[ietf-dkim] Re: Requirements comment: Bigbank example description
william at elan.net
Thu Aug 10 06:54:06 PDT 2006
Most legitimate Resent cases are such where last sender (person listed
in Resent-From) would very likely be be on the personal whitelist of
the recipient because of extensive prior communication.
BTW - this means that absolute/strict policy may not be absolute in
practice where the filter manufacture may decide to use system with
whitelisting overriding any results from policy check (including
those from very strict only-me, no 3p policy).
On Thu, 10 Aug 2006, Frank Ellermann wrote:
> Stephen Farrell wrote:
>> must be an earlier version Mike sent me.
> Okay, I started to wonder if what I have might be already old.
> Maybe draft-ietf-dkim-ssp-requirements-00.txt still needs your
> approval for its "official" publication, because it's a "-00".
>> I'm quite sure that some requirement will contradict some
>> aspect of each of the current proposals, but we shouldn't
>> worry about that too much for now.
> I like to have it clear in the requirements, that IFF a future
> SSP won't work for such "resent" scenarios, then that has to be
> explicitly stated.
> It's relevant for a hypothetical 2822bis, if 2822bis tries to
> decree that all forms of Resent-* are obsolete and should be
> replaced by ordinary forwards as MIME message/rfc822 parts.
> I vaguely recall that an MMS2SMTP gateway RFC already goes in
> that direction, either get rid of or ignore any Resent-*.
More information about the ietf-dkim