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

[8023-CMTF] Status of P802.3ar

I have good news and bad news.
First, the good news.
The P802.3ar TF held a brief, yet successful meeting in San Diego last week. We met Tuesday afternoon and resolved both comments received on D1.3. We then looked ahead to the 802.3 WG closing plenary and the votes on the revised objectives, PAR, and 5 critieria. In addition, we hoped to move to WG ballot. Note: We received no comments on the revised PAR and 5 criteria from other 802 "dot groups" during the week.
Now the bad news.
The 802.3 WG closing plenary was a very difficult meeting. After presenting the progress and results of the P802.3ar TF between March 2006 (Denver) and July 2006 (San Diego), I presented the revised TF objectives for 802.3 WG approval. The motion failed as follows: Y: 24, N: 10, A: 27. The motion required >= 75% and only achieved 71%. The following were some of the concerns raised during the debate on the motion to approve the revised objectives:
1) 802.3 shouldn't vote on the objectives until we've seen the PAR and 5 criteria
2) P802.3ar doesn't have broad market potential
3) 802.3x (PAUSE flow control) is not in wide use, why would another rate limiting project be different?
4) Rate limiting increases frame delay and frame delay variation
5) The current draft specifies 3 mechanisms not 1 mechanism as stated in the objective.
Following this defeat, I then proceeded to present the revised PAR (as directed by the TF during Tuesday's meeting). Several objected to voting on the PAR until the 5 criteria were approved. I then withdrew the motion on the revised PAR.
I then presented the revised Broad Market Potential criterion. The motion to approve also failed. Y: 19, N: 11, A: 27. It received just 63% (needing 75%). Many of the earlier concerns were reiterated. In addition, at least one chip vendor objected to rate limiting as an example of an "optional feature" that invariably becomes a hard requirement for customers if only for "check-list compliance".
At this point in the meeting, a motion from the floor was offered. It read:
"Move that 802.3 WG withdraw the IEEE P802.3ar PAR and return it to NESCOM, by authorizing the WG chair to fill out the appropriate form and presenting it to 802 EC"
After some debate, a motion to postpone consideration (of the prior motion) until the November 2006 plenary was approved (Y: 38, N: 5, A: 3).
- - -
The observation was made that the motions to approve the revised TF objectives and revised criteria received a large number of abstentions. It was pointed out that if 802.3ar has any chance of proceeding, it should focus on educating those who abstained during this meeting.
(I will cover the discussion on the September interim meeting in a separate post.)
Kevin Daines
Chair, P802.3ar TF