]

François,

On Monday, December 08, 2003 2:53 AM you wrote:
> ... 
> But before we write a library, we make a lot of test
> in an interactive mode. And after a copy/paste makes
> the new library. I dislike syntax modification at this
> point.

I agree with you. I think these differences are mostly a
legacy of Axiom's long history. A little effort applied
to resolving these differences at this stage of a new
release of Axiom would probably be very worthwhile.

It might well be possible to remove some of the syntax
differences (for TeXmacs users at least) by simple changes
to the TeXmacs-Axiom interface program tm_axiom. For
example, tm_axiom could do the conversion from "pile"
notation to ( ) block notation. Then at least the syntax
would be much closer.

There are differences however even in the semantics when
you write library routines that are intended to be compiled.
For example you must provide much more explicit type
information since you can no longer depend on the interpreter's
heuristics for type inference. I am not sure how ambitious
one should be in attempting to resolve such differences.

As a user, if you have the Jenks and Sutor book, I think
you would find it worthwhile to read Chapter 11 on how to
create packages.

If you would like to continue this discussion about possible
program changes we should probably shift the discussion to

  address@hidden

Regards,
Bill Page.