On Tue, Oct 21, 2003 at 01:08:47PM -0500, Dustin Trammell wrote:
If the mail bounces, it will return to
mailer-daemon(_at_)forwardingserver which should then recognize that it
forwarded the original message (perhaps by looking at the from:
line, a header line that it added when forwarding, etc. and perhaps
cross referencing it's logs) and then forwarding the bounce to the
original sender.
Logically that makes sense to me, seems a lot less kludgey than SRS,
and may require less code change to existing MTA's than implementing
SRS will.
I like Dustin's idea of adding multiple headers instead of
modifying the addresses via SRS. If a message bounced, the headers
could be used to backtrace the real path of the message. The header
could include and HMAC-SHA1 "signature" to prevent spammers from
generating bounces.
Here's an example of what I'm thinking of (and there may be a
much better way to do this).
Let's say that a(_at_)a(_dot_)com sends a message to b(_at_)b(_dot_)com
which gets
forwarded to c(_at_)c(_dot_)com(_dot_)
Here are some of the headers form the first "iteration" of the
message:
------
Return-Path: a(_at_)a(_dot_)com
From: a(_at_)a(_dot_)com
To: b(_at_)b(_dot_)com
------
The message now arrives at b.com, and is forwarded as follows:
------
Return-Path: b(_at_)b(_dot_)com (or should this be
bounce-manager(_at_)b(_dot_)com?)
From: a(_at_)a(_dot_)com
To: c(_at_)c(_dot_)com
Spf-Forward: <forward-count> b(_at_)b(_dot_)com c(_at_)c(_dot_)com <hmac-sha1
hash>
------
<forward-count> is a counter that increments for each
Spf-Forward header that is added. That way, they can be backtraced in
the proper order.
If the message bounces at c.com's server, it would go bact to
b.com's server, which could then look at the header, validate the
hmac-sha1 hash, and bounce the message back to a.com's server.
Comments anyone?
This approach might scale better across multiple forwards than
SRS does. I have a college alumni address that forwards to an address
that forwards to my inbox. My college uses that address on it's
mailing list, so that is three levels of forwarding. I would imagine
that three levels of forwarding is not uncommon in the real world, and
more happens occasionally.
As for the envelope sender, that could agree with the
Return-Path, and not the from addresses.
Izzy
______________________________________________________________________
The pursuit of perfection is the enemy of progress.
-------
Sender Permitted From: http://spf.pobox.com/
Archives at http://archives.listbox.com/spf-discuss/current/
Latest draft at http://spf.pobox.com/draft-mengwong-spf-02.txt
To unsubscribe, change your address, or temporarily deactivate your
subscription,
please go to
http://v2.listbox.com/member/?listname(_at_)©#«Mo\¯HÝÜîU;±¤Ö¤Íµø?¡