The floating point standard is really good at pretending to be real numbers on a computer.

Every problem is the mathematical consequence of its features.