Issue #3365 has been updated by Roberto Tomás Collins McCarthy. I'm confused. how does that match either of: the system's own handling of doubles (which the ruby api docs say it uses) or the IEEE754 specification (which would require to have a minum of 1 more zero in the tail) ---------------------------------------- http://redmine.ruby-lang.org/issues/show/3365 ---------------------------------------- http://redmine.ruby-lang.org