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

Re: [EFM] Re: [EFM-Copper] Long Reach Copper Presentations in Vancouver





As I said in my e-mail earlier I think it is important to hear Barry's
presentation.
I know how to rank each proposal technically on their own terms and the
market segment they are addressing.
From what I have seen I know which one of the two proposals (A or S) gives
better results for a given bundle composition and what are majority of
binder composition look like.

As for your desire to know how I vote, I don't think it is proper to discuss
voting before hearing all the conversations on the floor . I do not want to
influence anybody's preference one way or another and I do not share this
kinds of things publicly until at the meeting. If anybody wants to know how
I view things I would be happy to share them privately.

This is the last you will hear from me on this e-mail tread

Happy new year everybody.


-br


----- Original Message -----
From: "Matt Squire" <MSquire@HatterasNetworks.com>
To: "Behrooz Rezvani" <brezvani@ikanos.com>; "Hugh Barrass"
<hbarrass@cisco.com>
Cc: "Jonathan Thatcher" <Jonathan.Thatcher@worldwidepackets.com>; "Howard
Frazier" <millardo@dominetsystems.com>; <stds-802-3-efm-copper@ieee.org>;
<stds-802-3-efm@ieee.org>
Sent: Saturday, December 28, 2002 5:25 AM
Subject: RE: [EFM] Re: [EFM-Copper] Long Reach Copper Presentations in
Vancouver



Once you start talking about Markovian differential thingamajigs, you lose
me quick.

From a technical view, the proposals are completely independent.  I don't
think there's an argument there.

From a political view, any two things can be tied together.  All we have to
do is look at how spending on pig intestines can be tied to a defense bill
in congress.  But the things have nothing to do with eachother.  Its just
that sometimes people decide not to vote on the technical merits and instead
want something in return.

But you deftly avoided answering the original questions.  Given that you
support (AnnexJ + SHDSL) as the long-reach PHY, then
1) On the technical merits, would you vote for AnnexJ as a long-reach PHY?
2) On the technical merits, would you vote for SHDSL as a long-reach PHY?

Clearly you believe (AnnexJ+SHDSL) is the best option.  I have no problem
with that.  But given that position, would you vote 'yes' on either proposal
independently, even if you think (a) that the independent proposals are not
optimal, or (b) that the combined technologies have a better chance of
passing.  I'm not asking what's best or what will pass, but simply on the
merits of each PHY independently, would you vote yes or no on them?

And I guess all of this is off of your original point about Barry's
presentation.  Sorry for the left turn.

