On 1/25/07, Peter J. Holzer <hjp-asrg(_at_)hjp(_dot_)at> wrote:
> > good point. thanks. that would have to be fixed first of course;
> > for instance DATAFIRST connections only allowed from reputable
> > peers.
>
> I agree with Claus. Me too, I don't want to receive messages for non
> existent recipients.
>
> A way to fix this may be a negotiation between the server and the
> client, probably it's the server who shall ask for DATAFIRST.
It has to do that anyway in the EHLO reponse. It could advertise
DATAFIRST to "reputable" peers only. But I don't think that helps much,
because you specifically want it for the less reputable hosts: Where
some messages will be accepted by some users but rejected by others.
in the e-postage scenarios driving DATAFIRST, you would want to provide
the service only for peers participating in the e-postage system. The advantage
over deferred deliviry status reports becomes clear if you are trying
to accept or
refuse a two-cent e-mail for a million customers whose individual
gatekeeper policies
have the option of accepting, refusing, or haggling. Instead of
trying more than a million
times to send the message with one recipient at a time, DATAFIRST
provides a shortcut
to saying "same data as last time."
As such it is not a spam reduction technology in itself, but is an
optimization to make
an unimplemented e-postage scheme work more smoothly.
I brought it back up in response to Eric Hall's 2004 proposal, which
has come under
discussion again.
http://www1.ietf.org/mail-archive/web/asrg/current/msg12808.html was
the beginning
of the discussion of it, last summer. I guess I've got just a few
more months to
try to patent that aspect of my particular e-postage scheme in order
to woo my angels :)
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg