IAB issues

[RJ Atkinson]

On Wednesday, Nov 6, 2002, at 14:42 America/Montreal, avri doria wrote:
> We have spent a lot of time in this discussion talking
> about the need for an architecture.
> Or perhaps for a framework.  At any rate something that
> guides the IETF clearly.  If we do need something like
> that, who should be responsible?  Nominally at least,
> I thought that is was the job of the Internet Architecture
> Board.

IAB does (and has) offered advice, on its own initiative
and when asked (which is rare and most commonly from IESG).

Such advice does not constitute an edict that IETF is
required to obey, of course.

> I have read many of the IAB RFCs.  And some have been
> very helpful.  But they always appeared to me to be
> individual contributions of IAB members more then
> the view of the IAB.

If published as an IAB document, then it reflects smooth
consensus of the IAB as a whole.  Somehow we must not be
clear enough about that.

> As for the workshops they are
> interesting to hear the reports on, but I am not sure
> what their real affect has been.

Example:		IAB Character Set Workshop

	Conclusions from the workshop tended to drive the
actual IETF policy on character sets that later was published
as a standards-track document.  However, the IETF as a whole
approved that policy document and the IAB Workshop output
was just advice.

> I have also had conversations with IAB members who decry
> the shape of the IETF and the cluelessness of the
> participants.  And yet, in my opinion, if anyone is
> responsible for leading the IETF membership out of our
> cluelessness it should the IAB.

What more would you suggest that we do ?

> I guess I am looking for the IAB to have a real job
> to do in relation to the architecture.  I may be alone
> in that.  If that is the case, we should rename the body
> and stop pretending that they have a real role in
> architectural  oversight.  If the role is not being
> filled, then we should be clear about that.
>
> I am not looking for the IAB to dictate from
> on high.  What I am looking for them to do is understand
> the architectural necessities of the engineering task that
> needs to be done, and to explain those in such a way as to
> focus the work of the organization on the areas that need
> the most help.

If you have specific ideas, we are all ears.

> I am not advocating a return to the days when the IAB ruled.
> But I do think the pendulum did swing
> a bit too far in the other direction. And perhaps it is
> an educational and guiding role instead of a controlling
> role that I am looking for.  But if we are going to fix
> the problems we see, e.g. the lack of architectural coherence,
> and we are going to do so without the creation of new layers
> of management, then we have to use the structures and the
> talents we have to their best advantage.

I'm not arguing necessarily with anything you've said.
That noted, there would need to be clear community consensus
before IAB's role could change drastically relative to the IETF
processes.  I'm not sure quite what specific changes you
think would be appropriate, but I'm eager to listen.

One approach to such a change might be for an interested party
to write up an individual submission I-D with a proposal
and let that undergo public discussion to see where consensus
lies relative to that specific proposal.

Ran
rja@extremenetworks.com