Re: (ngtrans) Re: IPv6 dns

[Robert Elz <kre@munnari.OZ.AU>]

    Date:        Fri, 19 Jan 2001 11:35:14 -0500 (EST)
    From:        Jim Bound <seamus@bit-net.com>
    Message-ID:  <Pine.OSF.3.95.1010119112928.7043G-100000@www.bit-net.com>

  | A6 is PS not DS.  If it will not work as is we have every right to oppose
  | it.

All of IPv6 is DS (or less), not full std, so the same could be said.

  | But if we
  | reduce the levels of hierarchy of the A6 DNAMES to max 3 it may work.

Yes - but the problem is that it is very hard to enforce.  Just like the
NS may not name an alias, or MX may not, or aliases may not refer to
other aliases ... they are all supposedly DNS rules, but widely ignored.

  | But if anyone does 10 it will not perform for sometime.

The good thing about this is that people who do that are making it
difficult for themselves - just as people who set up stupid NS
architectures.   Sure it makes extra work for everyone else in the
attempts to resolve the names, but ultimately it is the site with
the bogus setup that suffers most.

The DNS could really do with a good recipe book, explaining to people
what works and what doesn't (not the theory, packets, ... just the how
to stuff).

  | I also think the
  | processing of A6 should not be done on clients but on the servers.

That argument was made (by you, and others) back when A6 was being
debated.   If you care to look back over the archives (IPNG, and
perhaps DNSIND) you will see why things are the way they are.
I'm sure you will recall that it wasn't just that no-one thought
of the possibility...

  | I can't see my IPv6 PDA or my international gaming device having to cache
  | 10 levels and retransmit to get my DNS records.  It just will not work.

Then have your implementation simply give up before going so far.
That's not going to bother anyone with a sensible setup.

kre