Crypto elsewhere

[Richard Clayton]

In article <Qec1fGAieRE1Ewfa@prs5rr8t.prestel.co.uk>, T Bruce Tober
<octobersdad@reporters.net> writes

>The following is copied from the SWATCH list. I've only just got back
>online and haven't had a chance to check it out yet myself.
>
>===================FORWARDED MESSAGE==========================
>Emitting from Loughborough this time. Be afraid. Be very afraid. ;)
>
>ftp://ftp.ietf.org/internet-drafts/draft-hamilton-fix-dns-00.txt

I think you will find that this is really just an "in-joke" which
probably explains why the description of the system is so haphazard and
so limited...

It arose, as I heard it, from a late night discussion over some beers
about a way to run DNS by consensus rather than using a central
authority. The reason for the discussion was the Magaziner proposal
(comment period nearly complete!). The discussion, as I recall, was
reported to the IETF mailing list in general detail, this seems to be
just a document to write it up for posterity.

The model Hamilton & Knight choose is Netnews where in theory it is all
chaotic but in practice nothing happens in the Big8 without a PGP signed
message from tale and nothing happens in uk.* without a PGP signed
message from control@usenet.org.uk etc etc

Basically, the draft proposes a scheme whereby top level DNS domains
would be promulgated by signed messages - and it would be down to
individual DNS server managers to decide what to honour...

That's pretty much what we have at the moment - most people honour the
domains in the a. b. etc root name servers and ignore the AlterNIC. All
this does is codify the way it really works (people trust people) but
they add the potential to increase chaos in the future. The example they
gave originally was what would happen if a government decided they owned
part of the namespace  *.uk  for example, but the rest of the Internet
disagreed with them...  The risks of fragmentation of name space are,
however, far too horrible to contemplate seriously.

So, bottom line: it's an amusing idea, but I don't view it as a serious
suggestion for solving any real problems. In particular I don't see the
people who would currently be trusted to sign messages currently being
interested in an ordered anarchy.

Mind you, what do I know ?

>(I suspect that it's shooting itself in the foot early in paragraph
> 3 which suggests that escrowed private keys would be about as 
> much use as... well, read it... )

read it by all means, but I don't think you need to give it much mind.

USG green papers on DNS are doubtless of interest to many on this list,
but not the topic of this list, so I shall say no more

-- 
richard                      richard.clayton    @    T U R N P I K E .com
 http://www.demon.net/news/features/crypto/  for Demon's views on crypto
"Assembly of Japanese bicycle require great peace of mind" quoted in ZAMM