Re: [Axiom-developer] Re: interval.spad INTRVL

[Mike Dewar]

On Sun, Aug 31, 2003 at 03:10:36AM -0400, Bill Page wrote:
<snip>
> This reminds me to ask about the other ALDOR files that are
> included in the build but apparently not compiled. Are you
> planning that these will at some point be compiled by ALDOR?
> (which is now a separate non-BSD licensed? package) How
I think the only interesting Aldor files apart from interval.as are for
producing Open Inventor geometry files, and whether you need them or not
depends on what your plans for graphics are.

> difficult is it for a user to encorporate ALDOR if they so
> wished? There would likely be a considerable speed and memory
> advantage for certain routines. Right?
Not necessarily because the Aldor compiler generates Lisp code which is
then incorporated into a running Axiom in the usual way.  The
alternative would be to use dynamic linking to link an Aldor shared
library into a running Axiom, which might well give the kind of
improvements that you suggest.  However to do this efficiently you might
need to tweak the two garbage collectors (Axiom's and Aldor's) to work
together, so that you could handle e.g. an Aldor list of Axiom objects.
There is some work being done on this at ORCCA, except that the "host
system" is Maple not Axiom.

> Another thing that has happened since Aldor separated from
> Axiom is that a very large part of the algegbra has been
> re-written and apparently improved. Do you think that in
> the medium to long term the Axiom and Aldor algebra libraries
> should be made compatible? How practical (and legal?) would
The new Aldor libraries aren't really drop-in replacaments for the Axiom
ones, and are designed with particular purposes in mind (namely
Polynomial arithmetic and ODE solving).  There is a fundamental
difference between Aldor and Axiom which is that the former allows for
post-facto extensions of constructors, and that effects the design of
the libraries.  In Axiom you have to try to design the final version of
anything because once implemented it is set in stone, while in Aldor you
can always add new exports as required.  So I suspect that you might
find that the existing Aldor libraries are missing things which Axiom
re;ies on.

> should be made compatible? How practical (and legal?) would
> it be for Axiom (at least as a option) to implement the
> same library. Is it difficult to convert to spad. Or is
> there any advantage if ALDOR can be easily linked with
> Axiom?
Its possible but non-trivial.  We had a long-term goal of re-writing
parts of Axiom so that it could act as an interactive front-end to Aldor
libraries.  It turned out that to build an empty Axiom interpreter was
quite hard since there were lots of hidden dependencies on algebra code.
However I don't think any of the problems would have been
insurmountable.

> > re: SIGNEF
> > 
> > The nature of that problem is that SIGNEF uses INTRVL.
> > 
This is all my fault since I wrote that code :-)  We had a long-standing
bug that Axiom couldn't integrate expressions with floating point
coefficients and this code fixed that.

Mike.

________________________________________________________________________
This e-mail has been scanned for all viruses by Star Internet. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________