WG starting point - WG query

[Dave Crocker <dcrocker@brandenburg.com>]

At 10:40 AM 2/4/2001 -0500, Hollenbeck, Scott wrote:
>Let me be clear about what I mean: the requirements draft has been out for
>broad IETF review for 10+ months.  If people are only now taking notice
>because we are forming a WG, it would be more productive to stop complaining
>and get comments in _now_ than to insist that there hasn't been ample time
>for review.

And let ME underscore that the Requirements document very much HAS been 
reviewed by others.

         When it came out, did many people assume that it was biased 
towards NSI?  Yes.

         When they read it did they change their minds?  Yes.

         Did the same thing happen with the protocol specification?  Yes.

         Does this mean that starting with those documents means that the 
working group would have no work to do?  Not at all.

The philosophical difference that is creating the current set of tensions 
is whether this IETF working group should

         a) move very aggressively, and

         b) whether it can take advantage of pre-existing work by Scott to 
finish much more quickly would otherwise be possible.

         -----

So I am personally requesting feedback from the others on this list, 
concerning consensus about use of Scott's existing work:

         My sense is that the Requirements document has already received 
extensive community review and has general support.

         My sense is that Scott's protocol specification has already 
received extensive community review and has general support.

         The terms of a working group charter need to have consensus of the 
proto-working group.  In other words, it is not just produced by the chair 
and the IESG.  It has to have support by those who will be participating in 
the working group.

         Taking existing documents as a starting point for a working group 
is an entirely acceptable practise in the IETF... IF there is consensus to 
do so.  The benefit is making progress MUCH more quickly.

         This does not prevent making changes.  When taking existing 
documents, the point of the working group is to REFINE things.  Hence, the 
working group takes over the document(s) and may make whatever changes it 
deems appropriate.

         Language in the charter often specifically cites such documents 
and limits the nature of the changes, in order to ensure better WG 
focus.  Hence, the charter language is intended to create a bias against 
big changes.  However let me repeat that strong working group consensus 
retains authority to make whatever changes the working group deems appropriate.

So my question to the working group:

         1.  Shall the working group take Scott's Requirements document and 
Scott's protocol specification is direct input, and then seek to refine them?

Or:

         2.  Shall the working group start from scratch and seek full 
development of Requirements and full development of a protocol 
specification, including possible choice among competing specifications?

In my opinion, alternative #2 usually costs a working group no less than 
6-12 months EXTRA.  Taking as much as 2 years EXTRA is not unusual!

d/


----------
Dave Crocker   <mailto:dcrocker@brandenburg.com>
Brandenburg Consulting   <http://www.brandenburg.com>
tel: +1.408.246.8253;   fax: +1.408.273.6464