John Levine wrote:
One of SPF's many problems is that it posits a model of the
e-mail world that is a lot simpler than the real world.
KISS is a feature and no problem.
In the real world, there are whole lot of remailers and
forwarders,
Remailers (like real mailing lists) have no problem at all
with SPF. Some arguments "against" SPF from say the EFF
are far beyond surreal. Any anonymous remailer is by its
own definition automatically "SPF ready".
no matter how desperately some domains might want to argue
that nobody's allowed to forward, etc.
These arguments are nonsense, you can forward as much as you
like. You only have a problem if you check SPF behind your
"border MTA", but the good news is, this is _your_ problem.
People want their bank statements or whatever, even if the
address they give to the bank is their permanent forwarding
address from their alma mater.
Something on their side has to give. It's not a problem of
the senders, they only publish which MTAs send their mail to
third parties, that's their border.
bounces vs. rejects are an increasing problem, and I'm having
trouble figuring out if there's any way I can do bounces
reasonably
You could try SPF, if the HELO resp. MAIL FROM sender policy
results in a FAIL simply reject the mail. You're only forced
to bounce if you forward mail and the next hop rejects it.
Otherwise it should be possible to come to a decision about
"reject" vs. "accept" during the SMTP dialogue. In theory.
BTW, something I've seen in MASS, you said that "most" mail
today contains HTML. Is that incl. or excl. UBE / UCE ?
If 4 of 5 mails are spam and other unsolicited stuff today,
and 4 of 5 spam mails contain HTML, then 64% of all mail
are both unsolicited and contain HTML. But it's not clear
how often HTML is used in solicited mail, I doubt that the
majority contains HTML.
Bye, Frank