Re: [Axiom-developer] aldor/axiom interoperability

[Gregory Vanuxem]

Tim,

Many thanks for your interest but as I said the code is not interesting.
The code that saves/restore the Axiom environment is lost (with other
things), other parts of code are purely experimental (can't publish them
as they are, in fact I do not like them in general), other parts are not
portable (interface to BLAS/LApack for reals, require new "column major
order" matrices with storage contiguous, a problematic thing). The rest
is pure and _simple_ optimisation to know what Axiom/"underlying Lisp
system" can do, applied or I don't remember it. Gaby asked me the same
thing in the past and I refused. Working on a project like Axiom
requires time (thing that I lack actually) and deep knowledge in a lot
of domains (other thing that I lack too). Axiom needs to be driven by
theories and my knowledge is often more practical. So I prefer to help
sometimes, if I can, and contemplate with admiration you and others :-)


Greg

PS : The code that displays memory utilisation works only on CMUCL,
Clisp and SBCL. GCL produces C code, it's not possible to know precisely
the memory utilisation, as far as I know at least.

Le samedi 15 mars 2008 à 13:57 -0500, root a écrit :
> Greg,
> 
> Can you make your code tree accessible somewhere? I can give you a
> userid on axiom-developer.org and you can upload it there? I'd like to
> pick up the changes you've been making and merge them back into the
> main tree.
> 
> Tim
>