Re: Integers and Mathematical Correctness

[wilhelm.spickermann@t-online.de]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=us-ascii, Size: 1177 bytes --]


On 01-Nov-01 Wes Groleau wrote:
> If your c_integer package is going to do something
> else, I would advise the following comment at the
> beginning of it:
> 
> -- NOTE: This package redefines integer division
> --       for the benefit of folks who think the
> --       RM definition is stupid.  Do not expect
> --       division of these types to behave like
> --       other integer types!
> 

I think, thats going too far. I, for instance, do not not believe
the "RM definition is stupid" -- it just reflects the abilities
of normal hardware and that is a *good* idea (thanks to Robert
Dewar for pointing me on that fact). I will not use the
c_integer_package, but only because it cannot make the old
division operator vanish completely and I don�t want to have two
of them with the same name and a different semantics.

But: It would be much more practical, if the division operator
(and the hardware) would be the way Chris Danx called
"mathematically correct". Most programs with integer division
- avoid negative numbers or 
- have to do additional operations which would not be necessary
  with mathematical."/" and "mod" or
- would be equally complex.

Wilhelm