[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

numeric speed

(James E. O'Dell) writes
>Of course what I meant was, "as fast as." I firmly believe that to be
>competitive in your target market that Dylan will have to have a mode
>that is at least as
>fast as FORTRAN. Not only on isolated FLOPs but on arrays of floats and
>fixnums. To do any less will relegate the language to company of basic
>and hypercard.

I don't see any way of doing this without allowing for the
declaration of multidimensional fixed size arrays.  The "convert to
simple vector and compute your own indices" alternative is less than


Richard Ferrante