Re:0.30000000000000004
by faragon on Sunday April 02, 2017 @06:41PM
(#14019456)
Base 2 floating point is perfectly OK. The problem is people using floating point not knowing that IEE754 floating point representation is binary, and not decimal, so most decimal constants will not have an exact match. Said that, many implementations do buggy binary <-> decimal conversions, not because of the limit accuracy in the conversion, but because of "creative" rounding criteria.
[ Parent
| on HN ]
There are no comments on this comment yet.