On Tue 29/Apr/2014 14:55:52 +0200 Patrik Fältström wrote: > > The problem exists if A is publishing such a policy, B is > acknowledging the policy, B is generating a bounce, and the > bounce is hitting the mailing list provider. > > I do not understand why a bounce should be generated (and not the > incoming mail to B would be tagged as spam and/or null-routed). There is also a p=quarantine policy. Yahoo! and AOL, as well as PayPal, LinkedIn, Facebook, etcetera use p=reject, though. > That said, the result of the above is that B is unsubscribed from > the mailing list due to large number of bounces, but that is > because B is recognizing the policy A is publishing. B's domain need to do so in order to reject scams claiming to be from banks or social networks. > Do I get this right? More or less. Many people set up forwarding to an address at a domain different from theirs, typically large providers, so there are multiple places where the bounce may occur. Ale