[ga] GA representation on the Names Council

[John C Klensin <klensin@mci.net>]

I've just realized (being a bit slow) that a lot of the
frustration that has periodically filled the GA list with noise
and name-calling is due to the perception that the NC was
composed of people appointed by Board-authorized constituencies
and that there was no possibility for people who were not part
of those constituencies to have a voice.  It it equally clear
that recognition of the (an) IDNO won't solve the problem --
someone will always feel left out.

I also realized that this has some similarity to the problem the
IETF faced some years ago in trying to figure out how to select
a (quite powerful) nominations committee in an organization that
has no membership list.

So, a proposal, independent of the chair selection process or
anything else...

(i) the NC be immediately expanded by three members.   If I
correctly recall the bylaws, they can't be given the vote
without ICANN Board action, but it should be possible for the NC
to seat them as observers (with the same standing to participate
in discussions) on its own initiative while the formalities are
being pursued.

(ii)  All three people selected to these positions will serve
for one year.  If any resign or decline to serve, the
replacement will be only for the duration of the original year.
The intent it to get this mess straightened out during the year
-- turning the GA into another constituency is not, IMO, a
desirable long-term approach-- and replace it with  whatever
permanent arrangements are needed/appropriate.

(iii) Those eligible to serve will be the entire
contents/membership of the -announce and -ga lists as of
(ideally) last Friday.  Selection of a date in the recent past
prevents "stacking" by a rash of subscriptions.   If capturing
last Friday's list is not feasible, the list contents should be
captured as of the time this note is received at the
secretariat.  In the interests of fairness, the secretariat
should add everyone who has been excluded from the list within
the next month for antisocial behavior back into the pool.

(iv) The _sole_ qualifications for these seats shall be 

     (a) Membership in the GA, as defined above by list
membership
     
     (b) Willingness to provide the secretariat and the NC a
     potentially-authenticatable name (e.g., one that might
     appear on a driver's license, passport, or national
     identity card, rather than a network persona), postal
     mailing address, telephone number, and other reasonable
     information to establish that the emailing address belongs
     to a person.

(v)  Within that pool of qualified names, an ordering will be
established by random selection (reference below to a procedure
that is known to be tediously fair; let's not waste a lot of
energy discussing this or other ways to get randomness).   The
first three names chosen will be seated as Names Council
members.   If one or more decline to serve, or subsequently
resign, he or she will be replaced by the next person on the
list.  Beyond those seated, the ordering of the names will not
be revealed in order to prevent gaming the system or one person
resigning in favor of another (to preserve randomness, people
should be encouraged to serve by any means necessary, including
fear that they might be replaced by someone they would consider
unacceptable).

Now, the weakness in (v) is that someone (or some very small
number of people) have to be trusted to do the computations and
then keep the list.  For convenience, I'd favor turning this
over to ICANN staff  or the board, and letting Mike or Esther do
it.  But some of those who are feeling least represented
obviously don't trust them.  It probably should be someone who
is not actively involved in the current DNSO fray -- perhaps we
could try to pull Tamar Frankel out of semi-retirement on this
subject (warning: I haven't consulted her about willingness to
serve -- this proposal will come as more of a surprise to her
than it does to you).   Or someone might have a better idea. But
I'd personally be reasonably comfortable having the NC or Board
make the choice, with the main requirement being integrity.

Anyway, does that appeal to anyone?  Comments from NC readers of
this list?  Other suggestions?

If we can't trust elections, maybe we can trust Gauss.

     john

Reference: the current randomization procedure used in the IETF
is described in

http://www.ietf.org/internet-drafts/draft-eastlake-selection-04.
txt
It has been nit-picked extensively by experts (on both
randomization and nit-picking), for whatever that is worth.