| From: Brian Mastenbrook <[email protected]> | Date: Tue, 22 Sep 2009 13:58:43 -0500 | | On Sep 22, 2009, at 10:23 AM, John Cowan wrote: | | > 2) Fixnums + flonums (limited range and precision types only) | > | > 5) Fixnums + flonums + compnums (2 with inexact complex numbers) | | Can someone please explain to me why bignums are so onerous a | requirement that implementations can't be expected to implement them? | I'm not understanding why conses, vectors, strings, etc. can all be | required, but bignums are problematic.
If all you had were bignums (no fixnums), or converted every fixnum to a bignum before operating, then it wouldn't be so much work. Coding for every binary and nary argument mixture of numeric types bloats the code. Naive division, GCD, and LCD generate a lot of garbage. _______________________________________________ r6rs-discuss mailing list [email protected] http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss
