Of course to be complete we should take into account both the positive and negative NaN (not-a-number) values...

a = 1/0
b = -1/0
? a == b  && gives .F.
? a == (1/0)  && gives .T.
? a == (10000000/0)  && gives .T.
? b == (1/0)  && gives .F.
? a>b  && gives .T.
? a<b  && gives .F.
? (1/a) == (1/b)  && gives .T.

Hmm.. I don't know if there is a bult-in function to check for NaN values.