Date: 23 Jan 2001 19:29:23 -0000
From: "D. J. Bernstein" <[EMAIL PROTECTED]>
Message-ID: <[EMAIL PROTECTED]>
| Renumbering is an example of the ``every problem can be solved by
| another level of indirection'' principle. I've been asking why this
| indirection should be in the _client_ rather than the _server_.
I am aware of that.
| So far, all I've heard is that (1) this might save more queries than it
| wastes, even though the analogous assertion for MX records is not true,
| and (2) this would occasionally save some DNSSEC CPU time.
Whether it ends up being true or not depends upon how one configures
the various records.
And it isn't DNSSEC CPU time that is the issue, but the zone key.
| I'm sorry, but I don't see server-side indirection discussed anywhere in
| the IPNG archives. What exactly were the dates of the discussion? What
| benefits were attributed to client-side indirection?
As I said in the message to Jim, I don't even recall which WG it was
discussed in. I'm away from home at the minute, and don't have my
mailing list archives available - and where I am at the minute general
net access is bad enough that the web rarely works (there's an enforced
proxy cache which mostly times out before it manages to retrieve web
pages...)
It is possible it was discussed at a face to face WG meeting, rather than
on the list (a list), though I really don't recall (but it was back in the
middling days of when A6 was just an internet-draft).
kre
--------------------------------------------------------------------
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]
--------------------------------------------------------------------
