> Actually, if you said 1 you are right - because you have your test > backwards! :-) > > I think the problem is that string literals are of type char * even though > they can't be modified.... Borland gives 1 (I guessed 0 expecting const char * ) JB