meruby / gmail.com wrote: > Sorry for typo in previouse post: > Shouldn't Float::MIN = -Float::MAX Float::MIN is the smallest number > 0 that can be represented by a Float. Usefull I guess if you need to represent really small values. Float objects are basically a wrapper around a C double. on my system it is: irb(main):001:0> Float::MIN => 2.2250738585072e-308 which is in the float.h header as #define __DBL_MIN__ 2.2250738585072014e-308 -Charlie