Has there been a concensus (like in the proposed standard) as to the scope of a typedef? I looked in k&r2, and there was nothing to indicate it must be local to the file in which it was, uh, typedef'ed, but that is the only logical thing that I could imagine. Eh? rob