Excerpts from none's mail of  7 Jul 2005 (EDT):
> I am not sure how we might proceed from here.

Here's my suggestion. Let's spend a little time fleshing things out on
Ara's wiki. When we get to a reasonable state, let's jump right in to
RubyForge. Write some very basic high-level code, to be fleshed out
later, push design specifics out to the mailing list and start working
it all out.

I think the trick is to get something simple working as soon as
possible. Then people have something tangible and it's easy to get
involved and excited.

> I think we *need* to get away from thinking about what libraries are
> available and how to cobble them together.  Instead, we need to
> *imagine* and *specify* what we would want in a perfect scientific
> language.

I agree completely. Forget about the specifics of GSL and Narray and
MRplot and everything else for a moment and just focus the ideal
interface. 

It's also important at this point to get as many different use cases in
our heads as possible. I'm coming from a dataframe-oriented, stats-heavy
R perspective; you're doing linear algebra... are people interested in
the symbolic computation aspects of Octave? Other things?

What do you think?

-- 
William <wmorgan-ruby-talk / masanjin.net>