Floating-point numbers are not real numbers in the mathematical sense. Real numbers have infinite precision; floating-point numbers have fixed precision, and resemble "badly-behaved" integers.