From: jamesd@netcom.com (James A. Donald)
Date: Sun, 4 Sep 94 11:24:38 PDT
To: hfinney@shell.portal.com (Hal)
Subject: Re: Problems with anonymous escrow 2--response
In-Reply-To: <199409040350.UAA07837@jobe.shell.portal.com>
Message-ID: <199409041824.LAA14498@netcom7.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


Hal writes
> > > What is this stuff, reputation capital?  What does it look
> > > like?  How can it be measured?  How much is it really
> > > worth?
 
jamesd@netcom.com (James A. Donald) writes:
> > Obviously none of these questions are answerable:  So what?
 
Hal writes
> No, my questions were not rhetorical at all.  I do think
> that various people have come up with ideas for what they
> call reputation capital that are much more formalized and
> structured than what you are referring to. This doesn't
> mean that they are right and you wrong, just that there are
> a lot of different concepts floating around under this
> umbrella of a term.
 
No no:
 
What is floating around are proposals for the structured
handling of the *information* on which reputations are
based, not proposals for the structured handling of
reputations.
 
Digital credentials, not digital reputations.
 
Structured handling of *reputations* would be catastrophic.
 
If you *define* reputations to be something formal and
explicit, and say that the system will work because people
guard their reputations, then reputations become something
that can be most efficiently granted and withdrawn by some
centralized authority.
 
And then, as with fiat currency, the value of those
"reputations" would in the end need to be backed by force
in order to make the system work.
 
I am complaining about dangerous carelessness in your use
and definition of words.  Your use of the word reputation
is as fraught with frightful consequences as Marx's use
of the word "value".
 
Reputation based systems work for freedom, and coercion
based systems work for centralized government, for obvious
reasons that all of us agree upon.
 
*Define* reputations to be something other than
reputations, and you are kicking the crucial foundation out
from under freedom.
 
You are defining the foundation of freedom away, in a way
precisely analogous to the way Marx defined the basis of
capitalism away, though he did it maliciously and
knowingly, and you are doing it accidentally.
 
> As one example, consider how signed endorsements could be
> used to create and validate a reputation.
 
True.  But signed endorsements are *not* a reputation.
 
> > > I think this concept needs to be clarified and examined if
> > > it is to serve as one of the principle foundations of
> > > pseudonymous commerce.
 
> > No it should not be "clarified and examined".
> > ...
> > We already know what reputations are.  "Defining" them is
> > going to make them into meaningless nominalist hot air.
 
> On the contrary, I think that a pseudonymous/anonymous
> world calls for a re-examination of the concept of
> reputations.  Today there is no implementation of a
> transferrable credential, where I could for example prove
> that company XYZ considers me a good credit risk, without
> XYZ linking my present nom de guerre with the one I used
> when with them.
 
This is an illustration of the danger of redefining
"reputation" as you appear to be doing.
 
Obviously a blind signed credential transferable between
digital pseudonyms would have no value to support a
reputation, but by abandoning the correct usage of the
word "reputation" you have obscured that fact from
yourself.
 
But what would have value was a credential whereby some
authority signed *one* private key that you possessed at a
certain time, without knowing either the private key or the
public keys associated with that private key.
 
But if you used that key to support multiple identities,
you would then be stuffed because it would then create a link
between Joe Robertson, software benchmarker, and Mike
Hardcase, purveyor of underage Ceylonese virgins.  Worse,
it might create a link between Mike Hardcase and Joe
Whatsyourpleasure, purveyor of Filipina whores, thereby
substantially reducing the value of the Ceylonese virgins
purveyed by Mike Hardcase, even though both Mike Hardcase
and Joe Whatsyourpleasure both had excellent reputations
until their reputations became linked.
 
> > If your on line personality is selling something, it would
> > seem highly desirable to have dossiers built up about it.
 
> Do we just say that "of course" dossiers of people's credit
> history and banks' lending history are the desirable and
> correct way to solve this problem, as we have today?  I
> would prefer to see whether solutions could be derived in
> which more privacy is provided to the participants. 
> Obviously total anonymity would make such lending virtually
> impossible, but perhaps there is some middle ground between
> that and a system of total identification. This is where
> Chaum is coming from with his credentials.
 
Exactly so:  And Chaum talked of digital credentials, and
the reputations of digital credentials.  He *did not* talk
about digital reputations.
 
Use the word *credentials*, not the word *reputations*. 
If we were to start using the word *reputations* in the
way that you have been using it, we will make errors with
vastly more serious consequences that the errors that you
have made.
 
-- 
 ---------------------------------------------------------------------
We have the right to defend ourselves and our
property, because of the kind of animals that we              James A. Donald
are.  True law derives from this right, not from
the arbitrary power of the omnipotent state.                jamesd@netcom.com