[Axiom-developer] RE: Boot vs. Lisp - various replies

[Bill Page]

On November 1, 2005 6:47 AM Bertfried Fauser wrote to
Tim Daly:

> 
> Perhaps I may add, that as a _user_ of AXIOM, I am totally 
> ignorant to its internal struggles (bild process from
> scratch, historical accidents, etc), but that I need to
> focus my resources.
>

I think we need to distinquish several different kinds of
"users":

Axiom user I (beginner):

  - uses the Axiom interpreter to evaluate expressions
  - writes single-line and multi-line functions
  - Ref. Axiom book chapters 1 - 7.

Axiom user II (advanced):

  - writes .input files containing imperative statements
    such as if, for, and while.
  - Ref. Axiom book chapters 8, 9.

Axiom user III (library developer):

  - writes Axiom packages, domains and categories using
    SPAD and/or Aldor.
  - needs to understand the logical/mathematical structure
    of the Axiom library and how to extend and build apon it.
  - Ref. Axiom book, chapter 11, 12, 13.

Axiom system developer:

  - maintains and enhances the underlying tools such as
    SPAD and the Axiom interpreter.
  - writes in boot and lisp.
  - needs to understand the internal design of Axiom.
  - Ref. Tim's new book volumes 4 and 5

Axiom support programmer:

  - writes in lisp (e.g. gcl), C/C++ and other languages
  - maintains the underlying tools such as lisp (gcl)
  - debugs problems directly related to lisp and/or the
    operating system
  - advises on porting Axiom to different systems, etc.

> I am not a programmer, but need to be able to develop own
> functionality. So I would vote for a system having
> 1 internal language (lisp, C++, ...) and probably one 
> toplevel language (Aldor, spad, ...). Any further mixing
> prevents a _user_ from even making first steps.

>From the categories above I think that most likely you would
be a user of the third kind (Axiom user III). In this case
you are mainly concerned with what you call the toplevel
language (SPAD), it's relationship to the interpreter and to
other existing domains in the Axiom library. Any internal
language such as boot, lisp or C would be invisible to you.

Making the first steps in Axiom usually involves just
evaluating complex expressions using existing operations,
i.e. a user of the first kind (Axiom user I). In this case
no programming is involved, but of course some mathematical
knowledge of Axiom operations is necessary.

Generally I expect people will progress from I to II and
maybe to III. Some highly motivated user of the third kind
with additional programming experience might be also work
as Axiom system developers.

Some Axiom support programmers might become Axiom system
developers.



> 
> Hence I would at least like to have a winner of the
> lanuage war, alas such stories usually will have no ...
> and they lifed happyly ever after.
> 

I think there is a good chance that even if there is no
winner, we will live happily ever after - ever after we
get some serious users of Axiom, that is!

Cheers,
Bill Page.