spf-discuss
[Top] [All Lists]

Re: [spf-discuss] Bounce-Spam and SPF-Ignorant ISPs - it is time to retaliate?

2005-11-29 05:38:49
On 11/29/2005 07:16, paddy wrote:
On Tue, Nov 29, 2005 at 11:53:25AM +0000, Julian Mehnle wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

David Woodhouse wrote:
[...] the single rule should be 'just don't generate bounce messages'.

You should never, during normal operation, accept a mail by SMTP unless
you're actually going to deliver it.

Do _all_ your checking, whether it includes SPF or not, before giving a
successful response to the end of the DATA. If you don't want the mail,
then just say so. Don't accept it and then find yourself stuck with the
task of bouncing it.

Very good advice indeed.

Actually even forwarders/relays should try to deliver a message to the
remote MTA in real time before accepting it themselves (sort of a
semi-proxy mode).

I've been toying with the same idea recently.

I'm a little wary, for the following reasoning:

some things are MUST, others are only nice to have.
Surprisingly often, when you get into the details,
the MUSTs preclude the nice to have.

store and forward would seem to be a MUST of email.
failure is always an option :)

I installed qpsmtpd the other day with the intention of looking at it
from angles like thisi (when I get a round tuit), and I'm keen to learn
how other MTAs support or can be coaxed into the kind of semi-proxy
behaviour?

sendmail recipe anyone?

Only if the remote MTA cannot be reached should the
message be accepted and stored for later delivery.

yes, a key reason why its hard to see why bounces should go away entirely.

Yes, but very much a corner case at most.  Border MTAs should have all the 
information necessary to determine if a message is acceptable to deliver 
(e.g. list of valid users).  The only time this might be appropriate is in 
the case of multi-hop e-mail accross administrative boundries.

Scott K

-------
Sender Policy Framework: http://www.openspf.org/
Archives at http://archives.listbox.com/spf-discuss/current/
To unsubscribe, change your address, or temporarily deactivate your 
subscription, 
please go to 
http://v2.listbox.com/member/?listname=spf-discuss(_at_)v2(_dot_)listbox(_dot_)com

<Prev in Thread] Current Thread [Next in Thread>