Re: [DNSOP] [[e-mail]: I-D Action: draft-grothoff-iesg-special-use-p2p-names-00.txt]

[Stephane Bortzmeyer <[e-mail]>]

On Sun, Dec 01, 2013 at 11:56:43AM -0500,
 Ted Lemon <[e-mail]> wrote 
 a message of 7 lines which said:

> Why do you support it?   Why is registering these names a good idea?

Good question but you'll have to be patient, the answer will take
time.

A note about IETF procedures, first: RFC 6761 does not say the use of
the "special" domain must be a "good idea". Just that it has to be
documented in a RFC on the standards track *or* approved by IESG.

There are many people who believe that the DNS has some weaknesses:
risks for privacy (see draft-bortzmeyer-dnsop-dns-privacy), and political
risks around the management of the root (the DNS being organized in a
tree, a specific node, the root, has too much power). Some of the
solutions proposed do not change the behavior of the programs and are
therefore out-of-scope for RFC 6761. For instance, the alternative
roots are just a political choice (see RFC 2826) and have zero
technical consequence.

Other solutions have been proposed that are different
technically. Sometimes, they are completely separated from the DNS
(and are out-of-scope for dnsop) and sometimes there is some gateway
with the DNS. My personal opinion is that most of these solutions are
just vague "back of the enveloppe" ideas and will not see any serious
deployment.

On the other hand, there are other proects that are technically
interesting and/or widely deployed. For the cases mentioned in 
draft-grothoff-iesg-special-use-p2p-names:

.onion is used a lot by Tor users, to access "hidden services"
<http://en.wikipedia.org/wiki/Tor_(anonymity_network)#Hidden_services>
It requires special handling (send the name to Tor and not to regular
DNS) and therefore deserves a registration as special.

.gnu and .zkey are used in Gnunet
<http://en.wikipedia.org/wiki/GNUnet>, the first one for
human-readable names and the second-one for cryptographical
names. GNUnet is probably much less used than Tor but is one of the
few serious (proper documentation and "running code") and
comprehensive attempts to reach the objectives of the Vancouver
meeting ("hardening the Internet").

[I let .i2p, I have no strong opinion.]