On the abc conjecture, II