Subject: Re: Spam
From: rpw3@rpw3.org (Rob Warnock)
Date: Fri, 23 Jan 2004 07:17:06 -0600
Newsgroups: comp.lang.lisp
Message-ID: <ZRWdnRjzpO5Pg4zdXTWc-w@speakeasy.net>
Tage Stabell-Kul� <task@ifi.uit.no> wrote:
+---------------
| rpw3@rpw3.org (Rob Warnock) writes:
| > Unfortunately, that assumption stopped being true about 3-4 years ago.
| 
| Well, I used Clouddmark up until mid December.  I hadn't paid so much
| attention to spam since I received only about 10 per day.  When I had
| to let go of Cloudmark I was flooded by 100-200 spams every day; I
| simply hadn't realized how effective Cloudmark was.
| 
| I believe that for the moment, sending 1 email to N addresses (where N
| >> 1000) takes an order or two less time than sending N * (1 email to
| 1 address).  So although in the long run it might be a dead end, it
| worked amazingly well four weeks ago.
+---------------

Then you've been lucky. Your assumption of "sending 1 email to N addresses"
is no longer true for the worst spam, which is now using "innocent"[1]
systems hijacked by viruses/worms to do their SMTP sending for them,
thus becoming a form of DDoS[2] attack by thousands of machines which
can easily spend the tiny amount of time needed to "customize" (randomize)
each message.

An examination of my current inbox confirms this. Where it was previously
common to find spam with dozens or even hundreds of addressees per message
[with my address burried in the middle of other addresses], now almost all
of them have only one recepient: me.


-Rob

[1] If every Microsoft user whose machine was used by a virus to attack
    or send spam to another machine was fined, say, $1000.00 per outgoing
    connection, betcha that stuff would stop quickly enough, eh? But in
    the current world, "it ain't gonna happen"...  (*sigh*)

[2] Distributed Denial of Service.

-----
Rob Warnock			<rpw3@rpw3.org>
627 26th Avenue			<URL:http://rpw3.org/>
San Mateo, CA 94403		(650)572-2607