[co-Author of the DNS discovery draft hat on]
I agree with your assessment that this proposal has missed its window of opportunity. We could probably do a very interesting post-mortem analysis why in the context of the problem-statement list, but this is not the topic for today's discussion.
Now that DHCPv6 is eventually going to be published and implemented,
the pressure for an alternate solution specifically for DNS is much lower
and I would agree we should now step back and look
at the more general problem from a fresh start.
There are actually two different topics that could be discussed in a bof:
- is DHCP enough for general parameter configuration or is
there a significant enough problem space where its not
applicable/desirable? For example, for address configuration,
we have DHCP and stateless autoconf. Does this dichotomy
between server-helped and self-derived configuration applies to
other services?- What does it mean to use 'well known voluntary ambiguous' addresses
for service configuration?
This is a discussion I had with Steve Deering before he went
on sabbatical. This is the equivalent of the '911' service:
wherever you are, you dial 911 and you are connected to the emergency
service. Can we use this paradigm for services that are
supposed to be somehow 'universal'?
For example, you want a NTP server, try '777::1', you want a DNS server,
try '411::1',... Charge to the network to make sure the 'call' connects
to the right place. The devil is, of course, in this last sentence.
- Alain.
-------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
