Re: [AMaViS-user] OT: Where to find info on DNS based spam prevention

[Robert L. <rj...@re...>]

At 01:27 2004/04/15, Jo...@id... wrote:
>At 11:17 15-4-2004 +0300, you wrote:
>>I had recently heard of a new attempt to stop spam by broadcasting the
>>hosts on which *your domain* sends email from; like for egenet.com.tr , I
>>could define a handful of servers that transmit email out, and each site
>>receiving mail from egenet.com.tr verifies that the email actually
>>originated from one of these hosts, and not some &*^*&^* sending stuff
>>from blah-dsl.verizon.net pretending to be from egenet.com.tr..
>
>I think you are referring to SPF: Sender Permitted From

Close--they changed the name to "Sender Policy Framework" back in 
February.  The official site is http://spf.pobox.com/.

There are at least two other competing proposals which do similar things:

Microsoft's "Caller ID for E-mail" 
(http://www.microsoft.com/mscorp/twc/privacy/spam_callerid.mspx)

Yahoo!'s "Domain Keys" (specification still in development)

Basically, SPF and its competitors offer an additional bit of useful 
information for the domains that implement them.  If a domain publishes SPF 
records, you can use that information to determine whether the peer 
connecting to your MTA is authorized to send mail from that domain.  If the 
domain does *not* publish an SPF record, you can't draw any conclusion at 
all.  The upcoming SpamAssassin 3.0 includes support for SPF, incidentally.


Robert LeBlanc <rj...@re...>
Renaissoft, Inc.
Maia Mailguard <http://www.renaissoft.com/maia/>