Boost logo

Glas :

[glas] Re: GLAS

From: Simon Perreault (nomis80_at_[hidden])
Date: 2005-02-09 19:31:49

On February 9, 2005 18:13, you wrote:
> Where the functionalities are already provided by BLAS and LAPACK, it
> makes sense for GLAS to be just a wrapper. This provides the user
> with flexibility, choice of the implementations of BLAS and LAPACK
> available, and the ability to leverage on existing high performance
> implementations of BLAS and LAPACK.

I totally agree with you, but I think you may be seeing GLAS from a
perspective that is too much practical. I heard that GLAS was supposed to
define the state of the art of generic linear algebra algorithms. Such an
experimental, let's-see-what-we-can-do library isn't usually very practical.
What I was asking was, isn't GLAS supposed to be something more than a
wrapper? Wouldn't it go into areas overlapping BLAS's jurisdiction, like how
to design the linear algebra algorithms themselves using generic programming?

Simon Perreault <nomis80_at_[hidden]> --