Re: OONSTD: Re: I got your stuff

Gabriel Dos Reis (Gabriel.Dos-Reis@dptmaths.ens-cachan.fr)
Thu, 16 Jul 1998 23:41:32 +0200 (MET DST)

>>>>> «ERT», E Robert Tisdale <edwin@maxwell.hrl.hac.com> wrote:

ERT> How does Roscoe know that he picked the best possible representation?
ERT> Apparently, Kent Budge regrets the adoption of his valarray<> proposal.

Kent Budge made a proposal. It was adopted in ISO C++ even though
there was no sound implementation at the moment (yes I know Daveed
implemented an approximation).
But that was not the first time the committee did invention.

ERT> What happens if valarray<> is deprecated.

Too bad. Too late. valarray<> should _not_ be deprecated. It should
be *improved*.

ERT> If computer architecture evolves to favor a different representation,
ERT> all the application programs that rely on access to the underlying
ERT> representation would suddenly become obsolete.

I think it is a trivial statement that there cannot be a universal
matrix representation. Given that, one needs an alias-free 1D
array. For one dimensional array, valarray<> is good enough. I do see
valarray<> as a fondamental building block. I don't consider
valarray<> in the same way as string. I expect compiler writers to
follow the standard wording.

ERT> If Roscoe were a commercial library developer,
ERT> he wouldn't be able to sell any more packages
ERT> and he would probably have a large number
ERT> of very angry customers attempting to sue him.

If he uses standard types, he will continue to sell his
products. valarray<> is a standard type.

--
Gabriel Dos Reis                    | École Normale Supérieure de Cachan
INRIA, Unité de Recherche de        | Centre de Mathématiques et de Leurs
Sophia Antipolis                    |          Applications
Projet SAGA                         |       Équipe de Géométrie