Re: [BP] [bp] 802.3 approves Backplane Ethernet PAR
Yong:
There is nothing that requires a 1:1 relationship between CFIs and SGs,
nor SGs and PARs. A significant percentage of CFIs fail to engender
enough support for a SG, and 802.3 initiates some projects without
either a CFI or SG. Though this exception is I believe only appropriate
for reaffirmation projects, revision projects and amendment projects as
part of our ongoing Maintenance process. (There was no CFI for
P802.3REVam that was recently approved by the Standards Board, it came
directly from the Maintenance TF.)
There is nothing that requires that a Study Group produce a PAR though
that is its primary purpose. The Study Group can recommend that no PAR
be initiated (or as has happened, die from lack of progress). There
also is no requirement that a Study Group only produce one PAR, it study
can conclude that two or more PARs is the appropriate way to do the
desired work (just as a TF can decide its work should be split into
different projects). During the Backplane Ethernet CFI I discussed the
possibility that a Backplane Ethernet Study Group if formed might
recommend multiple projects. It is the documented recommendation of the
BESG that work be divided (the motion to form a new SG for congestion
management). The January BESG meeting had MAC client sublayer
enhancements as a continuing item for discussion in March, and the 802.3
meeting announcement highlighted continued work on objectives in March.
The BESG recommendation to split the work though did not come until the
March meeting. Had it been a decision of the January meeting, then two
PARs would have been pre-circulated to 802.3 and the EC for
consideration in March. Because it was a March decision to split the
work into two projects, the alternatives to continue work on both
backplane PHYs and rate/congestion management in May were to either: 1)
delay approval of the P802.3ap draft PAR while a P802.3ar draft PAR,
criteria and objectives were prepared within the BESG for
precirculation, 2) approve renewal of the Backplane Ethernet study group
including continued study of rate management after P802.3ap PAR
approval, or 3) to immediately form a study group to continue work on
the MAC Client layer enhancements and create a draft PAR, criteria and
objectives (the later two only being a difference in SG name).
This is not an issue of openness (openness is defined as the ability to
participate and no one was prohibited from participating in the SG or
802.3 meetings). I don't know of any P&P requirement for
"transparency", nor am I confident would there be a consistent
definition. The CFI discussion that the SG might investigate
rate/congestion management (or possibly defer for future work), and the
January SG discussion of the topic and report that the topic was one for
continued consideration in March is certainly transparent by my
definition. I will also note that my presentation to the closing March
EC highlighted this controversy and another member of the EC that was in
attendance at the closing 802.3 meeting reiterated his reservation about
the SG being formed without a CFI. The EC though overwhelmingly
approved the SG formation.
The CFI process is an 802.3 concept (also now used by other 802 WGs), it
is not required by IEEE 802 nor IEEE-SA. It therefore becomes the
purview of 802.3 to decide if a new CFI is required under its rules. As
Adam pointed out, it was discussed, and some members believed a new CFI
should be required, but a majority did not believe another CFI was
required. I was clear as Chair that I did not believe another CFI was
required, allowing the motion (as I recall, the two past Chairs in the
room split on the need for another CFI). The WG is the ultimate
enforcer of its rules, and clearly if it disagreed with my decision,
802.3 could have disposed of the motion differently (tabled it, voted it
down, etc.)
The threshold to progress from SG to project will be 75%, and a
commitment was made to conduct a tutorial on the rate management topic
before 802.3 votes on approving a PAR.
Bob Grow
Chair, IEEE 802.3 Working Group
bob.grow@ieee.org
> -----Original Message-----
> From: owner-stds-802-3-blade@LISTSERV.IEEE.ORG
> [mailto:owner-stds-802-3-blade@LISTSERV.IEEE.ORG] On Behalf
> Of Yong Kim
> Sent: Wednesday, March 31, 2004 3:05 PM
> To: STDS-802-3-BLADE@LISTSERV.IEEE.ORG
> Subject: Re: [BP] [bp] 802.3 approves Backplane Ethernet PAR
>
>
> Adam,
>
> Thanks again for your detailed response. I am beginning
> to agree with you that the new study group formation without
> following the CFI/Advance Notice may be .3 issue. I'll take
> it up w/ Bob Grow.
>
> I am more concerned because:
> - may be out-of-order per .3 rules
> - judging from vote results, the issue was contentious,
> therefore any latitude given may not be "transparent"
> and "open".
>
> best regards,
>
> Yong.
>
> -----Original Message-----
> From: Healey, Adam B (Adam) [mailto:ahealey@agere.com]
> Sent: Wednesday, March 31, 2004 1:43 PM
> To: Yong Kim; STDS-802-3-BLADE@listserv.ieee.org
> Subject: RE: [BP] [bp] 802.3 approves Backplane Ethernet PAR
>
>
> Yong,
>
> 802.3 did not vote to split the PAR. Rather, per the
> direction given by
> the Backplane Ethernet Study Group, we...
>
> (1) ...presented a PAR, 5 Criteria, and Objectives list for
> the Physical
> Layer aspects of Backplane Ethernet, and these were approved by 802.3.
>
> (2) ...requested that a new Study Group be formed to look into Layer 2
> enhancements. The purpose of this Study Group is to generate
> a new PAR,
> 5 Criteria, and Objectives specific to Layer 2 enhancements. The
> question of whether this new study group should have been
> preceded by a
> CFI was a point of discussion in 802.3. In part, it depends on your
> view of whether the original CFI for Backplane Ethernet
> "covered" Layer
> 2 enhancements or not.
>
> The vote taken within the BESG was to form a new Study Group, and not
> initiate a new CFI. No motion is necessary to initiate a
> CFI. Rather,
> interested parties can submit a request to a have CFI at the next WG
> plenary (35 day advance notice is required). At the closing plenary
> session, those parties may wish to request that 802.3 form a new Study
> Group. That request comes in the form of a motion much like
> the one the
> BESG recommended I put before 802.3.
>
> Thank you,
> -Adam
>
>
>
> -----Original Message-----
> From: owner-stds-802-3-blade@LISTSERV.IEEE.ORG
> [mailto:owner-stds-802-3-blade@LISTSERV.IEEE.ORG]On Behalf Of Yong Kim
> Sent: Monday, March 29, 2004 3:13 PM
> To: STDS-802-3-BLADE@LISTSERV.IEEE.ORG
> Subject: Re: [BP] [bp] 802.3 approves Backplane Ethernet PAR
>
>
> Dear Adam,
>
> Thanks for your detailed (and I am sure time consuming)
> response. This helps a lot. I am puzzled though on your item
> 2 subpart c): New study group needing simple majority.
>
> IF this is new study group THEN
> 802.3 Operating Rule 4.2 - Formation - CFI scheduled
> 35 days prior to WG Plenary.
> ELSE IF this is not a new study group but a split THEN
> ???
> ENDIF
>
> If I were to fill in ???, all my prior .3 experience, a vote
> to split a PAR (close one I could remember was 1G copper, and
> exactly not the same) was NOT simple procedural (and no separate
> CFI was done to do the split). I recall the votes taken within
> BP SG was to recommend a new study group, which requires 35 day
> and CFI. Am I missing something? Please advise/clarify.
> Thanks in advance for your efforts.
>
>
> best regards,
>
> Yong.
>
> -----Original Message-----
> From: Healey, Adam B (Adam) [mailto:ahealey@agere.com]
> Sent: Monday, March 29, 2004 11:58 AM
> To: Yong Kim; STDS-802-3-BLADE@listserv.ieee.org
> Subject: RE: [BP] [bp] 802.3 approves Backplane Ethernet PAR
>
>
> Yong,
>
> My apologies for the delay in my reply.
>
> First, if you would like a detailed account of closing
> plenary, I would
> refer you to the plenary meeting minutes, which will be available at:
>
> http://ieee802.org/3/minutes/index.html
>
> I have been told the minutes from the March plenary meeting will be
> posted later this week.
>
> If you cannot wait for the minutes to be posted, you may also refer to
> the Backplane Ethernet closing plenary report, which you can
> find posted
> at:
>
> http://ieee802.org/3/bladesg/public/mar04/closing_plenary_0304.pdf
>
> In direct answer to your questions:
>
> 1. "please consider for approval under continuous process"
>
> Information on the continuous processing can be found at:
>
> http://standards.ieee.org/faqs/contproc.html
>
> To summarize, the next meeting of the New Standards Committee (NesCom)
> and the Standard Board is not until June. Under the regular process
> this would be when our PAR is reviewed and the implication is that we
> would not meet as a Task Force until the July plenary, at the
> earliest.
>
> Under continuous processing, e-mail and the web are used to
> expedite to
> process, giving us the chance of having our PAR approved in
> time for the
> May interim meeting. The study group has requested to have the PAR
> forwarded to NesCom for review. The addition of the "continuous
> processing" phrase simply expedites that process.
>
> 2. Ammendments to Study Group Motions at the 802.3 closing plenary
> session.
>
> There were three changes made via motions to amend, which I will
> enumerate below:
>
> (a) Change the objective "Support BER of 1E-12" to "Support BER of
> 10^-12 or better".
> The motion to amend passed unanimously (voice vote). This is superset
> of the original objective. A solution that operates at exactly 1E-12
> and a solution that operates at some BER below 1E-12 can both be
> considered to have satisfied this objective. There was also some
> confusion regarding scientific notation.
>
> (b) Changes second bullet of Distinct Identity Criteria from "The
> standard will define a single PHY for each speed for
> operation" to "The
> standard will define at most one PHY for 1Gb/s operation and
> at most one
> PHY for 10Gb/s operation" .
> The motion to amend passed unanimously (voice vote). If I recall
> correctly, the intent of the amendment was to allow for the
> possibility
> of one PHY that supported both speeds.
>
> (c) The motion to form a new study group for congestion
> management/rate
> control was changed from "...for Layer 2 enhancements for congestion
> management for Backplane Ethernet" to "...for Layer to
> enhancements for
> congestion management including the needs of Backplane Ethernet".
> Motion to amend passed Y:14,N:10,A:0. The amended motion passed:
> 23/18/15. Note that the motion to form a new study group is
> procedural,
> and only requires a simple majority to pass. The needs of Backplane
> Ethernet are considered in either form of the motion case. The
> amendment was made in the interest of not limiting the solution to be
> backplane-centric.
>
>
> Thank you,
> -Adam
>
>
>
>
> -----Original Message-----
> From: owner-stds-802-3-blade@listserv.ieee.org
> [mailto:owner-stds-802-3-blade@listserv.ieee.org]On Behalf Of Yong Kim
> Sent: Monday, March 29, 2004 1:53 PM
> To: STDS-802-3-BLADE@listserv.ieee.org
> Subject: Re: [BP] [bp] 802.3 approves Backplane Ethernet PAR
>
>
> Adam,
>
> 2nd message. I am still curious about what happened.
> Would you enlighten me/us? Thanks.
>
>
> best regards,
>
> Yong.
>
> ============================================
> Yongbum "Yong" Kim Direct (408)922-7502
> Technical Director Mobile (408)464-6386
> 3151 Zanker Road Fax (408)922-7530
> San Jose, CA 95134 Main (408)501-7800
> ybkim@broadcom.com www.broadcom.com
> ============================================
>
> -----Original Message-----
> From: Yong Kim
> Sent: Friday, March 26, 2004 9:20 AM
> To: STDS-802-3-BLADE@listserv.ieee.org
> Subject: Re: [BP] [bp] 802.3 approves Backplane Ethernet PAR
>
>
> Adam,
>
> I am not clear on how the words in parenthesis got
> there, i.e.
> "(please consider for approval under continuous process)". Also,
> I had heard that some votes reversed study groups' votes that required
> 75% with mere 50% vote in 802.3. Would you clarify what
> happened during
> the backplane Ethernet agenda in .3 closing plenary? Sorry, I had to
> leave right before your agenda item to catch evening flight back.
>
>
> best regards,
>
> Yong.
>
>
> -----Original Message-----
> From: Healey, Adam B (Adam) [mailto:ahealey@agere.com]
> Sent: Sunday, March 21, 2004 4:04 PM
> To: stds-802-3-blade@ieee.org
> Subject: [bp] 802.3 approves Backplane Ethernet PAR
>
>
>
> Colleagues,
>
> I am pleased to report that 802.3 voted to approve the PAR and forward
> it to the SEC and NesCom. The motion:
>
> 802.3 WG approve the Backplane Ethernet PAR, as per par_0104.pdf, and
> forward the PAR to the 802 SEC and NesCom for approval
> (please consider
> for approval under continuous process).
>
> TECHNICAL (75%)
> Moved - A. Healey on behalf of the Study Group
> Second - N/A
> 802.3 Voters (Y/N/A): 55/0/0
> MOTION PASSES
>
> Under continuous processing, there is a chance that our first
> Task Force
> meeting will be in May. Even if this does not occur, we will still be
> meeting as a Study Group. The meeting details are still being worked
> out, but it will take place the week of May 24 in either Monterey or
> Long Beach, CA. Meeting details and a work plan will be
> posted as soon
> as they are available.
>
> Thank you all for the great work.
>
> Adam Healey
> Chair, Backplane Ethernet Study Group
>