Re: [tlug] TLUG spam?

[Brett Robson <b-robson@example.com>]

Stephen J. Turnbull wrote:

>The only way to beat that is to go "AI", and assess spamminess and/or
>topicality of content.  Which is what I said in the first place.
>  
>

There is no silver bullet, and AI certainly isn't one. Any system based 
on content is only going to have a short run before it's rules are 
defeated; how many ways can v1agra be mis-spelt? Content based filtering 
is especially difficult when you have to accept all languages. We have 
real news stories lifted from news sources coming in with an image 
carrying the ad which are extremely difficult to detect. Plus content 
filtering is very CPU intensive.

Stopping spam needs layered protection: network level (eg refusing 
connections from dialup pools), mail protocol (eg increasing waits for 
each bad address), content (eg Spam Assasin), honey pots, and of course 
active monitoring. There is the other issue too of preventing abuse of 
your own network.


>Also, some people I trust to know what they're doing guesstimate that
>all network-based checks are expensive in bandwidth. 
>

Anything that stops thousands of emails from a single source is 
definitely not expensive in bandwidth.