Closed tomhoule closed 4 years ago
This is not the right solution long term, but with this change, any single-precision float that gets into the database will come out the same. In very rare cases, we will instead treat an f64 as a f32 and lose precision.
This is not the right solution long term, but with this change, any single-precision float that gets into the database will come out the same. In very rare cases, we will instead treat an f64 as a f32 and lose precision.