On Feb 15, 2004, at 6:19 PM, Dean Anderson wrote:
What's always amazing to me is the absolute certainty and zeal that
people
have that overly simple tests work. One should mark things you _think_
might be spam, as spam, and deliver them, or discard them. One should
not
create a mailbombing system. Self/foot/trigger.
I take some (very) small comfort from the thought that the implementors
of such systems will be ever less likely to communicate meaningfully
with each other about their own filters. If A asks B whether or not he
filters words like "viagra", B won't reply if the answer is yes. In
terms of natural selection, this reminds me (somewhat) of the voluntary
infertility of the Shaker movement (http://tinyurl.com/3a5k6) --
Nathaniel