I've added support in s7 for multiprecision integer, ratio, real, and complex numbers, including all the numeric functions. Use the configure flag --with-gmp. It uses the gmp, mpfr, and mpc libraries. Default float precision is 128 bits, but can be set higher via bignum-precision. Still a week or so of testing to do, but it currently passes the tests in s7test.scm. If this works out as I hope, I'll extend the mp arithmetic to clm and snd.
_______________________________________________ Cmdist mailing list [email protected] http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