- Matt

> -----Original Message-----
> From: Behrooz Rezvani [mailto:brezvani@ikanos.com]
> Sent: Saturday, December 28, 2002 3:45 AM
> To: Matt Squire; Hugh Barrass
> Cc: Jonathan Thatcher; Howard Frazier; stds-802-3-efm-copper@ieee.org;
> stds-802-3-efm@ieee.org
> Subject: Re: [EFM] Re: [EFM-Copper] Long Reach Copper Presentations in
> Vancouver
>
>
> actually I believe you are not only bright but also very clever :)
>
> The reason I don't agree with your analysis because this is
> not purely an
> independent process, it is more of conditional probability or
> Markov process
> with memory (partial differential equation in case of motion
> madness where
> things are moving really fast)
>
> you can skip the section between the dashed lines if you want
>
>
> ---------------------------------------------------------
> <<consider a  a card game:
>
> let's assume (black Jack) there are two independent player
> (ADSL and SHDSL)
> mutually exclusive and each of them are sitting at 13
> (votes), and the third
> player (ADSL+SHDSL) hand is sitting at 16. There are 8 more
> cards to draw,
> and based on luck we know that there are 3 10s, 4 9s and 1 5 in the
> remaining deck. If the 3rd guy wins he has promised he will share the
> winning with the other 2 guys. The safest bet for any player
> is of-course to
> bet against your opponent, meaning encouraging him to draw a
> card (or vote
> YES in this case). while that player will not draw(vote NO
> for all options)
> the best bet for any player is not to draw). This ensures
> favorable outcome
> for player 3 and he then shares the winning with 1 and 2.
> However if one guy
> wants to whole pie, i.e. an SHDSL guy, or ADSL, then he
> encourages the 3rd
> guy [only (ADSL+SHDSL) guy, since the other one (A or S guy)
> will not do
> that] to draw a card (or vote YES), At this point 3rd guy (A+S) should
> actually vote "NO" to all options to keep status quo if he
> wants to get some
> sense into the other 2 guys (when the 3rd guys is purely
> independent) or
> when there are some preferences (partially independent or dependent
> probability) then he votes YES for his favorite dependent
> choice and NO for
> not so favorite choice>>
> --------------------------------------------------------------
>
>
> For those who were present at CU STF for comment resolution
> of CL63, you may
> recall this is exactly what happened. Each group voted down
> (NO) of any
> changes for the other group.
>
> ANYWAY going back to my original point: I just wanted to hear Barry's
> presentation in favor A+S.  I think we can gain by listening
> to Barry's
> presentation. Maybe the outcome of the votes will not change
> or it may have
> some positive impact.
>
> BestRegards
>
> -br
>
>
>
>
> ----- Original Message -----
> From: "Matt Squire" <MSquire@HatterasNetworks.com>
> To: "Behrooz Rezvani" <brezvani@ikanos.com>; "Hugh Barrass"
> <hbarrass@cisco.com>
> Cc: "Jonathan Thatcher"
> <Jonathan.Thatcher@worldwidepackets.com>; "Howard
> Frazier" <millardo@dominetsystems.com>;
> <stds-802-3-efm-copper@ieee.org>;
> <stds-802-3-efm@ieee.org>
> Sent: Friday, December 27, 2002 7:11 PM
> Subject: RE: [EFM] Re: [EFM-Copper] Long Reach Copper Presentations in
> Vancouver
>
>
>
> I'm clearly not the brightest person on the list because I'm
> very confused.
> Each person likely has one of four opinions
>
> a) AnnexJ good, SHDSL bad
> b) SHDSL good, AnnexJ bad
> c) AnnexJ good, SHDSL good
> d) AnnexJ bad, SHDSL bad
>
> Its a pretty simple binary relationship on two variables.  In
> reality, the
> variables are independent, as one can like/dislike either
> technology on its
> own merits without regard to another technology.
>
> If your position is option (3) in the embedded notes (or
> option (c) above),
> then I assume you would vote thumbs up on both technologies if voted
> independently, correct?
>
> One may feel that in order to provide the broadest
> applicability, you need
> to have both technologies (and possibly infinite other DSL
> variants already
> invented or to be invented).  But each technology be voted
> independently,
> and if enough people feel the same way, then multiple
