Paul Libbrecht a écrit :
I would agree with Phil,
as long as Math is not modularized (which turned out to be quite a
hassle for jelly but probably pays on the long run, especially now that
maven 1.1 handles reactor cleanly) Nabla should not be part of it. At
least for the dependencies, it's quite different.
Yes. Math is a general library for widespread use (analysis, algebra,
statistics, geometry ...). Nabla is oriented toward analysis only and a
sub-part of analysis (but a very useful one, especially for simulation
purposes and physics).
Will Nabla use math? Would there be some common part? The functions
interfaces?
Maybe that is a question that needs to be answered soon.
No. Nabla will not use math at all. The maths it uses are hard-coded
symbolics computations (how to differentiate a sum, a product, a
reminder, an inverse hyperbolic cosine ...). There are no solvers, no
approximation, no converging series and of course no geometry and no
statistics.
Luc
paul
Le 14 avr. 08 à 03:37, Phil Steitz a écrit :
I agree with Torsten on this one. Why doesn't this belong as a part
of math (or maybe a submodule if math wants to have one)?
Um...maybe for the same kind of reason that [functor] does not belong
inside [collections] ;)
We don't want to add it to [math] (at least immediately), because it
does not really fit with the current focus of [math] and [math] is
getting large, and will get larger in 2.0. Nabla also should not
depend on [math] either and could be used by itself.
As I said, i think it is worth a try and in any case, we should be
able to start it in the sandbox.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
