> As was mentioned in a previous message to this group,
> what is really needed is not so much yet another class library as a standard,
> but rather an architectural specification (or interface specification)
> for how linear algebra libraries should behave.
Can you elaborate a little on this "architectural specification" for us?
It seems to be a vague and fuzzy notion which no one is able to articulate very
well.
Are you prepared to propose one? How will it be different from the standards
like the one I have proposed for example?
> Since one class library has been tossed into the mix
> for discussion as a possible starting point for a standard,
> let me toss in an alternative, based on the points above.
> We have a work in progress that we call the Matrix Template Library (MTL)
> in deference to STL as a motivating force behind it.
> The basic features are genericity, performance, and functional richness.
I don't like it. All the operations are implemented as void functions
which means that the application programmer can't write expressions
and must manage all of the temporary results explicitly. Application programs
will require two to three times as many lines of code as equivalent SVMT code
and will be much harder to read, understand and maintain than SVMT.
Your examples look like FORTRAN 77 code.
Is there something about your approach that requires a design like this?
Bob Tisdale <edwin2@gte.net>