RE: [Nea] use of a design team to develop requirements

["Stephen Hanna" <shanna at juniper.net>]

A better time for this discussion would have been at the NEA WG
meeting or during the month since then. I sent email to the NEA
list on November 8 explaining that a requirements design team
was being formed and an email to the SAAG list on November 9
reiterating this. Several more follow-up emails to the NEA list
were sent over the next few weeks. The design team is well under
way now. But I suppose it's never too late for an objection.

Certainly, the intent of using a design team is not to exclude
mailing list discussions and participation. It's to prepare a
reasonable starting proposal for discussions. This is not an
unusual process in the IETF. Working groups often create a
design team to solve a problem or create a first draft of
a specification. In fact, many specs are initially drafted
by one person and then revised in response to discussions
on the mailing list. As long as the initial draft from the
design team is regarded as only an initial proposal subject
to revision or rejection by the working group, this is a
perfectly acceptable process. In fact, RFC 2418 describes
this process explicitly and positively in section 6.5.

We have asked the design team to deliver the first draft
of the requirements document in early January. That's a
very short timeline and there are a lot of holidays there.
If we abandon the design team process and instead hold an
unstructured discussion of requirements on the mailing list
during this holiday period, I'm concerned that it will not
be a productive discussion. I'm especially concerned that
many working group participants will be absent during this
period and one or two people will end up dominating the
discussion. I have already seen tendencies in this direction
in recent NEA WG discussions.

I suggest that we allow the design team to finish their work
on the first draft of the requirements document and deliver
it to us in early January. Then we can have a free-ranging
discussion about the document, any assumptions implicit in
the document, what should be done instead, etc. We'll have
all hands on deck during January and can have a much more
productive and inclusive discussion then.

Thanks,

Steve

-----Original Message-----
From: Khaja Ahmed [mailto:khaja at windows.microsoft.com] 
Sent: Friday, December 08, 2006 8:49 PM
To: Keith Moore; Stephen Hanna; Susan Thompson
Cc: Sam Hartman; nea at ietf.org
Subject: RE: [Nea] use of a design team to develop requirements

I agree with Keith.  It is probably fine to have a 'design team' in the
sense of a designated set of people who have committed to contribute to
the development of requirements.  However I would prefer to see the
actual work and related discussion conducted on the list.

Khaja

-----Original Message-----
From: Keith Moore [mailto:moore at cs.utk.edu] 
Sent: Friday, December 08, 2006 2:57 PM
To: Stephen Hanna; Susan Thompson
Cc: Sam Hartman; nea at ietf.org
Subject: [Nea] use of a design team to develop requirements

Stephen & Susan,

I really have to question the use of a design team to develop 
requirements for the NEA WG.  A lot of the strength of the IETF process 
has traditionally been from its inclusiveness.  Experience suggests that

this is most beneficial in two areas - one is determining requirements 
and design considerations (the more people who have input into these, 
the more chances there are to spot important considerations that might 
otherwise have been overlooked) and another is in reviewing documents 
(the more reviewers, the greater the chance that a flaw in the document 
is discovered).

It's generally a good thing to have a small team to be selected as 
editors of a document, but IMHO, not so good to make that team 
responsible for actually determining requirements.  If that is really 
the intent, I think this will tend to impair the ability of the WG to 
produce a sound set of requirements, particularly when there are a 
number of dubious assumptions already included in the WG's charter.

Particularly given that the list has essentially been silent since the 
design team was announced, and also given that the design team was 
formed before much discussion had taken place here, it's hard to escape 
the impression that the parameters of the solution are being dictated 
rather than determined by WG consensus, and before the problem is even 
well understood.  I do not believe it is sufficient to have a design 
team develop requirements in isolation and then to have the WG ratify
them.

Keith


_______________________________________________
Nea mailing list
Nea at ietf.org
https://www1.ietf.org/mailman/listinfo/nea

_______________________________________________
Nea mailing list
Nea at ietf.org
https://www1.ietf.org/mailman/listinfo/nea