Long-float variant giving more precision to numerical computations.

Robert Dodier wrote:
> On Thu, Feb 14, 2008 at 8:39 AM, Raymond Toy (RT/EUS)
> <raymond.toy at ericsson.com> wrote:
> 
>>  A quick fix for that, and everything compiles.  I have to say it is
>>  rather nice to see all those extra digits.  There are a few issues
>>  though.  Printing large floats doesn't print out all the digits.
> 
> Yeah. There is a literal constant 16 in EXPLODEN.
> I guess that should become a function of the float size,
> s.t. like ceiling((mantissa #bits)/(log(10)/log(2))).
> Or maybe there is a way to specify a format directive
> which just means "print all available digits".

A new constant was added, see float.lisp
(defmvar $maxfpprintprec (ceiling (log (expt 2 (float-digits 1.0)) 10.0))
   "The maximum number of significant digits printed for floats.")

This is used as the default value for $fpprec, and in 'exploden as the default
when $fpprintprec is zero.  This is 16, as before, for double-floats, and for
the CMU CL double-double floats this is currently 32.

Regards
Douglas Crosher