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

Re: [EFM] What Ethernet are we talking about?

Howard Frasier wrote:
> The Project Authorization Request (PAR) for 802.3ah defines the scope
> of the project as:
>      Define 802.3 Media Access Control (MAC) parameters and
>      minimal augmentation of the MAC operation, physical layer
>      specifications, and management parameters for the transfer
>      of 802.3 format frames in subscriber access networks at
>      operating speeds within the scope of the current IEEE Std
>      802.3 and approved new projects.
> Note that there is no mention of adding priorities, extending VLAN
> tag fields, DOCSIS head end compatibility, tftp, encryption, security, 
> QoS, port authentication, interoperability with 802.17, voice, video,
> or many of the other issues that have been raised on this reflector
> lately.

Carlos Ribeiro wrote:
> For now, we have a pretty complete 'service requirement' draft, that does
> not depend on the particular technology being proposed; specific
> requirements for EPON; and to a lesser extent, requirements for EP2P. We
> can finish out the details pretty quickly given that we have a direction.
> Please reply to us privately either or to the group if you have any
> position on the issues above.


	Given the shortness of time, a public reply.  First of all, I
want to thank you for the high quality of your contributions to this
discussion; they have been well reasoned and well written.

        Second of all, I want to respectfully register dissent, with
the 'carrier' draft position in regards to the scope of this standards
project and with the unstated assumption that your group represents
all segments of the carrier market (apologies in advance if I am
reading too much into this.)

        When I read your comments, it appears to me that on at least
an implicit level, the draft requirements specify something pretty
close to circuit switched Ethernet networks.  To me, looks and smells
a lot like ATM or OSI. Traditionally, Ethernet has been much closer to
packet switching. There are those who think that Ethernet's packet
switched heritage is the major reason for its success.

        We are packet switch folks and don't think much of circuits;
we are also a non-legacy carrier.  We disagree with almost every
assumption telcos, cable companies and other legacy carriers make
about data communications.

        Circuit switched vs packet switched is something of a
religious argument.  The forum for such disagreements should be/is the
marketplace, not a standards committee.  As a group, we should not
embed our assumptions about how business should be done in a low level
protocol specification such as 802.3ah.  Rather, let us keep to the
existing project authorization, and add any facilities we need for our
particular business method to auxiliary specifications.  That way, you
can use the equipment to do business your way, and we can use the same
equipment (at least the same CPE equipment) to do things our way.
This is a win for all.

        If legacy carriers want to block peer-to-peer networking as
your group as proposed, they should be able to.  However, we should be
able to allow peer-to-peer and use it as a potential advantage against
you all.  If legacy carriers want to do VLANs and all that, you should
be able to, and use that to compete with us.  But, if it turns out the
VLAN cost is more than the value added, we should not pay the cost.
We all should be able to differentiate our services and let the best
network win.

	I am afraid my dissent comes across as more confrontational
than I really mean; I don't write well.  I do want to stress again
that I respect your position and your reasoning for that position.
However, I do want to make clear that there is an alternate view,
which is also well reasoned, starting from a different set of
assumptions.  During the 1980's, I watched the OSI standards process,
to its detriment, ignore the data packet switching point of view.
From 1989-1992, I watched the ATM standards process, to its detriment,
ignore the data packet switching point of view.  If either of those
standards had been a success, we wouldn't be having this discussion.
I really hope it doesn't happen again to 802.3ah.

thank you again,
Fletcher Kittredge
Great Works Internet
8 Pomerleau St.
Biddeford, ME 04005