Re: C++ wrapper [stuff to do]

[Gerard Jungman <jungman at lanl dot gov>]

Ivo Kwee wrote:
> I would be OK to code something up but I need some ready established
> interface specification.
> Earlier, SVMT has been mentioned. Is this the "best" one to go for?
> Anyone knows "better" or more "accepted" standards?

Unfortunately, there is none. SVMT is not perfect. I don't know
what happened to the standardization discussion at oonumerics;
there hasn't been any activity there for some time. It was
probably just decided (implicitly?) that the world was not
ready yet to adopt a standard, especially since people are
still working hard on their template-happy implementations.

> I was thinking
> myself if MatLab-like conventions would be better, but they are not
> really OO style. Suggestions?

OO-style as such (I guess I'm not sure exactly what you mean) is not
so important. What is important is that you don't expose implementation
details. Some people have argued here that this would be difficult
to do with the GSL vector/matrix stuff. I don't think it would be
that hard to do well, but there are some issues.


I don't want to be discouraging. But I think the whole
question of interface standards needs serious work.
I don't think it is time yet to start wrapping.


Steve ROBBINS wrote:
> I haven't used it myself, but I believe that Octave comes with C++
> wrappers around a bunch of numerical libraries.  Mostly linear algebra,
> I'd guess, but there should be some others.   I'd be interested in
> comments from folks that have looked into using this in their own C++
> programs (as opposed to using octave proper).

This is interesting. I guess I have heard this too.
Who wants to volunteer to figure out what they have
done? Maybe one of the Octave people can point us
to the executive summary...


-- 
G. Jungman