Serialization of floating point numbers with binary
is fantastically slow and incorrect if you’re using NaN
s, see
I recently spent half a day debugging performance problems because of this, and since backwards compatibility with older formats is required, this problem is probably not going to be solved.
We decided to switch to cereal
for this reason. With some patches cereal was 30x faster for the data we were serializing (scientific computing, mostly Double
s packed in nested records containing vectors).
The size of the serialized data is also roughly 3 times smaller – with binary
a Double
takes at least 25 bytes of space instead of 8. With Float
it’s even worse, 25 bytes instead of 8.