Yes, that is a much simper way, but then the trouble is that one can't trivially write one numerical function
and use different precisions later on. Using CLOS to shadow #'+, #'- ...(and math functions) is probably going
to be quite slow, and one would then have to resort to writing macros which can check for types and then do
compile-time "dispatch". I wonder if MOP can be tweaked to achieve this ?

Akshay


On 19 April 2013 23:24, Stephan Frank <crimsonf@cs.tu-berlin.de> wrote:
Hi Akshay,

but wouldn't a library with bindings to MPFR (using CFFI) not make more
sense? Maxima (and I suppose matlisp as well) have to implement their
own bigfloats data type anyway whereas the GMP integration has the
chance of being tightly and transparently integrated with the underlying
bignum implementation of SBCL. I could only see an advantage when
tightly coupling it with CMUCL's quad-floats.

Regs,
Stephan