Hello everybody,
> > Of the items you listed, generalized functions seem most important
> > and useful to me. Diffierential equations take second.
> > Numerical stuff comes after that.
> >
> > My general feeling is that numerical stuff is less important than
> > symbolic stuff, just because so many other packages already provide
> > numerical solutions. But it is still worthwhile to work on numerical
> > stuff.
>
> For what it's worth, I also agree with you here. There are already a lot
> of other libraries that can do numerical computations, and I don't think
> that numerical methods are going to bring many people to Maxima. But,
> maybe I'm wrong.
Yes, I belive this is wrong. It seems was one of the reason why Macsyma
lost. See the Richard Petti's "Macsyma Saga"
http://www.math.utexas.edu/pipermail/maxima/2003/005861.html
The engineering aspect is the most important in the modern world,
independently of what we are thinking about it.
So I would give the priority to numerics and to more tight integration with
python or octave. In particular, it would be nice to integrate
http://www.femlisp.org/ to maxima. Also, the current stage of the linear
algebraand special functions stuff in maxima make possible to implement
spectral(pseudospectral) methods for PDE.
best regards,
Valery