Why it is impossible to test a program completely?
With any software other than the smallest and simplest program, there are too many inputs, too many outputs, and too many path combinations to fully test. Also, software specifications can be subjective and be interpreted in different ways.