Re: "IETF consensus" in IANA considerations [was Re: Last Call: CR-LDP Extensions for ASON to Informational ]

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

    Date:        Thu, 30 Jan 2003 15:45:13 -0500
    From:        Thomas Narten <narten@us.ibm.com>
    Message-ID:  <200301302045.h0UKjDF05128@cichlid.adsl.duke.edu>

I had been avoiding reading this set of messages, because I couldn't
really see discussions of what was required to get IANA to assign a
number in some (irrelevant to me) code space could possibly be interesting.

But ...

  | In practice, "IETF Consensus" (as defined in 2434) means that
  | procedurally the IESG has to sign off on a document before IANA
  | assigns code points for it.

This is certainly true.   Unfortunately, it seems that the IESG sometimes
believes that this is all that is required (for IETF Consensus as a general
term).

That's nonsense.   Nor is "publication as a standards track RFC" or
anything similar the definition of what is "IETF Consensus" - that would
be absurd, as IETF Consensus is required for that to happen - that is,
publication as a standards track RFC is general evidence of IETF
consensus, but is not a necessary product of IETF consensus.

There can be IETF Consensus not to publish a document, just as there can
be to publish it.

IETF Consensus is the IETF's way of agreeing to some action.   The IETF
makes that decision, the IESG (currently anyway) is tasked with determining
whether or not the IETF has actually made a decision.   IETF Consensus
clearly requires a last call be made.   It does not (of itself) require
publication of anything at all - publication may be made to record the
consensus, it does not cause it, and is not required to obtain it.

kre