Re: [Asrg] Re: Why SPF? (was: Anti-spam laws do work, FYI. There's proof.)

["Alan DeKok" <aland at ox.org>]

der Mouse <mouse at Rodents.Montreal.QC.CA> wrote:
> If you check into it, I suspect you'll find that a lot of secondary-MX
> sites do a lot more than just adding Received: headers.  (Any that run
> sendmail, for example, will add missing message-IDs, add domains to
> from-lines that don't have them, that sort of thing.)

  And I would argue that such behavior is "non-optimal", and very
likely wrong.

> This sounds as though you're saying "the envelope has changed" in a
> rather complicated way: that an "SMTP forwarder" is something that
> doesn't change the envelope and an "email forwarder" is something that
> does.  Would that be an accurate summary?

  Er, pretty much, yes.

> Perhaps it would.  Feel free to try it.  Try to set up an email
> forwarder that maintains sufficient state to DTRT with responses.
> 
> Myself, I think the state storage requirements put it out of reach,
> especially given how long the email state needs to persist.  Perhaps
> I'm wrong; that's why I suggest trying it.

  I agree.  That's what research & protocol design is for.  It's
possible to mark up the forwarded email with information about the
return path.  That information can be signed, encrypted, whatever, so
that the email forwarder doesn't have to keep state.

  But the act of asking the recipient to send bounces to a third
party, rather than have it follow the immediate return path is most
likely wrong.

  I also think that using SMTP for bounces is most likely wrong, too.
But that's another story.

  Alan DeKok.


_______________________________________________
Asrg mailing list
Asrg at ietf.org
https://www1.ietf.org/mailman/listinfo/asrg