Using strconv.FormatFloat(val, 'g', 10, 64) to format floats means that very large or very small numbers will be rounded. For example, a value of 123456789012345680 will be saved as 123456789000000000. By using -1 for precision means the smallest number of digits will be used to represent a float exactly.
This is to resolve issue 51 https://github.com/shadowspore/t38c/issues/51
Using
strconv.FormatFloat(val, 'g', 10, 64)to format floats means that very large or very small numbers will be rounded. For example, a value of 123456789012345680 will be saved as 123456789000000000. By using -1 for precision means the smallest number of digits will be used to represent a float exactly.