"Fred L. Drake, Jr." <fdrake@acm.org> writes: > Except that the compatibility version defined in minicompat is being > used in this case, and it does accept a tuple as the second arg. I didn't notice minicompat is that powerful :-) Regards, Martin