nan-preserving-float 0.1.0

(Mostly) drop-in replacement for floats that preserves the signalling bit in NaNs under conversions and unary operations
Documentation