From: Wei Dai <weidai@eskimo.com>
Date: Thu, 22 Feb 96 00:39:41 PST
To: Robert Hettinga <rah@shipwright.com>
Subject: Re: "and two forms of ID"
In-Reply-To: <v02120d04ad51b2b33a96@[199.0.65.105]>
Message-ID: <Pine.SUN.3.91.960222002325.26468C-100000@eskimo.com>
MIME-Version: 1.0
Content-Type: text/plain


On Thu, 22 Feb 1996, Robert Hettinga wrote:

> >I'm sufficiently impressed with the arguments against name credentials
> >that Carl Ellison has made that I'm looking seriously into systems
> >that don't do any sort of conventional certificate binding at all...
> 
> ... and I bet, Wei Dai's contentions to the contrary, that they'll be
> *cheaper* to use than those which do certificate binding, all other things
> being equal.

You got my position completely backward on this.  I've always supported 
Carl's arguments in the past on this issue (for example see the tread 
"subjective names...").  You may be thinking of what I said about the 
cost of defeating traffic analysis.

The natural state of the Net seems to be a kind of semi-anonymity.  
Trying to push it in either direction (complete traceability or 
anonymity) is costly.

Wei Dai