On Thursday 25 March 2004 01:20, Guillaume Filion wrote:
> Tim Meadowcroft wrote:
> > I started off thinking that I'd suggest adding basic tar-pitting
> > (http://www.gordano.com/kb.htm?q=1112) to selected plugins[...]
>
> About tarpitting. It seems to me that any half technical spammer will
> use a multi threaded program that opens thousands of simultaneous SMTP
> connections. A tar pit would only stall one of those connections, using
> a few KB of RAM. Hundreds of tar pits would only slow the spammer by a
> fraction of a percent.

Well, first off the "half-technical" bit is doubtful ;^)

Then remember that I'm looking to stop the major source of who's joe-jobbing 
me, and these seem to home cable/DSL machines infected with trojans, and so 
many of them are going to low-end Win9x or similar, with little memory, a 
poor TCP/IP stack (yes, I did nmap quite a number of them when investigating 
who was joe-jobbing me - dangerous I know but I notified the abuse 
departments of the ISP of my actions), and crap threading, so stalling 
individual threads will slow those machines down - these aren't multi-CPU 
Solaris servers with efficient context switching and loads of RAM.

Plus, these machine may have 500kbits/sec download, but are usually limited to 
something like 64kbits/sec upload, so by making them send the full virus 
payload (a 64k MIME encoded .scr or pif file) or whatever email they're 
sending before denying them, you choke and waste their upload bandwidth.

That's why sending them a huge amount of text in return doesn't hurt that 
much, but choking their upload does. 

It's all about changing the economics of spam - like I said, refusing a 
connection early just makes life CHEAPER for the spammers - you're saving 
them money by rejecting their email before they've sent it, basic economic 
theory says that you should make rejections cost MORE than acceptance if you 
want to stop it. A typical stat is that a spammer sends 1 million spams for 1 
sale - if you make 10% of those rejections cost less, his profit margin goes 
up. Make those same 10% cost 10x as much, and you've just doubled his costs.

Cheers

--
Tim

Reply via email to