Re: [wg-d] WG Principles

[Kent Crispin <kent@songbird.com>]

On Mon, Aug 02, 1999 at 04:43:54PM -0400, Bret Fausett wrote:
> One of the mandates of this Working Group is to "Prepare Working Group 
> and Public Consultation Procedures."  To that end and to get the 
> conversation started, I thought I would circulate the following list of 
> principles that I believe the DNSO's internal operating procedures on 
> Working Groups should follow. 
> 
> I'm sure this is not complete, but I thought if we could agree on some 
> fundamental principles, then drafting proposed rules would be much 
> easier. 
> 
>         -- Bret

Bret -- it's a good list of principles, but I can't help but notice 
that for the most part these are exactly the characteristics of IETF 
working groups.  There are a couple of points I think should be 
made.  

First, WGs are not representative structures; they are, as the name
suggests, groups that are gathered to do work.  Confusing them with
representational structures would be a terrible mistake: they are 
not and cannot be such.

Second, the key to the WG is the charter.  It is a document that
represents an agreement between the WG members and the NC -- if you
will, a contract of sorts.  Generally in the IETF the area director
and the parties proposing the WG will iterate several times over the
charter, to be sure that the charter describes a parcel of work that 
can meaningfully be performed.

Third, the role of the chair is greatly misunderstood.  In the IETF
the chairs are always approved by the area directors, because the
role of the chair is to be responsible for the successful completion
of the work.  That is, a chair is not an elected representative of
the WG members, ever (*).  A far more accurate model would be to
consider the WG chair as an *employee* of the IESG who has the *job*
of bringing the WG to a successful conclusion.  This is, of course,
consistent with the model of a WG as a group that does work. 

Good WG chairs are highly valued in the IETF, and it takes a bit of 
participation in WGs to realize the unusual skills that are 
involved.

It may be that your idea of using the term "moderator" will be 
helpful -- it could help dilute the expectations generated by the 
term "chairman".  But I suspect that the term "moderator" will carry 
its own, but different, baggage -- the fact is that the role is 
unique, and conventional terms simply don't capture what is involved.

There are two areas where the DNSO is significantly different than
the IETF: 

1) top-down problem definition -- the ICANN board may present
problems to the DNSO, and ask for policy suggestions or guidance. 
This very rarely happens in the IETF -- almost all WGs are organized
entirely bottom up. 

2) the need to capture conflicting opinion.  By that I mean that the
output of a DNS WG may not include a consensus result -- it may
include two or three very divergent opinions.

In any case, the "legislative history" clearly indicates that IETF 
processes are an underlying model for the development of policy in 
the DNSO; I think it is necessary to study the applicable 
documents describing IETF structures, and use them as a starting 
point.

================================================================

(*) This point is worth a bit of comment.  IETF WGs are completely
open -- there is no defined membership of a WG, people enter and
leave as their interest and free time dictates.  Thus, there is never
at any time a meaningful "voters roll".  Voting simply does not work in
an open collaborative structure like a WG.  Hence the emphasis on 
consensus processes, and novel thermometers of consensus like the 
famous IETF "hum".

-- 
Kent Crispin                               "Do good, and you'll be
kent@songbird.com                           lonesome." -- Mark Twain