[lug] More Spam Please?

John E. Koontz koontz at boulder.nist.gov
Tue Jun 3 09:11:22 MDT 2003


Just discovered an "unfilterable" piece of spam.  It was in html format and 
had been salted at random intervals with strings of <k...> where ... is 
random characters, and the length of k... was 15.  Not quite sure what the 
mechanics were, but perhaps it takes advantage of some anomaly of the kbd 
directive.  I filter it out with a filter on opening-angle-bracket k.   I 
haven't experimented with this further.

At 11:33 PM 6/2/2003 -0600, you wrote:
>The Matt wrote:
>>  Thus,
>>without the 1000s of pieces of spam needed, it'll be a good 9 months or
>>so before enough spam reaches me (I get 6 or so a day) to train the
>>filter.
>
>But if you flood the filter with thousands of "unrelated" spam, it will 
>still take 9 months for it to learn the specifics for you anyway.  If not 
>longer, since there's now much more noise in the system.
>
>I know, this assumes that your spam is somehow different than what the 
>rest of us get.  That seems unlikely... except if you're only seeing six 
>pieces of spam per day it may be true.  I see about that much in an 
>hour... after dropping about half of the traffic because the sender has no 
>A or MX records.
>
>_______________________________________________
>Web Page:  http://lug.boulder.co.us
>Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug
>Join us on IRC: lug.boulder.co.us port=6667 channel=#colug

John E. Koontz
NIST OCIO CASD (182)
303-497-5180

N39° 59' 42.1" W 105° 15' 49.7"




More information about the LUG mailing list