[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [wg-c] about the consensus call
Dave and Kent have already covered a boatload of issues posed by the text
and method of the proposed consensus call (the Sheppard/Meuller proposal).
There is one additional point I'd like to raise, and one suggestion I wish
to offer. Both are long, I've a weakness of writing to that effect. This is
my 4th item of mail for the day, I don't think Jon's late-night assessments
are the best we can do, and time to get a report is quite limited. I've a
tribal political issue (the "squaw controversy" in Maine) so I'll be lucky
to even read email for the rest of the week.
I couldn't appreciate back in December why Kent chose to abstain from the
bitterly contentious super-majority vote on the previous 6-10 consensus,
a vote triggered in my view by the co-chair's accommodation to the poorly
judged claim made by a non-participant in WG-C that there was no consensus
on the 6-10 question .
After writing up my several commentaries on who voted and how , some of
which circulated privately, and were reflected in my mail this week  on
historical trivia -- the apparently fatal three-way (or four if the camp
opposed to the IAB Technical Comment on the Unique DNS Root  are to be
counted), it has come to me -- the limitation of the decision making
process WG-C is capable of, the point behind Kent's decision not to vote.
Consensus calls can only serve to make some party the ass of the moment,
(and voting has the same net effect).
If the proposal is to have policy (marks, human rights, non-profit, Position
Papers A, C, D and E), then Martin is the ass. He and his are vile
If the proposal is to have no policy but market forces (converse of the
above, Position Paper B), then Jon, Caroline, Kent and Eric are the asses.
They and theirs are the vile obstructionists.
If the proposal is to have new gTLDs (6-10), then Caroline (Position Paper C)
is the ass. She and hers are the vile obstructionists.
If the proposal is to have no more marks abuse (Position Paper C), then
Jon, Milton, Kent and Eric are the asses. They and theirs are the vile
And not to leave any stones unturned ...
If the proposal is to have discretionary capacity to act in the interests
of the Unique DNS Root then Tony and Karl are the asses. They and theirs are
the vile obstructionists.
That about fills out the barnyard watering trough and braying chorus.
For those unfamiliar with obstinante equines, think of a circular firing
We can't get to a consensus position. It is futile for the co-chair, even
if not a party to a position, to seek the impossible or the merely absurdly
difficult process goal of consensus on an issue more substantive than whether
tea is a popular beverage in China.
What we can get to is simple majority and super-majority votes, which are
as Kent observed, divisive, but at least offer clear process boundaries.
We risk of course ballot stuffing -- the point that Kent's response to
Jamie Love made, and has been made w.r.t. the number of warm bodies from
the marks interests earlier.
What we can also get to is a delicate balance point, the principled armistice
that our co-chair has tried to steer our process toward -- I simply think
he erred in projecting Sheppard/Meuller onto the results of the recent
straw poll. My reading of the straw poll and the balance of forces leads
me to offer the following:
o 6-10 in principle and specifically
o a chartered gTLD with a specific operator (.PPE), and
o a chartered gTLD with a reserved operator (.tbd), and
o a non-chartered gTLD with specific non-profit operators
o a non-chartered gTLD with specific for-profit operators
o a non-chartered gTLD with reserved non-profit operators, and
o a non-chartered gTLD with a reserved for-profit operator, and
o a chartered gTLD with an ICANN operator (.ICANN)
[Notation: PPx == Position Paper x, a placeholder, e.g., PPE == .NAA]
o Marks in principle and specifically total surrender of the problem
to the IPC during the first year of the roll-out period, extensible
to the second and subsequent years of operation upon simple request
by the DNSO's IP Constituency, subject to confirmation by super
majority of the ICANN Board. [They have to have a working solution
eventually, and my weekend reading is the set of IPC documents to
the Cairo meeting  to attempt to understand their current best
thoughts on the subject. I encourage everyone to take the time to
understand the IPC position in detail.]
o Selection of registry operators by informal means through a new
DNSO working group selected by the DNSO, for both chartered and
While favoring the positions advanced by Dave, Kent and I and about two
score of other WG-C participants, this does test the limits of patience of
the position articulated in Position Paper B, I can only offer to Martin,
Chris, et al the hope that a) we're honest and b) that more gTLDs will be
added to the root and that for some part of the DNS market that market
principles will prevail. This also tests the limits of caution of the
position articulated in Position Paper C, I can only offer to Caroline et
al the hope that a) we're honest and b) that they control the abuse issue,
and that this is a better solution than Sheppard/Kleiman, as I pointed out
in , and as others have it is subjective to the point of inutility.
To accommodate the desires of the Position Paper B co-signers, who I hope
will refrain from knifing each other as well as everyone else, I suggest
they (and anyone else with a substantive record of participation) are
simply allocated one gTLD to operate as a for-profit monolithic registry
(shared equity), with themselves as registrars (or that also combined into
the registry, subject only to the Marks principle (above), to be further
partitioned into a second and subsequent gTLDs, not necessarily "shared"
in operational or equity forms, subject to the Selection principle (above)
for so long as that informal process is in place.
To accommodate the desires of the Position Paper D co-signers, who I know
will play nice with others and not run with scissors, the same as above
but in non-profit form.
To accommodate the desires of the many who want a specific chartered gTLD,
I suggest that we get on with our oldest hypothetical, the museum example,
and give Kent free rein to find an NGO to operate that gTLD, as the charter
is about as simple as that of .EDU. Do it and go.
To accommodate (or channel) the desires of the carrier giants and be on the
leading edge of a technical problem (for a change), I suggest that we drop
the task of finding an entity to operate a gTLD to support IP telephony on
Mark Sportack (AT&T) and let the VoIP players write a reasonable charter and
get on with the problem of solving the E.164<->IP address mapping problem.
The roadmap I suggest is:
NAA in 2000 -> other human rights gTLDs in subsequent years
PPB in 2000 -> fraction/copy to other gTLDs in subsequent years
PPD in 2000 -> fraction/copy to other gTLDs in subsequent years
MUS in 2000 -> other cultural gTLDs in subsequent years
VoIP in 2000 -> other technical gTLDs in subsequent years
WG-C in 2000 -> more, possibly copies of the above gTLDs, to 6-10
WG-C process -> other gTLDs in subsequent years
A predicate is some degree of enlightened self-interest, commercial brand
trust. Most everyone makes some of the currency of their preference, and
most of us look like rational statespersons to our respective constituencies,
and the knives are put away until next time, when all of us may know better
than to engage in Mutual Assured Frustration, and ICANN itself proved to be
better than the IAHC at the problem, based upon results.
Nappi, the Siksika hero-fool, Coyote to the Lakota, Raven to the Salmon
Peoples, always says after his foolishness becomes apparent, "that was
the one thing I forgot". I'm sure I've forgotten something, I hope it is
Oh. The principle of geographic diversity! Participation in the first
round of PPB and PPD forms of non-chartered gTLDs may need to meet that
criteria, so we may need to have two or three of each to both meet the
principle of geographic diversity and not be complex to the point of
being impossible to equtize or operate.
Oh2. The possibility that the Marks issue is not solved or at least is not
considered tractible by the IPC in the first year. We all take Vixie's
suggestion and use random label assignments. The SLD/3LD/... strings will
be goofy, but may be tractibly mapped back to non-Marks strings after the
IPC has a solution, thus no injury to Marks and no bar to registry roll-out.
User asks for "foo", gets mumblefratz and post-IPC-solution has only the
registry policy (if any) to negotiate to get back to "foo".
Anyone have any alternatives that work out at least getting past the
 Mike Heltzer, INTA