Comment by gucci-on-fleek

What do you recommend instead? Standard floating-point ("float"/"double"), fixed-point arithmetic with thousandths (or smaller) of the minor unit, arbitrary-precision decimal numbers, or something else entirely?