On Mon, 2003-10-27 at 11:32, Raymond Toy wrote:
> And, for a program like maxima, I doubt there's much improvement to be
> had with safety 0, because most things aren't declared anyway, so
> type-checks are needed. Perhaps the plotting routines and some of the
> numerical routines can benefit from safety 0 (and speed 3), but not
> much else.
>
> Of course, I didn't check this supposition with real tests.
You make some very good points. I would love to see the results of some
optimization experiments with Maxima. I think the optimization settings
all go back to (at least) Schelter's time.
Anyone willing to try some tests?
--Jim