Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[802SEC] IEEE-SA CAG considering sponsoring competing standards to IEEE 802

Colleagues:

 

Some recent activities on Ethernet have raised concerns again about the IEEE-SA competing against itself by spawning competing standards activities through different organizations.  Many of you will remember when an ISTO affiliated consortium was formed that IEEE 802.16 perceived as direct but non-interoperable competition to IEEE 802.16.  This time, an Ethernet related activity has been discussed as an IEEE-SA Corporate Advisory Group (CAG) sponsored standards development project.  Please note that at this time, nothing formal has happened, but initial discussions raise some more general questions in the prevue of the Executive Committee.

 

As background, I provide my understanding of the relevant parts of CAG operation from "Operating Procedures for The Corporate Advisory Group as a Standards Development Sponsor", “Maximizing Your IEEE-SA Corporate Membership” and discussions with CAG leadership and IEEE staff.

 

  1. The "CAG is a committee of the IEEE-SA Board of Governors (BoG), constituted to provide operational and strategic planning advice on matters affecting the interests of the IEEE-SA corporate and organizational members."  In some ways, the CAG is similar to the LMSC EC in that it is a supervisory body representing its members.  The CAG represents IEEE-SA Corporate Members, only a subset of which have representatives with seats on the CAG.
  2. The CAG has 10 seats with 6 currently filled.  There are 46 entities listed as IEEE-SA Corporate Members on the web site (last modified September 2003).
  3. The CAG can function as a sponsor for IEEE standards projects (as can other societies within IEEE, like our sponsor the IEEE Computer Society).
  4. Before sponsoring a standards development project, the CAG must give existing societies 45 days to consider being the sponsor for the proposed project.  If the sponsor takes on the project, balloting must be as entities.
  5. At least three IEEE-SA Corporate Members must support doing a standards project, and the CAG must approve a PAR before forwarding to NesCom if it is to be the sponsor.
  6. The CAG is not required to get a supporting vote of the corporate membership, nor even solicit comment from the corporate membership about a proposed project.  The support by the requisite number of entity members is assumed to be sufficient indication of entity interest.

 

Recent activities in and outside 802.3 will highlight why I feel current CAG policies are an important issue for the LMSC Executive Committee to consider.  In general, my concern is that the current procedures of the CAG can be used to undermine the decision making authority of established working groups, and destroy industry respect for and support of IEEE-SA as a standards development organization.

 

Because the EC’s responsibility is mostly to process, I present the following as background information so that you can easily envision similar situations within your working groups.  Hopefully we can deal with this particular proposal as a case study, and independent of the technical merits and liabilities of the proposed work, focus on the more general strategic implications.

 

  1. In July 2003, a proposal was made to the 10GBASE-T study group for multi-rate operation (2.5, 5 and 10 Gb/s).  This was discussed on its technical merits as well as within the context of the Five Criteria.  After three hours of discussion, the SG declined to include an objective for multi-rate operation.  The  motions to change criteria text to allow multi-rate operation generated 23% or less support, and the individual votes on each of the Five Criteria passed by 78% or more.
  2. Also in July 2003 at the closing 802.3 plenary meeting, the proponents brought the issue directly to the floor during the study group report.  After more than an hour of discussion, and it being a late hour, no motion was made on objectives for the emerging P802.3an (10GBASE-T) PAR, and a motion to forward the PAR and criteria failed Y: 30, N: 16, A: 15.
  3. In November, efforts to promote 2.5 Gb/s were separate from the proposed 10GBASE-T project.  Motions to approve and forward the 10GBASE-T Five Criteria and PAR passed with little or no opposition.
  4. Also in November, proponents of 2.5 Gb/s Ethernet held a Call for Interest.  It was one of three CFIs held in sequence Tuesday evening.  Attendance was about 175 and seemed fairly consistent for all of the CFIs.  Two of the CFIs were successful in generating support for formation of a study group, but the 2.5 Gb/s CFI was not.  The straw poll question presented at the CFI was:

Should IEEE 802.3 form a Study Group to develop a project proposal for 2.5 Gbps Ethernet?

