2 comments

[ 4.5 ms ] story [ 16.4 ms ] thread
It's hard to take someone seriously when they ascribe floating point rounding error to a language defect.
Sure, but isn't it a language design choice to only support floating-point numbers?