Merely as a participant: +1 to Dave's reply to this. On one particular point:
On Mon, Nov 9, 2020 at 5:51 AM Doug Foster <fosterd= 40bayviewphysicians....@dmarc.ietf.org> wrote: > If an evaluator detects a DMARC record without a policy tag, it MAY reject > the record as invalid or it MAY treat the record as equivalent to p=none. > Consequently, domain owners SHOULD include a p= tag, as the recipient > action is otherwise unpredictable.. > <ad> When it comes to requirements language in standards track documents, and particularly with respect to SHOULD, be prepared to answer this question: Under what circumstances might an implementer deviate from the recommendation presented by the SHOULD? If the text doesn't make that clear already, you can expect I'll ask that question when it comes up for AD Evaluation. If there's no solid answer to that, then you might consider MAY or MUST, or omit the requirements language entirely. In this case, why might an operator consciously omit "p=", and to what end? If there's no good answer to that, I'd consider making "p=" mandatory. </ad> -MSK
_______________________________________________ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc
