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

Re: FW: [EFM] EFM Requirements

On Fri, 17 Aug 2001 18:34:59 -0700  "ramu" wrote:
> Harry, regarding your comment: "The (DOCSIS) protocol is rock solid
> and it works in millions of CM's deployed."
> Rock solid? You have a funny definition of solid. There have been over
> 300 ECNs (engineering change notices) accepted against it in the past
> few years, and it is hardly dropping off to a trickle today.

Please lose the sarcasm.  There may be some forums where it is an
acceptable rhetorical device; a standards discussion is not one of
them.  Sarcasm is good for scoring points; but is counter productive
in reaching informed consensus.

> I am all for reuse if it makes sense. However I have yet to see such
> an argument that spells out DOCSIS specifics. Clearly you cannot be
> endorsing DOCSIS to the letter, so what exactly are you endorsing
> about DOCSIS? Details please, not high-level concepts.

Any time the qualifier "clearly" is used in a statement, it is a
warning signal of a weak argument.  If the statement was really
"clear", then there would be no need for the qualifier.  "Clearly" is
used when the argument is far from clear and the writer is trying to
cover over the weakness using "proof by vigorous assertion".

I may have missed an email, but reviewing Harry's statements in the
messages I have, I don't believe he ever stated that he endorsed
DOCSIS in any way; he was offering a correction on how many DOCSIS
chip vendors there are.  If one's goal is to reach consensus, it is a
mistake to attribute straw dog statements to one's fellow consensus
seekers... especially just to shoot them down.

I agree with your focus on concrete details, it is a good way to reach
an informed consensus.  I would also suggest we try to base arguments
on verifiable facts.  Harry's stated one verifiable, concrete fact ( a
list of four DOCSIS chip vendors); your reply did not contain many
concrete, verifiable facts.  Your reply did contain a large amount of
opinion, vigorously stated.  Vigorous opinion is a weak foundation for
a standard (and yes, I am aware most standards have weak foundations.)

Here is my current argument on compatibility with DOCSIS:

1) As of 3/1/01 there were 4.7 cable modems deployed in the US,
   compared to 2.2 DSL modems [ "Get Wired: Why cable will beat the
   Bells in the race to wire your home"; Jonathan R. Laing;
   Barrons 8/20/01;]

2) That article also states a variety of different analysts opinions
   that future growth will favor cable modems over DSL.  I have always
   found the analysts opinions not worth the bits they are written on-- 
   I didn't believe them when they thought that DSL was a winner and
   I believe that telcos have a chance to change and catch up-- but
   let us say that cable will be at least as effective as DSL.  Now
   DSL has a 32% market share with CM having a 68% share.  In a few
   years, when EFM is ready to roll out at all data rates, it is likely
   that at least 50% of the installed base will be DOCSIS based.

From the above, I infer that:

1) DOCSIS is either an advantage or at least not a disadvantage to CMs.

2) DOCSIS 1.0 is a mature protocol with the largest market share, >65%,
   of installed devices.

3) DOCSIS chips have and will maintain economies of scale.

4) It would be an advantage to a device intended to supplant DOCSIS
   CMs to work well with DOCSIS based backend systems.
From these inferences, I would suggest the following requirements:

1) That EFM be DHCP compatible.  If there is any configuration of the
   client premises equipment (CPE), it be via DHCP.  the CPE should
   support the features that DOCSIS supports, such as "relay agent".

2) That if a configuration is loaded for the CPE, the loading follow
   the DOCSIS protocol state machine and message formats and tftp be

3) That if the CPE is capable of being monitored, that SMNP be used
   and that the DOCSIS MIB be supported.

None of the above requirements force the CPE to have any
intelligence.  My personal current opinion is that it would be
better to have the CPE be a very simple Ethernet bridge device.  In
that case, these requirements can be met by doing nothing.  However,
for providers with a strong DOCSIS base, it means all their backend
systems will have to be replaced.  Think of all the dial-up providers
that wanted to do DSL using their legacy dial-up backends.  Radius
auth, PPPoE and PPPoA are a greater abomination that DOCSIS will ever
be.  Ooophs, apologies for slipping the opinon vigorously stated as
fact in there....

I am sincerely interested in your response.

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