 

Attendees:        Y: 53, N: 64, A: 39

802.3 Voters:     Y: 20, N: 29, A: 21

 

5.   Also in November, the proponents brought essentially the same question before the 802.3 closing plenary.

 

Motion:  802.3 WG authorizes the formation of 2.5Gbps Study Group

 

Y: 17, N: 31, A: 17  Motion Failed

 

6.   In January 2004, Mr. Nikolich received a request for cooperation on a 2.5 Gb/s project from the Chair of the IEEE-SA CAG, initiating discussions among CAG leadership, IEEE 802/802.3 leaders and IEEE Staff.  Mr. Nikolich gave his personal opinion to the Chair of the CAG that executing this project through sponsorship by the CAG "was not advisable due to conflict with the 802.3 position", but he also requested more information about the CAG and corporate IEEE-SA participation before offering any formal response to the request and requested that 802 and 802.3 leadership be involved in any discussion by the CAG on this potential project.  To date, that request has been honored.

  1. Following this communication, a significant amount of misinformation began to circulate within the Ethernet community, including misrepresentations of Mr. Nikolich’s position on the proposal, the status of the proposal with the CAG, etc.
  2. Consequently, a number of email, telephone and conference calls resulted in IEEE staff arranging a 27 January call with:

IEEE 802 – 1st Vice Chair, 802.3 Chair, 802.3 Vice Chair, 10GBASE-T Chair

CAG – Chair and Vice Chair

IEEE staff

President of the IEEE-SA

In addition to the above information about CAG operation, the conference call indicated:

a)   The CAG leadership was committed to support this proposed project.

b)   A PAR if proposed would only require approval by the CAG to forward to NesCom. 

e)   IEEE 802 participants felt the absence of communication with the corporate membership about the advisability of sponsoring a project lessens the value proposition for becoming a corporate member.  “Why should I recommend my company join IEEE-SA when it would not be guaranteed any mechanism for disapproving proposed new work?”

f)    It was expected that a 2.5 Gb/s project would be submitted to NesCom for March consideration.  [Though it is believed that this will not happen.]

g)   The CAG leadership would allow the 45-day period for other societies to respond to overlap with the NesCom submittal. 

 

 

The similarities of this to IEEE 802.16’s problems with ISTO are obvious to those that were involved.  With ISTO, we argued about whether it was appropriate to call their output IEEE standards or specifications.  The major difference here is that standard versus specification isn’t a question with a CAG sponsored project.  A CAG sponsored project will be an “IEEE Std xxx” document.

 

The ability of the CAG to sponsor projects is a worthwhile activity for IEEE-SA to support.  Without proper constraint though, it can be used to undermine the decisions of 802 and other established working groups.  As an example, most of our groups have lively discussions about encoding/modulation techniques.  What if one of our WGs makes a decision and the proponents of a losing proposal recognize they only need two entity “friends” to go get their own standard through the CAG.  Consider what this particular case would mean as a precedent to IEEE Std 802.11, IEEE Std 802.15, IEEE Std 802.16, etc.  What if the losing side on a bridging or security proposal didn’t like the 802.1 decision.

 

I would like the EC to consider a position statement to the BoG on additional considerations for CAG sponsored standards projects.  I believe that a motion should state that:

 

The IEEE LAN/MAN Standards Committee (LMSC) is supportive of Corporate Advisory Group (CAG) standards activities where existing sponsors are disinterested in a proposed activity.  The LMSC is strongly opposed to the CAG sponsoring projects where there is strong interest in the proposed activity.  The LMSC disagrees with any presumption that rejecting a proposed standards activity is equivalent to disinterest in the activity.  CAG and IEEE-SA process must consider an established working group’s position that a particular standards project is within its area of work and it that it should not be approved.  The LMSC requests that CAG P&P be modified to support this method of operation.

 

To have a position ready for the BOG meeting on February 25-27, I will be making a motion similar to that above this week.  In the meantime, any wordsmithing suggestions will be appreciated.

 

Thank you for your attention to this lengthy but important message.

 

Bob Grow

 

SEC 2-5 letter.ZIP