> Our spec needs to fix the evaluators, and their products, not the sender policy.
No: it should be doing both. Let’s look at it this way: Suppose a general-use email provider called “Hooya” promulgated a policy that said receiving domains should bounce any message from a hooya.com sender that was fanned out to mailing list recipients. I think that, worded that way, we would all agree it’s a bad idea. If we were writing a policy spec of some sort and we were addressing this, I think we should say BOTH that such a policy is inappropriate in that would damage mailing list operation AND that receiving domains should not be accommodating that policy. Hooya’s use of p=reject would amount to announcing that policy, given how mailing list software works and has worked for decades. Barry
_______________________________________________ dmarc mailing list [email protected] https://www.ietf.org/mailman/listinfo/dmarc
