The list headers - List-Id, List-Unsubscribe and similar - are covered by DKIM2 signature and are observable by any receiver writing policy, so the messages are not quite indistinguishable from arbitrary single-hop mail.
You may be right that receivers will, in practice, treat our privacy relay's messages as ordinary single-hop mail. That is precisely the outcome we want. The BCP request is not asking receivers to treat re-originating intermediaries differently - it is asking the BCP to explicitly confirm that a valid single-hop originator signature without a prior chain is the expected and correct output of this pattern. If ordinary receiver behavior is already what you describe, the sentence costs the WG nothing. If a receiver later writes policy that treats "no chain plus threading markers" as a suspicious signal, the normative text is what forecloses that interpretation before it spreads. The DMARC history is the cautionary case. p=reject worked as intended, the harm to mailing lists was foreseeable, and the absence of normative BCP protection for the mailing list pattern left no lever to push back on. A one-sentence statement is cheap insurance against the same dynamic. William Weiner Weiner Advanced Development LLC, maker of EMail Parrot (emparrot.com) From: John R Levine <[email protected]> To: "William Weiner"<[email protected]>, "Emanuel Schorsch"<[email protected]> Cc: "ietf-dkim"<[email protected]> Date: Mon, 13 Jul 2026 16:11:08 -0600 Subject: Re: [Ietf-dkim] Re: DKIM2 with message encryption > On Mon, 13 Jul 2026, William Weiner wrote: > > > You are right that I overstated the case. Blocking mailing list mail under > > p=reject was the expected behavior from the spec, not an irrational local > > policy. The better lesson is that the spec worked exactly as intended and > > still produced collateral harm to a legitimate use case that wasn't > > protected by the BCP - and the workarounds required after the fact were > > worse than prevention would have been. > > > > You say you don't see harm in writing it down. That is all we are asking. > > Put me in the once-bitten-twice-shy camp: I would rather have the > > normative statement and not need it than need it and not have it. > > Your use case is to put it mildly rather exotic. There's nothing wrong > with that, but I don't think we can expect people to do anything special > to ensure that it works. > > Unless I'm missing something, your messages look like any other locally > originated one-hop messages. You and the final recipient may know that > there's history wrapped up inside it, but so long as the single signature > validates and the recipients don't complain, recipient mail filters will > treat it like any other single hop mail. > > R's, > John > _______________________________________________ Ietf-dkim mailing list -- [email protected] To unsubscribe send an email to [email protected]
