John Levine writes: > Here's a draft that puts the forwarding thing into DKIM, with the > dread version bump.
I agree that this proposal needs a version bump, for the reasons given in Section 5 of the document. I still think that these semantics are better put into the -Delegate header because it's simpler and requires no version bump[1], but the version bump approach seems tenable, and is more general, I suppose. I think the requirement that To and Cc "SHOULD" be signed when "t" and "c" are present in the fs parameter should be promoted to a "MUST". It's easy to satisfy the requirement at the signing stage, adds more complexity to the verifying stage, and the flexibility is redundant. (If desired the same semantics can be achieved by specifying the addresses in To: and/or Cc: explicitly, and not signing the corresponding field.) Footnotes: [1] I don't "dread" it, I just don't think it's necessary or useful. _______________________________________________ dmarc mailing list [email protected] https://www.ietf.org/mailman/listinfo/dmarc
