[e2e] JOIN FOR FREE! ... Learn and Earn

Vernon Schryver vjs at calcite.rhyolite.com
Wed Feb 6 10:16:37 PST 2002


> From: Joe Touch <touch at ISI.EDU>

> As I'm sure you're aware, it is difficult to automate the removal of 
> spam altogether. We decided against limiting posts to 'members only' 
> because of the complexity incurred by users with multiple accounts 
> (subscribe from Y would mean submissions must be sent only from Y - 
> while there are hueristics to help avoid this, they aren't reliable).

There are other tactics that limit spam, including blacklists of
various sorts and mechanisms that detect bulk mail.

Blacklists for mailing lists such as this would have obvious problems
or would at least be controversial.

Mechanisms that detect bulk mail might be useable.  For example, as
I recall that recent junk message was listed by the public Distributed
Checksum Clearinghouse (DCC) servers as having been seen "MANY" (or
millions) of times before I received my copy via this list.  A good
case can be made that most mailing lists ought to use something like
the DCC to reject (to a log) any message whose body has been seen more
than a few times.  (Given the many, often repeated copies of calls
for papers that seem to be welcome on this mailing list, the definition
of "few" might be controversial.)

See http://www.rhyolite.com/anti-spam/dcc/ for lots of words about
the DCC as well as source and other stuff.

Proofs showing the DCC is wonderful, perfect, impossible, illegal,
fattening, or simply useless should be discussed elsewhere, where
they might not be off-topic.


Vernon Schryver    vjs at rhyolite.com



More information about the end2end-interest mailing list