I see.  How would that work?  Can you give me an example of a routine that would, for example, add arrays of long integers and could handle the 3 cases that I mentioned?  Would it also handle arrays of bytes? arrays of words?  arrays of reals?  Would I need to repeat the same for multiplication?  for xor?  for string concatenation?  for modular arithmetic?  for 96-bit arithmetic?  for quaternions?  for division rings?
 
A general vector facility would be a great work saver, even if it did nothing regarding parallel execution.
 
 
In a message dated 8/6/2010 3:16:27 A.M. Eastern Daylight Time, felipemonteiro.carvalho@gmail.com writes:
On Fri, Aug 6, 2010 at 8:35 AM,  <Contestcen@aol.com> wrote:
> Except that you have to code  every single case separately, 2x6, 3x10,
> 4x4x4, etc.  You need to write a  separate routine for every shape of array you
> want to use.

Or you could use dynamic arrays instead of static ones. Then you can
write only one case.

--
Felipe Monteiro de Carvalho