exact decimal type, was: 12.3*70.95 a little bit weird



On Wed, Nov 30, 2011 at 6:43 AM, Rudy J. Richardson <rjrich at umich.edu>wrote:

> Robert, Dan,
>
> Scientific notation could be used to denote the exact signficant figures.
>
>
By default, maxima prints out all digits, except trailing zeroes are not
printed.   The format is essentially C's %g, except as many digits as
needed are pritned.  If maxima prints out ...00001, then maxima (the
underlying Lisp implementation, actually) thinks that is the printed
representation that, when read back in, will produce exactly the same
floating point number.

This is rather tricky to achieve, but I know cmucl and sbcl try very hard
to achieve that.

Ray