> technologies will
> achieve consensus.
>
> I guess my questions to you are
> (a) would you vote thumbs up/down on AnnexJ?
> (b) would you vote thumbs up/down on SHDSL?
>
> Given your preference for option (3), I have to assume you'd
> vote 'yes' on
> each choice, correct?
>
> - Matt
>
> > -----Original Message-----
> > From: Behrooz Rezvani [mailto:brezvani@ikanos.com]
> > Sent: Friday, December 27, 2002 7:07 PM
> > To: Hugh Barrass
> > Cc: Jonathan Thatcher; Howard Frazier;
> stds-802-3-efm-copper@ieee.org;
> > stds-802-3-efm@ieee.org
> > Subject: [EFM] Re: [EFM-Copper] Long Reach Copper Presentations in
> > Vancouver
> >
> >
> >
> > Hi Hugh,
> >
> > I think it is important to take few minutes of our colleagues
> > time at .3ah
> > to clarify what are the real options, given that there are
> > over 1 Billion
> > lines of copper out there.  I know there is pressure for us
> > copper lads to
> > move along, but I don't think the issues are very clean like
> > it is in other
> > areas such as Optics, and by the way they have been able to
> > settle down on
> > multi-port types for a given objective. If we all feel we
> > cannot have enough
> > time to discuss important copper issues then one option
> that has been
> > discussed a couple of times before in the .3ah task force is
> > whether we need
> > to de-couple copper from optics similar to what was done at Ge.
> >
> > Regarding your comment which is it that I am advocating:
> > In my opinion choice 3 is the only realistic way to move
> > forward. To see why
> > we need to see Barry's presentation which will act as
> > catalyst to bring some
> > of neutral opinions into A+S option. I am not sure if someone
> > who votes for
> > A also votes for S and vice-versa. For example it is not in the best
> > interest the person who likes ADSL(SHDSL) to vote for
> > SHDSL(ADSL). I think
> > if option A or option B don't pass the only chance we have is
> > to convince
> > the enthusiasts on both parts to compromise and let the
> other one also
> > address the Ethernet market thru (A=residential market) or
> (S=business
> > market)... some sort of equilibrium.
> >
> >
> > Best regards
> > Behrooz
> >
> >
> > ----- Original Message -----
> > From: "Hugh Barrass" <hbarrass@cisco.com>
> > To: "Behrooz Rezvani" <brezvani@ikanos.com>
> > Cc: "Jonathan Thatcher"
> > <Jonathan.Thatcher@worldwidepackets.com>; "Howard
> > Frazier" <millardo@dominetsystems.com>;
> > <stds-802-3-efm-copper@ieee.org>;
> > <stds-802-3-efm@ieee.org>
> > Sent: Friday, December 27, 2002 2:54 PM
> > Subject: Re: [EFM-Copper] Long Reach Copper Presentations
> in Vancouver
> >
> >
> > > Behrooz,
> > >
> > > I don't think you understood what Jonathan was saying. I
> > think you are
> > also
> > > forgetting that we, copperheads, are imposing on the Task
> > Force to take up
> > their
> > > time to discuss this issue. Most people in the room will not be
> > passionately
> > > interested in copper issues and I think we should optimize
> > the use of
> > their
> > > time. What you are suggesting is that we should ask people
> > to sit through:
> > >
> > > 1. Presentation saying "rah, rah, ADSL is great"
> > > 2. Presentation saying "rah, rah, SHDSL is great"
> > > 3. Presentation saying "rah, rah, ADSL is great, SHDSL is great"
> > >
> > > It seems obvious that number 3 is redundant.
> > >
> > > The job of the ADSL presentation is to prove to the Task
> > Force that ADSL
> > meets
> > > all of the criteria and there is need for ADSL as a PHY
> within EFM.
> > >
> > > The job of the SHDSL presentation is to prove to the Task
> > Force that SHDSL
> > meets
> > > all of the criteria and there is need for SHDSL as a PHY
> within EFM.
> > >
> > > After the presentations, you may vote for ADSL (or not) and
> > vote for SHDSL
> > (or
> > > not). Clearly, the Task Force may approve just one PHY,
> both PHYs or
> > neither
> > > PHY. Are you suggesting that we should allow a "vote for neither"
> > presentation
> > > also?
> > >
> > > Hugh.
> > >
> > > Behrooz Rezvani wrote:
> > >
> > > > Jonathan, All
> > > >
> > > > Happy New Year
> > > >
