Semipredicate problem

In computer programming, a semipredicate problem occurs when a subroutine intended to return a useful value can fail, but the signalling of failure uses an otherwise valid return value. The problem is that the caller of the subroutine cannot tell what the result means in this case.