J. Gomez writes:

 > Hello, I have this comment: What do you think are the probabilities
 > that YAHOO/AOL, etc, would go through the trouble of identifying
 > and declaring the many third parties needed to be allowed to DKIM
 > re-sign their messages, so that YAHOO/AOL users could painlessly
 > post to mailing lists in a world where DMARC happened to be widely
 > deployed?

First, in the usual case the mailing list is explicitly mentioned in
the addressee list at the time it reaches the destination domain, so
it automatically becomes a delegate.  So there would be no need to
specify "t=" in those cases.  Nor do I see a serious risk of
exploitation, since the implicit list is restricted to the explicit
(non-bcc) addressees.  DKIM sign To and Cc and replay attacks become
quite difficult, I should think.  I guess these considerations are
what Dave meant by "the protocol doesn't require that", but you'd have
to ask him.

Do you have a specific reason to think explicit delegate lists would
commonly be needed?

For your actual question, I can't speak to probability, but I think
that it would be reasonably easy to delegate the identification task
to the users.  Just as many webmail services allow you to "whitelist"
senders such as your bank, they could have a separate "delegate" list
(probably labeled "mailing lists").  Then the MUA module would
generate the "t=" parameter of the DKIM-Delegate header based on the
user's whitelist and the addressee fields.  I don't see a problem with
this from the mailbox provider's point of view, and it would be
greatly appreciated by many mailbox users.

_______________________________________________
dmarc mailing list
dmarc@ietf.org
https://www.ietf.org/mailman/listinfo/dmarc

Reply via email to