🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: sci.crypt
Article #810

Re: User authentication

#810
From: D. J. Bernstein
Date: Sun, 30 Aug 1992 04:52
46 lines
2734 bytes
In article <1992Aug28.140033.466@sniap.mchp.sni.de> frank@D012S436.sniap.mchp.sni.de () writes:
> So if this is interesting, how come more people don't post anonymously?
> Why is your message signed Dan Bernstein, not Guess Who?

Where do you get the idea that people don't often post anonymously?

I have forged all but a small fraction of the messages I have ever
posted. Many of them appear under the name ``Dan Bernstein''---which,
you should note, is shared by at least three different real people who
post to USENET. Many of them do not, and if I don't choose to point out
that those other names are controlled by the same real person, then you
will never know. Who cares?

Public-key cryptography gives you the opportunity to maintain a key
which everyone knows but only you (and whoever you share needles with,
so to speak) can use. This gives a *continuity* to keys which neither
requires nor benefits from an association between keys and ``real
people.''

> I assume that you would be peeved
> if 10,000 people started placing "secure" posts associated
> with the name Dan Bernstein, for example.  

If public-key cryptography weren't kept out of the public's hands by a
small group of people, I would be able to choose my own keys and have
them be recognized as *mine*. I'm not peeved about the existence of
multiple people with the same name---that's just a fact of life.

(In the July 1992 Bulletin of the AMS, Shreeram Abhyankar says that
Peter Neumann said that William Snow Burnside, who worked on the theory
of equations, is not the same as William Burnside, who worked on the
theory of groups of finite order. Both obtained a D.Sc. from Dublin
around 1890.)

> But I also see applications for secure newsgroups based on
> real world identities.

I do not. In fact, unless you have an application which depends on the
difference between one real person with two names and two people who
happen to cooperate (overtly or secretly), an obvious isomorphism
argument shows that your application *cannot* depend on ``real world
identities.'' Except for certain government functions (such as voting) I
see no such applications; and those functions won't be carried out on
newsgroups any time this millenium.

---Dan
Technical Details
Message-ID: <18293.Aug3004.52.1592@virtualnews.nyu.edu>
Path: arc1-ca.pugleaf.net!sparky!uunet!pmafire!news.dell.com!swrinde!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!ucbvax!virtualnews.nyu.edu!brnstnd
References: <19253.Aug2700.31.1692@virtualnews.nyu.edu> <1992Aug28.140033.466@sniap.mchp.sni.de>

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!pmafire!news.dell.com!swrinde!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!ucbvax!virtualnews.nyu.edu!brnstnd
From: brnstnd@nyu.edu (D. J. Bernstein)
Newsgroups: sci.crypt
Subject: Re: User authentication
Message-ID: <18293.Aug3004.52.1592@virtualnews.nyu.edu>
Date: 30 Aug 92 04:52:15 GMT
References: <19253.Aug2700.31.1692@virtualnews.nyu.edu> <1992Aug28.140033.466@sniap.mchp.sni.de>
Organization: IR
Lines: 46