> > > > I had a chance to look at Barry's presentation and I
> > think it has solid
> > > > arguments and very clear thinking on merits of having
> > both line code as
> > a
> > > > vehicle to deliver Ethernet in the access network and it
> > specifically
> > also
> > > > addresses some of the questions you suggested.
> > > >
> > > > I think the issues in the public access network is a
> > fairly complex one.
> > > > <There are couple of examples to remember. The success of
> > wireless LAN
> > and
> > > > not so successful Wireless Access (MMDS/LMDS type)>
> > > >
> > > > I think the TF 802.3ah would benefit by hearing that discussion.
> > > >
> > > > Behrooz
> > > >
> > > > ----- Original Message -----
> > > > From: "Jonathan Thatcher"
> <Jonathan.Thatcher@worldwidepackets.com>
> > > > To: "Behrooz Rezvani" <brezvani@ikanos.com>; "Howard Frazier"
> > > > <millardo@dominetsystems.com>; <stds-802-3-efm-copper@ieee.org>
> > > > Cc: <stds-802-3-efm@ieee.org>
> > > > Sent: Wednesday, December 18, 2002 3:51 PM
> > > > Subject: RE: [EFM-Copper] Long Reach Copper Presentations
> > in Vancouver
> > > >
> > > > All,
> > > >
> > > > I would also like to commend Howard on a well thought out
> > note and his
> > > > leadership here.
> > > >
> > > > Behrooz,
> > > >
> > > > I fear that this has the potential to confuse the issue. The key
> > question
> > > > remains: Do both, one, or neither of the two proposals
> meet the 5
> > criteria
> > > > and satisfy the agreed upon objectives? This question
> > needs our full
> > focus
> > > > and attention.
> > > >
> > > > jonathan
> > > >
> > > > | -----Original Message-----
> > > > | From: Behrooz Rezvani [mailto:brezvani@ikanos.com]
> > > > | Sent: Wednesday, December 18, 2002 3:35 PM
> > > > | To: Howard Frazier; stds-802-3-efm-copper@ieee.org
> > > > | Cc: stds-802-3-efm@ieee.org
> > > > | Subject: Re: [EFM-Copper] Long Reach Copper
> > Presentations in Vancouver
> > > > |
> > > > |
> > > > |
> > > > | Howard,
> > > > |
> > > > | Very well put.
> > > > |
> > > > | One note: Assuming fairly equal enthusiasm and charm on both
> > > > | sides of this
> > > > | discussion, would it be possible for some one (i.e. Barry or
> > > > | Nelson) to give
> > > > | a short presentation (10 minutes) on the advantages of having
> > > > | both solutions
> > > > | (option 3). This is because the first 2 presentations (Artman
> > > > | and Jackson)
> > > > | are focusing on how good individually are on their own in
> > > > | their respective
> > > > | markets.
> > > > |
> > > > | Thanks
> > > > | Behrooz
> > > > |
> > > > |
> > > > | ----- Original Message -----
> > > > | From: "Howard Frazier" <millardo@dominetsystems.com>
> > > > | To: <stds-802-3-efm-copper@ieee.org>
> > > > | Cc: <stds-802-3-efm@ieee.org>
> > > > | Sent: Wednesday, December 18, 2002 10:09 AM
> > > > | Subject: [EFM-Copper] Long Reach Copper Presentations
> > in Vancouver
> > > > |
> > > > |
> > > > | >
> > > > | > Dear Members of the IEEE 802.3ah EFM Copper Sub Task Force,
> > > > | >
> > > > | > There has been considerable discussion about the Long
> > Reach Copper
> > > > | > presentations that are being planned for the upcoming
> > meeting in
> > > > | > Vancouver.  I previously announced that we would reserve the
> > > > | > entire day on Monday, January 6th for an "all hands"
> > meeting of
> > > > | > the 802.3ah EFM Task Force to consider several "big
> > ticket" items
> > > > | > that require the attention and involvement of all of
> > the Task Force
> > > > | > members.  One of these "big ticket" items concerns
> > our long reach
> > > > | > copper PHY objective.
> > > > | >
> > > > | > Clearly, we need to resolve the question of how we
> > are going to meet
> > > > | > the long reach objective. Adhering to the motion that we
> > > > | passed last July,
> > > > | > that limits our consideration to those proposals based on
> > > > | the Artman and
> > > > | > Jackson presentations (advocating PHYs based on ADSL
> > Annex J and
> > > > | > g.shdsl, respectively), the task force has a finite
> > set of choices:
> > > > | >
> > > > | > 1) Adopt the ADSL Annex J proposal (with
