Hello,

will be there any concerns for sending slim forensic DMARC reports (ruf=) on 
failed DKIM validation, if

* between sender and recipient there are no intermedates/aliases/redirecting 
providers,
* the third MIME part of multipart/report is cut (contrary to 
https://tools.ietf.org/html/rfc5965#section-2 bullet d),
and
* in the message/feedback-report part
  - either the Original-Envelope-Id is included,
  - or Original-Message-Id is included

(where Original-Message-Id will be defined to be the Message-Id of the message 
that is reported)?

The Original-*-Id identifiers do not expose privacy information, but let the 
sending server identify for which message
the DKIM signing/validation do not match.  Whether the sending user has deleted 
the message in the meantime does not
matter.  Knowing which message is problematic is a huge improvement compared to 
the current situation.  First, the
sender can validate with different implementations whether they all produce the 
same signature for that message.

Second, if the message in question is sent over a mailing list, the From: was 
changed by the MLM, the DKIM signature was
added after the mail left the MLM but before leaving the MLM-mail-server, then 
this very message is likely to be
distributed to several mail providers.  If one provider does not validate the 
signature, and the other providers
validate the signatures, (or all mail providers do not validate), then somebody 
can take some actions so that the cause
for the failure is resolved and does not happen again in the future.  A clear 
plus for all DMARC-users.

Regards
  Дилян

_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc

Reply via email to