I quote from someone else: "standard IEEE floating values, are highly inaccurate for scientific calculations." I you sure you want to use floating point math for your library ? :)