<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7638.1">
<TITLE>Re: ISP wants to stop outgoing web based spam</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Hello Hank:<BR>
<BR>
<BR>
On 8/9/06 3:28 AM, "Hank Nussbacher" <hank@efes.iucc.ac.il> wrote:<BR>
<BR>
><BR>
> Back in 2002 I asked if anyone had a solution to block or rate limit<BR>
> outgoing web based spam. Nothing came about from that thread. I have an<BR>
> ISP that *wants* to stop the outgoing spam on an automatic basis and be<BR>
> a good netizen. I would have hoped that 4 years later there would be<BR>
> some technical solution from some hungry startup. Perhaps I have missed<BR>
> it. What I have found so far is:<BR>
><BR>
> Detecting Outgoing Spam and Mail Bombing<BR>
> <A HREF="http://www.brettglass.com/spam/paper.html">http://www.brettglass.com/spam/paper.html</A><BR>
> SMTP based mitigation - thing on HTTP/HTTPS<BR>
><BR>
> Stopping Outgoing Spam<BR>
> <A HREF="http://research.microsoft.com/~joshuago/outgoingspam-final-submit.pdf">http://research.microsoft.com/~joshuago/outgoingspam-final-submit.pdf</A><BR>
> Research paper - nothing practical<BR>
><BR>
> Throttling Outgoing SPAM for Webmail Services<BR>
> <A HREF="http://www.ceas.cc/papers-2005/164.pdf">http://www.ceas.cc/papers-2005/164.pdf</A><BR>
> Research paper - nothing practical<BR>
><BR>
> ISPs look inward to stop spam - Network World<BR>
> <A HREF="http://www.networkworld.com/news/2004/071204carrispspam.html">http://www.networkworld.com/news/2004/071204carrispspam.html</A><BR>
> Bottom line - no solution<BR>
><BR>
> So I am trying once again. Hopefully someone has some magic dust<BR>
> this time around.<BR>
><BR>
> Thanks,<BR>
> Hank Nussbacher<BR>
> <A HREF="http://www.interall.co.il">http://www.interall.co.il</A><BR>
><BR>
<BR>
My answer is based on the word "startup" so I'm assuming "no money" but I<BR>
could be "wrong". :-) We use the standard SpamAssassin, ClamAV setup both<BR>
on ingress and egress. On egress we set the detection levels and divert and<BR>
save anything that is marked as Spam rather than sending it on with headers<BR>
and subject modifications.<BR>
<BR>
We've found this to be very effective in reducing our scores with Comcast<BR>
and AOL in particular and it's pretty much stopped our being blocked by<BR>
those services, even using a fairly loose setting for SpamAssassin. As a<BR>
service provider that forwards tons of mail to addresses on those networks<BR>
(previously un-scanned so we forwarded everything, including Spam) we've<BR>
found it essential to put these filters in place to guarantee (as much as<BR>
anyone can) service for our email customers.<BR>
<BR>
Regards,<BR>
<BR>
Mike<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>