Well, at the high end of the spam game they do tests to see
if an address
is valid or not.. such things as image links in email that
are actually
links to a server side script at their end that is passed an
id matching
the email address that that particular message was addressed
to.

Also some record if the message was bounced or not, and this
is where
bouncing the spam back can be handy.. because the email
address will
not be validated as accepted and therefore it might be
removed from the
spammers Database.

Since many spammers use fake yahoo address's to send a ton
of mail
out in a short time, overloadign their accounts with bounced
messages
is a good thing too.

It will have a likewise beneficial effect on ISP's like the
russian spam
servers, whereby a huge stream of bounce messages uses up
some of the
servers bandwidth, and costs them money.. so they would be
more inclined
to do something about it.

A last benefit is the ego... its great to feel that you are
being active
rather then reactive with regards to spam.. filtering just
doesnt' feel
as good as rejecting spam... :-)

Right now this sort of thing is the best method we have for
reducing spam
that is unintrusive (unlike challenge response methods).

They guys that come up with stuff like postfix proxy are
legends (and WV,
the creator and maintainer of postfix (and tcp wrappers) is
a surprisingly
nice guy, (unlike D Beirnstein (qmail), who is an
egocentrical ellitist))
both are undoubtably brilliant though.


regards

Franki



-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Martin
Fahrendorf
Sent: Wednesday, 23 July 2003 5:55 PM
To: [EMAIL PROTECTED]
Subject: Re: [expert] Using Postfix to send mail.


Am Mittwoch, 23. Juli 2003 10:49 schrieb Joerg Mertin:
> Hi Martin,
>
> thx for the hint. I might give it a try.
> However - what buzzes me here is that if you use the proxy
method to
> identify spam - you have to get the spam anyway through
it - don't you ? So
> - the spam will use your bandwidth to get analyzed by the
"proxy"
> application - and the proxy application then returns a
Spam-Detected
> message which will be interpreted by the postfix process
and which will
> make that one reject the message definitly.
>
> IMHO - the only difference is that the remote side will
get a reject
> message if I understood correctly the process. Please
correct me if I'm
> wrong.

Jepp, that's right.

>
> Do you think this reject message will inhibit spammers to
send you more
> mail ?

Hm, that is wild guessing. I think spamers dont want to
waste bandwith, they
want to get their mails read. If you silently delete the
spam the spamer
don't know if their mails get read or not. So they assume
thei can send their
mails again.

If the spam get rejected they know that you don't accept the
mail. It is up to
the sending server to handle the rejection.

Do they send spam again? Yes, I fear they see it as a kind
of sport to get
their spam trough. But doing nothing is no sollution.

>
> NOTE - the actual spamassassin/postfix/anomy method
enables you to actually
> get the Mail in, spamassassin checks it through
spamc/spamd - and if it's
> beeing detected a SPAM - you can tell the delivery script
to delete it or
> move it to a local-file for laer analysis ...
>
> Cheers
>
>       Joerg
>

Martin
--
------------------------------------------------------------
H E L I X Gesellschaft f�r Software & Engineering mbH
------------------------------------------------------------
Hanauer Landstrasse 52              Telefon (069) 4789 35-30
D-60314 Frankfurt am Main           Telefax (069) 4789 35-44
------------------------------------------------------------
http://www.helix-gmbh.net                [EMAIL PROTECTED]
------------------------------------------------------------


Want to buy your Pack or Services from MandrakeSoft? 
Go to http://www.mandrakestore.com

Reply via email to