[Ecls-list] Using Femlisp with ECL / Bug reports

[Nicolas Neuss]

"Goffioul Michael" <goffioul at imec.be> writes:

>> Hmm, I wanted to ask if Matlisp has been ported to ECL, but Michael has
>> already answered this.  Indeed, Femlisp once depended on Matlisp, but I
>> found that really inconvenient.  If Matlisp was more widely used and
>> would provide a more complete interface to BLAS/LAPACK/..., the
>> situation would probably be different.
>
> Well, MatLISP provides a good basis for a BLAS/LAPACK interface with its
> set of classes and macros. The current interface is far from complete,
> but you can still add things quite easily (for our own purpose, I added
> support for dggsvd without much effort).

I also think the interface which Matlisp provides is quite OK and I mostly
followed it in Femlisp.  The problem is how it is implemented.  It should
not be "easy" to add things (which I also did once), it should be more or
less automatic.  I think this could be done very often, e.g. by issuing
something like (matlisp:get-lapack-function 'dggsvd) or
(matlisp:get-lapack-function 'gsvd 'double-float) in the case you
mentioned.  [The real problem is writing the CLOS layer, of course, but
this could be improved as well.  There is much code duplication in those
parts of Matlisp.]

Nicolas.