Re: [Internetgovtech] Guiding the Evolution of the IANA Protocol Parameter Registries

[S Moonesamy <[e-mail]>]

Hi Russ,

First of all, I would like to thank the IAB for 
requesting feedback before taking a position.

At 09:28 24-02-2014, IAB Chair wrote:
>Existing IETF and IAB consensus concerning Internet registry functions
>and IANA are documented in a variety of RFCs and IAB communications
>[RFC2860,RFC6220,IAB1,IAB2].  Since registry functions and IANA are
>likely to be the subject of discussion in a number of venues outside the
>IETF over the coming months and years, the IAB is seeking community
>feedback about operating principles to use when they find themselves
>involved in those discussions.

I suggest adding RFC 3172 to the above list of RFCs.

>= = = = = = = =
>
>Principles Guiding the Evolution of the IANA Protocol Parameter Registries
>
>1. The IETF protocol parameters function has been and continues to be
>capably provided by the Internet community.
>
>The strength and stability of the function and its foundation within the
>Internet community are both important given how critical protocol
>parameters are to the proper functioning of IETF protocols.
>
>We think the structures that sustain the protocol parameters function
>needs be strong enough that they can be offered independently by the
>Internet community, without the need for backing from external parties.
>And we believe we largely are there already, although the system can be
>strengthened further, and continuous improvements are being made.
>
>2. The administration of the protocol parameters function by ICANN is
>working well for the Internet and the IETF.
>
>We are pleased with the publication and maintenance of the protocol
>parameter function and the coordination of the evaluation of
>registration requests through the IANA function provided by ICANN.

I suggest not mentioning Point 2.

>3. The IETF protocol parameters function requires openness,
>transparency, and accountability.
>
>Existing documentation of how the function is administered and overseen
>is good [RFC2860,RFC6220], but further articulation and clarity may be
>beneficial. It is important that the whole Internet community can
>understand how the function works, and that the processes for
>registering parameters and holding those who oversee the protocol
>parameter function accountable for following those processes are
>understood by all interested parties. We are committed to making
>improvements here if necessary.
>
>4. Any contemplated changes to the protocol parameters function should
>use the current RFCs and model as the starting point.
>
>The protocol parameters function is working well, and as a result
>wholesale changes to the role of the IETF vis a vis the function are not
>warranted. The IETF/IANA Memorandum of Understanding [RFC2860] is a good
>model to work from in the event that other parties do want to contemplate
>changes. Put quite simply: evolution, not revolution.

If I am not mistaken the stated position has been 
that the protocol parameters registry is an IETF 
matter.  I would suggest caution before even 
considering any contemplated 
changes.    Evolution tends to have unintended consequences. :-)

>5. The Internet architecture requires and receives capable service by
>Internet registries.
>
>The stability of the Internet depends on capable provision of not just
>IETF protocol parameters, but IP numbers, domain names, and other
>registries. Furthermore, DNS and IPv4/IPv6 are IETF-defined protocols.
>Thus we expect the role of the IETF in standards development, architectural
>guidance, and allocation of certain name/number parameters to continue.
>IP multicast addresses and special-use DNS names are two examples where
>close coordination is needed.  The IETF will continue to coordinate with
>ICANN, the RIRs, and other parties that are mutually invested in the
>continued smooth operation of the Internet registries. We fully
>understand the need to work together.
>
>6.  The IETF will continue its direction and stewardship of the protocol
>parameters function as an integral component of the IETF standards
>process and the use of resulting protocols.

The word "stewardship" has been overused for a while.  According to RFC 6220:

    "the IETF asserts authority and responsibility for
    the management of all of its protocol parameters and their
    registries, even while it generally remains isolated from the
    selection of particular values once a registration is approved."

Point 6 might sent the wrong message.

>RFC 6220 specifies the role and function of the protocol parameters
>registry, which is critical to IETF standards processes and IETF
>protocols.  We see no need to revisit or reconsider our current approach
>with regard to protocol parameters, including the ability to delegate
>operational responsibility for registries to other organizations.

The responses to the NOIs mention "IANA Functions".  From the ICANN bylaws:

   "The mission of The Internet Corporation for Assigned Names and Numbers
    ("ICANN") is to coordinate, at the overall level, the global Internet's
    systems of unique identifiers, and in particular"

   "1. Coordinates the allocation and assignment of the three sets of unique
       identifiers   for the Internet, which are"

      "c. Protocol port and parameter numbers."

The above is inconsistent with what is written in 
RFC 6220.  According to an (old) contract ( 
http://www.icann.org/en/about/agreements/iana/iana-contract-17mar03-en.htm ):

   "C.2.1.1 DoC NTIA has a requirement for a Contractor to maintain the
    operation of the Internet by performing the IANA functions. In performance
    of this purchase order, the Contractor shall 
furnish the necessary personnel,
    material, equipment, services, and facilities 
(except as otherwise specified),
    to perform the following IANA requirements."

      "C.2.1.1.1 Coordinate the assignment of technical protocol parameters
       – This function involves the review and assignment of unique values to
       various parameters (e.g., operation codes, port numbers, object
       identifiers, protocol numbers) used in various Internet protocols.
       This function also includes the 
dissemination of the listings of assigned
       parameters through various means (including on-line publication) and the
       review of technical documents for consistency with assigned values."

It wopuld be good if the IETF considers the above 
as it is inconsistent which its stated position.

Regards,
S. Moonesamy