Re: [Maxima-commits] CVS: maxima/src rand-mt19937.lisp,1.1,1.2



--- Raymond Toy  wrote:

>     Robert> For GCL, route all floats through the double-float
>     Robert> function since GCL does not distinguish between 
>     Robert> single and double floats.
> 
> While strictly correct, I think this might be a bit misleading.  GCL
> has a rather odd arrangement that its single-float, double-float, and
> long-float types are all IEEE double-precision floats.  But GCL
> supports short-floats which are IEEE single-precision floats.
> 
> Given that maxima really only wants double-floats, maybe we can
> remove all support for anything besides double-floats?

I guess cutting out any floating point stuff other than
double floats is OK by me. But ideally the non-Maxima part 
of the rng would be preserved so that anyone could lift it 
and cut off the Maxima stuff. Maybe I'll add a GCL-only 
test to branch to %random-single-float if the argument
is a short float. I don't really have any strong feelings 
one way or another.

For what it's worth,
Robert Dodier


		
__________________________________ 
Do you Yahoo!? 
Yahoo! Small Business - Try our new resources site!
http://smallbusiness.yahoo.com/resources/