Dave Mitchell wrote: > Given the amount of time (i.e. too much) that I've already spent on > reworking the AASSIGN op, I think I'm going to leave undef By which I think you mean returning the rhs if the lhs is immortal. I would say yes, let sleeping dogs lie; since it is not clear how things *should* behave, just preserve the existing behaviour. > and \(@a) > lvalue behaviour alone for now. I'll still merge the f(($a,@b, $c) = ()) > fix though if people agree. I may be too late, but what would the new behaviour be? If I under- stand correctly, it would be that the number of elements on the rhs is returned in scalar context, but the elements on the lhs are returned in list context. Is that right? Is that what we want?Thread Previous | Thread Next