Abstract of proposed Internet Draft for Best Current Practic e (please comment)

David Barak thegameiam at yahoo.com
Mon Mar 10 15:52:46 UTC 2003



--- "Dr. Jeffrey Race" <jrace at attglobal.net> wrote:
> Please read the details in the text.  It is all
> spelt out there.

Well, when I read the abstract, I also got the
impression that the approach recommended was to
"filter first, ask questions later."

There were a few flaws in the assumptions:
1) Profitable, well-run companies always use BCPs.
   Simply put, unless you define "profitable and
well-run" to mean "uses BCPs" what you will find is
that some networks which have an admirable adherence
to best practices afte often not profitable, while
many networks which do not are quite profitable.

2) A consensus-based emergent governance will solve
the problem.  
   
   This is a fallacy, as Douglas Hofstadter
demonstrated in Metamagical Themas: he ran an iterated
prisoner's dilemma-type game among logicians (!) to
see whether they would all demonstrate cooperative
behavior, and to his surprise (but not to anyone
else's), they did not.  In fact, if the ISP community
were to sign on to such a system in the same ratio as
the logicians did in Hofstadter's experiment, we would
have the exact same solution which we have now: which
is a multitude of small-scale approximate solutions to
the same problem.
  The second piece of the fallacy is the idea that
emergent systems can be predicted: there is a reason
they're called "Emergent" systems - they arise from
the discrete actions of the constituent members rather
than from design.  Any attempt to build a system to
have a certain outcome is by definition not an
emergent system.

3) The abstract does come across as overzealous - if
you are proposing a "draft," you should expect to
receive positive and negative comments.  In fact, you
should probably modify the draft to answer the
negative comments received.

4)The area of the "spamverse" would steadily shrink in
importance and relevance.

   Even assuming that everything else were as you
describe, and that the BCP ISPs would be willing to
suffer the customer pain of not being able to
communicate with the non-BCP ISPs, what would keep the
spammers from finding another technological solution? 
Why would you think that they would not employ their
(considerable) technical skill to getting around any
mechanism we employ?  
  The Cantor Diagonal method could apply here: it is
impossible to predict every possible situation.  The
spammers are highly motivated (big $$), and the simple
fact which most people ignore is this:

  Spam works

If Spam didn't generate money, people wouldn't do it. 
I think that any solution will have to involve getting
people to not respond to spam, and steadily raising
the cost of sending mass email, so that there will be
a financial disincentive to do so.

Just my $.02.


=====
David Barak
-fully RFC 1925 compliant-

__________________________________________________
Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more
http://taxes.yahoo.com/



More information about the NANOG mailing list