RE: WG Consensus Call: draft-swallow-gmpls-overlay-00.txt

[Kireeti Kompella <kireeti@juniper.net>]

Hi Monica,

On Wed, 11 Dec 2002, Lazer, Monica A, ALASO wrote:

> Generally speaking, I do not think it is a good idea to proliferate
> multiple options supporting the same thing. It will make
> interoperability efforts quite painful. So, regarding you points, I
> would like to bring the following.
> OIF produces interoperability agreements,

Implementation agreements, to be pedantic.  And these are certainly
useful, but they are not standards.

In fact, if you look at the OIF FAQ, you'll see:

| Question: How will the work of the OIF differ from that of the other
| various standards bodies?

| Answer: The OIF will concentrate on interlayer issues that specifically
| pertain to data networking elements natively accessing the optical
| layer.  Most standardization efforts are defining characteristics within
| the optical network rather than addressing how it interacts with data
| networks. Because of this, the OIF expects to be completely
| complementary to work in the standards bodies.

In the case of GMPLS Overlay, a standardization body (namely, the
IETF) is addressing the interaction of the optical network with data
networks.  Does this mean that the OIF will recognize this and stop
any further work on the OIF UNI?  This is why I suggested that Khuzema
bring up this issue with the OIF.

> it is a forum recognized by
> ITU-T,

Can ITU-T recommendations make normative references to OIF implementation
agreements?  In what sense does the ITU-T recognize the OIF: is there
any recognition beyond the fact that liaison statements are exchanged?

Since this issue has come up several times, it would be useful if
someone speaking on behalf of the ITU-T would comment on the nature
of the relationship between the ITU-T and the OIF, the fact that the OIF
is not a standards forum, and that their documents are implementation
agreements.

> and until recently also by IETF.

I wasn't aware of this.  Can you point me to any document or public
statement that substantiates this?  That would be very helpful.  As
far as I know, there is no formal liaison (see
http://www.ietf.org/ietf/1ietf-liaisons.txt)

I would really like closure on this.  I see a *lot* of misleading
statements (e.g., supposed formal relationships between the OIF and the
IETF, that the OIF develops standards, etc.) used to support arguments,
and this is doing a real disservice to the community.

> ITU-T produces recommendations. So would one than state that they are
> not standards?  IETF produces RFC (request for comments?) Then these are
> also not standards?

This is what the ITU says about its Recommendations:

| ITU-T Recommendations
|
| ITU-T produces high-quality standards (Recommendations) on technical,
| operating and tariff questions.

As for RFCs, read 2026, section 4.1.1.

Both bodies recognize this and can make normative cross references.

As for the OIF, read their mission statement, or the FAQ above.

Kireeti.

PS: you should know from the tutorial I gave at the T1X1.5 meeting that
    RFCs are not really Requests for Comments.  Comments are welcome
    right up to Last Call, but once an RFC is published, it is immutable.