[SpamCop-List] Re: A mailman opt-in plus confirmation mailing list is spam?

[Mike Easter]

Don Wannit wrote:

> This means that a useful spamtrap address can never be
> any of the following:

Without getting into any of the specifics about spamtrap addresses which
are known by many around here, I once asked in this newsgroup about the
philosophy of spamcop spamtraps, whether they should be very random
usernames so as to 'never' occur in a so-called dictionary attack, very
common usernames so as to 'routinely' occur in so-called dictionary
attacks -- and similar 'extremes'.

At that time the answer from Ellen was 'yes'.  That is, that there are
all different philosophical kinds of spamtrap addresses.

The only requirement as I understand it is that the addy has never been
used by anyone for any purpose, so that its 'exposure' has never been to
subscribe to anything, including free-for-all or anything else.  The
fact that a spamtrap address may have been found by a miscreant and used
to forge subscribe to anything is not eliminated from the rack of the
wide range of possibilities for such spamtrap addies.

I don't think that spamtraps are manually eliminated by deputies who
find them forge subscribed in confirmation hits.  In fact, I don't think
spamtrap addies are manually eliminated for any reason -- even if the
reason might be that the spamtrap addy does not appear to be a 'secret'
any more.

My concept of a dictionary attack is that the 'dictionary' is made up of
many many usernames scraped from various places including millions CDs
coupled with alternative domainnames scraped from similar very many such
places.  The dictionary is /not/ made of dictionary type words.

-- 
Mike Easter
kibitzer, not SC admin