They were just talking about this on the Postfix list today, as well.
Wietse Venema is the developer of Postfix.  Attached is a question regarding
the Message ID, and his response.  Interesting that this issue came up on
both lists today.

Bill
----- Original Message ----- 
From: "R. Scott Perry" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, June 03, 2003 1:43 PM
Subject: RE: [Declude.JunkMail] Non-unique MessageID vs. BADHEADERS ?


>
> >So the HELOBOGUS will apply the same logic and NO longer check for
"BOGUS"
> >host names?
> >
> >If not - then why doesn't the same logic apply?
>
> The HELOBOGUS test gets around this by only checking the HELO/EHLO if the
> sender is not a local user.
>
> >Frankly, I rather prefer to have a test that does it advertises to do
(e.g.,
> >check for "BAD HEADERS"), and then let ME decide via "weights", how
highly I
> >want to rate this.
>
> The problem here is that if the BADHEADERS test catches 50% of legitimate
> E-mail, nobody will use it.
>
> If you'd like to take this on with the mail client vendors, that would be
> great.  But this is one battle we're not going to try taking on.
>
> >Remember: Nobody is forced to use the BADHEADERS test
>
> True.  But given that it can catch about 40%-50% of spam and virtually no
> legitimate E-mail, it's one of the best tests in Declude JunkMail.
>
> >  but those who do should be able to 'rely' on it discovering
non-compliancy.
>
> It isn't designed to be a foolproof test of the headers (such as one that
> mail client vendors could use to say "Our headers are 100%
> RFC-compliant).  It's designed to detect headers that are common in spam
> that are not sent by standard mail clients, and which are not
RFC-compliant.
>
> >For all other purposes you have the SPAMHEADERS test that is
> >designed/advertised to be "flexible" and which is expected to "adopt"
based
> >on occurrence of certain issues in the "wild" - so THERE it would make
sense
> >to leave the MessageID FQDN check out of SPAMHEADERS.
>
> The qualifications for the two tests are:
>
> BADHEADERS: The header [1] must be common in spam, [2] must not be sent by
> most legitimate mail clients, and [3] must be non-RFC-compliant.
>
> SPAMHEADERS: The header [1] must be common in spam, [2] must not be sent
by
> most legitimate mail clients, and [3] is RFC-compliant.
>
> The problem is that the non-FQDN in a Message-ID: header violates #2 -- it
> is commonly sent by legitimate mail clients.  Therefore, it isn't
> appropriate for these tests.
>
>                                                     -Scott
> ---
> Declude JunkMail: The advanced anti-spam solution for IMail mailservers.
> Declude Virus: Catches known viruses and is the leader in mailserver
> vulnerability detection.
> Find out what you have been missing: Ask for a free 30-day evaluation.
>
> ---
> [This E-mail was scanned for viruses by Declude Virus
(http://www.declude.com)]
>
> ---
> This E-mail came from the Declude.JunkMail mailing list.  To
> unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
> type "unsubscribe Declude.JunkMail".  The archives can be found
> at http://www.mail-archive.com.
>
--- Begin Message ---
M. Yamaura:
> Hi,
> 
> I want to reject incoming mail without message-id. I saw
> the some sources and it seems that Postfix can reject the content
> of header below but I couldnot find how to reject when no message-id mail
> is comming. 

The Message-ID header is not required.

A week ago I added some code to require Message-ID (or Received:
or From:  or Date:) and removed this code after a few days. This
feature stops too much mail.

        Wietse

--- End Message ---

Reply via email to