Re: [Texmacs-dev] Could this be interesting for TeXmacs?

[Henri Lesourd]

Alvaro Tejero Cantero wrote:

> http://code.google.com/apis/v8/intro.html
>
>
>  
One would need to find a way to implement
a gateway Guile<->V8, in such a way that
the v8 datastructures could be read from
Guile, and the Guile datastructures could
be read from V8.

It would be better to have only one core
engine, able to manipulate C++ datastructures,
and translate the syntax of the different
languages to this core engine. In such a
case using the V8 engine would be a good
solution to get rid of Guile.

The second problem is being able to emulate
the Scheme's APIs on top of which TeXmacs's
Scheme code is built. If you use another
engine, then you must reimplement the
corresponding libraries on top of this
engine (e.g., on top of V8).

On the other hand, you could think about
a more conservative approach, namely,
translate the desired scripting languages
to Scheme.

This would be very useful to bring other
communities closer to TeXmacs, for it is
a matter of fact that quite often, people
don't know (and are repulsed) by Scheme
philosophy.

Especially: Python and Javascript should
be considered. Python because the combination
C/C++ plus Python is the approach of choice
when people need to perform resource intensive
scientific calculations (e.g. simulations).

Javascript should be considered because
very clearly, scripting a browser is
usually done in JavaScript, thus offering
it would attract people from this (huge)
community.

This being said, I don't know when the
resources will be available for tackling
such tasks...