this change uses fround when available to make a better decision on
if a number is a 32 or 64 bit floating point number. This means that
we can more safely use floats in msgpack without losing precision.
In the case where Math.fround is not defined (I'M LOOKIN AT YOU, IE10),
this code defaults to using a float64 so that we don't accidentally lose
precision like we did with the previous implementation
this change uses
fround
when available to make a better decision on if a number is a 32 or 64 bit floating point number. This means that we can more safely use floats in msgpack without losing precision.In the case where
Math.fround
is not defined (I'M LOOKIN AT YOU, IE10), this code defaults to using a float64 so that we don't accidentally lose precision like we did with the previous implementation