[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Axiom-developer] RE: algebra Makefiles with explicit dependencies,

From: Stephen Wilson
Subject: Re: [Axiom-developer] RE: algebra Makefiles with explicit dependencies, bootstrap, fixed-points etc.
Date: Sun, 9 Jan 2005 01:22:04 -0500
User-agent: Mutt/1.5.6+20040907i


On Sat, Jan 08, 2005 at 10:21:34PM -0500, Bill Page wrote:
> I am not sure I understand what you mean by a `compilation model'
> in the case of a language like Axiom. Could you suggest some
> relevant references?

I dont have any references, unfortunately. The thought I have in mind
is really no more than an axiom specific `make'. In allowing the compiler
do the bookkeeping, we could automagicly resolve dependencies and have
a trusted logic to achieve the `fixed point' state. I used the term
`compilation model' since, however the process proceeds, we need a tight
definition of what actions the compiler will perform, how it finds
source files, etc. We might only need enough logic in the compiler
itself to produce appropriate error messages which could be captured
and parsed by the interpreter or some other process. 

No matter how I think of the problem of dependencies in terms of a
`build' procedure, I'm always compelled to regard it as a language
issue. We are currently defining mutually recursive structures in a
language which, strictly speaking, does not support them!

> Here are some references on mutual recursion that seem
> quite relevant to me;

Thank you very much for pointing these out! I will look them over this

I regard axiom as a particular implementation of a general philosophy
on how to do computational math. Trying to work towards a new
compiler/language, I think, has some great benefits. It would be great
fun :) It motivates one to study the existing implementation, which in
turn motivates maintenance and improvement of the existing system. And
of course there is all the benefits to the scientific community
(despite the apparent lack of funding for such work, I dont consider
the area of CAS design a solved problem! :).


reply via email to

[Prev in Thread] Current Thread [Next in Thread]