Floating point number serialization ignores leading zeroes after the decimal point

```C
jim_float(&jim, 3.0014, 4);
```

Results in: `3.14`

Edit: Also, whould be nice if there was a way to do automatic precision.