> appropriate updates)
> > > > | > 2) Adopt the g.shdsl proposal (with appropriate updates)
> > > > | > 3) Adopt both proposals
> > > > | > 4) Adopt neither proposal
> > > > | >
> > > > | > It is obvious to me that choice # 4 above is the least
> > > > | desireable outcome.
> > > > | > It is also the default outcome, because the first three
> > > > | choices require a
> > > > | > positive vote, while # 4 represents the status quo ante.
> > > > | > In the hope that the Task Force can reach a >= 75%
> > concensus on
> > > > | > one of choices # 1-3, I request that we invest all of our
> > > > | efforts in the
> > > > | > task of producing EXCELLENT material in support of
> > ADSL Annex J,
> > > > | > and EXCELLENT material in support of g.shdsl.
> > > > | >
> > > > | > Each of these proposals must stand on its own, and
> > must satisfy the
> > > > | > 5 Criteria. Each proposal must demonstrate that it has a
> > > > | Broad Market
> > > > | > Potential, that it is Compatible with 802.3 and 802,
> > that it has a
> > > > | > Distinct Identity, that it is Technically Feasible,
> > and that it is
> > > > | > Economically Feasible.
> > > > | >
> > > > | > I have heard some individuals argue (quite eloquently) that
> > > > | both proposals
> > > > | > must be adopted in order to satisfy the Broad Market
> > > > | Potential criterion.
> > > > | > In my opinion, this is not the best argument to put
> > forward. Neither
> > > > | > 802.3ah
> > > > | > nor 802.3 will adopt a proposal that fails to satisfy
> > all of the 5
> > > > | > Criteria, and
> > > > | > I fear that by saying that both proposals are required to
> > > > | satisfy the
> > > > | Broad
> > > > | > Market Potential criterion, we imply that neither
> > proposal alone is
> > > > | > sufficient to
> > > > | > satisfy it.
> > > > | >
> > > > | > May I therefore strongly urge the proponents of each of the
> > > > | two proposals
> > > > | > to concentrate on putting forward the best possible
> > > > | arguments in support
> > > > | > of their proposal.  If the Task Force concludes that
> > both proposals
> > > > | satisfy
> > > > | > the 5 Criteria, and that both proposals should be adopted,
> > > > | then the Task
> > > > | > Force will vote accordingly.  I do not intend to entertain
> > > > | a "shoot out",
> > > > | > "choose one and only one" motion (though I may conduct a
> > > > | "beauty contest"
> > > > | > type of straw poll, where I ask the Task Force members to
> > > > | indicate their
> > > > | > favorite).  I intend to entertain motions on each of
> > the proposals
> > > > | > individually,
> > > > | > in the hope that the Task Force casts a >= 75% vote in
> > > > | favor of choice
> > > > | > 1, 2, or 3, above.
> > > > | >
> > > > | > One last note about the interpretation of our long reach
> > > > | objective: I
> > > > | > interpret our long reach objective, as we adopted it last
> > > > | July, to permit
> > > > | > only ONE PHY for long reach copper.  This would seem
> > to eliminate
> > > > | > choice # 3 as an option. Based on past history, I
> don't think
> > > > | > that we can successfully argue that choice # 3 really
> > > > | represents only
> > > > | > one PHY.  As I have said before, our Task Force
> > members may not each
> > > > | > possess a Ph.D. in digital signal processing, but they can
> > > > | all count to
> > > > | > two!
> > > > | >
> > > > | > Therefore, if we adopt choice # 3, I believe that we will
> > > > | have a follow on
> > > > | > task to justify the choice, and to modify our objective(s)
> > > > | accordingly.
> > > > | If
> > > > | > we adopt choice # 3 on Monday, January 6th, I will assign
> > > > | an action item
> > > > | > to the Copper Sub Task Force to carry out this task, and we
> > > > | will review
> > > > | > their work on Thursday, January 9th in general session. We
> > > > | will then have
> > > > | > to present the change(s) to the 802.3 Working Group
> > when it meets in
> > > > | March.
> > > > | >
> > > > | > Howard Frazier
> > > > | > Chair, IEEE 802.3ah EFM Task Force
> > > > | >
> > > > | >
> > > > | >
> > > > | >
> > > > |
> > > > |
> > >
> >
> >
>
>
>