I'm declaring this one to be the winner!
(Not that my opinion counts for anythin, but I'm
just impressed: a1, b1 = zip(*ab)
(or a1, b1 = apply(zip,ab)?)
BTW: this process isn't 100% inverse.
It has the side effect of turning lists or strings
(non-tuple sequences) into tuples, if they are supplied
as the original sequences.