James Amundson <amundson@users.sourceforge.net> writes:
> On Sun, 2004-04-18 at 18:21, C Y wrote:
>
>> What about creating, say, integration, matrix, ODE, etc packages? Or
>> is there no need/advantage for something like that?
>
> Although it might be nice to have the various mathematical packages
> delineated, I don't see any compelling need to do so.
At least, we have to get rid of the (DECLARE-TOP (SPECIAL ...))
(DECLARE-TOP (UNSPECIAL ...)) mess, and creating packages, as Cliff
suggests, might be the right solution.
In particular, the compile-time environment and the run-time
environment are confusingly different due to DECLARE-TOP and we
actually do have rather bizarre bugs (which appear or disappear by
just reloading the code in question) because of this form. Recall
http://www.math.utexas.edu/pipermail/maxima/2003/005974.html
where Valery reported an apparent integration bug, the reason of which
was explained later in that thread (the autoloading bug he reported
has a similar explanation).
Wolfgang