Floats formatted with "correct" precision

Conversion float->string ensures that, for any float f,
tonumber(tostring(f)) == f, but still avoiding noise like 1.1
converting to "1.1000000000000001".
3 files changed