Overview
Appendix
The constant number is not a binary64 floating-point value. A closest value number will be used and unambiguously printed number. [inexact-float,parsing]