Re: [802SEC] Proposed action on 802.15.3a PAR withdrawal
Geoff,
I spent a great deal of time attempting to unravel the situation in
.15.3a from the inside. Let me offer some further detail.
66% approval - There is a fundamental parlimentry question here. As the
approval rate goes up, it clearly indicates a greater unanimity of the
body. But, if the body is divided, it makes it easier for the minority
to control the forward progress of the body. In theory at least, if
committees are showing evidence of frequent jamming, it is probable that
the approval level has been set too high and it has now become
profitable for groups to force a split standard rather than attempting
to compromise.
I believe that you are correct that the problem is not only present at
the task group level. If the task group level is set to 66%, the
problem may well work its way up the decision chain. Another possible
outcome is that people find the strategy to stalemate too tenuous and
become more willing to compromise. Your choices are: a) let the
situation continue b) reset all technical decisions to 66% b)experiment
by altering the task group position and evaluating the effect on the
process.
Entity balloting - The ideal of a democratic organization is
commendable. In the trenches, this is a level of idealism that is
fairly disconnected from practice. In practice, company members vote
together.
If a company has a significant vested interest in the process, it is
simply too attractive for them not to send a body of voters in an effort
to obtain a favorable outcome in the voting. This is the reality.
While I fully recognize that this suggests that a fix may well require a
fundamental reconsideration of the organization's founding principals
and/or a whole lot of effort to find other paths, this is a key factor
in much of the trouble that exists.
Let me give a simple example. If a startup has produced first prototype
silicon and is almost through round A funding at the time that proposals
are considered, it is extremely critical for their business to have
their approach largely if not entirely accepted. Millions of dollars
and entrepeneurial dreams are on the line. Would it be suprising to
find that sending a contengient of 20 or more people who have been
encouraged to vote positively (and who have stock options riding on the
outcome)is considered acceptable and sufficiently ethical?
Forced division of the PAR
Personally, I take the concept of a STANDARD rather literally. To me it
generally means that there is one approach that is recognized by the
body. Split standards, while in many cases expedient, are not really
beneficial to the market. They cause unnecessary confusion and in cases
of radio systems, interference.
I think that we should be working toward a process that reliably and
consistently generates single solution standards.
I would further argue that if we are experiencing the frequent use of
split standards to solve or avoid excessive delays, it is probably
another indication that your approval percentage is set too high. If a
group is in a minority position, it is an attractive option for them to
hold out for the split standard. They get a second bite at the apple in
the open market.
If the rules essentially force a split PAR approach, we would have
created an incentive. Getting to #2 is the same as a win. No further
compromise would be needed. This type of approach would seem to
severely diminish the likelihood of single solution standards.
-----Original Message-----
From: Geoff Thompson [mailto:gthompso@NORTEL.COM]
Sent: Thursday, January 26, 2006 9:28 PM
To: STDS-802-SEC@listserv.ieee.org
Subject: Re: [802SEC] Proposed action on 802.15.3a PAR withdrawal
Mat-
(2nd try, the list server rejected my 1st attempt)
At 09:38 AM 1/26/2006 , Sherman, Matthew J. (US SSA) wrote:
> Bob / All,
>
>
>
> For the sake of brainstorming here are some possible resolutions:
>
>
>
>1) Reduce consensus level to 66%
Ultimately will not work and just delays the problem. If you can't get
75%
at WG ballot what makes you think you can get 75% at Sponsor Ballot.
> 2) Entity balloting
Possibility but probably takes it out of 802 under current rules. In any
case it would require a PAR modification along with the attendant
approvals.
> 4) Task Group membership
>
> 4) Mandatory PAR split if predefined stalemate conditions occur
I don't quite understand how these allege to solve the problem, although
a
negotiated split has resolved some problems in the past.
Geoff
>
>
> I don't advocate one particular route or another,
> but all of these have been suggested as potential parts of a
solution.
>
>
>
> Another issue is cross task group / cross working group / sponsor
level
> influence. If you have two Task Groups working on potentially
competing
> technologies can they block each other? Also, people seem more apt
to
> cross working groups lately if they see activities in that other
group
> as competing with their own. In the past they were willing to live
and
> let live. This seems less true today. Also, even if you get a
standard
> through at the WG level, there is still the sponsor ballot. Any
> interested party can participate there and kill something if they
feel
> it competes with a different solution that they advocate.
>
>
>
> Personally I advocate a process which is more procedural and less
> judgmental. We have set ourselves up through the PAR / 5 criteria
and
> WG processes as judge and jury as to whether a project has merit, and
> which sets of technologies should be allowed to proliferate. I'd
rather
> see us merely be an instrument and forum for the development and
> documentation of solid specifications, and let the market decide.
The
> specifications that succeed in the market will be come the standards
> that further generations of specifications are based on. We need a
> process that reduces the potential gains available (or ability) to
block
> work on a basis of market addressed, or technology used.
>
>
>
> At least that's my view.
>
>
>
> Regards,
>
>
>
> Mat
>
>
>
> Matthew Sherman, Ph.D.
>
> Senior Member Technical Staff
>
> BAE SYSTEMS, CNIR
>
> Office: +1 973.633.6344
>
> email: matthew.sherman@baesystems.com
>
-----Original Message-----
From: Bob Heile [mailto:bheile@ieee.org]
Sent: Thursday, January 26, 2006 11:46 AM
To: STDS-802-SEC@listserv.ieee.org
Subject: [802SEC] Proposed action on 802.15.3a PAR withdrawal
Dear EC Members
As you have all probably seen in the press, Task Group 3a in 802.15
voted
to recommend the withdrawal of its PAR. The action was unanimously
affirmed by the full Working Group. The PAR was due to expire in
December
of this year anyway and given that one of the parties in the stalemate
adamantly refused to compromise, this is a sensible action. I will be
pre
submitting this recommendation for conditional inclusion on the NesCom
March meeting agenda subject to approval by the EC in Denver in a few
weeks. Just wanted to give everyone a heads up.
Press has been generally neutral to slightly positive regarding the
action. The IEEE has not been blamed. That is the good news.
Something
we may want to discuss though is the fact that this collision of
factions
is happening more and more. We seem stymied on how to deal with blocks
and
voter influence through economic pressure. Proving this stuff is
difficult
and the level of proof required is unclear. Does taking action expose us
to
lawsuits? Our 75% consensus requirement makes it relatively easy to
block
an action. These are topics we should perhaps discuss in Denver.
Bob
Bob Heile, Ph.D
Chairman, ZigBee Alliance
Chair, IEEE 802.15 Working Group on Wireless Personal Area Networks
11 Louis Road
Attleboro, MA 02703 USA
Mobile: +1-781-929-4832
email: bheile@ieee.org
----------
This email is sent from the 802 Executive Committee email reflector.
This list is maintained by Listserv.
----------
This email is sent from the 802 Executive Committee email reflector.
This
list is maintained by Listserv.
----------
This email is sent from the 802 Executive Committee email reflector.
This list is maintained by Listserv.
----------
This email is sent from the 802 Executive Committee email reflector. This list is maintained by Listserv.