RE: Unified PMD vs. Unified PHY
Roger, (all, please read):
You made the comment:
> Standards groups (again IMHO) should define standards, not
> implementations. I.e.
> define what is on the wire, not how it gets there.
This is quite correct. That is a priority.
But, I think that you would agree that it is not the only consideration. We
also have a responsibility to the 5 criteria
(http://grouper.ieee.org/groups/802/3/ae/criteria.pdf).
We also have a responsibility to the objectives
(http://grouper.ieee.org/groups/802/3/ae/objectives.pdf).
Ethernet has historically been very successful because previous committees
have been able to appropriately balance these requirements. While the word
"implementation" is not used explicitly in these documents, it is strongly
implied. If you will, a successful standard cannot become successful in the
market while ignoring practical implementations (no, I don't think that
Roger implied anything different than this, I just wanted to make it
explicit).
These "equations" must be simultaneously solved with a minimum of a 75%
consensus. If you will excuse another metaphor, that is the signal we need
to find amidst the noise.
Now that the brainstorming phase of the work IS OVER, all of us need to do
some soul searching. We have had a year to position our various proposals
and positions. Now is the time to come together. Everyone should take the
following imperfect, but invaluable documents and put them on a clean desk
in front of them:
1. The five criteria
2. The objectives
3. The PMD survey from the Dallas meeting
4. The PHY survey form the Albuquerque meeting
Then we should do the analysis to see what combinations of proposals gets us
to the finish line: 75%, July, meeting ALL of the objectives and criteria.
Any work or effort that deviates from this goal/objective is "noise."
jonathan