ietf
[Top] [All Lists]

Re: email and spam (was: Re: namedroppers, continued)

2003-01-17 10:54:42
From: Bob Braden <braden(_at_)ISI(_dot_)EDU>

...
Another thought on the spam problem and Frequently Proposed Solutions
in general: as a community we have become obsessed with ephemeral
information.  That is, we all sit in front of our terminals, read our
email, (re-)invent new ideas, and spew them instantly across the world;
these ideas are lost and forgotten in a week.

The web does go some limited way towards creating archival information
-- ideas that remain and can be used to inform later idea -- but it
does not go far enough.  Ultimately, the only way we can avoid
regurgitating the same ideas over and over is to invest some of our
effort in converting ephemeral information into truly archival
information. ...

Archiving is useless unless people consult those archives.  Those who
continually re-invent (non-)solutions for spam will not be detered by
archives any more than IPv8 advocates are bothered by the difficulties
some people claim to see in representing more than 4,294,967,296 things
with 32 bits.

For example, those who think that authentication could stop spam won't
change their minds because someone writes something more.  If written
words mattered, they'd know about RFC 2554, RFC 2487, RFC 2645, and
RFC 2476 instead of insisting that SMTP should be replaced with a mail
protocol that authenticates senders.

The stream of non-solutions to spam not only ignores the existing
literature, but also ignores itself.  If the people pumping out the
stuff would pause and look just little at other proposals, they'd
notice their wonderful "new" idea flying by.  That's the crux of the
anti-spam excitement.  We're watching a dot-com boomlet of hopes and
dreams to get rich or at least famous by Finally Stopping All Spam.
As with b-to-b, b-to-c, disintermediation, CMR, multi-media, and the
other SuperHypeWay crazes, people are rushing to market regardless of
history, archives, or any other consideration.


Vernon Schryver    vjs(_at_)rhyolite(_dot_)com