Write a program to print the positive floating point value closest to 0 that is possible to represent accurately with [[IEEE doubles]]. Also print a value slightly closer to show that it becomes 0. Does the language produce an error or simply a warning on underflow?