Gmane
Favicon
From: Joe Emenaker <joe <at> emenaker.com>
Subject: Rs: SpamAssassin 3.0.0-rc3 RELEASE CANDIDATE available!
Newsgroups: gmane.mail.spam.spamassassin.general
Date: 2004-09-07 23:58:15 GMT (4 years, 42 weeks, 6 days, 21 hours and 51 minutes ago)
>
>
>On Friday 03 September 2004 09:17 pm, Theo Van Dinter wrote:
>
>>  - SpamAssassin now includes support for SPF (the Sender Policy Framework,
>>    http://spf.pobox.com/).
>
>Why bother with this?
>
>http://www.infoworld.com/article/04/08/31/HNspammerstudy_1.html
>

Although others have already given reasons why, I figured I'd toss in 
the analogy to explain why the dude from CypherTrust in the article is 
lacking in clue:

1 - Suppose that we have a problem with terrorists boarding planes and 
blowing them up.

2 - So, we make a list of all of the names of known terrorists and hand 
them to the security screeners.

3 - Then, the terrorists start boarding planes with fake ID's.

4 - So, we come out with some retina-scan ID system that can't be faked. 
Anybody still using the old, fakeable ID's will be treated to a cavity 
check, x-rays, and a bunch of hassle.

5 - So, the terrorists figure that their best shot is to use the new 
unfakeable ID's and hope that the security screener doesn't refer to the 
list of known terrorists.

The *lynchpin* to this whole scenario is that, if you don't check the 
names against the list of terrorists, the reliability of their ID card 
is pointless. The unfakeable ID card isn't a way of finding the 
terrorists, per se. It's a way of making *another* method more reliable.

By the same token, the point was never to be able to spot spammers by 
noting who isn't using SPF. Rather, the point is to make the blacklists 
more reliable. It is *only* when you use SPF in *conjunction* with 
blacklists/whitelists that you see any benefit from SPF.

- Joe

-- 
When freedom gives way to tyranny, it is not because tyranny comes
dressed as a wolf. Rather, it comes dressed as a shepherd,
pointing out other wolves. Go *read* the Patriot Act.

Attachment (smime.p7s): application/x-pkcs7-signature, 3198 bytes