The proposed text is in GitHub Issue #2 at 
https://github.com/dkim2wg/spec/issues/2, which also contains the proposed 
Section 5.x definition of re-originating services. The relevant Section 6.1 
addition is reproduced below.

The email summary you are responding to was imprecise on two of your points. 
"p=reject" was an inline shorthand and should not be in the normative text - 
"authenticated sending domain" is the right framing. "Functioning abuse 
contact" is also vague; "published abuse contact" is more precise and avoids 
the definitional problem. Both are fair corrections.

On local policy: the goal is not to override irrational policy - you are right 
that no BCP does that. The goal is to give rational implementers writing DKIM2 
receiver policy for the first time a normative anchor, before deployment 
patterns harden. The specific concern is messages that carry standard list 
headers (List-Id, List-Unsubscribe, etc.) alongside a single-hop originator 
signature. Without BCP guidance, a policy author might reasonably treat "list 
headers + no chain" as a suspicious pattern worth penalizing. The proposed 
SHOULD prevent that interpretation from taking root.

A re-originating service does not remove accountability - it substitutes its 
own domain and reputation for the original sender's. If it enables spam or 
phishing, the complaint signal lands on the relay's domain directly. The 
accountability mechanism is intact; it operates at the relay layer rather than 
the author layer."

Proposed Section 6.1 addition:

Receiver ADMDs SHOULD NOT treat a message bearing a single valid DKIM2 
originator signature as suspicious solely because no prior chain is present and 
the message carries threading markers (References, In-Reply-To). Threading 
markers without a prior chain is the expected and legitimate output of a 
re-originating intermediary (see Section 5.x). A valid single-hop DKIM2 
signature from a re-originating service should be evaluated on the same basis 
as any other Sender ADMD originator signature.

William Weiner
Weiner Advanced Development LLC, maker of EMail Parrot (emparrot.com)



From: Richard Clayton <[email protected]>
To: <[email protected]>
Date: Thu, 16 Jul 2026 00:21:26 -0600
Subject: [Ietf-dkim] Re: DKIM2 with message encryption

 > -----BEGIN PGP SIGNED MESSAGE----- 
 > Hash: SHA1 
 >  
 > In message <19f5c160bc3.1b25e9471433010.3436756724027830514@wadevelopmen 
 > t.co>, William Weiner <[email protected]> 
 > writes 
 >  
 > >The concern is not whether this constitutes a valid DKIM2 chain. It does. 
 > >The 
 > >concern is whether the BCP will give receivers normative guidance to treat 
 > >single-hop re-originator signatures on equal footing with any other Sender 
 > >ADMD, 
 > >or whether the absence of a prior chain becomes a negative signal under 
 > >local 
 > >policy. 
 >  
 > the vast majority of email goes from A to B and never goes anywhere else 
 >  
 > why would a BCP need to say anything about that ? 
 >  
 > >You note it would be "quite irrational" for local policy to penalize this 
 > >pattern. 
 >  
 > many local policies are irrational -- but they are local and no BCP will 
 > make any difference to that one way or another 
 >  
 > >The BCP 
 > >text I am proposing is that normative statement for re-originators, written 
 > >before local DKIM2 policy hardens rather than after delivery breaks. 
 >  
 > I'm sorry, I missed the text that you are proposing. Please could you 
 > repeat what it was...  if it was along the lines of "local policy MUST 
 > be something or other" that is clearly nonsense because that would make 
 > what is being described something other than "local policy" 
 >  
 > >The ask is one sentence in Section 6: a valid single-hop DKIM2 signature 
 > >from a 
 > >re-originating service with documented characteristics (p=reject, 
 > >functioning 
 > >abuse contact, stable domain reputation) SHOULD be evaluated on the same 
 > >basis 
 > >as any other Sender ADMD originator, without prejudice for the absence of a 
 > >prior chain. 
 >  
 > ah right you are proposing something ... but you seem to have missed out 
 > the definition of what a "re-originating service" might be, you are 
 > tying it to DMARC (which seems problematic), you have not explained what 
 > "functioning" means in terms of an abuse contact (whatever that is) 
 > [[and note that RIPE have been tying themselves in knots for years for 
 > this issue in respect of IP address allocation]] ... and finally I am 
 > unsure what a Sender ADMD originator might be. 
 >  
 > - -- 
 > richard @ highwayman . com                       "Nothing seems the same 
 >  Still you never see the change from day to day 
 >  And no-one notices the customs slip away" 
 >  
 > -----BEGIN PGP SIGNATURE----- 
 > Version: PGPsdk version 1.7.1 
 >  
 > iQA/AwUBalh4ZmHfC/FfW545EQKtowCglcXG6HukicWO+3wMbUbvIFrjnbEAn3i3 
 > 1NHMO8zRwi4O1f42UpVyxYtJ 
 > =8SO7 
 > -----END PGP SIGNATURE----- 
 >  
 > _______________________________________________
 > Ietf-dkim mailing list -- [email protected]
 > To unsubscribe send an email to [email protected]
 > 


_______________________________________________
Ietf-dkim mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to