Re: [wg-c] Agreement on method for consensus determination

["Mark C. Langston" <skritch@home.com>]

On 13 July 1999, Kent Crispin <kent@songbird.com> wrote:

>On Tue, Jul 13, 1999 at 05:20:34PM -0700, Mark C. Langston wrote:
>> Until such time as the Secretariat adds me to wg-c-1 and wg-c-2, I'll have
>> to post this here.
>
>I think general discussion, such as this, should be on the wg-c 
>list, and not posted three times to each of the other lists.
>[...]

I've already made this point, Kent.  But because this discussion seems
to be occurring on wg-c-1 regardless, I felt the need to address it.

>
>> Finally, who runs the voting mechanism/votebot?  I would very much
>> prefer that this be a neutral party.
>
>No trusted third party is necessary.  This is a small body, and a
>roll-call vote on the mailing list is completely adequate.  The list
>is archived; the vote is a matter of public record; every member can
>see their own vote in that record, and every other members vote; 
>every member can check the tally, and every external person can 
>check the tally.
> 
>Roll call voting is frequently used for committees and other small
>bodies. 
> 
>Here are sample rules for a formal vote: the chair opens a voting
>period of at least a predetermined minimum length (in our case a day
>or so would be sufficient), and will fabricate and mail a ballot to
>the list.  Each member sends their vote (a reply to the to the ballot
>email) with their choices marked. At the end of the voting period 
>the chair tallies the votes, and announces the result.  For rigor, 
>the tally may be posted for challenge for a day, so that if someone 
>disagrees they can challenge the result.
>
>In practice, however, the formal procedure should not be necessary
>very often.  In a working consensus management model, the chair
>proceeds by looking for dissent, not looking for agreement, because 
>you work from agreement to agreement continuously.  The chair 
>typically asks "is there any objection?"  rather than putting things 
>to a vote, and if any objection is voiced, the objection is 
>discussed until the issue is clear, in which case either a rough 
>consensus will be apparent, or a vote will be called.


This is all well and good, Kent, but I don't remember any decision 
being made regarding which mechanisms we'd be using.  You are just
assuming that the IETF ideas will carry over.  The IETF is top-heavy
with technical experts, so 'rough consensus and working code' is 
a good rule of thumb.  However, ICANN, the DNSO, the pDNC, and
specifically these working groups, are *NOT* filled to the brim
with technical experts.  We've got laypeople, lawyers, businesspeople,
and others, all vying to be heard.

Given that, I posit that we at least *review* the IETF methodology 
before people begin to assume that it's what we're working under.
There's been no formal decision made on this, as I've already
pointed out.  

The cultures are not the same, and what applies in the IETF does so
mainly because that's what works for a roomful of very talented,
highly-strung, lofty-minded technocrats who need some structure
within which they can work.  That culture is not what exists here,
and we should at least question this framework before we proclaim
it's what will guide our hand in these matters.

Finally, with as much suspicion as that being cast on the DNSO, and
the pDNC in particular, it would be prudent to forego these
fast-and-loose informal methods, unless you really relish the idea
of being called to account for the decisions made here over and 
over again for years to come.  At least with a formal structure,
there will be slightly less cause for criticism.

It can chafe now, or it can chafe later.

-- 
Mark C. Langston	     			Let your voice be heard:
mark@bitshift.org				     http://www.idno.org
Systems Admin					    http://www.icann.org
San Jose, CA					     http://www.dnso.org