Floating point trouble with x86's extended precision

[Jim Wilson]

On Thu, 2003-08-21 at 18:50, Bradley Lucier wrote:
> > The following computational model will give predictable results and 
> > will allow

This looks like a reasonable thing to try.  The question now is who is
going to implement it.  If you are waiting for someone else to implement
it, you may have to wait a long time.
-- 
Jim Wilson, GNU Tools Support, http://www.SpecifixInc.com