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

RE: [RPRWG] Some thoughts from the Chair




To the IEEE 802.17 Chair and Working Group:

To come up with a reasonable schedule, we need to evaluate where we are in
the process and to estimate remaining work.

What we had in January was none of the proposals that had been presented and
reviewed previously, but a whole new proposal, which had elements of earlier
proposals. While this was useful to create a consensus, it was by all
measures a whole new proposal. It was not possible for this new proposal to
be fully understood at the January meeting, when it was first presented.
However, the group agreed to adopt it as a preliminary draft in order to
meet the schedule, with the understanding that it required work in order to
turn it into a workable standard.

While the initial draft did contain a fair amount of text (borrowed from
earlier proposals), there was relatively less substance in terms of state
machines or other acceptable forms of explicit and concise specifications as
is customary for the IEEE 802 standards.

What has become clear in the last 6 months is the degree of incompleteness
of the initial draft and a number of major technical concerns. To be
specific:

- Inaccurate and confusing MAC Service Model
- Missing MAC data path clause
- Missing topology discovery clause
- No state machines for any of the clauses
- Reference model was not representative of the MAC
- Fairness clause was largely incomplete and (still) missing explicit
behavioral specification of different modes.
- No simulator available to test the performance of
  the fairness algorithm.
- Concerns with frame replication, duplication, and 
  misordering
- Concerns with transparent bridging behavior and spatial reuse property of
the MAC

Technical issues can be overcome, provided individuals are allowed and
encouraged to point to the issues and ask for clarification or suggest
remedies. I do not believe people are trying to revive any of the original
proposals.

It is the duty of the chair to recognize valid concerns and to bring those
with valid concerns together in order to facilitate the work that is before
us. Creating divisions by innuendos and labeling "Those who failed to secure
a consensus around their proposals" is neither appropriate nor productive. 

Respectfully,

Nader Vijeh

Member of IEEE 802.17 working group

-----Original Message-----
From: Mike Takefman [mailto:tak@xxxxxxxxx]
Sent: Monday, July 01, 2002 7:34 PM
To: dot17
Subject: [RPRWG] Some thoughts from the Chair



RPRWGers,

As we come up to our July meeting I wanted to share some
thoughts with all of you. During the meeting we will be
discussing our schedule. I believe it is obvious to people
that our current schedule is not achievable and that we 
need to adjust it. The adjustment has to take into 
account several things. Foremost is the timeliness
of the standard, naysayers in the industry will point
to the slip as a reason to discount RPR entirely. We
must set a date that is achievable in terms of the 
802 process and our ability to do the work.

Let us recall how we got to this particular place. The study 
group and working group met for over 2 years prior to finally 
making a decision on a proposal to move forward on. That process
was open to everyone, and every voice that wanted presentation
time received it. This process provided  more than enough time 
for anyone to bring forward proposals and build concensus 
around it. The text that we approved is by no means the
final answer, it is not cast in stone. But, it represents
a set of features, values and methods that the vast majority
of the WG approved as the base to move forward. 

The process that we followed required that people to write
detailed descriptions of their proposals. This had the advantage
of requiring a certain degree of seriousness on the part of presenters
to do the work to convince the committee to adopt their proposal.
It also was intented to insure that all complexity of an 
algorithm was presented to the committee. Clearly in November 
of 2001 the committee was deadlocked around 3 proposals. Work 
was done to converge 2 of the proposals in the hope of achieving 
concensus. The disadvantage of the process was that as concensus 
was built, the text lagged behind the slideware and hallway 
conversations.

In January, a concesus was achieved. Those who failed to secure a 
concensus around their proposals had a decision to make in January. 
Some have embraced the group decision and are working to refine and 
improve it. Others, are choosing to push an agenda to reverse that 
decision. 

My point, and I do have one, is that we have to remember the goals
of the WG. Creating a standard that increases the available market 
for all members of the working group in a timely manner. Standards
work is about compromise, we achieved a certain amount of momentum
and it is up to the committee to determine whether the current
draft is good enough or how to improve it in areas of deficiency.

Improving it first involves clarifying the current text to 
insure that everyone understands the algorithms, state diagrams, 
etc. In other words, it should be complete.

Continued improvement then involves identifying issues, determining
the severity of the issue, and then considering alternatives to 
closing the issue. It is up to the committee to determine the 
"best" alternative to close an issue. 

In closing, remember that my role as chair is to work
to complete the standard in a timely manner. I appreciate
all the hard work of people to help achieve this
goal.

See you in Vancouver,

mike

-- 
Michael Takefman              tak@xxxxxxxxx
Manager of Engineering,       Cisco Systems
Chair IEEE 802.17 Stds WG
2000 Innovation Dr, Ottawa, Canada, K2K 3E8
voice: 613-254-3399       fax: 613-254-4867